Nitrogen in environment

United States
Department of
Agriculture
Economic
Research
Service
Marc Ribaudo, Jorge Delgado, LeRoy Hansen, Michael Livingston,
Roberto Mosheim, and James Williamson
Nitrogen in Agricultural
Systems: Implications for
Conservation Policy
Economic
Research
Report
Number 127
September 2011

wwwww
.ers.usda.govoo
For additional information on nitrogen management and conservation
policies, see:
www.ers.usda.gov/briefing/agchemicals/
www.ers.usda.gov/briefing/conservationpolicy/
www.ers.usda.gov/briefing/agandenvironment/
Recommended citation format for this publication:
Ribaudo, Marc, Jorge Delgado, LeRoy Hansen, Michael Livingston, Roberto
Mosheim, and James Williamson. Nitrogen In Agricultural Systems:
Implications For Conservation Policy. ERR-127. U.S. Dept. of Agriculture,
Econ. Res. Serv. September 2011.
Visit Our Website To Learn More!
The U.S. Department of Agriculture (USDA) prohibits discrimination in all its
programs and activities on the basis of race, color, national origin, age,
disability, and, where applicable, sex, marital status, familial status, parental
status, religion, sexual orientation, genetic information, political beliefs, reprisal,
or because all or a part of an individual’s income is derived from any public
assistance program. (Not all prohibited bases apply to all programs.) Persons
with disabilities who require alternative means for communication of program
information (Braille, large print, audiotape, etc.) should contact USDA’s
TARGET Center at (202) 720-2600 (voice and TDD).
To file a complaint of discrimination write to USDA, Director, Office of Civil
Rights, 1400 Independence Avenue, S.W., Washington, D.C. 20250-9410 or
call (800) 795-3272 (voice) or (202) 720-6382 (TDD). USDA is an equal
opportunity provider and employer.
Photo courtesy of USDA, Natural Resources Conservation Service.

United States
Department
of Agriculture
www.ers.usda.gov
A Report from the Economic Research Service
Abstract
Nitrogen is an important agricultural input that is critical for crop production. However, the
introduction of large amounts of nitrogen into the environment has a number of undesir-
able impacts on water, terrestrial, and atmospheric resources. This report explores the use
of nitrogen in U.S. agriculture and assesses changes in nutrient management by farmers
that may improve nitrogen use efficiency. It also reviews a number of policy approaches for
improving nitrogen management and identifies issues affecting their potential performance.
Findings reveal that about two-thirds of U.S. cropland is not meeting three criteria for good
nitrogen management. Several policy approaches, including financial incentives, nitrogen
management as a condition of farm program eligibility, and regulation, could induce farmers
to improve their nitrogen management and reduce nitrogen losses to the environment.
Keywords
Reactive nitrogen, nitrogen management, fertilizer, water quality, greenhouse gas, economic
incentives, conservation policy, regulation
Acknowledgments
This report benefited from the insightful comments provided by Keith Fuglie, Marca
Weinberg, and Mary Bohman of USDA’s Economic Research Service, Ralph Heimlich of
Agricultural Conservation Economics, Roberta Parry of the U.S. Environmental Protection
Agency, Jerry Hatfield of USDA’s Agricultural Research Service, USDA’s Natural Resources
Conservation Service and an anonymous reviewer. Thanks also to John Weber for excellent
editorial assistance and to Curtia Taylor for the design and layout.
Marc Ribaudo, mribaudo@ers.usda.gov
Jorge Delgado
LeRoy Hansen
Michael Livingston
Roberto Mosheim
James Williamson
Nitrogen in Agricultural
Systems: Implications for
Conservation Policy
Economic
Research
Report
Number 127
September 2011

ii
Nitrogen In Agricultural Systems: Implications For Conservation Policy / ERR-127
Economic Research Service/USDA
Contents
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Chapter 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter 2
Environmental Implications of Nitrogen and Goals
for Agricultural Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 3
State of Nitrogen Management on Cropland . . . . . . . . . . . . . . . . . . . . . . . . . 9
Chapter 4
Policy Instruments for Nitrogen Reduction . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter 5
Implications for Nitrogen Management Policies . . . . . . . . . . . . . . . . . . . . 46
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Appendix 1
Estimating Water Treatment Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Appendix 2
Using NLEAP To Model Nitrogen Loses . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Appendix 3
Estimating Changes in Nitrogen Fertilizer Application Rate . . . . . . . . . . . 72
Appendix 4
Comparing Costs of Farms Using Different
Nutrient Management Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Appendix 5
Estimating Wetland Restoration Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

iii
Summary
What Is the Issue?
Nitrogen is an agricultural input that is critical for crop production. Human-
induced production and release of reactive nitrogen has greatly affected the
Earth’s natural balance of nitrogen, contributing to changes in ecosystems,
both beneficial and harmful, including increased agricultural productivity
in nitrogen-limited areas, ozone-induced injury to crops and forests, over-
enrichment of aquatic ecosystems, biodiversity losses, visibility-impairing
haze, and global climate change. Incentives for encouraging farmers to adopt
improved nitrogen management can take many forms, from purely voluntary
to regulatory. Designing a cost-effective policy requires that factors influ-
encing fertilizer use be fully understood. Also, an understanding of how
farmers are likely to respond to different incentives may help policymakers
assess potential environmental tradeoffs driven by nitrogen’s ability to
change forms and cycle through different environmental media.
What Did the Study Find?
• Emission of reactive nitrogen to the environment can be reduced by
matching nitrogen applications more closely with the needs of growing
crops. This can be achieved by adopting three “best management prac-
tices” (BMPs):
• Rate: Applying an amount of nitrogen at a rate that accounts for all
other sources of nitrogen, carryover from previous crops, irrigation
water, and atmospheric deposits.
• Timing: Applying nitrogen as close to the time that the crop needs it
as is practical (as opposed to the season before the crop is planted).
• Method: Injecting or incorporating the nutrients into the soil to reduce
runoff and losses to the atmosphere.
• Among all U.S. field crops planted in 2006 that received nitrogen fertil-
izers, 35 percent are estimated to have met all three of the nutrient BMPs.
For the remaining cropland, improvements in management are needed to
increase nitrogen use efficiency (i.e., reduce the amount of nitrogen avail-
able for loss to the environment).
• Corn is the most intensive user of nitrogen fertilizer, on a per acre basis
and in total use. Fertilizer applied to corn is least likely to be applied in
accordance with all three BMPs.
• Incentives for improving nitrogen use efficiency by adopting the rate,
timing, and method BMPs can come from policy or market forces:
• Government programs that provide financial assistance for adopting
BMPs can be effective if they encourage the participation of farmers
with land most in need of improvement and if the farmers choose the
most cost-effective practices. Data suggest that the amount of crop-
land needing improvement would require a substantial increase in the
current Federal budget devoted to nutrient management practices.

iv
• Including nitrogen management in compliance provisions for receiving
Federal farm payments could encourage farmers to adopt more
effective management practices. In 2005, producers of U.S. corn
received Government payments that were much higher than the cost
of improving nitrogen management. The strength of this incentive,
however, has declined in recent years because of increases in crop
prices and a decline in direct commodity payments.
• Emissions markets, such as water quality trading and greenhouse gas
cap-and-trade, could provide financial incentives to farmers to adopt
improved nitrogen management and produce nitrogen credits that can
be sold in these markets. The effectiveness of such markets would
depend on market design, including rules defining who can participate
and what needs to be done to produce credits.
• Onfield improvements to nitrogen use efficiency could be supplemented
with off-field practices, such as wetlands restoration and vegetative filter
strips that can filter and trap reactive nitrogen that leaves the field through
surface runoff and groundwater flow. Of the two practices, restored
wetlands can be more cost effective at removing nitrogen and provide
additional environmental benefits, but they are limited to areas with suit-
able soils and hydrology. Vegetative filters can be employed more widely
across the landscape but are not effective when existing tile drains bypass
the filters.
• Policies for increasing nitrogen use efficiency should recognize the poten-
tial environmental tradeoffs when addressing particular issues related to
reactive nitrogen. Focusing strictly on one issue, such as nitrate leaching,
could lead to increased emissions of other nitrogen compounds, such as
nitrous oxide, even when total reactive nitrogen emissions are reduced.
How Was the Study Conducted?
ERS researchers used an extensive literature review, modeling, and data
from USDA’s Agricultural Resource Management Survey (ARMS) of major
field crops. ARMS data provided information on nitrogen use, defined
by the rate, method, and timing application criteria. This, in turn, helped
researchers determine the types of management improvements needed the
most.
The following market forces and policy instruments were evaluated to
measure their influence on nitrogen management: nitrogen fertilizer taxes,
Federal financial assistance offered to farmers to adopt practices that improve
nitrogen use efficiency or filter and trap nitrogen runoff, emissions markets
such as water quality trading and greenhouse gas cap-and-trade, compliance
with nitrogen BMPs as a condition for receiving farm program benefits, and
regulation.
Because reactive nitrogen is mobile and able to transform into different
compounds, researchers used a field-level nitrogen loss simulator developed
by USDA’s Agricultural Research Service to track how improving nitrogen
use efficiency by meeting all three BMPs affects emissions of different reac-
tive nitrogen compounds. These interactions were taken into account when
evaluating alternative policy options.

v
Glossary
ARMS – Agricultural Resource Management Survey
BMP – Best management practice
CEAP – Conservation Effects Assessment Program
EQIP – Environmental Quality Incentives Program
NUE – Nitrogen use efﬁciency
N – Nitrogen
N2 – Gaseous nitrogen
NO3 – Nitrate
NOx – Nitrogen oxides
N2O – Nitrous oxide
NH3 – Ammonia
Nr – Reactive nitrogen
NRCS – Natural Resources Conservation Service (USDA)
VFS – Vegetative ﬁlter strip

1
Chapter 1
Introduction
Most of the cropping systems in the world are naturally deficient in
nitrogen, making nitrogen inputs necessary to produce the crop yields
needed to support human populations. Gaseous nitrogen (N2) is abundant
in the atmosphere, but it cannot be used by living organisms unless it is
first converted into useable forms. Leguminous plants and soil microorgan-
isms contribute significant amounts of nitrogen used by crops, but yields
necessary to support growing populations need more nitrogen than can be
provided by natural means.
The Haber-Bosch process converts “unreactive” gaseous nitrogen from the
atmosphere into a biologically useable “reactive” form. The development of
the process in the early 1900s led to the massive production of relatively inex-
pensive nitrogen fertilizer that boosted crop yields (Follett et al., 2010). The
increasing use of reactive nitrogen in agriculture also increased the potential
for nitrogen to be lost to the environment as ammonia (NH3), ammonium
(NH4), nitrogen oxides (NOx), nitrous oxide (N2O), and nitrate (NO3);
these compounds are all reactive forms of nitrogen (Galloway et al., 2003).
Excessive amounts of reactive nitrogen inputs can lead to imbalances in the
natural movement of nitrogen among atmospheric, terrestrial, and aquatic
nitrogen pools, leading to disruptions in ecosystem function and the supply of
valuable ecosystem services.
Reactive nitrogen directly affects species composition, diversity, dynamics,
and the functioning of terrestrial, freshwater, and marine ecosystems (Matson
et al., 1997; Vitousek et al., 1997). Human-induced increases in reactive
nitrogen emissions to the environment may contribute to the following
harmful changes to ecosystems:
• Ozone-induced injury to crop, forest, and natural ecosystems
• Acidification and eutrophication (nutrient enrichment) effects on forests,
soils, and freshwater aquatic ecosystems
• Eutrophication and hypoxia (oxygen depletion) in coastal and lake
ecosystems
• Harmful algae blooms
• Biodiversity losses in terrestrial and aquatic ecosystems
• Regional haze
• Depletion of stratospheric ozone
• Global climate change
• Nitrate contamination of drinking water aquifers
A variety of steps can be taken to reduce the relatively large share of nitrogen
that is lost from agricultural systems. Improved management of nitrogen
fertilizers, animal manure, and other agricultural inputs can improve overall
nitrogen use efficiency (NUE) and reduce the loss of reactive nitrogen to the
environment while maintaining crop yields.

2
Incentives for encouraging farmers to adopt improved nitrogen manage-
ment can take many forms, from purely voluntary to regulatory. Designing
a cost-effective policy requires that factors influencing fertilizer use be fully
understood. Also, an understanding of how farmers are likely to respond to
different incentives may help policymakers assess potential environmental
tradeoffs driven by nitrogen’s ability to change forms and cycle through
different environmental media.
This report takes a broad view of several questions related to nitrogen
management: (1) Why is nitrogen management so important? (2) How
many acres of cropland are not using nitrogen best management practices
(BMP)? and (3) What are the strengths and weaknesses of alternative policy
approaches for improving nitrogen management on those acres?
Ideally, alternative policies would be assessed on the basis of the cost of
achieving a particular level of NUE across U.S. crop production. However,
physio-economic models that would allow for this type of assessment are not
available on a national scale. Instead, this analysis uses survey data to help
identify the number of acres of cropland that would benefit from improved
management and to assess the characteristics of each alternative policy
approach. Policy approaches are assessed in terms of factors consistent with
cost effectiveness, including flexibility, ability to target, crop acres covered,
and implementation costs. These factors are assessed through original
research and an extensive review of the literature.

3
Chapter 2
Environmental Implications of Nitrogen and
Goals for Agricultural Management
Agriculture is the predominant source of reactive nitrogen emissions into
the environment. In the United States, agriculture contributes 73 percent
of nitrous oxide emissions (EPA, 2010a), 84 percent of ammonia emissions
(EPA, 2010a), and 54 percent of nitrate emissions (Smith et al., 1997). Most
losses from cropland are attributable to runoff, ammonia volatilization,
nitrification and denitrification, and nitrate leaching (see box, “Pathways for
Nitrogen Losses”).
Nitrogen’s impacts on water resources (Dubrovsksy et al., 2010; Bricker et
al., 2007; Rabalais et al., 2002a, b), atmosphere (Cowling et al., 2002; Follett
et al., 2010), and terrestrial resources (Galloway et al., 2008) are extensive.
Estimates of the economic value of these damages are lacking. Crutchfield et
Pathways for Nitrogen Losses
Soil erosion - Nitrogen can be lost from the soil surface when attached to soil
particles that are carried off the field by wind or water. Although wind and water
erosion can be observed across all regions, wind erosion is more prevalent in
dry regions and water erosion in humid regions. Overall, little nitrogen is lost
through erosion when basic conservation practices are in place (Iowa Soybean
Association, 2008b).
Runoff - Surface runoff of dissolved nitrogen (generally in the form of nitrate)
is only a concern when fertilizer and or manure are applied on the surface and
rain moves the nitrogen before it enters the soil (Legg and Meisinger, 1982; Iowa
Soybean Association, 2008b).
Ammonia volatilization - Significant amounts of nitrogen can be lost to the at-
mosphere as ammonia (NH3) if animal manure or urea is surface applied and
not immediately incorporated into the soil (Hutchinson et al., 1982; Fox et al.,
1996; Freney et al., 1981; Sharpe and Harper, 1995; Peoples et al., 1995). Addi-
tionally, warm weather conditions can accelerate the conversion of manure and
other susceptible inorganic nitrogen fertilizers to ammonia gas.
Denitrification and nitrification - When oxygen levels in the soil are low, some
microorganisms known as denitrifiers will convert NO3 to nitrogen (N2) and
nitrous oxide (N2O), both of which are gases lost to the atmosphere (Mosier and
Klemedtsson, 1994). Nitrogen gas is not an environmental issue, but N2O is a
powerful greenhouse gas. Denitrification usually occurs when nitrate is present
in the soil, soil moisture is high or there is standing water, and soils are warm.
NOx and N2O gases can also be produced through a process called nitrification.
Leaching - Leaching occurs when there is sufficient rain and/or irrigation to
move easily dissolvable nitrate through the soil profile (Randall et al., 2008).
The nitrate eventually ends up in underground aquifers or in surface water via
tile drains and groundwater flow. Tile drains may be a chief passageway by
which nitrogen moves from crop soils to surface water (Turner and Rabalais,
2003; Randall et al., 2008; Randall et al., 2010).

4
al. (1995) estimate that consumers in four U.S. rural areas would be willing
to pay between $73 million and $780 million per year (in 1995 dollars) for
reduced chemical concentrations (including nitrate) in groundwater tapped
by private wells. Dodds et al. (2009) estimate that consumers spend over
$800 million each year on bottled water due to nutrient-related taste and odor
problems.
Using data from water treatment plants, ERS estimates the cost of removing
nitrate from U.S. drinking water supplies is over $4.8 billion per year (see
app. 1). Based on the contribution of nitrate loadings from agriculture (Smith
et al., 1997), agriculture’s share of these costs is estimated at about $1.7
billion per year. Most costs are borne by the large utilities, due to the volume
of water treated. ERS findings indicate that reducing nitrate concentrations in
source waters by 1 percent would reduce water treatment costs in the United
States by over $120 million per year.
Managing Nitrogen for Agriculture
and the Environment
USDA’s Natural Resources Conservation Service (NRCS) defines nutrient
management as managing the amount, source, placement, form, and timing
of the application of plant nutrients to the soil (USDA, NRCS, 2006).
Optimizing nitrogen management both economically and environmentally
requires farmers to perform a juggling act: Applying too much nitrogen cuts
into financial returns and increases the likelihood of nitrogen escaping into
the environment; applying too little increases the risk of reduced yields and
lost income.
Crop production is characterized by uncertain and stochastic, or random,
weather and soil conditions that affect crop yields and nitrogen loss. To main-
tain viable operations, farmers may manage temporal variability in weather
and soil nitrogen by overapplying nitrogen to protect against downside risk
(i.e., use an “insurance” nitrogen application rate) (Sheriff, 2005; Babcock,
1992; Babcock and Blackmer, 1992). Additionally, farmers may take a
“safety net” approach to maximize economic returns by setting an optimistic
yield goal for a given field based on an optimum weather year to ensure that
the needed amount of nitrogen for maximum yields is available (Schepers et
al., 1986; Bock and Hergert, 1991). Thus, during the years in which weather
is not optimal for maximizing yields, nitrogen will be overapplied from an
agronomic standpoint. Almost by definition, optimal conditions are infre-
quent, so farmers overfertilize crops in most years.
The following hypothetical example helps illustrate the reasoning behind
a farmer’s decision to apply a certain amount of fertilizer. Assume that a
farmer applies 179 pounds of nitrogen (N) per acre to his or her cornfield.
Under ideal conditions, the farmer might produce 170 bushels of corn per
acre. In most years, however, conditions are not ideal and production averages
148 bushels per acre. This yield requires only 165 pounds of N per acre, but
at this level, the farmer will miss out on an extra 22 bushels in the event of
ideal weather conditions. Assuming a fertilizer price of $0.50 per pound of
N, the extra N applied in an average year costs $7 per acre. Assuming a corn
price of $4.50 per bushel, the benefit from having enough nitrogen available
to take advantage of optimal conditions would be $99 per acre. In most years,

5
however, the extra fertilizer is not used by the crop and is available to leave
the field and affect environmental quality.
Definitions of Nitrogen Use Efficiency
Researchers calculate nitrogen use efficiency to assess the effectiveness of
nitrogen management. The NUE of a cropping system is the proportion of
all nitrogen inputs that are removed in harvested crop biomass, contained
in recycled crop residues, and incorporated in soil organic and inorganic
nitrogen pools (Cassman et al., 2002) (fig. 2.1). Nitrogen not recovered in
these nitrogen sinks is lost to the environment. Increases in NUE reduce
the share of nitrogen left in the soil and available for loss to water or the
atmosphere. Increased NUE is treated as a goal of environmental policy
throughout this report.
Recommended Input Rate and Nitrogen Credits
The nitrogen application rate has a major effect on NUE (Bock and Hergert,
1991; Meisinger et al., 2008; Freney et al., 1995; Power et al., 2001). Nitrogen
losses have been shown to increase rapidly when N inputs exceed assimila-
tion capacity (Vanotti and Bundy, 1994; Schlegel et al., 1996; Dobermann et
al., 2006; Bock and Hergert, 1991). Reducing application rates reduces the
losses of all forms of reactive nitrogen.
Figure 2.1
Nitrogen balance and nitrogen use efficiency
Nitrogen balance consists of N inputs of fertilizer and manure/legume N (NF
) and
miscellaneous atmospheric deposition (NMISC); outputs of crop harvested N (NCH), N
leaching (NL), erosion (NE), and gaseous losses (NG); and internal N pools of crop residue
N (NCR), soil organic N (NSON), soil inorganic N (NSIN), and net N mineralization (NMIN).
Nitrogen use efficiency is the proportion of all N inputs (NF
and NMISC) that are removed in
harvested crop biomass (NCH), contained in recycled crop residues (NCR), and incorpo-
rated into soil organic matter (NSON) and inorganic N (NSIN) pools. The remainder is what
is lost to the atmosphere through gaseous emissions (NG), leaching (NL), and erosion
(NE). The goal of nitrogen management is to reduce these losses through reductions in
fertilizer inputs and through soil, water, fertilizer, and crop management that affects the
cycling of nitrogen in the soil.
Source: USDA, Economic Research Service using data from Meisinger et al., 2008.
NG
NL
NF
NFNMISC
NCH
NE
NCR
NSON
NSIN
NMIN

6
The effectiveness of nitrogen management may be raised by accounting for
all nitrogen sources when determining a nitrogen fertilizer application rate.
Depending on the region, such sources may include inorganic nitrogen levels
in the root zone, soil organic content, previous crop (e.g., leguminous crop),
manure applications, irrigation water, and atmospheric deposition (Cassman
et al., 2002; Meisinger et al., 2008; Iowa Soybean Association, 2008a).
Method/Placement
The goal of appropriate method and placement of fertilizer is to provide
nutrients to plants for rapid uptake and to reduce the potential for losses
to the environment. Studies have shown that NUE can be doubled under
some conditions by placing fertilizers in the soil rather than “broadcasting”
them on the surface (Malhi and Nyborg, 1991; Power et al., 2001). Liquid or
gaseous forms of nitrogen can be injected directly into the soil with special-
ized equipment that is consistent with low-till systems. Solid forms can be
broadcast on the surface and immediately incorporated into the soil with
tillage equipment. Such placement reduces the risks of losses to the atmo-
sphere and through surface runoff. The method of application can also reduce
losses of nitrogen stemming from ammonia volatization (Meisinger and
Randall, 1991; Peoples et al., 1995; Fox et al., 1996; Freney et al., 1981).
The impact of fertilizer placement on nitrous oxide emissions is less certain.
Liu et al. (2006) found that injection of liquid urea ammonium nitrate at
deeper levels resulted in 40-70 percent lower N2O emissions than the rate
associated with shallow injection or surface application. Some studies,
however, have reported that incorporation into the soil increases N2O emis-
sions (Flessa and Beese, 2000; Wulf et al., 2002; Drury, 2006). Injection or
incorporation could also increase nitrate leaching, especially where soils are
coarse textured (Abt Associates, 2000).
Timing
The research on improving NUE in crop production emphasizes the need
for greater synchronization between crop nitrogen demand and the supply
of nitrogen from all sources throughout the growing season (Doerge et al.,
1991; Cassman et al., 2002; Meisinger and Delgado, 2002). Balancing supply
and demand implies maintaining low levels of inorganic nitrogen in the soil
when there is little plant growth and providing sufﬁcient inorganic nitrogen
fertilizer during periods of rapid plant growth (Doerge et al., 1991; Alva et
al., 2005). For example, the corn plant’s need for nitrogen is not very high
until about 4 weeks after it emerges from the ground, which typically falls in
June through July in the major corn-producing States (Baker, 2001). Ideally,
to ensure that growing crops have adequate N and to minimize losses from
the soil, a farmer could split nitrogen applications or “spoon feed” nitrogen
when using center-pivot sprinkler irrigation systems from June through July-
August, using information from soil tests and/or advanced remote sensing
techniques (Bausch and Delgado, 2003). Though splitting nitrogen applica-
tions is seen as an effective way to increase NUE and reduce nitrogen losses
to the environment, several factors must ﬁrst be considered: workload,
seasonal fertilizer price differences, the risk associated with not being able
to apply at the right time, application costs, the possibility of compacting the
soil, and possible damage to growing crops (Doerge et al. 1991; Westermann

7
and Kleinkopf, 1985; Westermann et al., 1988; Alva et al., 2005; Delgado and
Bausch, 2005).
Form
NUE is also influenced by the form of nitrogen fertilizer (Raun and Schepers,
2008; Freney et al., 1995). Some of the more widely used nitrogen fertilizer
forms include anhydrous ammonia (gas), urea (solid), UAN (liquid), and
manure (solid). These forms vary in how quickly they can be transformed
into nitrate, which is what crops actually use. The closer in time the fertilizer
is applied to when the crop needs it, the faster it needs to be transformed into
nitrate. A mismatch of fertilizer form with appropriate timing can lead to
large environmental losses and poor yields.
Manure Effects
Manure is an important source of N, but it poses challenging management
problems (Eghball et al., 2002; Kirchmann and Bergstrom, 2001; Davis et
al., 2002). The nitrogen content of manure depends on the animal type and
the method of manure storage (Davis et al., 2002; Eghball et al., 2002), and
nitrogen content may be inconsistent from batch to batch (Davis et al., 2002).
Manure is more difficult to handle than inorganic nitrogen fertilizers, and,
if in solid form, is difficult to apply uniformly. Most of the nitrogen content
of manure is in the organic form and has to be mineralized before crops can
use it. Since the transformation process depends on manure type, soil, and
weather conditions, it is more difficult to control soil nitrate levels relative
to crop needs when manure is applied than when other forms are applied
(Eghball et al., 2002; Power et al., 2001). Consequently, controlling environ-
mental losses from manured fields is more difficult than from fields using
commercial fertilizer.
Off-Site Practices That Capture Nitrogen
Off-field conservation measures can be used in conjunction with onfield
nitrogen management to either capture reactive nitrogen in biomass or convert
it to inert N2 through denitrification. Examples of off-site practices include
vegetative buffers or filters and restored and constructed wetlands (Hefting
et al., 2003; Jacobs and Gilliam, 1985; Lowrance et al., 1984). Buffers and
wetlands reduce nitrogen loads to water through plant uptake, microbial
immobilization and denitrification, soil storage, and groundwater mixing
(Pionke and Lowrance, 1991; Lowrance et al., 1997; Hey et al., 2005; Mayer
et al., 2005).
Buffers can remove nitrogen from both surface flow and groundwater (Mayer
et al., 2005; Angier et al., 2001; Randall et al., 2008; Mitsch and Day, 2006).
The effectiveness of vegetative buffers depends on the size of the buffer,
the density of vegetation, and hydrologic conditions within the buffer zone
(Dosskey et al., 2005; 2007). Based on a wide range of studies, Mayer et
al. (2005) estimate that buffers can remove about 74 percent of the nitrogen
passing through the buffer root zone. However, in many areas of the country
where tile drains are used to control the water table, especially in the Corn
Belt, subsurface flows pass below the root zone and are not filtered by vegeta-
tive buffers.

