Summary of Life Cycle Inventory Data for Cotton

(Field to Bale – version 1.1 – 2 July 2009)

Table of Contents
Abbreviations Used in this Report .......................................................................................................................... 2

Introduction ............................................................................................................................................................ 3

Land Use .................................................................................................................................................................. 5

Water Use ............................................................................................................................................................... 8

Energy Use............................................................................................................................................................... 9

Greenhouse Gas (GHG) Emissions ........................................................................................................................ 10

Energy Use and GHG Emission Data Sources ........................................................................................................ 11

Ecosystem Services ............................................................................................................................................... 12

Ongoing improvements......................................................................................................................................... 15

References............................................................................................................................................................. 19

Appendix A. Summary of units of measure and U.S. cotton data sources ........................................................... 22

Appendix B. Data sources used to derive energy use and GHG emissions estimates for U.S. cotton
production ............................................................................................................................................................. 23

Appendix C. Methods used to derive energy use and GHG emissions estimates for U.S. cotton production ..... 25

Land Preparation and Planting ........................................................................................................ 25

Within-Season Management ........................................................................................................... 26

Harvest and Ginning ..................................................................................................................... 26

Soil Nitrous Oxide Emissions .......................................................................................................... 27

Relationship between per area and per mass basis ........................................................................... 27

Additional References for Appendix C .............................................................................................. 31

Page 1 of 31
Abbreviations Used in this Report

Bt Bacillus thuringiensis

Ce Greenhouse gas emission estimated in carbon equivalents (= CO2e divided by 3.667)

CO2e Greenhouse gas emission estimated in carbon dioxide equivalents

GHG Greenhouse gas

GJ 1 Gigajoule= 109 Joules = 1,000 MJ = 947,867 British Thermal Units (BTU)

Ha 1 Hectare = 10,000 square meters = 2.47 acres

LCI Life Cycle Inventory

Mg 1 Megagram = 1,000,000 grams = 1 metric tonne = 2205 U.S. pounds

MTR Energy embodied in Manufacturing, Transportation, and Repair of machinery

USDA United States Department of Agriculture

USEPA United States Environmental Protection Agency

USGS United States Geological Survey

Page 2 of 31
Introduction
This document is a summary of the publically available information and data on land, water and
energy use, greenhouse gas emissions and ecosystem services for cotton production in the United
States. The data are derived from U.S. government sources, United Nations documents and from
publications in established, peer-reviewed scientific journals. This self-evaluation of cotton
production’s impact on the environment conducted by Cotton Incorporated forms the basis of a life
cycle inventory (LCI) of the agricultural production of cotton, a prerequisite for conducting the more
comprehensive life cycle analysis (LCA) of a cotton product.

All data sources and information used in the calculations throughout this paper are provided as
Appendices A, B and C. Appendix A includes a summary of key resource impact metrics in different
measurement units (English and metric), Appendix B contains additional information on the energy
and GHG emissions data sources, and Appendix C provides a detailed explanation of how one set of
the energy and GHG emissions estimates were derived.

Cotton Incorporated continuously reviews and works to improve cotton’s environmental footprint.
By its funding of agricultural research programs throughout the U.S., Cotton Incorporated assists
growers in adopting technologies and practices that conserve natural resources and enhance grower
efficiency. The research programs are highly valued by growers and have contributed significantly to
cotton’s environmental gains throughout the past 40 years. By way of example, based on the findings
outlined in this summary, additional funding was directed toward the development of technologies
and strategies to improve nitrogen and water use efficiencies during the cotton production. The
results of these studies along with any additional data or information that we become aware of will
be incorporated as revisions to this document.

Cotton Incorporated’s summary of life cycle inventory data show:

Page 3 of 31
Land – Cotton is very land efficient, meeting 36% of the world’s textile needs on 2.5% of
agricultural land resources. Cotton is expected to meet future increases of fiber demand on
fixed land resources as a result of continued yield increase (more cotton per unit of land).

Water – In the U.S., 64% of cotton is grown with naturally occurring rainfall. Most of the
remaining acres receive partial or supplemental irrigation during low rainfall periods.
Globally, cotton production accounts for 3% of the world’s agricultural water use, an amount
proportional to cotton’s global agricultural land use of 2.5%. Cotton is a naturally drought and
heat tolerant crop and provides many smallholder growers around the world with a source of
income in places where few other crops can be grown.

Energy – From field to bale, cotton produces more net energy than is required for its
production due to energy stored in the cottonseed.
Greenhouse Gas Emissions – Cotton has a neutral greenhouse gas (GHG) footprint when the
carbon in the fiber and that stored in the soil is accounted for; when credit is given for the
energy stored in the cottonseed, cotton actually captures more GHG emissions than those
emitted.
Ecosystem Services - Cotton fields supply oxygen, provide habitat for wildlife and capture and
store carbon. Cotton producers' are actively involved in numerous programs designed to
protect ecosystems and safeguard wildlife habitat.

Page 4 of 31
Land Use
According to the United Nations Food and Agricultural Organization (FAO, 2005), cotton fields
represent approximately 2.5% of the world’s agricultural land (Figure 1), and on this small bit of land
36% of the world’s textiles are produced (Figure 2). This is possible because of cotton’s consistent
record of yield increase (production per area of land) throughout the last 40 years. In fact, the
amount of land needed to produce 1Mg of fiber (1 Mg = 1 metric ton) is almost half of that required
20 years ago (Figure 3). Additionally, despite steep rises in demand over the last 40 years, cotton’s
global land requirements have essentially remained unchanged (Figure 4).

Although cotton’s global land use has remained fairly stable, cotton yields and the land used for
cotton production can vary depending on the region and the year. Environmental conditions, crop
management practices and genetic potential contribute to this variation in yield. Figure 5 illustrates
cotton yields (production per unit area of land) based on average production levels from 2005 to
2008 (Meyer et al., 2008; USDA, 2008a) for the overall world average, and then average yield levels in
the U.S.

But cotton is more than just a fiber. Cottonseed, a valuable byproduct of fiber production is used in
numerous food and industrial applications. For every kilogram of fiber produced, 1.5 kilograms of
cottonseed are also produced (USDA, 2008a). In terms of land use it is difficult to properly value the
benefits of this important byproduct because of the diversity of cottonseed uses and potential
metrics used to assess their contribution (mass, biofuel potential, farm gate value, processed oil
value, meal feed value, etc.). However, as illustrated in subsequent sections of this report, capturing
the value of cottonseed is more straightforward when cotton production’s energy use and
greenhouse gas (GHG) emissions are evaluated.

Because of cotton’s history of consistent technological advances in productivity (yield), cotton is

expected to continue to meet demand now and in the future, despite the predicted limited
availability of land.

Page 5 of 31
 Cotton
2.5%
Fruit &
Veg.
7%

Other
Crops
Corn
35%
11%
Rice
11%
Other
Cereals
27%
Soybean
7%

Figure 1 - Percentage of global agriculture land devoted to cotton and other crops
(FAO, 2005)*
*totals more than 100% due to rounding

Rayon /
Acetate** Wool Other
4% 2% 1%

Cotton
Synthetics* 36%
(poly, nylon,
etc.)