8
Restored wetlands have been shown to be effective at reducing the transfer of
nitrogen from agricultural land to water bodies (Jansson et al., 1994) and have
been proposed as a technique to remove reactive nitrogen from the environ-
ment (Hey et al., 2005; Mitsch and Day, 2006). Wetland vegetation uptakes
nitrogen, and wet soils enhance denitriﬁcation. The effectiveness of wetlands
as a ﬁlter of reactive nitrogen depends on their size, seasonal weather condi-
tions, and hydrologic characteristics. Wetlands also provide a host of other
ecosystem services that are valued by society, such as wildlife habitat and
carbon sequestration.

9
Chapter 3
State of Nitrogen Management
on Cropland
Nitrogen Management on U.S. Cropland
Data on the nutrient management practices of U.S. producers of barley, corn,
cotton, oats, peanuts, sorghum, soybeans, and wheat (table 3.1) are derived
from USDA’s Agricultural Resource Management Survey (ARMS) (see
box, “Agricultural Resource Management Survey”). The basic practices for
improving NUE are agronomic application rate, appropriate timing of appli-
cations, and proper placement (USDA, NRCS, 2006). For the purposes of this
analysis, these practices are defined as follows:
• Rate. Applying no more nitrogen (commercial and manure) than 40
percent more than that removed with the crop at harvest, based on the
stated yield goal, including any carryover from the previous crop. This
approach is consistent with a more traditional approach for estimating
N rate recommendations (Millar et al., 2010) and is also the criterion
used by NRCS in its assessment of conservation practices in the Upper
Mississippi Basin (USDA, NRCS, 2010). Crop uptake coefficients are
from NRCS (Lander et al., 1998, table 3.1). This agronomic rate accounts
for unavoidable environmental losses that prevent some of the nitrogen
that is applied from actually reaching crops.
Table 3.1
Crops, ARMS Phase II reference years, States surveyed, commodities, and nitrogen uptake
per unit of crop yield
Crop
Reference
year States surveyed Commodity
Lbs N
per unit Unit
Barley 2003 CA, ID, MN, MT, ND, PA, SD, UT, WA, WI, WY grain 0.9 bushel
Corn 2005
CO, GA, IL, IN, IA, KS, KY, MI, MN, MO, NE, NY, NC,
ND, OH, PA, SD, TX, WI
grain 0.8 bushel
silage 7.09 ton
Cotton 2003 AL, AZ, AR, CA, GA, LA, MS, MO, NC, SC, TN, TX lint plus seed 15.19 bale
Oats 2005 IL, IA, KS, MI, MN, NE, NY, ND, PA, SD, TX, WI grain 0.59 bushel
Peanuts 2004 AL, FL, GA, NC, TX
nuts with
pods
0.04 pound
Sorghum 2003 CO, KS, MO, NE, OK, SD, TX
grain 0.98 bushel
14.76 ton
Soybeans 2006
AR, IL, IN, IA, KS, KY, LA, MI, MN, MS, MO, NE, NC,
ND, OH, SD, TN, VA, WI
beans 3.55 bushel
Wheat
Winter
Other spring
Durum
2004
CO, ID, IL, KS, MI, MN, MO, MT, NE, ND, OH, OK, OR,
SD, TX, WA
grain 1.13 bushel
grain 1.39 bushel
grain 1.29 bushel
Notes: N = nitrogen. ARMS = USDA’s Agricultural Resource Management Survey. The nitrogen uptake coefficients are from Lander et al.
(1998). The coefficients for soft (1.02 lbs/bushel) and hard (1.23 lbs/bushel) winter wheat were averaged because the type of winter wheat
produced was not available. To download estimates based on these data, or to learn more about the surveys, go to www.ers.usda.gov/data/
arms/beta.htm.
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06) and Lander et al.
(1998).

10
• Timing. Not applying nitrogen in the fall for a crop planted in the spring.
• Method. Injecting (placing fertilizer directly into the soil) or incorpo-
rating (applying to the surface and then discing the fertilizer into the soil)
nitrogen rather than broadcasting on the surface without incorporation.
Form also plays a role in nitrogen management for improving NUE.
However, available data do not allow for an assessment of the form of
nitrogen fertilizer applied.
In this report, we evaluate nitrogen management only during the survey year
covered by ARMS data. The loss of nitrogen to the environment in a partic-
ular year is mostly a function of current and not past management decisions.
However, current management decisions have to account for past manage-
ment, such as planting of a legume. The amount of commercial nitrogen
applied is readily available from the ARMS responses; however, the amount
of manure nitrogen must be estimated. We base these estimates on the quan-
tity of raw manure applied, the form of the manure (liquid or solid), and the
animal source of the manure. We also note whether the previous crop was a
legume so as to account for the potential carryover of nitrogen.
A farm can fall into one of eight nitrogen management categories, defined by
the three management decisions in a particular year:
1. All of the criteria are followed.
2. The rate and timing criteria are followed.
3. The rate and method criteria are followed.
4. The timing and method criteria are followed.
5. Only the rate criterion is followed.
6. Only the timing criterion is followed.
7. Only the method criterion is followed.
8. None of the criteria are followed.
Agricultural Resource Management Survey
USDA’s Agricultural Resource Management Survey (ARMS) is an annual survey
of farm and ranch operators administered by ERS and the National Agricultural
Statistics Service (NASS). ARMS gathers data on field-level production prac-
tices, farm business accounts, and farm households. ARMS is a multiple-phase
survey. In the fall, NASS interviews producers of major commodities, such as
feed grains, food grains, or cotton, to collect information about production prac-
tices and land use on select fields. In the spring, NASS re-interviews farmers
who successfully completed the fall survey. Spring data collection focuses on
the structural and economic characteristics of the farm business and farm op-
erator household. This approach helps link commodity production activities and
conservation practices with the farm business and operator household.
Each phase of ARMS contains multiple versions of the survey questionnaire. The
commonality of questions across versions provides one facet of data integration.
In the fall data collection, the target commodity distinguishes questionnaires.

11
How Many Acres Treated With Nitrogen Met the Criteria for Best
Management Practices?
Because the crops covered in the analysis were surveyed in different years,
we specify a reference year, 2006, to examine the extent to which best
nitrogen management practices are being followed. Weights are calibrated
so that the weighted sums of acres planted by the surveyed crop producers
match USDA’s published estimates of planted acres for 2006 (USDA, NASS,
2008). This provides reasonable baseline estimates under the assumption that
the percentages of planted and treated acres and application rates by manage-
ment category were stable between the survey reference years (see table 3.1)
and 2006. We maintain this assumption throughout the analysis.
Sixty-nine percent of the 242 million acres planted to barley, corn, cotton,
oats, peanuts, sorghum, soybeans, and wheat in 2006 were estimated to be
treated with commercial and/or manure nitrogen (table 3.2). Corn accounted
for an estimated 45 percent of the 167 million crop acres treated with
nitrogen and 65 percent of the 8.7 million tons of nitrogen applied to these
crops during 2006.
The application rate criterion was not met on over 53 million acres treated
with nitrogen (32 percent). Cotton had the highest percentage of treated acres
not meeting the rate criterion (47 percent), followed by corn (35 percent).
However, corn accounted for 50 percent of all treated crop acres not meeting
the rate criterion.
The timing criterion was not met on over 40 million treated acres (24
percent). About 34 percent of treated corn acres received commercial and/or
manure nitrogen in the fall. These corn acres account for over 64 percent of
all treated crop acres not meeting the timing criterion.
Table 3.2
Planted and nitrogen-treated acres, nitrogen applied, and the shares of treated acres and applied nitrogen
that did not meet the rate, timing, or method criteria, by crop, 2006
Total Did not meet rate Did not meet timing Did not meet method
Planted
acres
Treated
acres Tons N
Treated
acres Tons N
Treated
acres Tons N
Treated
acres Tons NCrop
Thousands Percent
Barley 3,452 3,176 98 14 23 20 20 25 25
Corn 78,327 76,052 5,799 35 46 34 26 37 20
Cotton 15,274 12,566 591 47 61 18 11 32 24
Oats 4,168 2,748 93 33 49 28 32 42 41
Peanuts 1,243 737 14 1 7 16 11 39 29
Sorghum 6,522 5,370 220 24 31 16 16 27 21
Soybeans 75,522 16,827 248 3 31 28 56 45 43
Wheat 57,344 49,808 1,766 34 50 11 12 37 32
Total 241,852 167,285 8,829 32 47 24 23 37 24
Notes: N = nitrogen. These estimates are based on weighted sums, where the weights were calibrated so that the sums of planted acres for
each crop based on the survey data match published estimates of planted acres for 2006 (USDA, 2008).
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06), Phase II. See table
3.1 for details.

12
Nitrogen was not incorporated/injected on over 61 million treated crop
acres (37 percent). These acres received 24 percent of all applied nitrogen.
Soybeans (45 percent) had the highest percentage of acres not meeting the
method criterion. However, corn accounted for about 46 percent of all treated
acres not meeting the method criterion.
Corn acres make up nearly half of all acres that are in need of some type of
improvement in nitrogen management, in that at least one of the three criteria
is not met (fig. 3.1). Any policy aimed at improving nitrogen use efficiency
would have to consider the factors driving management decisions in corn
production.
From a regional standpoint, the Corn Belt and Northern Plains dominate in
terms of cropland not meeting the management criteria (figs. 3.2, 3.3). Not
coincidentally, these are the primary corn-growing areas in the United States.
However, in terms of nitrogen application in excess of the criterion rate, the
Corn Belt and Lake States receive the greatest amounts of excess nitrogen
(fig. 3.4).
As described in the previous chapter, NUE is highest when all three manage-
ment criteria are met. Table 3.3 shows the percentage of treated acres in each
nitrogen management category, as well as the degree to which excess nitrogen
is applied in relation to the rate criterion. About 35 percent (58 million
acres) of the treated acreage meet all three criteria. Corn has the smallest
percentage of treated acres meeting all three criteria (30.4 percent). Because
of the large amount of cropland planted to corn, this represents about half of
all crop acres needing improvement in nitrogen management (rate, timing,
or method). Only 4.2 percent of all treated acres do not meet any of the three
criteria.
About 47 percent of all treated crop acres meet the method and timing
criteria. Most of the acres exceeding the rate criterion do so by less than 50
Figure 3.1
Acres treated with commercial and/or manure nitrogen not using nitrogen best
management practices, 2006
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06),
Phase II. See table 3.1 for details.
Million treated acres
Barley Corn Cotton Oats Peanuts Sorghum Soybeans Wheat
0
10
20
30
40
50
60

13
percent. For example, about 14 percent of corn acres receive applications of
10 percent or less over the criterion rate. Reducing application rates on these
acres so that the rate criterion is met would mean that nearly 80 percent of
all corn acres would meet the rate criterion and that 35 percent of corn acres
would meet all three criteria.
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06), Phase II.
See table 3.1 for details.
Million treated acres
Figure 3.2
Acres treated with commercial and/or manure nitrogen not using nitrogen best management
practices, by region, 2006
0
5
10
15
20
25
30
35
40
Appalachia Corn
Belt
Delta Lake
States
Mountain Northeast Northern
Plains
Pacific Southeast Southern
Plains
Source: USDA, Economic Research Service.
Figure 3.3
USDA farm production regions
Pacific
Lake
Delta
Corn Belt
Mountain
Southeast
Northeast
Appalachia
Northern
Plains
Southern
Plains

14
It should be noted that our ﬁndings differ somewhat from those reported by
USDA’s Conservation Effects Assessment Project (CEAP) assessment of
the Upper Mississippi River Basin (USDA, 2010). The CEAP study reports
smaller percentages of cropland meeting the nitrogen management criteria.
The ERS study, however, examines nitrogen management for only the survey
year. The CEAP analysis looks at nutrient management practices over an
entire crop rotation, which may run from 2 to 5 years (see box, “CEAP
Analysis of Nitrogen Management in the Upper Mississippi River Basin”).
All three criteria had to be met in each year of the rotation for CEAP to
consider the cropping system as having met the nitrogen management goal.
The CEAP approach is stricter than that used by ERS.
Manure Use
Previous research has indicated that farms with animals tend to overapply
nutrients to crops, primarily because of the large amount of manure produced
on the farm needing disposal (Ribaudo et al., 2003; Gollehon et al., 2001).
ARMS data provide additional evidence that manure use is associated with
overapplication of nutrients. About 10 percent of crop acres treated with
nitrogen (treated acres) received manure. Ninety-three percent of treated
acres receiving manure did not meet all three criteria, compared with 62
percent of treated acres not receiving manure (table 3.4). Most of the cropland
receiving manure was used to grow corn (72 percent). Over 95 percent of the
corn acres receiving manure did not meet all three criteria, compared with 65
percent for corn acres not receiving manure.
1,000 tons excess nitrogen
Figure 3.4
Total nitrogen applications above criterion rate by region, 2006
Appalachia Corn
Belt
Delta Lake
States
Mountain Northeast Northern
Plains
Pacific Southeast Southern
Plains
Note: Criterion rate defined as nitrogen removed at harvest plus 40 percent.
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06),
Phase II. See table 3.1 for details.
36
298
1
185
7
44
84
1 5
18
0
50
100
150
200
250
300
350

15
Table 3.3
Percent treated acres by management category, crop, and degree of excess application, 2006
Rate criterion status
Timing or method criteria met
Timing and method Timing Method Neither
Percent of treated acres
At or less than criterion rate
Barley 52.0 16.4 13.0 4.3
Corn 30.4 15.0 12.0 6.2
Cotton 32.9 11.6 6.5 2.3
Oats 33.8 13.5 8.0 11.0
Peanuts 53.5 29.7 7.0 8.7
Sorghum 44.5 18.4 9.5 3.3
Soybeans 43.0 27.7 9.8 16.1
Wheat 36.8 22.2 5.2 1.7
Total 34.8 18.3 9.2 5.5
0 -10% over rate
Barley 1.8 1.1 0.7 0.1
Corn 4.6 2.0 4.2 3.3
Cotton 6.6 3.5 3.7 0.7
Oats 1.2 1.2 0.0 0.1
Peanuts 0.3 0.0 0.0 0.0
Sorghum 3.7 0.3 1.1 0.1
Soybeans 0.0 0.1 0.1 0.0
Wheat 4.4 2.5 0.8 0.1
Total 4.1 2.0 2.5 1.6
10-50% over rate
Barley 3.9 1.7 1.0 0.9
Corn 4.6 5.0 2.3 2.8
Cotton 10.4 7.2 1.8 0.8
Oats 6.4 2.3 0.5 1.6
Peanuts 0.1 0.0 0.0 0.0
Sorghum 6.6 1.9 1.7 0.3
Soybeans 0.0 0.2 1.1 0.6
Wheat 8.9 5.3 2.6 0.1
Total 5.9 4.5 2.1 1.5
50-100% over rate
Barley 1.1 0.3 0.1 0.2
Corn 0.7 1.0 0.4 1.1
Cotton 3.3 4.0 1.3 0.5
Oats 3.6 1.9 0.4 1.3
Peanuts 0.0 0.4 0.0 0.0
Sorghum 1.7 0.3 0.0 0.0
Soybeans 0.1 0.1 0.1 0.2
Wheat 3.9 3.1 0.1 0.0
Total 1.9 1.7 0.3 0.6
-- continued

16
Other Considerations
The environmental impacts of low nitrogen use efficiency on the environment
can be affected by different land management practices, such as the pres-
ence of underground tile drains and the use of filter strips or riparian buffers.
Tile drainage plays a role in nitrogen losses from fields (David et al., 2010).
Tile drainage lowers the water table, enabling fields that would otherwise be
wet part of the year to be intensively cropped. These drained soils tend to be
highly productive. Tiles, however, provide a rapid conduit for soluble nitrate,
effectively bypassing any attenuation that may occur in the soil. ARMS
data indicate that nearly 26 percent of treated cropland is tiled, most of this
in corn production (table 3.5). Of particular interest is the degree to which
nitrogen management on this vulnerable cropland is not using nitrogen BMP.
ARMS data indicate that about 71 percent of tiled acres do not meet all three
nitrogen management criteria. Most of these acres are in corn production.
Much of the tile-drained cropland is located in the Mississippi River Basin,
which has implications for hypoxia in the Gulf of Mexico.
Land management practices can mitigate nitrogen losses from fields. The use
of filter strips or riparian buffers can reduce the amount of nitrogen lost to
surface water bodies. Less than 10 percent of treated crop acres not meeting
the rate, timing, or method criteria have filter strips that could reduce losses
in runoff and subsurface flows (table 3.6). For corn, about 11 percent of acres
not using nitrogen BMPs have filter strips that could mitigate losses to water,
but significant improvements could still be made. Filter strips, however,
do not address atmospheric losses and may not be effective if not sited or
managed appropriately. In addition, buffers would be ineffective on the 26
percent of treated cropland that is tile drained.
Table 3.3
Percent treated acres by management category, crop, and degree of excess application,
2006 (continued)
Rate criterion status
Timing or method criteria met
Timing and method Timing Method Neither
Percent of treated acres
Greater than 100% over rate
Barley 0.2 0.0 0.1 0.2
Corn 0.6 0.3 1.2 0.8
Cotton 1.5 0.9 0.2 0.2
Oats 2.2 4.7 1.3 4.3
Peanuts 0.3 0.0 0.0 0.0
Sorghum 4.3 2.0 0.1 0.0
Soybeans 0.4 0.0 0.0 0.0
Wheat 0.2 1.7 0.1 0.1
Total 0.7 0.9 0.6 0.5
Total not meeting rate criterion 12.6 9.1 5.5 4.2
Notes: Figures in bold meet the rate criterion. See the notes to table 3.2.
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource Management Survey (2003-06), Phase II. See
table 3.1 for details.

17
Nitrogen Management on U.S. Corn
A high percentage of crop acres meet at least some of the nitrogen manage-
ment criteria (see table 3.3). Corn, however, meets all three criteria least
often. Corn is the most intensive user of nitrogen and the most widely planted
crop. Improvements in rate, timing, and/or application method are needed
on 70 percent of corn acres to improve NUE. In addition, growth in corn
demand due to the biofuels mandate suggests that corn acreage may increase
CEAP Analysis of Nitrogen Management in the
Upper Mississippi River Basin
Our assessment of nitrogen management on cropland using data from USDA’s
Agricultural Resource Management Survey (ARMS) has some similarities
with the assessment of nutrient management on cropland in the Upper Missis-
sippi River Basin (UMRB) conducted by the Conservation Effects Assessment
Project (CEAP). The two studies also have some important differences. CEAP
was initiated by USDA’s Natural Resources Conservation Service, Agricultural
Research Service, and Cooperative State Research, Education, and Extension
Service (recently renamed the National Institute of Food and Agriculture). The
goal of CEAP is to estimate conservation benefits from conservation invest-
ments and to provide research and an assessment on how to best use conser-
vation practices in managing agricultural landscapes to protect and enhance
environmental quality. The assessment of cultivated cropland in the UMRB is
the first of a series of studies that will cover major crop-producing areas of the
United States. Findings from the UMRB study are available at www.nrcs.usda.
gov/technical/nri/ceap/umrb/index.html.
Both analyses assess baseline nitrogen management on cropland according to
three criteria: rate, timing, and method. The definitions we used for each are
based on those used in the CEAP analysis. Both studies used a survey to col-
lect data on nutrient management practices. The major difference between our
analysis using ARMS data and the CEAP analysis is how the criteria were ap-
plied. ARMS collects information about cropping practices during a single
crop year. Our analysis, therefore, based the assessment of nitrogen manage-
ment on practices used to produce the crop sampled by the survey. The CEAP
analysis focused on cropping systems, which could be up to 5 years in length
and contain several different crops. Data were collected on production practices
used each year of the crop rotation. CEAP used these data to evaluate the entire
rotation, not just the crop grown during the year the survey was conducted. If
the rate, timing, or method criteria were not met during any year of the rotation,
then that sample point was identified as not meeting the nitrogen management
criteria. This approach is more stringent than the one used in our analysis. For
example, assume corn and soybeans were on a 2-year rotation and that corn was
grown during the year the ARMS and CEAP surveys were conducted. In our
analysis, if the nitrogen application rate on corn met the rate criterion, then that
corn sample was identified as such. In the CEAP study, the nitrogen application
rate on both the corn and the previous year’s soybean crops were assessed. If
the application rate on corn met the rate criterion but excess nitrogen was ap-
plied to soybeans, then the rotation was identified as not meeting the criterion.
This leads to the CEAP assessment reporting a smaller percentage of crop acres
meeting the rate criterion than we would report. Overall, the CEAP analysis
reports fewer crop acres meeting the rate, timing, and method criteria than does
the ERS report.

18
in the future, along with the intensity of corn production. Together, these
factors could increase reactive nitrogen emissions to the environment unless
nitrogen use efﬁciency is improved.
An examination of an additional year of survey data collected during the
2001 growing season and disaggregated regionally helps determine if
management has undergone recent changes and if such changes vary by
region. The share of corn acres not meeting the rate criterion declined
from 41 to 35 percent between 2001 and 2005 (table 3.7). This ﬁnding is in
Table 3.4
Percent treated crop acres receiving commercial or manure nitrogen that did not meet the rate, timing,
and method criteria, by crop, 2006
Crop
Planted
acres
Treated
acres
Acres treated
with commercial N only
Acres treated with
commercial and manure N
Acres treated with
manure N only
Thousands
Percent of
all treated
acres
Percent
vulnerable1
Percent of
all treated
acres
Percent
vulnerable
Percent of
all treated
acres
Percent
vulnerable
Barley 3,452 3,176 94 45 4 96 2 89
Corn 78,327 76,052 84 65 14 96 2 91
Cotton 15,274 12,566 96 67 3 85 1 29
Oats 4,168 2,748 78 59 9 88 13 92
Peanuts 1,243 737 93 46 5 52 2 41
Sorghum 6,522 5,370 98 55 1 98 1 49
Soybeans 75,522 16,827 85 51 2 100 13 91
Wheat 57,344 49,808 99 63 1 92 0 28
Total 241,852 167,285 90 62 7 96 3 86
1Vulnerable acres are those not meeting the rate, timing, and method criteria.
Notes: N = nitrogen. See notes to table 3.2. These estimates were weighted by the total amount of nitrogen applied by management category.
3.1 for details.
Table 3.5
Nitrogen-treated acres with tile drainage that did not meet the rate, timing, or method criteria by crop, 2006
Treated acres
Crop Total With tile drains
Tile-drained acres that do not meet the
rate, timing, or method criteria
Thousands Percent
Barley 3,176 42 43
Corn 76,052 34,738 70
Cotton 12,566 583 71
Oats 2,748 216 66
Peanuts 737 40 44
Sorghum 5,370 46 43
Soybeans 16,827 5,690 69
Wheat 49,808 1,644 94
Total 167,285 43,000 71
Notes: See notes to table 3.2.
3.1 for details.

19
agreement with those of other reports on improving nitrogen use efficiency
based on steady application rates and increased corn yields (Turner et al.,
2007). Improvements in rate were seen in all regions except Appalachia and
the Southeast. Notable improvements were seen in the Corn Belt, Lake States,
and Northeast. Timing and method, however, did not show similar improve-
ments in the more recent data. For most regions, the percentage of corn acres
not meeting these two criteria increased.
Changing Management May Result in Environmental Tradeoffs
Changing management practices may improve nitrogen use efficiency, but
the environmental outcomes may not always be desirable. We use the new
Nitrogen Loss and Environmental Assessment Package with GIS (Geographic
Information System) capabilities (NLEAP-GIS) model to assess how changes
in nitrogen management practices on corn affect the losses of nitrate (to
water), nitrous oxide (to air), and ammonia (to air) (Shaffer et al., 2010;
Delgado et al., 2010a). Of particular interest is the extent to which tradeoffs
in environmental outcomes might occur as overall nitrogen use efficiency is
improved. See appendix 2 for more details on NLEAP.
Because NLEAP is a field-level model, we selected eight different soils in
four States (Arkansas, Ohio, Pennsylvania, and Virginia) to assess changes in
nitrogen emissions to the environment from management changes in nonirri-
gated corn production.1 Four of the soils are type A or B soils (well drained),
and four are type D soils (relatively poorly drained). For each soil, we
examined two rotations (corn-corn and corn-soybeans), two tillage practices
(conventional and no-till), and two sources of nitrogen (inorganic fertilizer
and inorganic fertilizer + animal manure). The slopes for these soils were 0
to 6 percent, with low erosion potential.
For each soil/rotation/tillage/nitrogen-source combination, eight different
scenarios were modeled with NLEAP, each representing one of the combi-
1These four States were selected
because they present a wide variation
in growing conditions and because the
data necessary for running NLEAP
were already developed.
Table 3.6
Nitrogen-treated acres not meeting the rate, timing, or method criteria
that have filter strips, by crop, 2006
Crop
Number of acres not
meeting rate, timing
or method criteria
No. of acres not meeting
rate, timing, or method
criteria with filter strips
% of acres with
filter strips not
meeting criteria
Barley 1,523 68 4
Corn 52,910 5,909 11
Cotton 8,432 397 5
Oats 1,818 99 5
Peanuts 343 42 12
Sorghum 2,983 64 2
Soybean 9,600 475 5
Wheat 31,475 2,530 8
Total 109,084 9,584 9
Notes: See notes to table 3.2.
Source: USDA, Economic Research Service using data from USDA’s Agricultural Resource
Management Survey (2003-06), Phase II. See table 3.1 for details.

20
nations of nitrogen management criteria outlined in table 3.4. Therefore, 64
different scenarios were modeled for each soil.
A recommended application rate was speciﬁed for each soil/cropping system
combination, based on local agronomic recommendations, as described by
Espinoza and Ross (2008) for Arkansas, Alley et al. (2009) for Virginia,
Beegle and Durst (2003) for Pennsylvania, and Vitosh et al. (1995) for Ohio.
For the purposes of this analysis, overapplication was set at 75 percent more
than the recommended rate (at the upper end of overapplication found in the
ARMS data and reported in table 3.3). For example, if the recommended rate
was 132 pounds of N per acre, the overapplication scenario used 231 pounds
(see app. 2).
The modeled policy goal is that all three nitrogen management criteria be
met. For demonstration purposes, we used the NLEAP results to assess the
potential emissions tradeoffs when method, timing, timing and method, or
rate BMPs are adopted by corn farmers. For example, to evaluate the trad-
eoff when timing is improved (rate and method criteria are already met), we
compare the NLEAP results for the rate and method BMPs with the results
for the rate, timing, and method BMPs. Each cropping system is evaluated
separately. Because of the volume of results for the eight soils modeled, we
present only those from the two soils in Ohio (tables 3.8a-d). Results for the
other States are similar, in terms of the direction of changes.
All the scenarios show the expected changes in total nitrogen losses, with
reductions indicating improvements in NUE. The NLEAP results were
consistent with the expectation that nitrogen emissions are minimized when
all three criteria are met. Since nitrogen cycles through different forms and
ecosystems, the long-term environmental beneﬁts of reducing total nitrogen
Table 3.7
Nitrogen-treated acres and the shares that did not meet the rate, timing, or method criteria for corn,
2001 and 2005
Region Treated acres Did not meet rate Did not meet timing Did not meet method
2001 2005 2001 2005 2001 2005 2001 2005
Thousands Percent of treated acres
Appalachia 1,925 2,118 52 66 12 16 56 78
Corn Belt 35,087 39,145 46 38 41 41 39 34
Lake States 12,965 13,958 46 34 37 41 36 30
Mountain 1,243 1,018 18 14 9 20 33 50
Northeast 2,696 2,477 42 32 39 40 54 53
Northern Plains 16,962 18,293 27 28 10 15 36 45
Southeast 280 286 39 50 27 29 41 55
Southern Plains 1,708 2,109 31 32 45 38 33 18
Total 72,868 79,404 41 35 32 34 38 37
Notes: In both years, corn producers were surveyed in Colorado (Mountain); Kansas, Nebraska, North Dakota, and South Dakota (Northern
Plains); Texas (Southern Plains); Michigan, Minnesota, and Wisconsin (Lake States); Illinois, Indiana, Iowa, Missouri, and Ohio (Corn Belt); New
York and Pennsylvania (Northeast); Kentucky and North Carolina (Appalachia); and Georgia (Southeast). These estimates are based on weight-
ed sums, with the weights recalibrated so that the weighted sums of planted acres for each crop based on the survey data match estimates for
2001 and 2005 (USDA, 2008).
Source: USDA, Economic Research Service using data from USDA’s 2001 and 2005 Agricultural Resource Management Survey, Phase II.

21
are clear. However, some of the tradeoffs between different forms of nitrogen
could pose environmental problems. Adopting injection/incorporation always
increases nitrate leaching, sometimes substantially (more than doubling
leaching in some cases). Similarly, shifting applications from fall to spring
(without changing application rate) reduces nitrate losses and total nitrogen
losses but increases N2O emissions as applications are shifted to generally
warmer, wetter conditions, which is consistent with the findings of Delgado
et al. (1996), Rochette et al. (2004), and Hernandez-Ramirez et al. (2009).
Because of concerns over greenhouse gas (GHG) emissions, this outcome
would have to be carefully considered when making recommendations to
improve nitrogen use efficiency.
Adopting both method and timing again produces mixed results. NH3
emissions are always reduced. Leaching is generally reduced, but in some
cases where manure is used, it may increase. N2O emissions almost always
increase, from 5 to 50 percent, depending on the situation. In agreement with
basic principles of nitrogen management, only reducing the application rate
guarantees that losses of all three forms of reactive nitrogen are reduced
(Mosier et al., 2002; Meisinger and Delgado, 2002). Based on these findings,
a recommendation could be that in areas where leaching to drinking water
sources is a concern, improvements in nitrogen use efficiency could focus on
application rate reductions or improvements in timing.
Table 3.8a
Changes in nitrogen losses resulting from improvements in nitrogen management, NLEAP estimates -
Ohio - Type A soil - conventional tillage
Management
improvement
Without manure
Criterion rate=132 pounds N per acre
With manure
Criterion rate=198 pounds N per acre*
Total NO3
5 N2O6 NH3
6 Total NO3 N2O NH3
Pounds of N per acre
Continuous corn
Method1 -32.8 7.0 -1.7 -38.1 -17.0 24.6 -1.2 -40.4
Timing2 -16.6 -17.4 0.8 + -16.6 -17.6 1.0 +
Method+timing3 -33.0 -9.1 0.4 -23.7 -18.6 11.4 0.8 -30.8
Rate4 -69.3 -50.6 -0.9 -17.7 -105.1 -81.0 -1.3 -22.9
Criterion rate=102 pounds N per acre Criterion rate=153 pounds per acre*
Corn-soybean
Method1 -16.6 0.4 -0.8 -16.2 -14.7 3.8 -0.4 -18.1
Timing2 -5.7 -6.0 0.3 + -5.2 -5.6 0.4 +
Method+timing3 -13.1 -4.2 0.1 -9.0 -13.8 0.5 0.3 -14.6
Rate4 -15.7 -8.6 -0.4 -6.8 -37.2 -26.0 -0.6 -10.6
Note:*Manure is applied every other year. Criterion rate is met on average over 2-year period. + indicates a positive but very small change.
N = nitrogen. NO3 = nitrogen trioxide. N2O = nitrous oxide. NH3 = ammonia.
1Timing and rate best management practices (BMP) are already in place.
2Method and rate BMPs are already in place.
3Rate BMP is already in place.
4No BMPs are in place.
5Nitrate loss to water (primarily through leaching but often ends up in surface water).
6Ammonia and nitrous oxide loss to atmosphere.

22
Summary
The survey data indicate that in 2006, all of the nitrogen management criteria
were met on an estimated 35 percent of the crop acres treated with commer-
cial and/or manure nitrogen.2 In addition, a high percentage of treated acres
met at least some of the nitrogen management criteria. Among all crops, corn
met the criteria the least, and corn accounts for 50 percent of the treated acres
upon which one or more improvements to management could be made to
improve nitrogen use efﬁciency. Improvements in rate, timing, and/or method
might be needed on 67 percent of corn acres.
NLEAP-GIS simulation results reported in the literature show that changing
timing or method of application could potentially increase the loss of one
type of nitrogen compound, even if total nitrogen emissions decline and NUE
increases. NLEAP modeling indicates that only reducing application rates
ensures that all nitrogen emissions decrease, in agreement with established
principles of nitrogen management.
2Recall that this adoption rate is
higher than that reported by the USDA-
NRCS CEAP analysis, which considers
adoption over multiyear rotations (see
box on page 17).
Table 3.8b
Changes in nitrogen losses resulting from improvements in nitrogen management, NLEAP
estimates – Ohio – Type A soil - no-till
Management
improvement
Without manure
With manure
Total NO3
5 N2O6 NH3
6 Total NO3 N2O NH3
Continuous corn
Method1 -29.6 5.6 -1.1 -34.1 -15.6 23.5 -0.3 -38.8
Timing2 -27.5 -28.6 1.1 + -16.2 -17.3 1.1 +
Method+timing3 -40.6 -20.8 1.1 -20.9 -27.3 0.2 1.3 -28.8
Rate4 -53.7 -37.3 -0.6 -15.8 -85.0 -63.8 -0.8 -20.3
Criterion rate=86 pounds N per acre Criterion rate=129 pounds N per acre*
Corn-soybean
Method1 -14.0 0.7 -0.8 -13.9 -12.7 4.7 -0.1 -17.3
Timing2 -9.9 -10.3 0.4 + -8.6 -9.0 0.4 +
Method+timing3 -14.9 -7.6 0.3 -7.6 -15.1 -2.2 0.5 -13.4
Rate4 -15.5 -9.5 -0.3 -5.7 -28.2 -18.7 -0.4 -9.1

23
Table 3.8c
Changes in reactive nitrogen losses resulting from improvements in nitrogen management, NLEAP
estimates – Ohio - Type D soil - conventional till
Management
improvement
Without manure
With manure
Total NO3
5 N2O6 NH3
6 Total NO3 N2O NH3
Continuous corn
Method -28.3 0.7 -5.0 -24.0 -20.0 12.9 -3.1 -29.8
Timing -8.1 -9.4 1.3 + -12.1 -13.5 1.4 +
Method+timing -20.2 -7.6 1.2 -13.8 -17.2 4.6 1.7 -23.5
Rate -56.3 -44.1 -1.8 -10.4 -91.3 -70.9 -3.0 -17.4
Corn-soybean
Method -14.7 0 -4.1 -10.6 -16.2 1.3 -2.2 -15.3
Timing -1.9 -2.5 0.6 + -2.7 -3.3 0.6 +
Method+timing -6.8 -2.1 0.5 -5.2 -12.5 -0.4 0.8 -12.9
Rate -9.3 -4.7 -0.7 -3.9 -27.8 -17.1 -1.4 -9.3

24
Table 3.8d
Changes in reactive nitrogen losses resulting from improvements in nitrogen management, NLEAP
estimates –Ohio - Type D soil - no-till
Management
improvement
Without manure
With manure
Total NO3
5 N2O6 NH3
6 Total NO3 N2O NH3
Continuous corn
Method -35.4 0.7 -1.4 -34.4 -25.8 13.6 -0.3 -39.1
Timing -21.4 -22.0 0.6 + -11.1 -11.8 0.7 +
Method+timing -38.8 -18.3 0.6 -21.1 -32.2 -4.1 1.2 -29.3
Rate -37.3 -20.4 -1.0 -15.9 -66.3 -44.2 -1.8 -10.4
Corn-soybean
Method -14.5 0.3 -0.8 -14.0 -16.0 1.6 0 -17.6
Timing -7.2 -7.4 0.2 + -6.2 -6.5 0.3 +
Method+timing -13.3 -5.9 0.2 -7.6 -16.7 -3.7 0.6 -13.6
Rate -10.1 -4.0 -0.4 -5.7 -20.4 -10.5 -0.7 -9.2
1Timing and rate best management practices (BMPs) are already in place.

25
Chapter 4
Policy Instruments for Nitrogen Reduction
Based on ARMS data, 65 percent of surveyed cropland, or 109 million acres,
is in need of improved nitrogen management. Given nitrogen’s effects on the
environment, improving nitrogen management on vulnerable lands is a policy
goal, both nationally and regionally.
Farmers adjust the management of their crops for a variety of reasons.
Economic factors, such as input or output price changes, may lead to more
(or less) careful use of nitrogen inputs. Farmers may also have to consider
various policy-based incentives for adopting practices that improve nitrogen
management. Over the years, policy instruments have been employed to
improve the management of agricultural inputs and resources. USDA conser-
vation programs rely primarily on subsidies for management practices and
education. USDA also employs compliance mechanisms to protect wetlands
and highly erodible soils. The U.S. Environmental Protection Agency (EPA)
is using regulations to address nutrient management on certain confined
animal feeding operations. A few States have used nitrogen fertilizer taxes
to raise revenue for nutrient management programs. Such policy approaches
may have a role to play in increasing the number of crop acres that meet the
three nitrogen management criteria described earlier.
Provide Information (Education)
A lack of knowledge about their performance may be preventing farmers
from using the most efficient nutrient management practices. Education is
used to provide producers with information on how to farm more efficiently.
Its success depends on alternative practices being more profitable to farmers
than current practices (Ribaudo and Horan, 1999). Two practices that can
lead to more efficient fertilizer use are soil testing and tissue testing. These
tests provide information that reduces some of the uncertainty surrounding
nutrient availability and enables producers to apply fertilizer at rates more
consistent with plant needs and high nitrogen use efficiency.
ERS research supports previous findings that nitrogen testing is having
the desired effect on nitrogen application rates for certain nitrogen users.
Data from the 2001 and 2005 ARMS indicate that about 21 percent of corn
farmers used a soil or tissue test as a basis for their level of nitrogen applica-
tion (table 4.1). Farmers who used commercial nitrogen followed the recom-
mendations closely. In our sample, their mean application rate of nitrogen was
136 lbs per acre, and the mean recommended rate based on a nitrogen soil
test was 137 lbs per acre (table 4.2).
Compliance with the soil test, however, was much worse for farmers who
used both manure and commercial fertilizer. In their case, the recommended
nitrogen application rate was 123 pounds per acre. And while farmers applied
only 85 pounds per acre of commercial fertilizer, total nitrogen application
rates were 175 pounds per acre when manure was added.
We compared nitrogen application rates of those farmers who use soil N and
tissue tests with those who do not using regression analysis that accounts

26
for a number of production, land, and operator characteristics (see app. 3).
Findings show that soil nitrogen testing has a statistically significant impact
on nitrogen application rates. In the case of farmers who use commercial
nitrogen exclusively, those who tested the soil applied 73.9 pounds per acre
less than those who did not, all else equal. Other studies have found soil tests
to be of similar effectiveness (Wu and Babcock, 1998; Musser et al., 1995).
An information-based approach can meet nitrogen efficiency goals only if
the information provided leads to increased profits for farmers (Ribaudo
and Horan, 1999). As long as there are expectations that more efficient
nitrogen management leads to increased risk or higher costs, then nitrogen
management goals are unlikely to be met with information alone. However,
information has proven valuable in support of other policy goals. Education
can reduce the cost of adopting nitrogen BMPs required by regulation or
funded through financial incentives. For example, Bosch et al. (1995) found
that education affected the outcomes associated with a regulation requiring
nitrogen testing in Nebraska. Producers did not use the information provided
by testing unless they received education assistance.
Financial Incentives
Financial assistance is an important tool used in many USDA conservation
programs to promote the adoption of BMPs. Program effectiveness depends
on how farmers respond to the incentive being offered. When a farmer
accepts a payment in return for adopting a management practice, he or she
is signaling that the payment at least represents the economic cost of imple-
menting the practice, sometimes referred to as the willingness-to-accept.
Generally, only the producer knows the true cost. This makes it difficult for
program managers to find the minimum payment rate that entices enough
producers into the program to achieve the particular environmental goal at
least cost.
Table 4.1
Factors influencing farmers’ nitrogen fertilizer application decision
Application used 2001 2005
Percent of farmers
Soil or tissue test 18.8 27.0*
Crop consultant recommendation 13.0 17.6*
Fertilizer dealer recommendation 28.7 41.2*
Extension service recommendation 3.2 4.6*
Cost of nitrogen and/or expected commodity price 11.4 17.3*
Routine practice 70.9 71.7*
Number
Observations 1,646 1,344
*Statistically different from 2001 at the 1-percent level, based on pairwise two-tailed delete-a-
group Jackknife t-statistics (Dubman, 2000)
Source: USDA, Economic Research Service using data from USDA’s 2001 and 2005
Agricultural Resource Management Survey, Phase II, Cost of Production Practices and Costs
Report.

27
USDA’s NRCS supports management practices that specifically address fertil-
izer application rate, timing, or method in their standards. The Environmental
Quality Incentives Program (EQIP) is the largest USDA program that
provides producers with technical and financial assistance for implementing
and managing BMPs on working farmland. Management practices supported
by EQIP that can influence nitrogen use efficiency include nutrient manage-
ment and waste utilization (for manure). Implementing a nutrient manage-
ment plan directly affects measures of stewardship. Nutrient management
planning addresses the amount, source, placement, form, and timing of the
application of plant nutrients and soil amendments (USDA, NRCS, 2006).
Further, the practice requires the application rate be based on an assessment
of plant-available nitrogen developed through Land Grant University soil
and tissue tests or recognized industry practices. Waste utilization guidelines
specify that rates of application must be compatible with the soil’s ability to
absorb and hold the waste, and methods of incorporation are prescribed for
liquid manure forms to prevent nutrients from rising to the surface.
Data from EQIP contracts in force for year 2008 show that participating
farmers accepted an average payment of $8.88 per acre for adopting nutrient
management (table 4.3). A higher per acre payment induced farmers to adopt
a waste utilization practice ($14.75). Relatively few corn farm operations have
livestock or a direct source of manure (organic) fertilizer, and, as reported
later, the practice can be more costly to farmers than using commercial (inor-
ganic) fertilizer.
A focus on the Corn Belt reveals variation in the accepted payments for
the two practices (table 4.3). The variation may stem from cost differences
within the region that are driven by local conditions, which, in turn, influence
the State-level payment rate for the practice. To examine how management
practices can affect a farm’s cost of operations, we estimate a cost function
using a generalized linear regression model estimated with 2001 ARMS data
(see app. 4).3 Model results show that several conservation practices have
3Because we are comparing 2001
costs with 2008 payments, we inflate
2001 costs using the U.S. Bureau of
Labor Statistics’ Consumer Price Index.
Table 4.2
Influence of soil/tissue nitrogen testing on fertilizer application rates for corn, with and without
manure use, 2001 and 2005
For farmers using a soil test
Required nitrogen
based on expected yield1
Soil test
recommended nitrogen
Commercial
nitrogen applied
Total nitrogen applied
(commercial + manure)
Pounds of nitrogen per acre
Commercial nitrogen with
manure
Observations = 154
152 123 85† 175†
Commercial nitrogen with-
out manure
Observations = 645
165 137 136 136
1Based on nitrogen removed in expected harvest plus 40 percent to account for unavoidable nitrogen losses.
†Means are statistically different from the recommended nitrogen amount at the 1-percent level, based on pairwise two-tailed delete-a-group
Jackknife t-statistics (Dubman, 2000).
Source: USDA, Economic Research Service using data from USDA’s 2001 and 2005 Agricultural Resource Management Survey, Phase II, Cost
of Production Practices and Costs Report.

28
little effect, on average, on the cost of operation relative to other methods of
management. For example, the difference in operation costs for farms using
nutrient management and for farms not using these practices is not statisti-
cally significant.
Based on results from our cost analysis, we also find that using manure as a
nitrogen source costs roughly $26.84 more per acre than using only commer-
cial fertilizer. However, we observe a national average per acre EQIP payment
for the waste utilization of $14.75, and only two States in the Corn Belt
(Illinois and Indiana) have payment levels that approach the estimated cost
figure. The results suggest that the EQIP rate is insufficient to entice farmers
who are not using manure to begin doing so in an environmentally sensi-
tive manner. However, farms with livestock or poultry need to dispose of the
waste. Therefore, rather than be a practice by choice, waste utilization may be
a practice that complements the necessary disposal of manure, and a payment
that covers increased production costs may not be a necessary condition for
the willingness to adopt the practice.
Not all farmers require a cost share to adopt conservation practices. Cooper
and Keim (1996) use farmer surveys to conclude that 12 to 20 percent of
farmers may be willing to adopt practices such as split fertilizer applications
and nutrient testing without financial assistance but do not do so because they
lack information or are uncertain about the practices’ economic performance.
However, they also find that the adoption rate would not increase beyond 30
percent unless subsidy rates were substantially increased. A farmer’s percep-
tion of the effectiveness of a practice can also influence the decision to adopt.
Evidence from Lichtenberg and Lessley (1992) suggests that farmers may
need more than a cost share to overcome perceptions of conservation prac-
tices and the state of environmental quality off-site.
In some cases, farmers are willing to adopt conservation practices that
reduce profits if they believe that others will benefit from the subsequent
change in environmental quality (Bishop et al., 2010; Chouinard et al., 2008).
For example, based on survey responses from the State of Washington,
Chouinard et al. (2008) conclude that farmers would be willing to forgo up
Table 4.3
Per acre average EQIP payments for conservation practices, 2008
Corn Belt
Practice All States Illinois Indiana Iowa Missouri Ohio
Dollars per acre
Nutrient management1 8.88 9.75 7.47 6.12 13.90 10.91
Waste utilization2 14.75 25.95 25.84 10.90 5.83
1Nutrient management planning addresses the amount, source, placement, form, and timing of the application of plant nutrients and soil
amendments.
2Waste utilization guidelines specify that rates of application must be compatible with the soil’s ability to absorb and hold the animal waste, and
methods of incorporation are prescribed for liquid manure forms to prevent nutrients from rising to the surface.
Notes: Blank cells indicate no contracts for such practice in that State.
Source: USDA, Economic Research Service using contract data from USDA’s Environmental Quality Incentives Program, fiscal years
1997-2008, payments made in fiscal year 2008.

29
to $4.52 (median value estimate) in per acre annual profits to implement soil-
conserving stewardship practices.
The scope of a program’s coverage is an important consideration for poli-
cymakers and program managers evaluating the adequacy of the financial
incentives offered to program participants. In 2008, the financial incentives
from EQIP encouraged farmers to enroll 4 million acres in the program’s
nutrient management practice. However, because participation in the program
is voluntary, it is not known if the cropland most in need of treatment was
enrolled.
We can use the data from EQIP and table 3.3 to estimate the cost to improve
nitrogen use efficiency on those acres needing additional treatment. About 35
percent of all crop acres meet all three criteria, which means that over 108
million acres of cropland are not using nitrogen BMPs. Applying the average
payment rate for nutrient management ($8.88 per acre) to all acres needing
improved management implies annual EQIP payments of $959 million.
However, the findings from Cooper and Keim (1996) suggest that higher
rates would be needed to entice a sizable percentage of farmers to voluntarily
enroll in a program. Assuming a payment rate 50 percent higher results in
program expenditures of $1.4 billion. This is roughly the current annual
budget for EQIP.
Given the potential cost of treating the entire 108 million acres of cropland
not using nitrogen BMPs, which groups might be most important to address
first? We previously reported that manure users generally apply much more
total nitrogen to the field than farmers who exclusively apply commercial
nitrogen. Providing financial assistance for nutrient management on the 7.7
million acres that received manure and failed to meet the rate criterion would
cost between $68.4 and $103 million per year.
Off-Site Filtering for Reducing Nitrogen
Losses From Fields
Similar to its efforts aimed at improving nitrogen use efficiency on working
lands, the Government can provide financial incentives for installing manage-
ment practices that capture nitrogen after it leaves a field, primarily nitrogen
in water. This analysis estimates and evaluates the cost effectiveness of
two such measures, wetlands restoration and vegetative filter strips (VFS),
assuming that funding is targeted to areas where nitrogen removal is likely to
be most effective.
The Costs of Nitrogen Capture by Restoring Wetlands
Our analysis of wetlands restoration focuses on the Glaciated Interior Plains
(GIP), where models of wetlands nitrogen removal have been developed. The
GIP includes major parts or all of Ohio, Minnesota, Wisconsin, Michigan,
Iowa, Illinois, and Indiana—major corn-producing States. This area is also an
important source of nitrogen that reaches the Gulf of Mexico and contributes
to the hypoxic zone (Goolsby et al., 2001; Robertson et al., 2009). Wetlands
in other parts of the United States can also reduce nitrogen loadings. But,
because of regional differences in ecosystems, we do not extrapolate our find-
ings to other areas.

30
Wetlands once made up a large portion of land on the GIP (fig. 4.1). Water
tables were lowered to facilitate crop production by installing underground
tile and surface drainage systems. Such drainage systems become conduits
for the rapid movement of nitrate from fields to water resources.
The costs of creating wetlands vary widely as do nitrogen removal rates on
wetlands. Costs are driven by the cost of the land and the cost of restoring
wetland ecosystems. Nitrogen removal depends on the rate of nitrogen inflow,
nitrogen concentration, seasonal variations in flow, wetland size, and other
factors.
We use the USDA Wetland Reserve Program (WRP) contract data for the
GIP to estimate multinomial land and restoration cost functions (see app.
5). With these functions, we generate county-level cost estimates throughout
the GIP. The objectives of the WRP are to enhance, restore, and preserve
wetlands. As of October 1, 2009, the WRP enrolled 2.18 million acres, with
wetlands in every State. Along with the land and restoration cost variables,
the WRP contract data contain information on the size and the county loca-
tion of each contract. The land (wetland easement) cost variable represents
the difference between the agricultural value of the land and the value of
the land with a wetland easement. The easement requires that the landowner
maintain the health of the ecosystem. Data for other variables in our analysis
come from the NASS agricultural census. Across the counties within the GIP,
wetland easement costs range from $1,490 to $3,030 per acre, as generated
by our estimated land cost function. Expected wetland restoration costs range
from $506 to $602 per acre. Annualizing over perpetuity with a discount rate
of 5 percent, we estimate that the median annual expected cost of restoring
and preserving wetlands is $153 per acre per year (table 4.4). Because
marginal costs are less than average costs, one can expect average per acre
Source: USDA, Economic Research Service using data from the 1997
National Resources Inventory.
Figure 4.1
Historical wetlands converted to cropland, by county, 1997
Former wetland acres
< 500
500-1,500
> 1,500

31
costs to be lower for larger wetlands and potentially more cost effective as a
nitrogen filter, all other things being equal.
Wetlands remove most nitrogen through denitrification (Crumpton et al.,
2008), which converts nitrate to nitrous oxide (N2O). However, there is a
general belief, supported by a limited number of studies, that N2O releases
are a very small portion of nitrogen removal, even in wetlands with elevated
nitrogen loadings (EPA, 2010b). Researchers estimate that N2O accounts
for between 0.13 and 0.30 percent of total annual wetland nitrogen loss
(Hernandez and Mitsch, 2006; Crumpton et al., 2008). The reported rates of
N2O releases by wetlands are similar to estimated releases on cropland in the
Midwest, so restoring wetlands is likely to have no net effect on N2O emis-
sions (Crumpton et al., 2008).
Crumpton et al. estimate that nitrogen loads to surface water could be
reduced by 30 percent (~500 million pounds) in the Upper Mississippi and
Ohio River basins with the addition of 0.5 to 1.1 million acres of strategically
placed wetlands, for an average per acre reduction of 450 to 1,000 pounds
per year. These removal rates assume an optimal placement of the restored
wetlands—areas with a high water flow with high nitrogen concentrations.
Mitch et al. (1999) estimate that wetlands in the Midwest remove 142 to
214 pounds per acre of nitrogen per year. The researchers assume that the
wetlands are well constructed and placed, but their estimates are based on
a wide range of nitrogen concentrations and hydrologic flows. Each study
includes multiple wetlands and a variety of flow conditions and nitrogen
concentrations.
The unit cost of nitrogen removal by wetlands, based on nitrogen removal
rates of 450 to 1,000 pounds per acre per year reported by Crumpton et al.
(2008), is $0.08 to $0.34 per pound (table 4.4). Based on the removal rates of
142 to 214 pounds per acre per year reported by Mitch et al. (1999), unit cost
ranges from $0.71 to $1.08 per pound.
The Costs of Nitrogen Capture Using Vegetative Filter Strips
Vegetative filter strips present another off-field option for capturing and
removing nitrogen from runoff and subsurface waters. The cost of a VFS
tends to be lower than the cost of wetlands restoration. The VFS cost has two
components: the opportunity cost of holding the land out of production and
the cost of establishing cover (e.g., grasses, trees, or both). Cropland rental
Table 4.4
Costs of nitrogen removal by wetlands
Wetland cost
N removal rate =
142 lbs/ac
N removal rate =
214 lbs/ac
N removal rate =
450 lbs/acre
N removal rate =
1,000 lbs/acre
$/acre $/lb of N removed by wetland
Marginal cost 77 0.54 0.36 0.17 0.08
Average cost 153 1.08 0.71 0.34 0.15
Note: Because marginal costs are less than average costs, per acre costs would be lower for larger wetlands. N = nitrogen.
Source: USDA, Economic Research Service using data from Mitsch et al., 1999 (142 and 214 pounds per acre) and Crumpton et al.,
2008 (450 and 1,000 pounds per acre).

32
rates are an economic measure of the opportunity cost of taking cropland
out of production. We assume that average cropland rental rates are equal to
the economic return to land converted to a VFS. Based on the distribution
of corn acreage reported in the 2005 ARMS and county-level rental data
provided by NASS, the annual opportunity cost of converting corn cropland
into a VFS is estimated at $94 per acre.
We assume that the cost of establishing vegetative cover is about the same as
establishing cover on land retired in USDA’s Conservation Reserve Program
(CRP). CRP data do not specify cost by cover type, but data do provide
insights on the range of costs. Across the 25th, 50th, and 75th percentiles,
cover costs are $16, $35, and $60 per acre. Because establishing forest cover
is more costly, the lower percentile costs likely reflect the cost of establishing
grasses.
The cover cost is a one-time investment. We annualized this cost by assuming
that it is to last for the foreseeable future and a 5-percent discount rate.
Together, the land and cover cost would total approximately $95 to $97 per
acre per year, with the higher estimate more likely representative of the use of
forest cover.
Mitch et al. (1999) tabulate several plot studies with a focus on the quantity
of nitrogen removed across varying sizes of filter strips and levels of nitrogen
inflow. They apply their findings to nitrogen runoff rates typical of those in
corn-producing areas and estimate that properly designed forested riparian
VFS will remove approximately 17.8 to 53.0 pounds of nitrogen per acre with
strips ranging in width from 10 to 50 feet (Mitch et al., 1999, pg. 47).
At an annual nitrogen runoff removal rate of 17.8 to 53.0 lbs per acre and a
forested VFS cost of $97 per acre, VFS nitrogen removal costs are estimated
to range from $1.83 to $5.45 per pound of nitrogen. The cost estimate is a
weighted average across the corn-producing areas of the GIP.
Results suggest that, within the GIP, wetlands can be much more cost effec-
tive at removing nitrogen than VFS, primarily because of their substantial
nitrogen removal rates. Within corn-producing regions, especially in areas
where fields are tile drained, water moves quickly through and passes under
root zones, rendering VFS ineffective. On the other hand, VFS can be estab-
lished in many landscape settings where wetlands cannot.
The wide range in nitrogen removal rates by wetlands reflects, at least in part,
the advantage of targeting wetlands to areas where they are likely to be more
effective—areas where wetlands capture large quantities of water with high
nitrogen concentration rates. But even the low nitrogen removal rates of 142
to 214 pounds per acre reported by Mitch et al. are three or more times the
removal rates of VFS. Additionally, the rich wetland ecosystems have the
potential of providing a greater array of environmental services than those
delivered by VFS.
Participation in Emissions Trading Programs
An alternative to publicly provided financial incentives for adoption of
conservation practices is for private markets to pay farmers to adopt

33
management practices that produce ecosystem services valued by consumers
(the public). Emissions trading uses markets to efficiently achieve pollution
targets. The development of markets for ecosystem services is characterized
by uncertainties about whether viable markets for public goods can exist,
but the EPA and USDA are promoting emissions trading markets for water
quality and greenhouse gases as a way of reducing the costs of meeting envi-
ronmental goals. Agriculture has a potential role to play in both markets.
Water Quality Trading Program
The promise of emissions trading, along with the real-world success of air
emissions trading, has led to the creation of water quality trading markets in
a number of impaired watersheds. Under the Clean Water Act, point sources
(e.g., factories, sewage treatment plants) were initially regulated through
a nontradable permit system. A permit specifies how much of a particular
pollutant the permit holder can discharge. Traditionally, permit holders were
required to meet their permit obligations through their own effluent reduc-
tions. EPA policy guidelines on water quality trading now allow point sources
to meet their Water Quality Based Effluent Limitation requirements through
discharge reductions from other sources under certain conditions, including
agricultural nonpoint sources (EPA, 2004). The guidelines encourage
States to consider agriculture as a source of offsets in water quality trading
programs, and a number of States are either implementing or considering
water quality trading programs that allow point/nonpoint source trading.
There appears to be many opportunities for point/nonpoint trading programs
to be established. Almost 7,000 water bodies impaired by nutrients (pollut-
ants produced by both point and nonpoint sources) have been listed under
Section 303(d) of the Clean Water Act (EPA, 2009). To date, over 4,000 Total
Maximum Daily Loads (TMDLs) have been developed to address 5,000
of these impaired waters. The presence of a TMDL is a basic requirement
for a trading program, as it creates the demand for credits (Ribaudo et al.,
2008). Agriculture is a major source of nutrients in most of the watersheds
containing impaired waters (Ribaudo and Nickerson, 2009). The marginal
cost of reducing nitrogen loss from cropland is generally less than the
marginal cost of reducing nitrogen discharges from point sources (primarily
sewage treatment plants) (Camacho, 1992; Shortle, 1990).
Forty water quality trading programs have been created in the United States
since 1990 (Breetz et al., 2004). Fifteen include production agriculture as a
potential source of credits for regulated point sources, most often for nutrients
(nitrogen and phosphorus). However, point/nonpoint trading has not been
very successful, at least in terms of the participation of potential traders and
the number of trades between regulated sources and farms (Breetz et al.,
2004).
Regulators designing point/nonpoint trading markets must contend with
uncertainty about sources and levels of emissions, the effectiveness of best
management practices, the water quality impacts of emissions from different
sources, and farmer willingness to participate in a market driven by regula-
tion (on point sources) (Hoag and Hughes-Popp, 1997; King, 2005; King and
Kuch, 2003; Woodward and Kaiser, 2002; Ribaudo and Gottlieb, 2011; Horan
and Shortle, 2011). The failure of current programs to perform as advertised
can largely be attributed to failures of market design and program rules to

34
adequately address these issues, or the high transactions from incorporating
uncertainties into market design.
One issue that has particular relevance for addressing nitrogen pollution
is the baseline used for calculating credits. The EPA defines a baseline
participation requirement as the pollutant control requirements that apply to
a seller in the absence of trading (EPA, 2007). EPA suggests that practices
generally accepted as good management define a baseline for agriculture,
under an assumption that all farms would eventually adopt these practices
voluntarily. Some practices that States have used in trading programs to
define a baseline include the use of filter strips or a nutrient management
plan (Wisconsin DNR, 2002; Pennsylvania DEP, 2008). However, the issue
is that our survey data indicate that very few crop acres would meet these
baseline requirements as the percentages of cropland with filter strips or
nutrient management plans are only 6.8 and 5.0, respectively, meaning that
most crop acres would not be able to participate in a trading program until
the baseline requirements were met. If the incentives from a credit market
are insufficient to induce farms that have not already voluntarily adopted the
minimum set of practices to incur the cost of meeting the baseline require-
ment, then these farms will continue unabated discharge. This entry cost
would therefore potentially limit participation and adversely affect the effi-
ciency of the market (Ribaudo and Gottlieb, 2011; Ghosh et al., 2011).
Greenhouse Gas Mitigation
Another emissions market that might influence nitrogen management deci-
sions in agriculture is an offset market for mitigating emissions of CO2 and
other greenhouse gases, such as nitrous oxide (N2O). Nitrous oxide is a
powerful greenhouse gas (310 times the global warming potential of CO2
over 100 years) and can be emitted from fields receiving nitrogen fertilizer
(see chapter 2). A trading program for nitrous oxide emissions would have
many of the same design and implementation issues of point/nonpoint trading
for water quality. One would expect that the use of models for predicting
reductions, based on field and management characteristics, would figure
heavily in any trading program.
We use NLEAP results and ARMS cost data to determine changes farmers
might make given the opportunity to participate in an offset market for N2O
reductions by producing credits and likely environmental tradeoffs. These
analyses were conducted across different management scenarios and general
hydrologic soils (e.g., well-drained soils with a large leaching potential versus
poorly drained soils with a low leaching potential) from selected counties in
Virginia, Ohio, Pennsylvania, and Arkansas.
For each soil, we identified the changes a farmer might make in nitrogen
management practices to produce N2O reductions (offset credits) at the lowest
cost while meeting a requirement that total nitrogen emissions (the sum of
NO3, N2O, and NH3 losses) not increase. In other words, trading rules do not
permit a management change that reduces N2O but increases total nitrogen
emissions. Changes in cost are defined as the difference in average variable
costs (chemicals, fuel, and electricity) and value of lost production (changes
in yields). We assumed farmers would maintain the same basic cropping
system and alter timing, method, or application rate only. A description of

35
NLEAP and the cost model and assumptions are presented in appendices 2
and 3.
Table 4.5 summarizes the nitrogen management systems that farmers evalu-
ated in the model would adopt to produce credits at the lowest cost, given
baseline practices. For example, of the 64 farm types not meeting any of the
criteria prior to a market (“None” in the baseline criteria column), 17 would
reduce the application rate to the criterion rate, 10 would reduce the rate and
inject/incorporate nitrogen, 1 would reduce the rate and apply nitrogen in the
spring, and 36 would adopt all three management choices. The choice depends
on the soil type, climate, rotation, tillage practice, and nitrogen source.
The results highlight the importance of meeting the application rate criterion
for reducing both N2O and total reactive nitrogen. For all farms not meeting
the rate criterion, reducing application rate either alone or in combination
with another practices was selected to reduce N2O. Method or timing was
never the sole practice adopted by farms to reduce N2O emissions. Model
results also indicate that 148 of the 512 farming systems will not be able to
reduce N2O emissions by meeting the rate, timing, or method criteria. For
example, none of the 64 farm types meeting the rate and method criteria at
the start of a market can reduce N2O emissions by also meeting the timing
criterion.
Table 4.6 provides more detail for one soil in Ohio. It shows the reduction in
N2O that would be generated for each decision a farmer in a particular base-
line situation could make and credit revenue earned assuming a carbon price
of $15 per ton of CO2 equivalent.4 The range of N2O reductions presented
here is similar to that found for the other soils modeled with NLEAP.
4Based on EPA analysis of the
American Clean Energy and Security
Act of 2009, H.R. 2454.
Table 4.5
Least-cost N management systems in corn production for reducing N2O emissions for 512 model
farms, assuming a credit price of $15 per ton of CO2 equivalent, based on NLEAP modeling
Criteria1 met after changing
management Method Rate Timing
Rate and
method
Rate and
timing
Timing and
method
Rate,
timing, and
method
Total
model
farms
Number of model farms
Criteria1 met in baseline
None 17 10 1 36 64
Method 16 17 3 28 64
Rate 19 42 3 64
Timing 63 1 64
Rate and method 64 64
Rate and timing 3 23 1 37 64
Timing and method 31 33 64
Rate, timing, and method 64 64
1Criteria are appropriate rate, timing, and method of nitrogen application (see chapter 3).
Note: N = nitrogen. NLEAP = Nitrogen Leaching Environmental Analysis Project. N2O = nitrous oxide. CO2 = carbon dioxide. A total of 512
cropping systems are evaluated with NLEAP, 128 each in Arkansas, Ohio, Pennsylvania, and Virginia. Each deﬁnes a soil type (A or D), a rota-
tion (continuous corn, corn soybeans), tillage practice (conventional, no-till), nutrient source (inorganic, manure+inorganic), timing of application
(before planting, at/after planting), method (inject/incorporate, broadcast) and application rate (meet criterion, 75% over criterion).

36
Table 4.6
How a corn farmer may change N management in a market for nitrous
oxide (N2O) greenhouse gas emissions with credit payments of $15/
ton of carbon dioxide equivalent, for a model Ohio farm on Ottoke soil
Baseline practice
Practices after
N2O credit offered
N2O reduction Credit revenue
Pounds per acre Dollars per acre
CC-CON-MF
M RTM 0.9 2.09
RM No change 0.0 0.0
R RM 0.3 0.70
RTM No change 0.0 0.0
RT RM 3.4 7.90
TM RT 3.0 6.98
T RT 4.4 10.23
NONE RTM 0.8 1.86
CC-CON-OF
M RTM 0.3 0.70
RM No change 0 0
R RM 0.6 1.40
RTM No change 0 0
RT RTM 2.7 6.28
TM RT 0.9 2.09
T RT 3.1 7.21
NONE RTM 0.8 1.86
CC-NT-MF
M RTM 0.2 0.46
RM No change 0 0
R No change 0 0
RTM No change 0 0
RT RTM 0.5 1.16
TM RT 3.3 7.67
T RT 2.8 6.51
NONE RM 0.9 2.09
CC-NT-OF
M R 1.1 2.58
RM No change 0 0
R RM 0.2 0.46
RTM No change 0 0
RT RTM 1.7 3.95
TM RT 1.4 3.26
T RT 2.8 6.51
NONE R 0.9 2.09
-- continued

37
Table 4.6
How a corn farmer may change N management in a market for nitrous
oxide (N2O) greenhouse gas emissions with credit payments of $15/
ton of carbon dioxide equivalent, for a model Ohio farm on Ottoke soil
-- continued
Baseline practice
Practices after
N2O credit offered
N2O reduction Credit revenue
Pounds per acre Dollars per acre
CS-CON-MF
M RTM 0.6 1.40
RM No change 0 0
R RM 0.2 0.46
RTM No change 0 0
RT RM 1.3 3.02
TM RT 1.6 3.72
T RT 1.7 3.95
NONE RTM 0.2 0.46
CS-CON-OF
M RTM 0.2 0.46
RM No change 0 0
R RM 0.3 0.70
RTM No change 0 0
RT RTM 1.2 2.79
TM RTM 1.1 2.56
T RT 1.2 2.79
NONE RTM 0.5 1.16
CS-NT-MF
M RT 0.2 0.46
RM No change 0 0
R No change 0 0
RTM No change 0 0
RT RM 0.8 1.86
TM RT 1.4 3.26
T RT 1.4 3.26
NONE RM 0.5 1.16
CS-NT-OF
M R 0.2 0.46
RM No change 0 0
R RM 0.2 0.46
RTM No change 0 0
RT RTM 1.3 3.02
TM RTM 1.1 2.56
T RT 1.4 3.26
NONE R 0.5 1.16
Note: N = nitrogen. CC = continuous corn, CS = corn-soybeans, CON = conventional till,
NT = no-till, MF = manure+inorganic N, OF = inorganic N, M = N incorporate d/injected,
R = N rate is less than 40% more than N removed at harvest, T = spring application.

38
Even though our sample of cropping conditions is very small, we believe
we can still make some inferences from the results. We found that if the
baseline system is not meeting the application rate criterion, application
rate will be reduced to produce credits, either alone or in combination
with timing or method; reducing the application rate is generally the most
cost-effective means of reducing N2O emissions. Adopting method and/
or timing BMPs alone cannot reduce N2O emissions or can do so only by
reducing overall nitrogen use efficiency, which is not permitted under our
simulated market rules.
Farms already meeting both the rate and method criteria will only be able to
reduce N2O emissions by reducing their application rate below recommended
rates. The NLEAP modeling indicates only small reductions in N2O when the
application rate is reduced to a level below the criterion rate. This is consis-
tent with field studies that indicate a nonlinear relationship between excessive
N application rates and N2O emissions (Jarecki et al., 2009; McSwiney and
Robertson, 2005). Excessive nitrogen inputs accelerate the rate of N2O emis-
sions. For example, reducing the application rate from the criterion rate to 25
percent below the recommended rate only reduces N2O by between .2 and 1.3
pounds per acre for the Class A (well-drained) soil in Ohio, depending on the
cropping system. Assuming a credit rate of $15 per ton of CO2 equivalent,
this translates into a payment of between $0.46 and $3.02 per acre. These
rates are insufficient to cover the 10-percent reduction in corn yields that
we assume would occur for such a reduction in N (Bock and Hergert, 1991).
Even for smaller N reductions, it is unlikely that revenue from GHG credits
would be sufficient to cover the increased risk from cutting N application
rates to something close to plant uptake. However, higher offset prices could
increase the incentive to cut application rates to reduce N2O emissions, even
when yields might be affected.
When we apply these results to the survey results summarized in table 3.3,
we conclude that farmers with treated corn acres meeting the rate, timing,
and method criteria or the rate and method criteria (about 42 percent of all
corn acres) will not likely participate in a GHG cap-and-trade program that
would allow farmers to sell offsets from N2O reductions. These farms cannot
make any management changes to reduce N2O without reducing overall
nitrogen use efficiency, which would violate a market rule. The treatment of
such “good stewards” in an emissions trading program is an important policy
issue.
The potential revenue from GHG credits produced by reducing N2O appears
to be quite small. In the Ohio example, only a few situations are capable of
producing credit revenue of over $5 per acre, assuming a credit price of $15
per ton of CO2 equivalent (and the results are similar for the other States
studied). These rates are less than the rates farmers could receive for nutrient
management from EQIP, which is a measure of farmers’ willingness to accept
payment for the practice (table 4.3). In general, farms overapplying nitrogen
and broadcasting fertilizer can produce the largest reductions in N2O.
However, only 8.3 percent of corn acres fall in this category (see table 3.3).
While we found that changes in operating costs after changing management
are near 0 or even negative in most cases, we did not consider short-term
adjustment costs, changes in risk, or the administrative costs of participating
in an offset program. In the case of farms that also have animals, we did not

39
consider the cost of moving manure produced on the farm to more acres (to
reduce application rates), or of moving excess manure off the farm entirely
(Ribaudo et al., 2003)—all of which would reduce farmer participation below
the rates estimated here.
One issue of concern is the possibility that reducing N2O could increase
nitrate losses to water. As described in chapters 2 and 3, changes in manage-
ment could change conditions in the soil so that gaseous forms, such as N2O,
are converted to highly soluble nitrate (NO3). It might seem that allowing
only management changes that do not increase total losses of nitrogen
would prevent this, but we found otherwise. In 25 percent of the cases
where management changes were made to reduce N2O, NO3 losses to water
increased, even though total nitrogen emissions fell. This occurred almost
exclusively when the rate criterion was already being met and injection/incor-
poration was adopted as an additional practice. While overall N2O and total
nitrogen losses decreased, water quality worsened. Such an outcome would
be a concern in regions trying to address water quality problems, such as the
Corn Belt, where corn production is the major source of nitrogen contrib-
uting to hypoxia in the Gulf of Mexico. Including these factors in the analysis
would likely further reduce the net value to society of producing GHG offsets
through N2O emissions reductions.
Response to Price Changes, and What It
Means for an Input Tax
Input prices can influence a farmer’s planning. For example, low fertilizer
prices can lead to “insurance” applications of fertilizer that reduce overall
nitrogen use efficiency. Increases in fertilizer prices relative to other input and
output prices through the use of an input tax would likely decrease fertilizer
use and reduce the number of acres receiving excessive rates. Several States
have levied fertilizer taxes in the past but only at low levels that had little
impact on use.
The effectiveness of an input tax in reducing excessive application rates
would depend largely on the responsiveness of farmers to changes in nitrogen
prices. Data from studies spanning several decades reveal that responses to
a price change (known as the price elasticity) can vary widely, depending on
the data source and time period covered, the type of econometric methods
used to analyze the data, the number of crops covered, and the type of crop
to which the nitrogen fertilizer is applied. While no true consensus exists,
study findings generally show that nitrogen demand was relatively insensi-
tive to price. Burrell (1989) provides a convenient summary of 14 empirical
demand studies through the 1980s. Of those 14 studies, only 4 report elastici-
ties greater than unity. Estimates were generally in the range of -0.20 to -0.70,
implying that a 10-percent increase in the price of fertilizer reduced demand
by 2 to 7 percent (see, for example, Griliches (1958); Carman (1979); Ray
(1982); and Shumway (1983)).
Denbaly and Vroomen (1993) use cointegrated and error-corrected models
with time series data from 1964 to 1989 to estimate short- and longrun
Marshallian elasticities. They report a shortrun Marshallian elasticity of -0.21
and a longrun elasticity of -0.41. Hansen (2004) estimates nitrogen fertilizer

40
demand of farmers in Denmark using an unbalanced panel spanning 1982-91.
He concludes that nitrogen demand is similarly insensitive to own-price, with
an elasticity of -0.45.
Not all studies found the price elasticity of demand for nitrogen fertilizer to
be inelastic. Carman (1979) examines the nitrogen demand in 11 Western
States and finds significant State-level variation in elasticities. Statistically
significant elasticity estimates in Carman’s study range from -0.55 to as large
as -1.84. His study shows that demand can vary significantly even within
a region. Roberts and Heady (1982) also use annual time-series data from
the United States, but spanning 1952-76, and find price elastic demand for
nitrogen applied to corn (-1.148). In a study of aggregate fertilizer, Weaver
(1983) investigates the demand in just two States, North Dakota and South
Dakota, and finds fertilizer demand to be highly elastic, ranging from -1.377
to -2.156.
Some evidence suggests that farmers may be becoming more sensitive
to changes in fertilizer prices. Using 2001 and 2005 field-level data from
ARMS, we estimate a demand elasticity of nitrogen fertilizer of -1.38 for
farmers who applied commercial nitrogen fertilizer to corn (app. 3). Stated
another way, if the price of nitrogen fertilizer was to rise by 10 percent,
farmers would reduce the amount applied by 13.8 percent. At the mean
amount of commercial nitrogen, such a change in price would result in a
decrease of 18.2 lbs of fertilizer per acre.5
Manure can also be used as a source of nitrogen nutrients, usually in conjunc-
tion with commercial nitrogen fertilizer. In the ARMS sample, slightly
less than a quarter of corn farmers applied manure to the field, and all of
them did so in conjunction with commercial nitrogen. When the analysis
is expanded to include these farmers, we find a demand elasticity of -0.67;
that is, for every 10-percent increase in the price of commercial nitrogen
fertilizer, farmers reduce their use of nitrogen (organic and inorganic) by
about 7 percent. The results are driven by farmers who use both manure and
commercial nitrogen; we find they are relatively less sensitive to the price of
commercial nitrogen fertilizer than farmers who apply commercial nitrogen
exclusively, which is consistent with the idea that manure and inorganic forms
of nutrients are imperfect substitutes. Also, manure management decisions on
farms with animals might be driven less by nitrogen prices than by the need
to dispose of manure (Ribaudo et al., 2003).
The estimates of price elasticity can be used to provide a rough estimate of
the tax that would be needed to reduce application rates so that more acres
meet the rate criterion. Figure 4.2 displays the distribution of the nitrogen
application rates that represent the criterion rate described in chapter 3. In
the case of farmers who used commercial nitrogen exclusively, we have esti-
mated an average criterion application rate at 170.8 lbs per acre for produc-
tion year 2005. Thirty-five percent of the 76 million corn acres treated with
nitrogen exceeded their criterion rate (26.7 million acres), and farmers who
exceeded their criterion rate had a mean rate of 185.5 pounds per acre. From
the distribution depicted in figure 4.3, the concentration of farmers near
zero indicates that most of the farmers who applied nitrogen at rates above
the criterion rate are situated near the threshold (also seen in table 3.3). In
5The mean commercial nitrogen
application rate in our sample was
129.72 lbs per acre.

41
1
Criterion rate defined as nitrogen removed at harvest plus 40 percent, based on the
farmer-stated yield goal.
Note: The kernel density, represented by the smooth line, is an estimate of the continuous
density using an Epanechnikov kernel.
Source: USDA, Economic Research Service using USDA’s 2005 Agricultural Resource
Management Survey.
Figure 4.2
Distribution of criterion rates1
for corn, based on reported expected
yield, 2005
Percent
Pounds per acre
5
10
15
20
25
0 100 200 30015050
0
250
1
Criterion rate defined as nitrogen removed at harvest plus 40 percent, based on the
farmer-stated yield goal.
Note: The kernel density, represented by the smooth line, is an estimate of the continuous
density using an Epanechnikov kernel.
Source: USDA, Economic Research Service using USDA’s 2005 Agricultural Resource
Management Survey.
Figure 4.3
Distribution of nitrogen fertilizer applied to corn that exceeded
the criterion rate,1
2005
Percent
Pounds per acre exceeding criterion rate
0
10
20
30
5
15
25
0 100 200 300 40050 150 250 350

42
fact, 50 percent of farmers who exceeded the criterion rate exceeded it by 19
pounds per acre or less.
Table 4.7 provides a summary of the input tax needed to reduce the excess
use of nitrogen by farmers who exceed their criterion rate, evaluated for
differing levels of demand elasticity. From the table it is evident that the more
elastic the demand, the less the price must change to reduce excessive appli-
cation rates. A highly inelastic demand for nitrogen, for example -0.20, would
require more than a 50-percent increase in the price to achieve a 50-percent
reduction in excess application. To achieve a reduction of 75 percent, the price
would have to more than double.
Based on our estimated elasticity of -1.38, if an input tax increased the price
of nitrogen by 7.4 percent, 50 percent (about 13.4 million acres) of the 26.7
million overtreated acres would then meet the rate criterion. Seventy-five
percent of heavy nitrogen users exceed the criterion rate by 43.4 pounds per
acre or less; thus, raising the price of nitrogen by 17 percent would reduce
cropland exceeding the criterion rate by 20 million acres. For context,
consider the mean price of nitrogen fertilizer in 2005 was 33 cents per pound;
therefore, a 7.4-percent change in the price equates to slightly more than 2.4
cents per pound, and a 17-percent change equates to less than 6 cents per lb.
As a policy instrument, a tax on inputs has some desirable characteristics
as well as some well-known drawbacks. First, a tax gives farmers flexibility
in how they reduce emissions. Farmers face heterogeneous costs, and a tax
enables farmers to tailor their input responses (nitrogen abatement) accord-
ingly (Ribaudo et al., 1999). In the case of nitrogen, an input tax directly
affects the farmer’s decision that has the largest impact on nitrogen losses to
the environment. It would also encourage a farmer to manage nitrogen more
carefully, which could lead to appropriate timing and method of application.
A tax does not require monitoring or enforcement, unlike a regulation. It
can also be easily adjusted if policy goals are not met or exceeded. Another
advantage of an input tax is that it raises revenue while reducing application
rates. The revenue could be used to reduce the tax burden of crop producers
through a system of lump-sum rebates to those producers who improve
Table 4.7
Fertilizer price increases needed to reduce excess nitrogen†
applications by 50 percent and 75 percent
Elasticity of
nitrogen fertilizer
demand
Reduce excess nitrogen application by:
50 percent 75 percent
Necessary
price change Tax
Necessary
price change Tax
Percent Dollars Percent Dollars
-0.20 51.2 0.169 117.0 0.386
-0.50 20.5 0.068 46.0 0.154
-0.70 14.6 0.048 33.4 0.110
-1.00 10.2 0.034 23.4 0.077
-1.38 7.4 0.024 17.0 0.056
Note: † Excess nitrogen application is defined as rate exceeding 40 percent more than
nitrogen removed at harvest (see chapter 3).

43
nitrogen use efficiency. Revenue can also be used to remedy damages caused
by nitrogen losses.
A tax on an input also has drawbacks. An input tax makes no distinction
between whether fertilizer is in excess or not. A tax on nitrogen may also
encourage increased use of untaxed manure, resulting in no discernable
change in nitrogen applications where manure is readily available.
The question of who bears the burden of the tax, also known as the incidence,
can have notable distributional consequences. Statutorily, the incidence of
the tax could fall on the wholesaler or retailer of nitrogen fertilizer; however,
the true, or economic, incidence is likely to be shared with the farmer. How
much so is an empirical question that relies on the relative sensitivity of
farmers to the price change, as well as the elasticity of the supply of nitrogen:
the more sensitive a farmer’s demand for nitrogen is, the less of a burden he
or she will bear, all else equal. The supply of nitrogen fertilizer is projected
to more than meet the demand over the near term; therefore, the standard
assumption is that the burden of the excise tax would be considerably shifted
to the consumer of the good, in this case the farmer (Fullerton and Metcalf,
2002; FAO, 2008). While corn production in the United States accounts for
39 percent of the world’s total corn production, the ability of U.S. farmers to
pass along the cost of the tax will depend on the relative elasticities of supply
and demand for corn (USDA, FSA, 2011). While a factor tax on nitrogen may
improve welfare from society’s point of view, ultimately, the tax will change
the functional distribution of income. The distributional impact may be miti-
gated if revenues raised by the tax are returned to the farmer in some manner,
for example, by supporting other conservation activities.
Nitrogen Compliance
Compliance provisions require farmers to meet some minimum standard of
environmental protection on environmentally sensitive land as a condition for
eligibility for many Federal farm program benefits, including conservation
and commodity program payments. Under current compliance requirements,
farm program eligibility could be denied to producers who:
• Fail to implement and maintain an NRCS-approved soil conservation
system on highly erodible land (HEL) (Conservation compliance)
• Convert HEL grasslands to crop production without applying an
approved soil conservation system (Sodbuster)
• Convert a wetland to crop production (Swampbuster)
Evidence suggests that the current compliance provisions have contributed
to a reduction in soil erosion and discouraged the conversions of noncropped
HEL land and wetlands to cropland (Claassen et al., 2004). A possible exten-
sion of the provisions could include nutrient management.
Crop producers are a major source of nitrogen. Assessments of the potential
efficacy of compliance must consider two key questions:

44
• To what extent do crop producers who have the greatest potential for
reducing nitrogen emissions also participate in farm programs?
• Are Government payments to these producers large enough to encourage
broad adoption of practices that improve nitrogen use efficiency and
reduce nitrogen emissions?
Claassen et al. (2004) estimate that 75 percent or more of cropland acres
with medium, high, or very high potential for nitrogen leaching or runoff
are located on farms that receive Government payments. We used data from
the 2005 ARMS corn survey to estimate Government payments received by
corn producers.6 We looked at all treated corn acres, as compliance provides
an incentive both for farmers already practicing good nitrogen management
and willing to continue and for farmers not using nitrogen BMPs and willing
to adopt them. Over 97 percent of corn acres receive Government payments,
averaging $51.39 per acre. This average is higher than our estimated costs
of improving NUE or of adopting NRCS practices. Eighty-eight percent
of treated corn acres receive Government payments in excess of $27 per
acre per year, which is more than the average EQIP payments for nutrient
management or waste use. (Note that for corn acres that are highly erodible
and subject to conservation compliance, it is the sum of erosion control and
nitrogen management costs that would be considered by the farmer.)
A drawback of compliance is that the strength of the incentive is dependent
on the level of Government payments. Current events present a good example.
Direct Government payments have been reduced by about 50 percent between
2005 and 2009 due to a number of factors, including higher crop prices and
smaller disaster payments (USDA, ERS, 2010). Assuming that average per
acre payments to corn producers were reduced by the same percentage, the
average estimated cost of the more expensive nitrogen management prac-
tices, such as waste utilization, would be greater than the program benefit.
Compliance would not be an effective tool in this case. The point is that
program payments can vary greatly, making compliance an unpredictable
policy instrument.
Regulation
Another policy approach for improving NUE is to legally require farms to
adopt and implement particular management practices. Such an approach
would be a major change in the way most of agriculture is treated under
current environmental laws. With few exceptions, agricultural operations
are exempt from regulation under the Clean Water Act and Clean Air Act. A
number of arguments have been used as justification. First, agriculture is so
diverse across the United States that the conventional regulatory approach of
applying uniform standards is impractical (Nanda, 2006). Second, due to the
nonpoint nature of agricultural pollution, individual polluters cannot be iden-
tified except at great cost.
Regulation can conceptually be placed on a continuum between performance
standards and design standards (Ribaudo et al., 1999). Performance standards
directly regulate emissions. Design standards dictate how producers manage
their operations, including practices that should not be used and/or BMPs that
should be adopted. Because of the nonpoint nature of agricultural pollution,
6The ARMS data do not enable us to
identify only those program payments
subject to compliance, but they are a
good approximation.

45
design standards are the only practical approach for addressing nitrogen
losses.
One approach would be to require that farmers adopt specific BMPs to
improve their nitrogen use efficiency. Generally, a practice-based regulation is
inefficient because it requires producers to adopt the same practice, whether
it is appropriate for their particular farm or not. It may be more effective to
define BMPs locally so as to allow flexibility and to account for agriculture’s
heterogeneous nature. For example, a nitrogen management plan is a flexible
practice that is based on a farmer’s resources and cropping system. However,
farmers may fail to implement the plan properly. The effectiveness of a regu-
lation therefore requires effective inspection and enforcement by a resource
management agency. Implementation costs would likely be high. Several
States, such as Nebraska and Maryland, have required farmers in particularly
vulnerable areas to adopt specific nutrient management practices to protect
ground or surface water (Ribaudo, 2009).
One of the few segments of the agricultural sector that has been subjected
to regulatory environmental measures at the national level is animal feeding
operations, reflecting heightened concern over pollution from animal waste
from the largest operations (USDA-EPA, 1999). Manure is estimated to be
a source of about 17 percent of nitrogen entering U.S. waters (Smith et al.,
1997). Clean Water Act regulations now require that animal feeding opera-
tions designated as Concentrated Animal Feeding Operations, or CAFOs, and
needing a National Pollutant Discharge Elimination System (NPDES) permit
(those CAFOs that discharge or propose to discharge to surface waters),
develop and implement a nutrient management plan to cover fields that
receive manure. Such a plan, which would meet NRCS standards, sets a limit
on the amount of nutrients that can be applied per acre of land and specifies
erosion control measures to prevent the loss of sediment and nutrients. Also
under the new regulations, CAFOs that are not required to have an NPDES
permit but that wish to claim the storm water exemption (the provision in
the Clean Water Act that exempts field practices from requiring a discharge
permit) for runoff from fields must develop and implement a nutrient manage-
ment plan to demonstrate that due care is being taken to minimize polluted
runoff from fields receiving manure. If a waterway becomes polluted with
animal waste from field runoff and a CAFO does not have an approved
nutrient management plan, this would be a violation of the Clean Water Act.
This approach sets a level of expected stewardship, namely the implementa-
tion of a nutrient management plan.
Requiring not just CAFOs but all animal feeding operations to adopt nutrient
management plans would be costly. ERS estimates that reductions in net
returns in the livestock and poultry sector would be about $1.4 billion per
year, and national economic welfare for producers and consumers would
decline almost $2 billion per year (Ribaudo et al., 2003). The benefit would
be improved air and water quality. Targeting the regulatory approach only
to those operations most susceptible to pollution problems would lower the
overall costs.

46
Chapter 5
Implications for Nitrogen
Management Policies
Nitrogen is critical for producing abundant food and generating high net
returns to producers, yet it has wide-ranging environmental impacts across
land, water, and the atmosphere. More careful management that reduces
environmental losses would address a number of environmental issues, such
as hypoxia in coastal estuaries and bays, the potential for global warming,
and nutrient enrichment of terrestrial ecosystems. Policymakers have a
number of tools at their disposal, each with its own strengths and weak-
nesses (table 5.1). No one policy approach can be considered “best,” and
a concerted effort to address the Nation’s nitrogen problems will likely
require a solution comprising a mix of policies. Our analysis provides some
guidance on determining which sectors of agriculture are most in need of
improved management, what are the potential pitfalls, and how might the
different policies be orchestrated in an overall policy framework.
Reducing Application Rates as a Priority Policy Goal
Reducing the application of nitrogen fertilizers appears to be the most effec-
tive BMP for reducing the emission of nitrogen into the environment. Based
on the literature, and confirmed by our NLEAP modeling, reducing applica-
tion rates is the one BMP that reduces all forms of reactive nitrogen, even
when the timing and method of application are not ideal. Improving timing
or method of application alone could increase one type of reactive nitrogen
(transmitted to the atmosphere, groundwater, or surface water) while still
reducing total nitrogen emissions. Reducing the application rate is therefore
conducive to an ecosystem approach to management that provides protection
to all ecosystem services and functions. Improving rate, timing, and method
of nitrogen application would produce the greatest environmental benefits.
Reducing application rates that are agronomically excessive may increase
the perceived risk of reduced yields. Farmers often use nitrogen fertilizer
to manage the downside risk due to uncertain weather and soil nitrogen.
Research on how farmers view risk and how they might respond to an incen-
tive payment for reducing application rates, coupled with the use of a risk
management instrument, could result in the development of a more effec-
tive approach for reducing nitrogen in the environment. Revenue or yield
insurance policies could be offered to protect the income of farmers who
adopt conservation measures that improve nitrogen use efficiency but may
decrease yields because of nitrogen insufficiency stemming from unfavor-
able weather conditions. Findings from other studies suggest that insurance
will likely lead to reductions in nitrogen fertilizer applications, but by how
much is uncertain (see Babcock and Hennessy, 1996; Mishra et al., 2005;
Smith and Goodwin, 1996).

47
Corn Is the Most Important Crop for Addressing
Nitrogen-Related Environmental Issues
Corn is the most widely planted crop in the United States and the most inten-
sive user of nitrogen. In 2006, corn accounted for an estimated 65 percent
of the total quantity of nitrogen applied to major U.S. field crops. Corn also
accounted for half of all nitrogen-treated crop acres that were not meeting the
rate, timing, or method of application criteria used in this analysis to define
acceptable nitrogen management. Land used to grow corn accounted for the
largest share of treated acres that had tile drainage in 2006. Although tile
Table 5.1
Summary of policy instruments for improving nitrogen use efficiency
Policy instrument
Characteristics Input tax Information
Financial
incentives Compliance
Emissions
market Regulation
Strength of
incentive
Depends on
level of tax and
price elasticity of
demand.
A farmer will
take action only
if management
practice
improves profits.
Depends on
level of subsidy.
Depends
on level of
Government
program
payments
subject to
compliance.
Depends on
level of demand
from regulated
sectors.
Strong.
Acres covered Covers all acres
that are treated
with commercial
nitrogen.
No guarantee
that acres
in need of
treatment will be
addressed.
No guarantee
that acres most
in need of
treatment will be
addressed.
May not cover
all acres.
May be limited
by geographic
scope of market
and baseline
rules.
Can cover all
acres that use
commercial
nitrogen or
animal waste.
Targets problem Directly
addresses
application rate,
but not timing and
method. Also,
does not address
application of
animal waste.
Information can
be targeted
to specific
problems.
Incentives can
be targeted
to specific
practices
and regions.
However,
important
to consider
potential
environmental
tradeoffs.
Strength of
incentive
may not be
correlated with
acres most
in need of
treatment.
Generally
limited to one
pollutant and
not overall
nitrogen use
efficiency.
Environmental
tradeoffs
a potential
problem.
Can target
all aspects
of nutrient
management.
However,
important
to consider
potential
environmental
tradeoffs.
Flexibility Very flexible
– farmers can
adjust in the most
cost effective way.
Flexible –
farmers act on
information that
is beneficial to
them.
Practice-based
incentives are
less flexible than
incentives on
environmental
performance.
Flexibility
depends on
how provisions
are defined.
Can be flexible,
but depends on
market rules.
Limited
flexibility, as
regulations
generally
require
specific
practices.
Implementation
costs
Easy to
implement,
and generates
revenues that can
be used to reduce
economic impacts
for farmers
who make
improvements.
Requires
research and
extension
outreach.
High costs to
taxpayers.
Enforcement
costs may be
high.
Transactions
costs can be
very high.
Enforcement
costs can be
high.

48
drains improve yields, they also increase the amount of nitrogen that is lost
to surface water. Tiled corn cropland not meeting all three nitrogen manage-
ment criteria would be a prime target for policies for improving nitrogen use
efficiency.
In addition, recent demand pressures due to the biofuels mandate, as well as
increasing international demand for feed grains, suggests that corn acreage
and the intensity of corn production are likely to increase. Together, these
factors increase the importance of raising the NUE in corn production in the
United States, especially on farms that raise livestock and apply manure to
their fields.
Which Policy Is Best?
This analysis provides some guidance on how different policies might
be orchestrated in an overall policy framework. The current approach to
improving nutrient management on cropland has relied primarily on financial
incentives and information. While years of financial and technical assistance
have resulted in some progress, operators of over 65 percent of U.S. crop-
land are still not implementing nitrogen BMPs. Higher payment rates would
encourage more producers to adopt practices that improve nitrogen use effi-
ciency, but the cost to taxpayers may be substantial. The level of financial
assistance that would be required to entice all farmers with cropland acres
needing improved management to enroll in a program would likely consume
most of the budget for EQIP. While nitrogen management is an important
conservation goal, EQIP and other USDA conservation programs address a
host of other issues. Any elevation of nitrogen management as a priority for
EQIP may result in fewer resources for other conservation issues.
Emissions markets, such as those for water quality or greenhouse gases,
could be a source of financial support for improving nitrogen use effi-
ciency. Markets for agricultural offsets shift the financial burden away from
taxpayers to regulated sectors of the economy. While emissions markets are
receiving much interest in efforts to improve water quality and to reduce
greenhouse gas emissions, their role in improving nutrient management on all
acres needing improvement is probably limited. Emissions markets generally
target particular geographic areas or particular practices, potentially limiting
the number of acres that might be affected. Market rules designed to ensure
the “additionality” of offsets by setting baselines consistent with a high level
of management may limit participation by farmers not using BMPs, even
though a market would benefit by their participation. In addition, the nonpoint
source nature of nitrogen emissions from agriculture greatly complicates the
design of markets and raises transactions costs.
If voluntary financial assistance programs or emissions markets are limited
in their ability to improve nitrogen management across all crop acres, what
other approaches might achieve improved nitrogen use efficiency at least
cost? The alternative approaches all result in increased costs for farmers. In
theory, cost-effective policy instruments target the problem, are flexible, are
easy to implement (low transactions costs), and limit costs to both farmers
and Government. A tax on nitrogen fertilizer would provide an incen-
tive to all users to manage commercial nitrogen more carefully. If farmers
are responsive to price, then this instrument may be an effective means of

49
reducing nitrogen losses. Our assessment of farmer price responsiveness
indicates that a relatively low tax may pay high environmental dividends.
However, if farmers are as unresponsive to nitrogen prices as generally
reported in the literature, a substantially higher tax would be necessary to
obtain the same environmental benefits. The burden on farmers would be
substantial. Another drawback of an input tax is that a tax would also be
paid on applications that are not excessive. A tax only on emissions would be
far more efficient, but such a tax is not practical since emissions cannot be
observed or easily measured. Finally, some means of addressing the applica-
tion of animal waste would have to be found, as a fertilizer tax would likely
encourage the substitution of manure for commercial nitrogen.
A nutrient management plan is an inherently flexible management practice
that is strongly encouraged by USDA but only required for animal feeding
operations that are designated as CAFOs. Requiring that all users of nitrogen
inputs (commercial and manure) develop and implement a nutrient manage-
ment plan would be a major change in the way the environmental perfor-
mance of agriculture is managed. The costs to crop farmers of implementing
a nutrient management plan may not be high, except for those managing
large amounts of manure produced on the farm. However, many aspects of a
nutrient management plan, such as application rate, are difficult to observe,
making enforcement difficult and costly.
Enforcement costs could also be high for a compliance approach to getting
farmers to adopt nutrient management plans. The effectiveness of compliance
would depend on the level of program payments received by farmers and a
coincidence of the incentive with those crop acres most in need of improved
management. A large share of crop acres in need of treatment receives high
levels of Government program payments. While the incentive level in 2005
was quite high, program payments have declined in recent years as crop
prices have risen. Continued high prices and general concerns about Federal
budget outlays may limit the strength of a compliance-type policy instrument
unless it is linked to a broader suite of payments than current compliance
requirements.
Improving nitrogen use efficiency reduces the amount of emissions from
cropland but does not eliminate them. In areas where even small levels of
emissions could cause environmental problems, offsite filtering could supple-
ment onfield management. The Government currently provides financial
incentives for creating and preserving wetlands and vegetative filter strips.
Though funds are not allocated solely for nitrogen capture and removal, there
may be reasons to do so. An economic comparison of the two types of filters
suggests that wetlands can be much more cost effective at removing nitrogen
than filter strips. While our analysis found that the cost of establishing a
wetland is greater than the cost of establishing filter strips, annual nitrogen
removal rates are several times greater for wetlands. Filter strips may also
be rendered ineffective where tile drains are present, while wetlands can be
strategically positioned in the landscape to filter drainage coming from tiled
fields. Wetlands also produce a number of other desirable ecosystem services,
such as wildlife habitat. Filter strips, however, can be established in landscape
settings where wetlands cannot. The choice will depend on geography, soil,
and hydrologic conditions.

50
While one single policy instrument does not emerge as a clearly supe-
rior approach to improving NUE across all cropland, a role can be seen
for each. Financial assistance could be made available to those producers
wanting to voluntarily improve nutrient management and to install vegeta-
tive filters or resore wetlands. Since commodity programs are important to
farmers, compliance can provide some incentive for those receiving program
payments. The level of incentive may vary from year to year, but it may be
effective for some farmers. Finally, in regions where nitrogen-related pollu-
tion is of particular concern, such as the Chesapeake Bay watershed and the
watersheds contributing nitrogen to the Gulf of Mexico, a regulatory back-
stop could be a measure of last resort for those unwilling to voluntarily adopt
nitrogen BMPs.
Information Supports All Policies
Information about the environmental and economic performance of improved
nitrogen management practices supports all policies aimed at improving
NUE. Reliable, timely information on soil and plant nitrogen reduces one
source of uncertainty that tends to encourage overapplication of nitrogen. Our
research supports previous findings that testing for nitrogen available in the
soil and contained in crops may result in lower application rates. Information
from testing can be incorporated into an adaptive management framework,
where a farmer evaluates his practices from the previous year (or even at
the start of the current growing season) to assess what options may be avail-
able to improve nutrient management while sustaining yields and reducing
nutrient losses to the environment. So, whether farmers are considering best
nitrogen management practices due to regulation, taxes, or financial incen-
tives, information on how to conduct and interpret nitrogen tests and how
to successfully implement new practices can reduce the overall costs and
increase adoption rates.
Potential Tradeoffs Are an Important Consideration
Reactive nitrogen is easily converted to forms that are readily transported
by hydrologic and atmospheric processes. Therefore, focusing strictly on one
issue, such as nitrate leaching, could lead to increased emissions of other
nitrogen compounds, such as nitrous oxide to the atmosphere, if nitrogen’s
characteristics are ignored. Even when total nitrogen emissions are reduced
by a policy, emissions of one or more nitrogen compounds might increase
and degrade environmental quality. This effect was predicted in the case of
the market for nitrous oxide offsets—farmers reduced total emissions but
increased nitrogen losses to water. These tradeoffs often depend on soils
and cropping practices, so it is difficult to develop general “rules of thumb,”
other than recommending that a holistic approach to management that
considers potential environmental tradeoffs be adopted. Reducing nitrogen
application rates is the easiest and most effective way to reduce all forms of
reactive nitrogen.

51
References
Abt Associates. 2000. Air Quality Impacts of Livestock Waste. Report
prepared for U.S. Environmental Protection Agency, Office of Policy.
Alley, M.M., M.E. Martz, Jr., P.H. Davis, and J.L. Hammons. 2009.
Nitrogen and Phosphorus Fertilization of Corn. Publ. 424-027. Virginia
Cooperative Ext., Virginia Polytechnic Institute and State University.
(Available online at www.pubs.ext.vt.edu/424/424-027/424-027.html.)
(Verified June 3, 2010.)
Alva, A.K., S. Paramasivam, K. Sajwan, A. Fares, J.A. Delgado, and D.
Mattos. 2005. “Nitrogen and Irrigation Management Practices To Improve
Nitrogen Uptake Efficiency and Minimize Leaching,” Journal of Crop
Improvement 15(2):369-420.
Angier, J.T., G.W. McCarty, T.J. Gish, and C.S. Daughtry. 2001. “Impact of
a First-Order Riparian Zone on Nitrogen Removal and Export From an
Agricultural Ecosystem,” TheScientific World 1(Suppl 2):642-651.
Babcock, B.A. 1992. “The Effects of Uncertainty on Optimal Nitrogen
Applications,” Review of Agricultural Economics 14(2):271-280.
Babcock, B.A., and A.M. Blackmer. 1992. “The Value of Reducing Temporal
Input Nonuniformities,” Journal of Agricultural and Resource Economics
17:355-347.
Baker, J. 2001. “Limitations of Improved Nitrogen Management to Reduce
Nitrate Leaching and Increase Use Efficiency,” in Optimizing Nitrogen
Management in Food and Energy Production and Environmental
Protection: Proceedings of the 2nd International Nitrogen Conference on
Science and Policy, TheScientificWorld 1(S2):10-16.
Bausch, W.C., and J.A. Delgado. 2003. “Ground-Based Sensing of Plant
Nitrogen Status in Irrigated Corn To Improve Nitrogen Management,” in
T. Van Toai, D. Major, M. McDonald, J. Schepers, and L. Tarpley (eds.),
Digital Imaging and Spectral Techniques: Applications to Precision
Agriculture and Crop Physiology, ASA Spec. Publ. 66, American Society
of Agronomy, Crop Science Society of America, Soil Science Society of
America. pp. 151-163.
Beckie, H.J., A.P. Moulin, C.A. Campbell, and S.A. Brandt. 1995. “Testing
Effectiveness of Four Simulation Models for Estimating Nitrates and
Water in Two Soils,” Canadian Journal of Soil Sciences 75(1):135-143.
Beegle, D.B., and P.T. Durst. 2003. Nitrogen Fertilization of Corn, Agronomy
Facts 12, College of Agricultural Sciences Cooperative Extension, The
Pennsylvania State University. (Available online at http://export.nbii.gov/
xml/web-resources/xmlfiles)

52
Bishop, C.P., C.R. Shumway, and P.R. Wandschneider. 2010. “Agent
Heterogeneity in Adoption of Anaerobic Digestion Technology:
Integrating Economic, Diffusion, and Behavioral Innovation Theories,”
Land Economics 86(5):585-608.
Bock, B.R., and W. Hergert. 1991. “Fertilizer Nitrogen Management,” in
H. Follet et al. (eds.), Managing Nitrogen for Groundwater Quality and
Farm Profitability, Soil Science Society of America, pp. 140-164.
Bosch, D.J., Z.L. Cook, and K.O. Fuglie. 1995. “Voluntary Versus Mandatory
Agricultural Policies To Protect Water Quality: Adoption of Nitrogen
Testing in Nebraska,” Review of Agricultural Economics 17:13-24.
Breetz, H.L., K. Fisher-Vander, L. Garzon, H. Jacops, K. Kroetz, and R.
Terry. 2004. Water Quality Trading and Offset Initiatives in the U.S.: A
Comprehensive Survey, Dartmouth College. (Available online at http://
www.dartmouth.edu/~kfv/waterqualitytradingdatabase.pdf)
Bricker, S., B. Longstaff, W. Dennison, A. Jones, K. Boicourt, C. Wicks,
and J. Woerner. 2007. Effects of Nutrient Enrichment in the Nation’s
Estuaries: A Decade of Change, NOAA Coastal Ocean Program Decision
Analysis Series No. 26, National Centers for Coastal Ocean Science.
Burrell, A. 1989. “The Demand for Fertilizer in the United Kingdom,”
Journal of Agricultural Economics 40:1-20.
Camacho, R. 1992. Chesapeake Bay Program Nutrient Reduction Strategy
Reevaluation. Financial Cost Effectiveness of Point and Nonpoint Source
Nutrient Reduction Technologies in the Chesapeake Bay Basin, Interstate
Commission on the Potomac River Basin Report No. 92-4.
Carman, H.F. 1979. “The Demand for Nitrogen, Phosphorous, and Potash
Fertilizer Nutrients in the Western United States,” Western Journal of
Agricultural Economics 4(1):23-32.
Cassman, K.G., A. Dobermann, and D.T. Walters. 2002. “Agroecosystems,
Nitrogen-Use Efficiency, and Nitrogen Management,” Ambio
31(2):132-140.
Chouinard, H.H., T. Paterson, P.R. Wandschneider, and A.M. Ohler. 2008.
“Will Farmers Trade Profits for Stewardship? Heterogeneous Motivations
for Farm Practice Selection,” Land Economics 84 (1): 66-82.
Claassen, R., V. Breneman, S. Bucholtz, A. Cattaneo, R. Johansson, and M.
Morehart. 2004. Environmental Compliance in U.S. Agricultural Policy:
Past Performance and Future Potential, Agricultural Economic Report
No. 832, U.S. Department of Agriculture, Economic Research Service.
(Available online at www.ers.usda.gov/publications/aer832)
Cooper, J.C., and R.W. Keim. 1996. “Payments To Encourage Farmer
Adoption of Water Quality Protection Practices,” American Journal of
Agricultural Economics 78(1): 54-64.

53
Cowling, E., J. Galloway, C. Furiness, and J.W. Erisman et al. 2002.
“Optimizing Nitrogen Management and Energy Production and
Environmental Protection.” Presented at the Second International Nitrogen
Conference. Bolger Center, Potomac, MD, October 14-18, 2001. (Available
at www.initrogen.org/fileadmin/user_upload/Second_N_Conf_Report.
pdf. (Verified May 10, 2010).
Crumpton, W. G., D. A. Kovacic, D.L. Hey, and J.A. Kostel. 2008. “Potential
of Restored and Constructed Wetlands To Reduce Nutrient Export From
Agricultural Watershed in the Corn Belt,” in Final Report: Gulf Hypoxia
and Local Water Quality Concerns Workshop, American Society of
Agricultural and Biological Engineers.
Crutchfield, S.R., P.M. Feather, and D.R. Hellerstein. 1995. The Benefits of
Protecting Rural Water Quality: An Empirical Analysis. Agricultural
Economic Report No. 701, U.S. Department of Agriculture, Economic
Research Service, January. (Available online at www.ers.usda.gov/
publications/aer701)
David, M.B., L.E. Drinkwater, and G.E. McIsaac. 2010. “Sources of Nitrate
Yields in the Mississippi River Basin,” Journal of Environmental Quality
39:1657-1667.
Davis, J.G., K.V. Iversen, and M.F. Vigil. 2002. “Nutrient Variability in
Manures: Implications for Sampling and Regional Database Creation,”
Journal of Soil and Water Conservation 57:473-478.
Delgado, J.A. 2001. “Use of Simulations for Evaluation of Best Management
Practices on Irrigated Cropping Systems,” pp. 355-381, in M.J. Shaffer, L.
Ma, and S. Hansen (eds.), Modeling Carbon and Nitrogen Dynamics for
Soil Management, Boca Raton, FL: Lewis Publishers.
Delgado, J.A., and M.J. Shaffer. 2008. “Nitrogen Management Modeling
Techniques: Assessing Cropping Systems/Landscape Combinations,”
p. 539-570, in R.F. Follett and J.L. Hatfield (eds.), Nitrogen in the
Environment: Sources, Problems and Management. New York: Elsevier
Science.
Delgado, J.A., and W. Bausch. 2005. “Potential Use of Precision Conservation
Techniques To Reduce Nitrate Leaching in Irrigated Crops,” Journal of
Soil and Water Conservation 60:379-387.
Delgado, J.A., A.R. Mosier, R.H. Follett, R.F. Follett, D.G. Westfall, L.K.
Klemedtsson, and J. Vermeulen. 1996. “Effects of N Management on N2O
and CH4 Fluxes and 15N Recovery,” Nutrient Cycling in Agroecosystems
46:127-134.
Delgado, J.A., C.M. Gross, H. Lal, H. Cover, P. Gagliardi, S.P. McKinney,
E. Hesketh, and M.J. Shaffer. 2010a. “A New GIS Nitrogen Trading Tool
Concept for Conservation and Reduction of Reactive Nitrogen Losses to
the Environment,” Advances in Agronomy 105:117-171.

54
Delgado, J.A., M. Shaffer, C. Hu, R. Lavado, J. Cueto-Wong, P. Joosse, D.
Sotomayor, W. Colon, R. Follett, S. DelGrosso, X. Li, and H. Rimski-
Korsavok. 2008a. “An Index Approach to Assess Nitrogen Losses to the
Environment,” Ecological Engineering 32:108–120.
Delgado, J.A., M.J. Shaffer, H. Lal, S.P. McKinney, C.M. Gross, and H.
Cover. 2008b. “Assessment of Nitrogen Losses to the Environment With a
Nitrogen Trading Tool (NTT),” Computers and Electronics in Agriculture
63:193-206.
Delgado, J.A., R. Khosla, W.C. Bausch, D.G. Westfall, and D. Inman. 2005.
“Nitrogen Fertilizer Management Based on Site-Speciﬁc Management
Zones Reduce Potential for NO3-N Leaching,” Journal of Soil and Water
Conservation 60:402–410.
Delgado, J.A., R.F. Follett, and M.J. Shaffer. 2000. “Simulation of NO3-N
Dynamics for Cropping Systems With Different Rooting Depths,” Soil
Science Society of America Journal 64:1050-1054.
Delgado, J.A., R. Ristau, M.A. Dillon, H.R. Duke, A. Stuebe, R.F.
Follett, M.J. Shaffer, R.R. Riggenbach, R.T. Sparks, A. Thompson,
L.M. Kawanabe, A. Kunugi et al. 2001. “Use of Innovative Tools To
Increase Nitrogen Use Efﬁciency and Protect Environmental Quality in
Crop Rotations,” Communications in Soil Science and Plant Analysis
32:1321-1354.
Delgado, J.A., S.J. Del Grosso, and S.M. Ogle. 2010b. “15N Isotopic Crop
Residue Cycling Studies and Modeling Suggest That IPCC Methodologies
To Assess Residue Contributions to N2O-N Emissions Should Be
Reevaluated,” Nutrient Cycling in Agroecosystems 86:383–390.
Denbaly, M., and H. Vroomen. 1993. “Dynamic Fertilizer Nutrient Demands
for Corn: A Cointegrated and Error-Correcting System,” American
Journal of Agricultural Economics 75(1): 203-209.
Dobermann, A., R. Ferguson, G. Hergert, C. Shapiro, D. Tarkalson, D.T.
Walters, and C. Wortmann. 2006. “Nitrogen Response in High-Yielding
Corn Systems of Nebraska,” in Proceedings of the Great Plains Soil
Fertility Conference, Denver, CO, March 7-8, Vol. 11, pp. 50-59, Potash
and Phosphate Institute, Brookings, SD.
Dodds, W.K., W.W. Bouska, J.L. Eitzmann, T.J. Pilger, K.L. Pitts, A.J. Riley,
J.T. Schloesser, and D.J. Thornbrugh. 2009. “Eutrophication of U.S.
Freshwaters: Analysis of Potential Economic Damages,” Environmental
Science and Technology 43(1):12-19.
Doerge, T.A., R.L. Roth, B.R. Gardner (eds.). 1991. Nitrogen Fertilizer
Management in Arizona, University of Arizona, College of Agriculture.
Dosskey, M.G., D.E. Eisenhauer, and M.J. Helmers. 2005. “Establishing
Conservation Buffers Using Precision Information,” Journal of Soil and
Water Conservation 62:349-354.

55
Dosskey, M.G., M.J. Helmers, and D.E. Eisenhauer. 2007. “An Approach for
Using Soil Surveys To Guide the Placement of Water Quality Buffers,”
Journal of Soil and Water Conservation 61:344-354.
Drury, C.F., W.D. Reynolds, C.S. Tan, T.W. Welacky, W. Calder, and N.B.
McLaughlin. 2006. “Emissions of Nitrous Oxide and Carbon Dioxide:
Influence of Tillage Type and Nitrogen Placement Depth,” Soil Science
Society of America Journal 70(2):570-581.
Dubrovsky, N.M., K.R. Burow, G.M. Clark, J.A.M. Gronberg, P.A. Hamilton,
K.J. Hitt, D.K. Mueller, M.D. Munn, L.J. Puckett, B.T. Nolan, M.G.
Rupert, T.M. Short, N.E. Spahr, L.A. Sprague, and W. G. Wilbur. 2010.
The Quality of Our Nation’s Waters - Nutrients in the Nation’s Streams
and Groundwater, 1992-2004. Circular-1350, U.S. Geological Survey.
Eghball, B., B.J. Wienhold, J.E. Gilley, and R.A. Eigenberg. 2002.
“Mineralization of Manure Nutrients,” Journal of Soil and Water
Conservation 57:470-473.
Espinoza, L., and J. Ross. 2008. “Fertilization and Liming,” pp. 23-27, in L.
Espinoza and J. Ross (eds.), Corn Production Handbook. Cooperative
Extension Service, University of Arkansas, Little Rock, AR. (Available
online at www.uaex.edu/Other_Areas/ publications/HTML/MP-437.asp.)
(Verified June 3, 2010.)
Flessa, H., and F. Beese. 2000. “Laboratory Estimates of Trace Gas
Emissions Following Surface Application and Injection of Cattle Slurry,”
Journal of Environmental Quality 29: 262-268.
Follett, J.R., R.F. Follett, and W.C. Herz. 2010. “Environmental and Human
Impacts of Reactive Nitrogen,” in J.A. Delgado and R.F. Follett (eds.),
Advances in Nitrogen Management for Water Quality, Soil and Water
Conservation Society.
Food and Agriculture Organization of the United Nations (FAO). 2008.
Current World Fertilizer Trends and Outlook to 2012, Rome.
Fox, R.H., W.P. Piekielek, and K.E. Macneal. 1996. “Estimating Ammonia
Volatilization Losses From Urea Fertilizers Using a Simplified
Micrometerological Sampler,” Soil Science Society of America Journal
60:596-601.
Freney, J.R., J.R. Simpson, and O.T. Denmead. 1981. “Ammonia
Volatilization,” Ecological Bulletin 33:291-302.
Freney, J.R., M.B. Peoples, and A.R. Mosier. 1995. Efficient Use of Fertilizer
Nitrogen by Crops, Extension Bulletin 414, Food & Fertilizer Technology
Center, Taipei, Taiwan. (Available online at www.agnet.org/library/
eb/414) (Accessed December 29, 2009.)
Fullerton, D., and G.E. Metcalf. 2002. “Chapter 26 Tax Incidence,”
Handbook of Public Economics 4:1787-1872.

56
Galloway, J.N., A.R. Townsend, J.W. Erisman, M. Bekunda, Z. Cai,
J.R. Freney, L.A. Martinelli, S.P. Seitzinger, and M.A. Suttan. 2008.
“Transformation of the Nitrogen Cycle: Recent Trends, Questions, and
Potential Solutions,” Science 320(May 16):889-892.
Galloway, J.N., J.D. Aber, J.W. Erisman, S.P. Seitzinger, R.W. Howarth, E.B.
Cowling, and B.J. Cosby, 2003, “The Nitrogen Cascade,” BioScience
53(4): 341-356.
Ghosh, G., Ribaudo, M., and J. Shortle. 2011. “Do Baseline Requirements
Hinder Trades in Water Quality Trading Programs?” Journal of
Environmental Management 92(8):2076-2084.
Gollehon, N., M. Caswell, M. Ribaudo, R. Kellogg, C. Lander, and D. Letson.
2001. Confined Animal Production and Manure Nutrients. Agricultural
Information Bulletin No. 771, U.S. Department of Agriculture, Economic
Research Service. (Available online at www.ers.usda.gov/publications/
aib771/)
Goolsby, D.A., W.A. Battaglin, B.T. Aulenbach, and R.P. Hooper. 2001.
“Nitrogen Input to the Gulf of Mexico,” Journal of Environmental Quality
30: 329-336.
Greene, W.H. 2000. Econometric Analysis, Fourth Edition. Upper Saddle
River, NJ: Prentice Hall.
Griliches, Zvi. 1958. “The Demand for Fertilizer: An Economic
Interpretation of a Technological Change,” Journal of Farm Economics
40(3): 591-606.
Grossman, M.R. 2007. “Agriculture and the Polluter Pays Principle,”
Electronic Journal of Comparative Law 11(3), 1-66.
Hansen, Lars Garn. 2004. “Nitrogen Fertilizer Demand From Danish Crop
Farms: Regulatory Implications of Heterogeneity,” Canadian Journal of
Agricultural Economics 52(3): 313-313.
Hefting, M.M., R. Bobbink, and H. de Caluwe. 2003. “Nitrous Oxide
Emission and Denitrification in Chronically Nitrate-Loaded Riparian
Buffer Zones,” Journal of Environmental Quality 32:1194-1203.
Henandez, M., and W. Mitsch. 2006. “Influence of Hydrologic Pulses,
Fooding Frequency, and Vegetation on Nitrous Oxide Emissions From
Created Riparian Marshes,” Wetlands 26(3): 862-877.
Hernandez-Ramirez, G., S.M. Brouder, D.R. Smith, and G.E. Van Scoyoc.
2009. “Greenhouse Gas Fluxes in an Eastern Corn Belt Soil: Weather,
Nitrogen Source, and Rotation,” Journal of Environmental Quality
38:841-854.
Hey, D.L., L.S. Urban, and J.A. Kostel. 2005. “Nutrient Farming: The
Business of Environmental Management,” Ecological Engineering
24:279–287.

57
Hoag, D.L. and J.S. Hughes-Popp. 1997. “Theory and Practice of Pollution
Credit Trading in Water Quality Management,” Review of Agricultural
Economics 19:252-262.
Horan, R.D., and J.S. Shortle. 2011. “Economic and Ecological Rules for
Water Quality Trading,” Journal of the American Water Resources
Association 47(1):58-69.
Hutchinson, G.L., A.R. Mosier, and C.E. Andre. 1982. “Ammonia and Amine
Emissions From a Large Cattle Feedlot,” Journal of Environmental
Quality 11:288-293.
Iowa Soybean Association. 2008a. “Managing Nitrogen for Economic and
Environmental Performance,” On-Farm Network Update (July):2-11.
Iowa Soybean Association. 2008b. “Understanding Soil Nitrogen Dynamics,”
On-Farm Network Update (October):2-11.
Jacobs, T.J., and J.W. Gilliam. 1985. “Riparian Losses of Nitrate From
Agricultural Drainage Waters,” Journal of Environmental Quality
14:472-478.
Jansson, M., R. Andersson, H. Berggren, and L. Leonardson. 1994.
“Wetlands and Lakes as Nitrogen Traps,” Ambio 23(6):320-325.
Jarecki, M.K., T.B. Parkin, A.S.K. Chan, T.C. Kaspar, T.B. Moorman,
J.W. Singer, B.J. Kerr, J.L. Hatfield, and R. Jones. 2009. “Cover Crop
Effects on Nitrous Oxide Emission From a Manure-Treated Mollisol,”
Agriculture, Ecosystems, and Environment 134:29-35.
Just, R. 1975. “Risk Aversion Under Profit Maximization,” American Journal
of Agricultural Economics 57(2): 347-352.
Khakural, B.R., and P.C. Robert. 1993. “Soil Nitrate Leaching Potential
Indices: Using a Simulation Model as a Screening System,” Journal of
Environmental Quality 22(4):839-845.
Khosla, R., K. Fleming, J. Delgado, T. Shaver, and D. Westfall. 2002. “Use
of Site Specific Management Zones To Improve Nitrogen Management
for Precision Agriculture,” Journal of Soil and Water Conservation
57:513-518.
King, D.M. 2005. “Crunch Time for Water Quality Trading,” Choices
20(First Quarter):71-76.
King, D.M., and P.J. Kuch. 2003. “Will Nutrient Credit Trading Ever Work?
An Assessment of Supply and Demand Problems and Institutional
Obstacles,” Environmental Law Reporter 33:10352-10368
Kirchmann, H., and L. Bergstrom. 2001. “Do Organic Farming Practices
Reduce Nitrate Leaching?” Communications in Soil Science and Plant
Analysis 32:997-1028.

58
Lander, C.H., D. Moffitt, and K. Alt. 1998. Nutrients Available for Livestock
Manure Relative to Crop Growth Requirements. Resource Assessment and
Strategic Planning Working Paper 98-1, U.S. Department of Agriculture,
Natural Resources Conservation Service, February. (Available online
at www.nrcs.usda.gov/technical/NRI/pubs/nlweb.html.) (Accessed
September 10, 2009.)
Legg, J.O., and J.J. Meisinger. 1982. “Soil Nitrogen Budgets,” in Stevenson,
F.J. (ed.), Nitrogen in Agricultural Soils, Agronomic Monograph 22,
American Society of Agronomy, pp. 503-557.
Lichtenberg, E. 2004. “Cost-Responsiveness of Conservation Practice
Adoption: A Revealed Preference Approach,” Journal of Agricultural and
Resource Economics 29(3):420-435.
Lichtenberg, E., and B. V. Lessley. 1992. “Water Quality, Cost-Sharing and
Technical Assistance: Perceptions of Maryland Farmers,” Journal of Soil
and Water Conservation 47(3): 260-264.
Liu, X., A. Mosier, A. Halvorson, and F. Zhang. 2006. “The Impact of
Nitrogen Placement and Tillage on NO, N2O, Ch4 and CO2 Fluxes From a
Clay Loam Soil,” Plant Soil 280(1):177-188.
Lowrance, R., L.S. Altier, J.D. Newbold, R.R. Schnabel, P.M. Groffman,
J.M. Denver, D.L. Correll, J.W. Gilliam, J.L. Robinson, R.B. Brinsfield,
K.W. Staver, W. Lucas, and A.H. Todd. 1997. “Water Quality Functions
of Riparian Forest Buffer Systems in Chesapeake Bay Watersheds,”
Environmental Management 21:687-712.
Lowrance, R.R., R.L. Todd, and L.E. Asmussen. 1984. “Nutrient Cycling in
an Agricultural Watershed: Streamflow and Artificial Drainage,” Journal
of Environmental Quality 13:27-32.
Malhi, S., and M. Nyborg. 1991. “Recovery of 15N-labelled Urea: Influence
of Zero Tillage, and Time and Method of Application,” Fertilizer
Research 28(3):263-269.
Matson, P.A., W.J. Parton, A.G. Power, and M.J. Swift. 1997. Agricultural
Intensification and Ecosystem Properties,” Science 277:504-509.
Mayer, P.M., S.K. Reynolds, Jr., and T.J. Canfield. 2005. Riparian Buffer
Width, Vegetative Cover, and Nitrogen Removal Effectiveness: A
Review of Current Science and Regulations. EPA/600/R-05/118. U.S.
Environmental Protection Agency, National Risk Management Research
Laboratory.
McSwiney, C.P., and G.P. Robertson. 2005. “Nonlinear Response of N2O
Flux to Incremental Fertilizer Addition in a Continuous Maize (Zea mays
L.) Cropping System,” Global Change Biology 11:1712–1719.

59
Meisinger, J.J., and G.W. Randall. 1991. “Estimating N Budgets for Soil-Crop
Systems,” in R.F. Follett et al. (ed.), Managing Nitrogen for Groundwater
Quality and Farm Productivity, American Society of Agronomy, p.
85-124.
Meisinger, J.J., and J.A. Delgado. 2002. “Principles for Managing Nitrogen
Leaching,” Journal of Soil and Water Conservation 57: 485-498.
Meisenger, J.J., J.S. Schepers, and W.R. Raun. 2008. “Crop Nitrogen
Requirement and Fertilization,” in J.S. Schepers and W.R. Raun (eds.),
Nitrogen in Agricultural Systems, American Society of Agronomy, Crop
Science Society of America, Soil Science Society of America.
Millar, N., G.P. Robertson, P.R. Grace, R.J. Gehl, and J.P. Hoben. 2010.
“Nitrogen Fertilizer Management for Nitrous Oxide (N2O) Mitigation in
Intensive Corn (Maize) Production: An Emissions Reduction Protocol
for US Midwest Agriculture,” Mitigation and Adaptation Strategies for
Global Change 15:185-204.
Mitsch, W.J., and J.W. Day. 2006. “Restoration of Wetlands in the
Mississippi-Ohio-Missouri (MOM) River Basin: Experience and Needed
Research,” Ecological Engineering 26:55-69.
Mitsch, W.J., J.W. Day, Jr., J.W. Gilliam, P.M. Groffman, D.L. Hey, G.W.
Randall, and N. Wang. 1999. Reducing Nutrient Loads, Especially
Nitrate-Nitrogen, to Surface Water, Ground Water, and the Gulf of
Mexico: Topic 5 Report for the Integrated Assessment on Hypoxia in the
Gulf of Mexico. Decision Analysis Series No. 19, U.S. Department of
Commerce, National Oceanic and Atmospheric Administration.
Mosier, A.R. and L. Klemedtsson. 1994. “Measuring Denitriﬁcation in the
Field,” in R.W. Weaver, S. Angle, P. Bottomley, D. Bezdicek, S. Smith,
A. Tabatabai, and A. Wollum (eds.), Methods of Soil Analysis: Part 2,
Microbiological and Biochemical Properties, Soil Science Society of
America, pp. 1047-1065.
Mosier A.R., J.W. Doran, and J.R. Freney. 2002. “Managing Soil
Denitriﬁcation,” Journal of Soil and Water Conservation 57:505-513.
Musser, W.N., J.S. Shortle, K. Kreahling, B. Roach, W. Huang, D.B.
Beegle, and R.H. Fox. 1995. “An Eonomic Analysis of the Pre-Sidedress
Nitrogen Test for Pennsylvania Corn Production,” Review of Agricultural
Economics 17(1):25-35.
Nanda, V.P. 2006. “Agriculture and the Polluter Pays Principle,” American
Journal of Comparative Law 54:317-339.
Pennsylvania Department of Environmental Protection, 2008, Final Trading
of Nutrient and Sediment Reduction Credits – Policy and Guidelines,
Appendix A, Doc. No. 392-0900-001, Harrisburg, PA.

60
Peoples, M.B., J.R. Freney, and A.R. Mosier. 1995. “Minimizing Gaseous
Losses of Nitrogen,” in P.E. Bacon (ed.), Nitrogen Fertilization in the
Environment, New York: Marcel Dekker, Inc., pp. 565-602.
Pionke, H.B., and R.R. Lowrance. 1991. “Fate of Nitrate in Subsurface
Drainage Waters,” in R.F. Follett, D.R. Keeney and R.M. Cruse (eds.),
Managing Nitrogen for Ground Water Quality and Farm Proﬁtability,
Soil Science Society of America, pp. 237-57.
Power, J.F., R. Wiese, and D. Flowerday. 2001. “Managing Farming
Systems for Nitrate Control: A Research Review From Management
Systems Evaluation Areas,” Journal of Environmental Quality 30
(November-December):1866-1880.
Rabalais, N.N., R.E. Turner, and W.J. Wiseman, Jr. 2002a. “Gulf of
Mexico Hypoxia, A.K.A. “The Dead Zone,” Annual Review of Ecology,
Evolution, and Systematics 33:235-63.
Rabalais, N.N., R.E. Turner, and D. Scavia. 2002b. “Beyond Science Into
Policy: Gulf of Mexico Hypoxia and the Mississippi River,” Bioscience
52(2):129-142.
Rajsic, P., A. Weersink, and M. Gandorfer. 2009. “Risk and Nitrogen
Application Levels,” Canadian Journal of Agricultural Economics
57:223-239
Randall, G.W., J.A. Delgado, and J.S. Schepers. 2008. “Nitrogen Management
To Protect Water Resources,” in J.S. Schepers, W.R. Raun, R.F. Follett,
R.H. Fox, and G.W. Randall (eds.), Nitrogen in Agricultural Systems.
Agronomy Monograph 49, Soil Science Society of America, pp. 911-946.
Randall, G.W., M.J. Goss, and N.R. Fausey. 2010. “Nitrogen and Drainage
Management To Reduce Nitrate Losses to Subsurface Drainage,” in J.A.
Delgado, and R.F. Follett (eds.), Advances in Nitrogen Management for
Water Quality, Soil and Water Conservation Society.
Raun, W.R., and J.S. Schepers. 2008. ”Nitrogen Management for Improved
Use Efﬁciency,” in J.S. Schepers, W.R. Raun, R.F. Follett, R.H. Fox,
and G.W. Randall (eds.) Nitrogen in Agricultural Systems, Agronomy
Monograph 49, Soil Science Society of America.
Ray, S.C. 1982. “A Translog Cost Function Analysis of U.S. Agriculture,
1939-77,” American Journal of Agricultural Economics 64(3): 490-498.
Ribaudo, M.O. 2009. “Non-Point Pollution Regulation Approaches in the
U.S,” in J. Albiac and A. Dinar (eds.), The Management of Water Quality
and Irrigation Technologies, London: Earthscan.
Ribaudo, M., and C. Nickerson. 2009. “Agriculture and Water Quality
Trading: Exploring the Possibilities,” Journal of Soil and Water
Conservation 64(1):1-7.

61
Ribaudo, M.O. and J. Gottlieb. 2011. “Point-Nonpoint Trading – Can It
Work?” Journal of the American Water Resources Association 47(1):5-14.
Ribaudo, M.O., and R.D. Horan. 1999. “The Role of Education in Nonpoint
Source Pollution Control Policy,” Review of Agricultural Economics
21(2):331-343.
Ribaudo, M., L. Hansen, D. Hellerstein, and C. Greene. 2008. The Use of
Markets To Increase Private Investment in Environmental Stewardship.
Economic Research Report No. 64, U.S. Department of Agriculture,
Economic Research Service. (Available online at www.ers.usda.gov/
publications/err64)
Ribaudo, M., N. Gollehon, M. Aillery, J. Kaplan, R. Johansson, J. Agapoff,
L. Christensen, V. Breneman, and M. Peters. 2003. Manure Management
for Water Quality: Costs to Animal Feeding Operations of Applying
Manure Nutrients to Land, Agricultural Economic Report No. 824. U.S.
Department of Agriculture, Economic Research Service. (Available online
at www.ers.usda.gov/publications/aer824)
Ribaudo, M.O., R.D. Horan, and M.E. Smith. 1999. Economics of Water
Quality Protection From Nonpoint Sources: Theory and Practice.
Agricultural Economic Report No. 782, U.S. Department of Agriculture,
Economic Research Service. (Available online at www.ers.usda.gov/
publications/aer782)
Roberts, R.K., and E.O. Heady. 1982. “Fertilizer Demand Functions for
Speciﬁc Nutrients Applied to Three Major U.S. Crops,” Western Journal
Robertson, D.M., G.E. Schwarz, D.A. Saad, and R.B. Alexander. 2009.
“Incorporating Uncertainty Into the Ranking of SPARROW Model
Nutrient Yields From Mississippi/Atchafalaya River Basin Watersheds,”
Journal of the American Water Resources Association 45(2):535-549.
Rochette, P., D.A. Angers, M.H. Chantigny, N. Bertrand, and D. Cote. 2004.
“Carbon Dioxide and Nitrous Oxide Emissions Following Fall and Spring
Applications of Pig Slurry to an Agricultural Soil,” Soil Science Society of
America Journal 68:1410-1420.
Roosen, J., and D.A. Hennessy. 2003. “Tests for the Role of Risk Aversion on
Input Use,” American Journal of Agricultural Economics 85(1):30-43.
Schepers, J.S., K. D. Frank, and C. Bourg. 1986. “Effect of Yield Goal
and Residual Soil Nitrogen Considerations on Nitrogen Fertilizer
Recommendations for Irrigated Maize in Nebraska,” Journal of Fertilizer
Issues 3:133-139.
Schlegel A.J., K.C. Dhuyvetter, and J.L. Havlin. 1996. “Economic and
Environmental Impacts of Long-Term Nitrogen and Phosphorus
Fertilization,” Journal of Production Agriculture (9):114-118.

62
Shaffer, M.J., and J.A. Delgado. 2001. “Field Techniques for Modeling
Nitrogen Management,” in Follett et al. (ed.), Nitrogen in the
Environment: Sources, Problems, and Management. Elsevier Science, pp.
391-411.
Shaffer, M.J., J.A. Delgado, C. Gross, R.F. Follett, and P. Gagliardi. 2010.
“Simulation Processes for the Nitrogen Loss and Environmental
Assessment Package (NLEAP),” in J.A. Delgado and R.F. Follett (eds.),
Advances in Nitrogen Management for Water Quality, Soil and Water
Conservation Society.
Sharpe, R.R., and L.A. Harper. 1995. “Soil, Plant, and Atmospheric
Conditions as They Relate to Ammonia Volatilization,” Fertilizer
Research 42:149-153.
Sheriff, G. 2005. “Efficient Waste? Why Farmers Over-Apply Nutrients and
the Implications for Policy Design,” Review of Agricultural Economics
27(4):542-557.
Shortle, J.S. 1990. “Allocative Efficiency Implications of Water Pollution
Abatement Cost Comparisons,” Water Resources Research 26:793-797.
Shumway, R. 1983. “Supply, Demand, and Technology in a Multiproduct
Industry: Texas Field Crops,” American Journal of Agricultural
Economics 65(4): 748-760.
Smith, R.A., G.E. Schwarz, and R.B. Alexander. 1997. “SPARROW Surface
Water-Quality Modeling Nutrients in Watersheds of the Conterminous
United States: Model Predictions for Total Nitrogen (TN) and Total
Phosphorus (TP).” (Available online at http://water.usgs.gov/nawqa/
sparrow/wrr97/results.html).
Turner, R.E., and N.N. Rabalais. 2003. “Linking Landscape and Water
Quality in the Mississippi River Basin for 200 Years,” Vol. 53 No. 6,
BioScience 563-572.
Turner, R.E., N.N. Rabalais, D. Scavia, and G.F. McIsaac. 2007. “Corn
Belt Landscapes and Hypoxia of the Gulf of Mexico,” in J.I. Nassauer,
M.V. Santelmann, and D. Scavia (eds.), From the Corn Belt to the Gulf:
Societal and Environmental Implication of Alternative Agricultural
Futures, Resources for the Future.
U.S. Department of Agriculture, Economic Research Service (USDA, ERS).
2010. “Direct Government Payments by Program, United States, 1996-
2009,” Farm Income: Data Files GP9609us. (Available online at www.ers.
usda.gov/data/farmincome/finfidmuxls.htm#payments)
U.S. Department of Agriculture, Foreign Agricultural Service (USDA, FSA).
2011. “Corn at a Glance.” (Available online at www.fas.usda.gov/htp/
CP2011/Corn-2011-Final.pdf)

63
U.S. Department of Agriculture, National Agricultural Statistics Service
(USDA, NASS). 2008. Agricultural Statistics 2008. (Available online
at www.nass.usda.gov/Publications/Ag_Statistics/2008/index.asp.)
(Accessed February 17, 2010.)
U.S. Department of Agriculture, Natural Resources Conservation Service
(USDA, NRCS). 2010. Assessment of the Effects of Conservation
Practices on Cultivated Cropland in the Upper Mississippi River Basin.
(Available online at ftp://ftp-fc.sc.egov.usda.gov/NHQ/nri/ceap/UMRB_
final_draft_061410.pdf)
U.S. Department of Agriculture, Natural Resources Conservation Service
(USDA, NRCS). 2006. “Natural Resources Conservation Service
Conservation Practice Standard: Nutrient Management,” National
Handbook of Conservation Practice.
U.S. Department of Agriculture – U.S. Environmental Protection Agency.
(USDA-EPA) 1999. Unified National Strategy for Animal Feeding
Operations. (Available online at http://cfpub.epa.gov/npdes/afo/ustrategy.
cfm)
U.S. Environmental Protection Agency (EPA). 1992. Another Look: National
Survey of Pesticides in Drinking Water Wells, Phase II Report, EPA
579/09-91-020.
U.S. Environmental Protection Agency (EPA). 2009. National Summary
of Impaired Waters and TMDL Information. (Available online at http://
iaspub.epa.gov/waters10/attains_nation_cy.control?p_report_type=T)
U.S. Environmental Protection Agency (EPA). 2010a. Inventory of U.S.
Greenhouse Gas Emissions and Sinks: 1990-2008, EPA 430-R-10-006.
U.S. Environmental Protection Agency, Office of Atmospheric Programs
(EPA). 2010b. Methane and Nitrous Oxide Emissions from Natural
Sources, EPA 430-R-10-001. April.
U.S. Environmental Protection Agency, Office of Wastewater Management
(EPA). 2007. Water Quality Trading Toolkit for Permit Writers, EPA
883-R-07-004, August.
U.S. Environmental Protection Agency, Office of Water (EPA). 2004. Water
Quality Trading Assessment Handbook: Can Water Quality Trading
Advance Your Watershed’s Goals? EPA 841-B-04-001.
Vanotti, M.B., and L.G. Bundy. 1994. “An Alternative Rationale for Corn
Nitrogen Fertilizer Recommendations,” Journal of Production Agriculture
(7):243-249.
Vitosh, M.L., J.W. Johnson, and D.B. Mengel. 1995. Tri-State Fertilizer
Recommendations for Corn, Soybeans, Wheat and Alfalfa. Extension
Bulletin E-2567, Michigan State University, The Ohio State University,
and Purdue University. (Available online at http://ohioline.osu.edu/ e2567/
index.html.) (Verified June 3, 2010.)

64
Vitousek, P.M., J.D. Aber, R.W. Howarth, G.E. Likens, P.A. Matson, D.W.
Schindler, W.H. Schlesinger, and D.G. Tilman. 1997. “Human Alteration
of the Global Nitrogen Cycle: Sources and Consequences,” Ecological
Applications 73(3):737-750.
Weaver, R.D. 1983. “Multiple Input, Multiple Output Production Choices and
Technology in the U.S. Wheat Region,” American Journal of Agricultural
Economics 65(1): 45-56.
Westermann, D.T., and G.E. Kleinkopf. 1985. “Nitrogen Requirements of
Potatoes,” Agronomy Journal 77:616-621.
Westermann, D.T., G.E. Kleinkopf, and L.K. Porter. 1988. “Nitrogen
Fertilizer Efﬁciencies on Potatoes,” American Potato Journal 65:377-386.
Williams, J.R., and D.E. Kissel. 1991. “Water Percolation: An Indicator of
Nitrogen-Leaching Potential,” in R.F. Follet, D.R. Keeney, and R.M.
Cruse (eds.), Managing Nitrogen for Groundwater Quality and Farm
Proﬁtability, Soil Science Society of America, pp. 59-83.
Wisconsin Department of Natural Resources, 2002. Agricultural Nonpoint
Performance Standards and Prohibitions, NR 151, Subchapter 2.
Woodward, R.T., and R. A. Kaiser. 2002. “Market Structures for U.S. Water
Quality Trading,” Review of Agricultural Economics 24:366-383.
Wu, J., and B.A. Babcock. 1998. “The Choice of Tillage, Rotation, and Soil
Testing: Economics and Environmental Implications,” American Journal
Wulf, S., M. Maeting, and J. Clemens. 2002. “Application Technique and
Slurry Co-Fermentation Effects on Ammonia, Nitrous Oxide, and
Methane Emissions After Spreading: II Greenhouse Gas Emissions,”
Journal of Environmental Quality 31:1795-1801.
Xu, C., M.J. Shaffer, and M. Al-kaisi. 1998.” Simulating the Impact of
Management Practices on Nitrous Oxide Emissions,” Soil Science Society
of America Journal 62:736-742.

65
Appendix 1
Estimating Water Treatment Costs
We estimated a treatment cost model with data from the 1996 American
Water Works Association (AWWA) survey of its members. There are only
52 usable observations for which utilities provided all required data. This
is the last survey in which data on costs and water quality (both raw water
coming into and finished water going out of the utility) were gathered at the
same time by AWWA. We assume this sample is representative of all water
treatment plants. The model is a variable cost function with two outputs (one
desirable (water) and one undesirable (nitrogen)); four inputs (three vari-
able and one fixed); and nine factors hypothesized to influence production of
drinking water (app. table 1.1).
The bootstrap method employed uses network density as the stratum—the
result of this stratification is a more homogeneous sample and hence a smaller
standard error.
Econometric specification of simple production
model and discussion
1 2
1 2 1 2
3 3 3
ˆ ˆ ˆ ˆˆ ˆln ln ln ln ln ln
w wV
y N K
w w w
οβ α α β β η
⎛ ⎞ ⎛ ⎞ ⎛ ⎞
= + + + + +⎜ ⎟ ⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠ ⎝ ⎠
2.01*** 0.80*** 0.03*** 0.62 *** 0.43*** 0.03**
(4.55) (81.50) (2.69) (16.82) (5.22) 2.27)
1 2 3 4 5 6 7
ˆ ˆ ˆ ˆ ˆ ˆ ˆ2 3 4netd public dww syssize loc loc locδ δ δ δ δ δ δ ε+ + + + + + +
-0.00004*** 0.14*** 0.08*** 0.05*** 0.21*** 0.22*** 0.15***
(1)
(-94.50) (4.65) (4.96) (4.19) (5.53) (4.77) (2.93)
Bootstrapped z in parenthesis. Significance level of 0.01 and 0.05 denoted by
*** and **, respectively.
The estimated variable cost function meets most of the theoretical regularity
conditions (i.e., it is monotonically increasing in desirable output as well as in
variable inputs). The only case in which the desirable theoretical properties
of inputs are not met is in the case of capital, which, in variable cost function
setting, should be negative. The explanation resides in overcapitalization of
water utilities—a phenomenon widely observed for regulated utility firms of
all kinds. Homogeneity in the cost function is imposed by dividing both input
prices and variable costs by price of chemicals. Consistent with the literature
on undesirable outputs, the presence of an undesirable byproduct in a produc-
tion process, in this case nitrogen, implies a higher cost to the utility which
it then abates either to meet regulation6 or more generally to reduce risk to
customers.
6EPA regulates nitrate in drinking
water (measured as nitrogen) at 10
mg/L.

66
As to the exogenous effects, network density has a negative effect on vari-
able costs as expected. Also, larger systems have higher variable costs. Public
utilities have higher variable costs than investor-owned utilities. This makes
sense from the perspective that public firms may have agency and control
problems relative to investor-owned enterprises. Operations that have only
a distribution function have lower variable costs than those that have both
waste water and distribution. All locations have higher variable costs relative
to New England.
Derivation of shadow cost of nitrogen abatement and discussion
1 2
1 2 1 2 1
3 3
2 3 4 5 6 7
ˆ ˆ ˆ ˆˆ ˆln ln ln ln
ˆ exp
ˆ ˆ ˆ ˆ ˆ ˆ2 3 4
w w
y N netd
w wV
public dww syssize loc loc loc
οβ α α β β δ
δ δ δ δ δ δ
⎛ ⎞⎛ ⎞ ⎛ ⎞
+ + + + +⎜ ⎟⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠= ⎜ ⎟
⎜ ⎟+ + + + + +⎝ ⎠
* 3w (2)
Appendix Table 1.1
Summary statistics and definitions
Definition (unit) variable Mean (Variance) Definition (unit) variable Mean (Variance)
Variable cost
(in $) VC
8,479,039
(13,477,167)
System type
(1 = Distribution and waste
water, 0 = Otherwise) dww
0.54
(0.50)
Annual water production
(in millions of gallons) y
14,449
(23,498)
System size
(1 = if population served
greater than 100,000, 0 =
Otherwise) syssize
0.48
(0.50)
Annual salary
(in $) w1
$34,353
($11,538)
Consumer structure
(ratio of residential to total
water delivered) cs
0.57
(0.23)
Nitrogen abatement
(in difference of raw-finished
nitrates in water) (in mg/L) N
0.98
(4.04)
Water system location
(New England) loc1
0.19
(0.40)
Electricity price
(in $ per kilowatt hour) w2
$0.05
($0.01)
(Northeast) loc2
0.21
(0.41)
Chemicals price
in $ per pound) w3
0.2
(0.0)
(South) loc3
0.15
(0.36)
Capital
(residual rate of return) K
$ 145,916,037
(217,806,925)
(Mid-west) loc4
0.21
(0.41)
Network density
(population served/length of
distribution main) netd
1176
(5608)
(West) loc5
0.23
(0.43)
Organizational type
(1 = public, 0 = otherwise) public
0.87
(0.34)
Source: USDA, Economic Research Service using data from 1996 American Water Works Association survey.

67
2
ˆ ˆˆ *N
V
SC V
N N
α∂
= =
∂
(3)
The shadow marginal cost of nitrogen abatement is derived in equation (3) by
taking the derivative of (2), estimated variable cost, which in turn was
derived by taking the exponential of (1). From equation (3), various addi-
tional derivations can be made: shadow marginal cost by millions gallons,
ˆV
N
∂
∂
/ y, estimated shadow total variable cost of nitrogen abatement (SVC),
ˆV N
N
∂ ×
∂
, and SVC per millions of gallons of water produced ( ˆV N
N
∂ ×
∂
)
/ y.
The results from the above derivations were used to estimate nitrogen
removal costs by system size (app. table 1.2).
Appendix Table 1.2
National estimates of nitrogen removal costs for community water systems, by system size
System size (SS)
[in millions of gallons per year]
(CWS population
in parenthesis)
Estimated average
production by CWS
(millions of gallons
per year)
Estimated average cost of
nitrogen removal
(variable cost per million
gallons per year per CWS)
Estimated total cost of
nitrogen abatement
(million $ per year for all
systems)
SS > 0 and SS <= 3,300
(42,624)
570
$34.2
[46 %]1
830
SS > 3,300 and SS <= 10,000
(4,871)
4,797
$25.55
[41 %]1
597
SS > 10,000 (4,156) 42,485
$19.18
[31 %]1
3,386
CWS = Community Water System.
1Percent of cost attributable.
Source: USDA, Economic Research Service using data from 1996 American Water Works Association survey.

68
Appendix 2
Using NLEAP To Model Nitrogen Losses
The Nitrogen Loss and Environmental Assessment Package (NLEAP)
(Delgado et al., 2010a; Shaffer et al., 2010) can be used to assess the potential
for management practices to increase nitrogen use efficiency and generate
nitrogen savings that can be traded in water and air quality markets (Delgado
et al., 2008b; 2010a). The NLEAP model has been used extensively across
national and international systems (Delgado et al., 2008b).
This tool is capable of simulating the effects of management practices and
generating reasonable assessment values that are similar to measured field
studies conducted across small-scale plots and large commercial field opera-
tions (e.g., water budgets, nitrate leaching, residual soil nitrate, crop uptake,
nitrogen dynamics, and N2O emissions; Beckie et al., 1995; Khakural and
Robert 1993; 2001; Delgado et al., 2001; Xu et al., 1998).
Detailed descriptions of NLEAP-GIS capabilities and limitations can be
found in Shaffer and Delgado, 2001; Shaffer et al., 2010; Delgado and
Shaffer, 2008; and Delgado et al., 2010a; 2010b. This improved version can
quickly evaluate multiple long-term scenarios across a large number of soils
and conduct assessments of the effects of BMPs on nitrogen use efficiency
and nitrogen losses via different pathways. The new NLEAP-GIS tool also
has a Nitrogen Trading Tool option (with GIS capabilities) (Delgado et al.,
2008a; 2008b; 2010a; 2010b).
General assumptions
NLEAP has been tested, calibrated, and used to accurately evaluate the
effects of management for cropping systems and risky landscape combina-
tions across national and international agroecosystems. In order to evaluate
these systems, users established basic assumptions to simplify the evalua-
tion process, which is very complex due to the nature of the nitrogen cycle
and management interactions with environmental factors (Shaffer and
Delgado, 2001).
Yields: It is well known that yield variability can impact nitrogen use effi-
ciency (Bock and Hergert, 1991). Instead of using the maximum yields at a
given site as traditionally done by farmers as a safety net approach to calcu-
lating nitrogen inputs (Bock and Herget, 1991), State average yields for corn
and soybeans derived from the USDA Census of Agriculture were used for
the NLEAP-GIS simulations.
We assumed that yields for no-till systems were 10 percent lower than those
for conventional tillage. Since we also evaluated excessive nitrogen input
scenarios and low nitrogen input (deficit) scenarios, we used the corn yield
and nitrogen input response curve from Bock and Herget (1991) to estimate
the average yields for these scenarios. It was assumed that for the excessive
nitrogen input rates, yields were increased by only 1 percent; however, for
the deficit nitrogen input scenario a 10-percent drop in average yield was
assumed (Bock and Herget, 1991). We believe that our approach of using
average yields to evaluate the effects of management on the nitrogen use

69
efficiency of commercial systems is a valid approach, as reported by Shaffer
and Delgado (2001), Delgado (2001), and Delgado et al. (2000; 2001).
Since the USDA Census of Agriculture does not report yields by soil type,
we assumed that yields for the soil types tested were similar. However, corn
yield can vary among soil type, with lower yields in the sandier, less fertile
soils that have higher nitrate leaching potential than those finer soils with
lower leaching potential (Khosla et al., 2002; Bausch and Delgado, 2003;
Delgado and Bausch, 2005; Delgado et al., 2005). Nonetheless, we still
believe that assuming average yields for a 24-year period being evaluated is
a valid approach to assessing the trends and effects of management practices
on these different soil types and produces results that are in agreement with
average measured values (Delgado et al., 2001; 2008b; 2010a). If additional
site-specific field information for a given farm is needed, spatial soil maps for
the given farm can be downloaded from USDA NRCS websites, and evalua-
tions using farmers’ inputs can be conducted.
Nitrogen Inputs and Uptake: For nitrogen rates, we used the recommended
best management practices for site-specific State and/or soil as described
by Espinoza and Ross (2008) for Arkansas; Alley et al. (2009) for Virginia;
Beegle and Durst (2003) for Pennsylvania; and Vitosh et al. (1995) for Ohio.
We calculated the recommended nitrogen (N) rate per bushel of corn derived
from each State’s recommended BMPs (Espinoza and Ross, 2008; Alley et
al., 2009; Beegle and Durst, 2003; Vitosh et al., 1995). A summary of the
nitrogen inputs simulated is presented in appendix table 2.1.
Since nitrogen fertilizer inputs were calculated based on yield, the no-till
systems received lower nitrogen fertilizer inputs than the conventional
systems. However, since a similar rate of uptake per unit of bushel was used
for both systems, the removal of nitrogen in harvested grain from the no-till
system was also lower than the removal of nitrogen in the grain from the
higher yield conventional system. Total nitrogen uptake by the plant was
calculated. Initial surface residue cover was simulated at 100, 90, 40, and 30
percent for no-till corn-corn, no-till corn-soybeans, conventional corn-corn,
and conventional corn-soybeans, respectively.
For the manure system, manure was applied every 2 years. For the corn-
corn rotation, manure was applied in the first year, and only fertilizer was
applied in the second year. The manure rate was calculated for each system
to match the fertilizer rate. However, since manures will have a large frac-
tion of organic nitrogen that is not immediately available (Davis et al., 2002;
Eghball et al., 2002), an additional 50 percent of the recommended rate was
added as inorganic nitrogen fertilizer. In other words, the total nitrogen input
during the first year of corn-corn rotation was 150 percent of the total appli-
cation rate of the inorganic nitrogen fertilizer scenario (app. table 2.1). The
corn-corn rotation did not receive any manure application in the second year,
and the corn received the same rate of nitrogen fertilizer as in the nitrogen-
fertilizer-only scenario. Thus, over the 2-year period, the manure scenario
for corn-corn received an average of 25 percent more nitrogen input per year.
The same relationships apply to the excessive and deficit nitrogen scenarios
(app. table 2.1).

70
For the corn-soybean rotation, there was no application of nitrogen fertil-
izer or manure for any of the scenarios during the soybean year (app. table
2.1). Additionally, for this rotation, the nitrogen cycling from the leguminous
soybean crop was credited, as is recommended for each State, so the calcu-
lated nitrogen inputs for the corn in the corn-soybean rotation was lower than
in the corn-corn system.
The excessive nitrogen fertilizer scenarios received 75 percent higher nitrogen
inputs than the State-recommended rate. For the deﬁcit nitrogen application
scenarios, nitrogen inputs were applied at a 25-percent lower nitrogen rate
than the best management practice scenario (app. table 2.1).
Soil Type Physical and Chemical Information: For each State, the county’s
soil chemical and physical information averages for the selected soils were
downloaded. To evaluate all of the management scenarios described above,
we selected a soil with a higher leaching potential (Hydrology A or B) and a
soil with a lower leaching potential (Hydrology C or D).
Long-Term Weather: Long-term USDA, Natural Resources Conservation
Service weather databases for each county were used to conduct the 24-year
assessment as described by Delgado et al. (2008b, 2010a) nitrogen trading
tool evaluations.
Other Best Management Practices Tested: For all the scenarios described
above, we evaluated the method of application. The best management prac-
tice for method of application was incorporation of nitrogen fertilizer and/
or manure. Surface application without incorporation was found to be a
poor management practice. We also evaluated time of application. The best
management practice for time of application was application of manure and/
or nitrogen fertilizer before planting, closer to the time of higher demand by
Appendix Table 2.1
Relationships used to develop yields and nitrogen (N) rates used across the study sites
Tillage Best management practice Excessive Deﬁciency
Yield (bushels per acre)
Conventional x1 x*1.01 x*0.9
No-till x*0.9 x*0.9*1.01 x*0.81
N rate for fertilizer-only scenarios (lbs N per acre)
Conventional x2 z*1.75 z*0.75
No-till y3 y*1.75 y*0.75
N rate for manure with N fertilizer scenarios (lbs N per acre)
Conventional z(org) + 0.5z(fert) 1.75z(org) + z(0.875) 0.75z(org) + z(0.375)
No-till y(org) + 0.5y(fert) 1.75y(org) + y(0.875) 0.75z(org) + z(0.375)
1The x values were 131, 101, 103, and 107 corn bushels per acre for OH, VA, PA, and AR, respectively. The x values were 40, 27, 37 and 27
soybean bushels per acre for OH, VA, PA, and AR, respectively.
2The z values were 132, 121, 100, 120, and 125 lbs of N per acre for OH, VA, PA, AR (Hydrology A) and AR (Hydrology D), respectively, for
conventional tillage.
3The y values were 116, 109, 90, 100, and 105 lbs of N per acre for OH, VA, PA, AR (Hydrology A) and AR (Hydrology D), respectively, for
conventional tillage.
Source: USDA, Economic Research Service using data from USDA, Agricultural Research Service.

71
the crop. The poor management scenario was application of manure and/or
fertilizer the previous fall, when the nitrogen is more susceptible to losses.
Long-Term Evaluations: All these scenarios were evaluated over the long
term. To conduct the long-term evaluations, we used a 24-year period using
long-term weather data for the given county. Similar to what was done with
the nitrogen trading tool, the ﬁrst 12 years were used to run the model, and
years 13 to 24 were used to evaluate the effect of management practices on
nitrogen use efﬁciency and on reactive losses to the environment (Delgado et
al., 2008b, 2010a).

72
Appendix 3
Estimating Changes in Nitrogen Fertilizer
Application Rate
This appendix describes the econometric model used to estimate changes
in nitrogen (N) fertilizer application rate. We estimate nitrogen application
rates using an instrumental variables (IV) approach to overcome identifica-
tion issues presented by farmer heterogeneity and endogenous soil N-testing.
Price plays an important role in the nitrogen management decision, and
the recent price growth of nitrogen has implications for nitrogen manage-
ment behavior and by extension, nitrogen use efficiency (NUE). Notably,
we instrument for nitrogen price using a cross-section of data by exploiting
exogenous spatial variation between domestic ammonia production plants
and cornfield locations.
Research using observational data presents econometric challenges, and this
is particularly true for research examining the effect of potentially endoge-
nous variables on a study population. For example, when estimating the effect
of N-soil tests on application rate, researchers do not know why two observa-
tionally identical farmers make different choices about testing the soil. The
underlying problem is the concern that unobserved farmer characteristics are
responsible for determining whether the farmer conducts a test. For example,
a farmer who tests the soil regularly may also have unobserved preferences
for land stewardship. If differences beyond observed field, farm operation,
and operator characteristics play a role in determining who conducts the
test and how the test is used, then the test may be endogenous to the amount
applied.
Nitrogen price also presents a challenge in a sample of microdata. Prices are
likely to embody an error-in-variables problem because in the case of ARMS,
they were created as a share variable that represents the nitrogen fertilizer’s
relative size of the total expenditures for all fertilizer (nitrogen, phosphorus,
and potassium). To see how this effects the estimation of nitrogen demand,
consider that we observe nitrogen price as a function of the true, unobserved
price plus a disturbance term, v.
(1) N NPrice Observed Price True* v= + .
Because the observed price on the left-hand side of equation (1) is a func-
tion of true price and v, an ordinary least squares (OLS) model of nitrogen
demand estimated with the observed price will include v and will cause the
estimate to be biased and inconsistent. Specifically, in the classic errors-in-
variables example, the coefficient in an OLS model will be biased toward
zero.7 Prices farmers pay may also change with their level of demand. For
example, if farmers receive quantity discounts when purchasing nitrogen
fertilizer and their application rate is correlated with total nitrogen demand,
then failing to account for this also results in bias.
To overcome the problem of mismeasured nitrogen prices and endogenous
soil testing, we employ an IV approach, which allows for the development
of consistent and unbiased estimates. In the case of endogenous N-soil
testing, we find a set of instruments that are correlated with N-soil testing
7See Greene (2000) for a formal
discussion of measurement error and
the resulting attenuation bias.

73
but uncorrelated with disturbance process: average annual soil percolation
and average annual precipitation. Because percolation facilitates nutrient
leaching (Williams and Kissell, 1991), we expect the greater soil percolation
to increase uncertainty about available nutrients, and, therefore, encourage
soil testing. Higher precipitation generally reduces the ability to conduct soil
test, therefore we expect annual average precipitation to be negatively related
to N-soil test.
We identify the nitrogen own-price effect on demand using three sources
of exogenous variation: distance between the field and domestic ammonia
fertilizer production; production capacity of nearby ammonia plants; and
distance from the field to New Orleans, LA, site of the majority of interna-
tional ammonia importation.8 Ammonia is increasingly being imported by
the United States, and a majority of shipments enter from the Gulf of Mexico,
and specifically, New Orleans; therefore, we also include a distance-to-New
Orleans measure. These variables are useful instruments because the distance
between the field and production capacity are arguably uncorrelated with the
behavior of the farmer or the placement of the field;9 therefore, the instru-
ments allow one to capture the exogenous variation in price and use it to esti-
mate application rates.
Instrumental variables model
We use an IV model specified with two endogenous variables to estimate a
partial-equilibrium static demand model derived from profit maximization
theory. The model assumes producers make immediate adjustments to quan-
tity demanded in response to changes in price, and that prices are known at
the time of production planning. These assumptions are reasonable given the
ability of farmers to enter into contracts that establish price for delivered corn
and inputs to production, such as forward or marketing contracts, and other
hedging instruments. Further, production technology is assumed known and
fixed. Since only two time periods separated by 4 years are used, technology
is unlikely to change. The most likely technological change is that of seed
technology—the use of biotech (Bt) corn; however, the model specification
controls for this. In 2001, 20 percent of corn acres were planted with Bt corn;
in 2005, the amount was slightly greater than 30 percent.
We characterize the problem posed to the farmer as one of profit maximiza-
tion with uncertainty, as evidenced by the nitrogen overtreatment, but the
decision of the farmer could also be conceptualized as a utility maximization
problem. In this case, the farmer chooses a level of output that maximizes the
farmer’s initial wealth plus expected profit from the operation. Under utility
maximization, a farmer considers not only expected profit but moments of
the profit distribution as well, and deviations from the recommended level of
nitrogen then depend on the farmer’s level of risk aversion. Evidence from
field trial suggests that risk-neutral farmers would be willing to overapply
nitrogen to increase profits during a year of “good” growing conditions
(Rajsic et al., 2009). On the other hand, risk-averse farmers will reduce their
nitrogen rate to reduce profit variance. In practice, our empirical results are
not dependent on the conceptual framework; in both cases, nitrogen prices
enter the profit function, and the identification strategy would not change.
Rather, the level of risk aversion primarily drives the differences. Some
8Ammonia production data come
from the North American Fertilizer
Capacity Annual Reports issued by the
International Fertilizer Development
Center. We calculate the distances from
the field to ammonia production using
the location of the plant and geocoded
corn field samples from USDA’s
2001 and 2005 Agricultural Resource
Management Survey. It should be
noted that these are sample points, and
they do not represent all corn produc-
tion in the United States; however,
when we estimate a model of nitrogen
demand, the sample points are weighted
to reflect total U.S. corn production.
9To test that the instruments are
uncorrelated with the residual compo-
nent in the second stage of the IV
model, or exogenous to the rate of
fertilizer application, we test overidenti-
fication restrictions using a Sargan test.
The test statistic is computed as n×R2
and has a χ2(k-r) distribution, where k is
the number of instruments and r is the
number of endogenous variables. The
results of the test are presented in the
results table.

74
research, however, suggests that risk-averse farmers are more responsive to
price because of profit risk (Just, 1975; Roosen and Hennessy 2003; Rajsic et
al., 2009), and, if farmers are on average risk averse, our elasticity estimates
will represent an upper bound.
Equation (2) is the outcome equation where Y represents the log transformed
per acre rate of nitrogen applied to the field of farm i in USDA production
region r at time t. Endogenous variables, Tˆ and Pˆ , are estimated N-soil
testing probability and nitrogen price from equations (3) and (4). The set of
excluded instruments for N-soil test are represented by ZT, and the excluded
instruments used to estimate nitrogen price are represented by ZP. The vector
X is a set of independent variables that includes characteristics of the oper-
ator, farm operation, and the field; the disturbance term is represented by ε.
irttrirtirtirtirt PTY ευφδλβα ++++++= 111111
ˆˆ X ,
irttrirt
T
irtirt ZT κυφδβα +++++= 22222 X ,
irttrirt
P
irtirt uZP +++++= 33333 υφδβα X .
A case can be made that countrywide trends over time affect the use of
nitrogen. Perhaps in response to outreach efforts to reduce fertilizer runoff
due to overuse, for example, environmental awareness campaigns that
communicate the benefits of reduced nitrogen in the environment, attitudes
about nitrogen rates have changed. We control for trends in nitrogen use that
change over time with a time effect term, υt. As well, use of nitrogen across
production USDA-defined regions may also affect application rates, therefore
we control for region-specific factors with a fixed-effect term, φr.
Data
The data are cross-sectional and come from USDA’s Agricultural Resource
Management Survey (ARMS). ARMS comprises responses to a series of inter-
views with farm operators designed to solicit information about production
practices, costs of production, business finances, and operator and household
characteristics. Commodity specific surveys are fielded on a rotating basis,
usually every 5 to 8 years. We focus on corn production because of its intense
use of nitrogen, for which ARMS last fielded surveys in 2001 and 2005.
We use data from two components of ARMS. The first component is the
Corn Production Practices and Costs Report, which surveys the farm enter-
prise’s costs of production and a host of production practices at the field level.
The second component is the Corn Costs and Returns Report, which collects
indepth financial information concerning the farm business and the house-
hold of the operator. The two components can be linked together to provide
a complete view of the farm operation from the farm’s representative field to
its financial statement, and we restrict the sample to farmers who completed
both surveys.
As covariates, we include the farmer’s age, education, and income earned
from work off the farm. We account for land quality and tenancy issues by
including the per acre annual value of production, the per acre value of the

75
land, and acres owned by the operator. We also control for environmental
characteristics of the field, for example, whether any part of the field is a clas-
sified as a wetland. The presence of livestock and a nutrient management plan
on the farm may indicate a greater reliance on manure, driven often by the
need to dispose of manure. We account for these with dummy variables as
well. The nutrient requirements of a current corn crop are also based, in part,
on the plant-available nutrients existing in the soil, and past cropping practice
can influence these nutrients. Therefore, we use a dummy variable to control
for crop rotation pattern of 3-year straight corn rotation.
The timing and method of application may also be important determinants
of application rate. A spring application is better timed to meet the plant’s
need for nutrients and reduces the risk of loss due to environmental factors
relative to a fall or winter application. On the other hand, farmers may opt to
apply nitrogen in the fall, when there are fewer time demands and prices are
often lower. In such a case, a nitrogen inhibitor is often used to further slow
the nitrification process, though average annual nitrate losses can still be 50
percent higher under fall application than under spring application (Randall
and Mulla, 2001). To counter this, in many cases, anhydrous ammonia is
injected into the soil because low temperatures at this time of year slow
the conversion of ammonia to ammonium and nitrate, reducing the loss of
nitrogen. We control for the method of application with a dummy variable
indicating whether the nutrient was incorporated or injected into the soil.
Technology and other management practices thought to affect nitrogen rate
are captured by explanatory variables indicating the use of field irrigation
and biotech (Bt) corn seed. Irrigation is an important component in nitrogen
management. Irrigation may be a necessary practice due to the climate, or
it may be another way of more precisely controlling growing conditions.
If water and nitrogen are complementary inputs, the presence of irrigation
should increase the rate of nitrogen application. The use of biotech seed is
driven by the associated cost reductions from the technology’s herbicide, pest,
or fungus resistance. We also include a dummy variable representing whether
the corn crop was grown for silage or corn. A full list of covariates and
summary statistics is presented in appendix table 3.1.
Outcome Measures
We estimate the application rate for four different permutations of nitrogen
fertilizer use. First, we estimate commercial nitrogen use by farmers who
exclusively apply commercial nitrogen—a group that accounts for a 78
percent of the farmers in our sample. We also examine the rate of total
commercial nitrogen use by all farmers, regardless of whether they used
commercial nitrogen exclusively or in conjunction with manure. The third
measure examines the sensitivity of commercial nitrogen use by farmers
who use manure in conjunction with commercial nitrogen—a group that
employs an imperfect substitute for commercial nitrogen. These farmers
make up a minority of the sample, 22 percent. Finally, we examine the effect
of our explanatory variables on total nitrogen application rate, which includes
commercial nitrogen and manure. It should be noted that all of the farmers
in the sample reported at least some use of commercial nitrogen fertilizer.
Estimates from the IV model are presented in appendix table 3.2.

76
Appendix Table 3.1
Summary statistics
Variable name Description Mean
95%
confidence Interval
Soiltestn Nitrogen soil test 0.21 0.18 0.24
Nprice Nitrogen price 0.328 0.324 .332
Dealerrec Dealer recommendation 0.32 0.29 0.35
Consultrec Consultant recommendation 0.14 0.12 0.16
Extrec Extension agent recommendation 0.04 0.02 0.05
Routine Routine practice 0.28 0.26 0.30
op_age Operator’s age 52.73 52.11 53.36
Retired Operator is retired from farming 0.04 0.03 0.06
College Operator holds college degree 0.35 0.31 0.37
Workoff Derive income from off-farm work 0.38 0.35 0.42
Anycropins Insurance participation rate 0.659 0.62 0.70
Prodvalpa Production value per acre $4, 372.57 $337.29
Landvalpa Land value per acre $1,616.55 $709.46 $2,523.64
Ownacre Acres owned 323.37 301.10 345.63
Corn_p Corn price 1.87 1.84 1.90
CCC Straight corn rotation (3 years) 0.25 0.21 0.28
Nutrient plan Nutrient plan in place 0.076 0.063 0.088
Irrigate Irrigate the field 0.063 0.0397 0.0853
Wetland Wetland on any part of the field 0.03 0.02 0.04
Tenure Years farming 27.61 26.89 28.33
Spring Spring fertilizer application 0.80 0.77 0.84
Inc Incorporated fertilizer 0.75 0.73 0.78
Inhibit Fertilizer applied with inhibitor 0.07 0.05 0.09
Bt_corn Biotech corn 0.34 0.30 0.38
Yldgoal Yield goal 173.62 166.31 180.94
Silage Corn for silage 0.11 0.09 0.13
Livestock Presence of livestock on the farm 0.576 0.55 0.602
Commercial nitrogen w/o manure Commercial nitrogen users only 129.72 125.67 133.77
Total commercial nitrogen Total commercial nitrogen use 118.42 114.42 122.42
Commercial nitrogen w/ manure Commercial nitrogen use by manure users 77.23 70.60 83.87
Total commercial nitrogen and manure use 137.59 132.16 143.02
Total nitrogen observations 2,874
Source: USDA, Economic Research Service using data from USDA’s 2001 and 2005 Agricultural Resource Management Survey.

77
Appendix Table 3.2
IV estimates of nitrogen application rate
Commercial
nitrogen:
nonmanure users S.E.
Total
commercial
nitrogen S.E.
Commercial
nitrogen: only
manure users S.E.
Total nitrogen
(manure and
nonmanure users) S.E.
Soiltestn -0.924** 0.290 -1.142** 0.336 0.333 0.742 -1.080** 0.308
Lognprice -1.347 0.715 -1.379* 0.630 0.531 1.408 -0.674 0.589
Dealerrec 0.131** 0.043 0.159** 0.047 0.155 0.099 0.157** 0.042
Consultrec 0.229** 0.078 0.291** 0.083 -0.004 0.171 0.303** 0.078
Extrec 0.084 0.086 0.143 0.084 0.239 0.156 0.163* 0.073
Routine -0.170** 0.065 -0.164** 0.063 -0.071 0.100 -0.136** 0.057
Op_age -0.011** 0.003 -0.008** 0.003 0.005 0.007 -0.008** 0.002
Retired 0.104 0.098 0.107 0.100 0.171 0.211 0.002 0.089
College 0.043 0.034 0.055 0.037 0.118 0.117 0.025 0.034
Workoff -0.091** 0.037 -0.0810* 0.0398 -0.124 0.094 -0.115** 0.037
Anycropins 0.061 0.054 0.1065* 0.0498 0.035 0.083 0.1203** 0.0468
Prodvalpa -7.84E-06 3.32E-05 -3.09E-05 2.50E-05 -4.38E-05 3.25E-05 4.64E-05** 1.98E-05
Landvalpa -4.93E-07 4.41E-07 -8.96E-07 7.65E-07 -5.21E-05 2.81E-05 -1.57E-06 8.10E-07
Ownacre 3.62E-05** 1.37E-05 3.92E-05** 1.48E-05 -6.32E-05 5.88E-05 2.95E-05 1.51E-05
logcorn_p 0.006 0.043 0.029 0.048 -0.032 0.112 0.034 0.045
Ccc 0.0315 0.055 0.092 0.052 0.192** 0.088 0.082 0.051
Wetland -0.081 0.118 -0.065 0.110 -0.187 0.326 -0.013 0.098
Nutrplan 0.167** 0.070 0.023 0.077 -0.280** 0.133 0.172** 0.072
Irrigate 0.527** 0.085 0.532** 0.089 -0.370 0.364 0.551** 0.084
Tenure 0.006** 0.002 0.005** 0.002 0.005 0.007 0.004 0.002
Spring 0.028** 0.041 0.013 0.048 -0.083 0.150 0.026 0.042
Inc 0.063 0.050 0.061 0.049 0.052 0.101 0.053 0.046
Inhibit 0.083 0.057 0.2239** 0.0590 0.556** 0.120 0.176** 0.056
Bt_corn 0.042 0.036 0.067 0.040 0.081 0.100 0.062 0.036
Yldgoal 0.001** 0.0002 0.0003 0.0002 -5.27E-06 2.23E-04 0.0001 0.0002
Silage -0.404** 0.093 -0.350** 0.078 -0.060 0.094 -0.098 0.076
Live -0.154** 0.046 -0.233** 0.051 -0.265 0.186 -0.142** 0.047
Observations 2253 2874 624 2874
F-Statistic 6.69
[<0.000]
11.87
[<0.000]
6.31
[<0.000]
7.82
[<0.000]
Source: USDA, Economic Research Service using data from USDA’s 2001 and 2005 Agricultural Resource Management Survey.

78
Appendix 4
Comparing Costs of Farms Using Different
Nutrient Management Practices
The goal of this analysis was to estimate the variable production costs for
farms using different nutrient management strategies. The results are used to
estimate the cost of changing from a less-efficient to a more-efficient nutrient
management strategy. We restricted our analysis to corn, given the large
acreage and its intensive use of nitrogen.
Data on corn are from USDA’s 2001 Agricultural Resource Management
Survey (ARMS). This is the last corn survey from which field-level cost of
production data are estimated for each observation. SAS General Linear
Model procedure (GLM) was used to estimate a model of variable production
costs as a function of management and resource-base variables. Least squares
means were used to compare the per acre variable production costs between
practices directly related to nitrogen management.
Total variable costs (TVC) were defined as the costs of seed, fertilizer,
manure, pesticides, custom work, and fuel lubricants. We specified a model of
TVC as a function of the following variables:
(1) Use of biotech or herbicide resistant corn
(2) Use of rotation with soybeans
(3) Use of nitrogen inhibitor
(4) Tillage (conventional till vs. reduced/no till)
(5) Timing (fall vs. spring application)
(6) Method (broadcast vs. inject/incorporate)
(7) Conservation cropping (contour or strip)
(8) Presence of nutrient management plan
(9) Use of variable rate technology
(10) Presence of irrigation
(11) Presence of highly erodible soils (yes or no)
(12) Presence of tile drains
(13) Growing season (northern tier, middle tier, southern tier)
(14) Farm size (total corn acres on farm)
(15) Yield goal
An interaction term for timing and method (fall/no fall – incorporate/
broadcast) was also included. The cost model was run separately for those
farms that do not use manure and for those farms that use both manure and
commercial fertilizer. About 16 percent of U.S. corn acres receive manure.
Since most of the variables are class variables, we used the SAS General
Linear Model procedure (GLM) to estimate the model. The R-Squares of the
no-manure and manure-cost models are 0.21 and 0.16, respectively, and the
models are significant at the 1-percent level. The majority of the explanatory

79
variables are statistically significant at the 5-percent level. Least-square
means of the production costs ($/acre) under the different management
systems are presented in app. table 4.1, along with an indication of whether
the difference is statistically significant. Of interest to this study is that the
cost under the preferred method/timing combination (spring/incorporate)
is significantly different from the costs under the less-preferred, alternative
combinations (at the 5- and 10-percent levels) for those farms that use only
commercial fertilizer (84 percent of treated corn acres). No significant differ-
ences in costs were found for farms that use both manure and commercial
fertilizer.
Part of the difference in costs observed with ARMS data is due to differ-
ences in chemical application rates. Since the NLEAP scenarios assumed the
management changes were independent, altering rate, timing, and method in
different combinations, we needed to separate out the nitrogen fertilizer cost
from the total of changing management. We ran the same models, but with
nitrogen application rate as the dependent variable. Both of the models were
Appendix Table 4.1
Variable cost per acre of management practices
Commercial
nitrogen only
Commercial
nitrogen and manure
Dollars per acre Pr > t Dollars per acre Pr > t
Management choice
Continuous corn
Rotation with soybeans
131.23
124.02
.0001 165.63
158.06
.1330
Conventional tillage
Reduce/no-till
128.79
126.46
.1554 128.79
126.46
.6671
Fall/broadcast
Fall/incorporate
Spring/broadcast
Spring/incorporate
127.84
128.39
132.54
121.74
.0582
.0557
.0001
158.89
155.25
158.53
174.70
.1587
.1867
.2078
No irrigate
Irrigate
133.33
121.92
.0009 164.11
159.58
.7292
No highly erodible soil
Highly erodible soil
124.92
130.34
.0088 157.98
165.71
.2013
No nitrogen inhibitor
Nitrogen inhibitor
125.34
129.92
.0832 153.80
169.88
.0441
No conservation cropping
Conservation cropping
131.83
123.42
.0004 163.25
160.44
.6278
No nutrient plan
Nutrient plan
127.87
127.39
.8475 157.63
166.06
.1461
No variable rate
technology
Variable rate technology
125.45
129.81
.1522 155.39
168.30
.2945
No tiles
Tiles
128.79
126.46
.2095 168.02
155.67
.0366
Source: USDA, ERS using USDA’s 2001 Agricultural Resource Management Survey.

80
signiﬁcant, with R-squares of 0.23 and 0.24. Differences in application rates
between the spring/inject and the other management combinations were posi-
tive (as expected) and signiﬁcant at the 1-percent level for farms using only
commercial fertilizer (app. table 4.2). The difference in nitrogen fertilizer
costs was subtracted from the cost difference derived from the cost model,
using a nitrogen fertilizer price of $0.30/lb. The cost of adopting appropriate
method (assuming no change in fertilizer application rate) was estimated to
be $7.35/acre, appropriate timing was $3.01 per acre, and both appropriate
method and timing were $1.86/acre. For farms using manure, we assumed no
differences in costs.
Appendix Table 4.2
Nitrogen application rates per acre by management practice
Commercial
nitrogen only
Commercial
nitrogen and manure
Pounds
per acre Pr > t
Pounds
per acre Pr > t
Management choice
Continuous corn
Rotation with soybeans
136
140
.1544 218
192
.0420
Conventional tillage
Reduce/no-till
137
139
.3583 202
208
.6433
Fall/broadcast
Fall/incorporate
Spring/broadcast
Spring/incorporate
143
141
140
129
.0001
.0042
.0001
191
201
220
208
.0420
.6433
.8382
No irrigate
Irrigate
129
147
.0002 210
200
.7692
No highly erodible soil
Highly erodible soil
139
137
.5955 222
188
.0353
No nitrogen inhibitor
Nitrogen inhibitor
135
141
.0017 189
221
.1354
No conservation cropping
Conservation cropping
143
133
.0004 175
235
.0001
No nutrient plan
Nutrient plan
137
139
.6002 197
213
.2950
No variable rate technology
Variable rate technology
138
138
.9899 224
186
.3175
No tiles
Tiles
139
137
.4957 214
196
.2384
Note: Parameter estimates from GLM model.
Source: USDA, Economic Research Service using data from USDA’s 2001 Agricultural
Resource Management Survey.

81
Appendix 5
Estimating Wetland Restoration Costs
The cost of restoring a wetland is the sum of the cost of the land and the cost
of restoring the land’s water-hold capability and the wetland ecosystem. We
generate wetland and restoration costs using cost functions that we estimated
using available wetland cost data. Sample observations lie in the Glaciated
Interior Plains (GIP).
The cost of the land to society is the difference in its value with and without
the wetland. The value of agricultural land without a wetland is assumed to
be a function of the net value of its output, but the potential for nonagricul-
tural use can play a role.
The USDA Wetland Reserve Program (WRP) sets wetland easement prices
equal to the difference in land values with and without a permanent wetland
easement. Therefore, WRP easement payments are well suited as a measure
of land cost.
Land cost is modeled as a function of the agricultural value and value
squared of the land in the contract (AgrValue and AgrValuesq), contract size
and size squared (Acres and Acressq), the potential for urban development
(Urban), and farm size (Fsize). Because a measure of the agricultural value of
the land is not available, we use the product of the county-average farmland
rental rate (Rent) and contract acreage as a proxy (it represents the annual
agricultural value of the land).
The adjusted R-square of the estimated land cost model indicates that the
estimated ordinary least squares model explains 90 percent of the variation in
WRP land costs. Variables are statistically signiﬁcant and have the expected
sign. With this cost function, we generate marginal and average land cost
estimates by county throughout the GIP. To generate average cost, we divide
total land cost (generated with our model) by the size of the contract—all cost
estimates are based on the median-size WRP easement. Across the counties
of the GIP, average per acre costs range from $1,490 to $3,030.
Second, we generate the marginal cost function (MCL) by differentiating the
estimated land-cost model with respect to Acres:
MCL = 925 + 4.32*Rent + 2.39(10-6)*AgrValue*Rent - 0.127*Acres.
For average-sized contracts, county-level estimates of MCL in GIP range
from $985 to $1,790 per acre with a median cost of $1,390.
Restoration costs are modeled as a function of the agricultural value of the
land, the size of the contract, and other variables. The agricultural value
is included as an explanatory variable because we believe that landowners
would spend more to drain more productive lands and assume that restoration
costs are positively correlated with drainage costs.
Approximately 15 percent of the WRP contracts of the GIP report zero resto-
ration costs. Because the dependent variable is truncated, we use the Tobit

82
procedure to estimate the restoration cost function. The Tobit procedure
simultaneously estimates the probability that the dependent variable is non-
zero and its expected size. Variables of the estimated model are statistically
signiﬁcant and have the expected sign.
The estimated model is used to generate expected restoration costs. By
dividing our model’s county-level expected cost estimates by contract size,
we generate estimates of expected average restoration costs. Costs range from
$506 to $602 per acre across counties.
Differentiating the estimated Tobit model with respect to the contract acres
generates the expected marginal restoration cost function (MCR):
MCR = (Z)*(0.888*Rent -2.12* AgrValue*Rent + 167)
where (Z) is the cumulative probability function and Z is the estimated Tobit
equation. For average-sized contracts, estimates of MCR across counties of
the GIP range from $101 to $210 per acre.

Nitrogen in environment

More Related Content

What's hot (20)

Similar to Nitrogen in environment (20)

More from Andrew Hutabarat (20)

Recently uploaded (20)

Nitrogen in environment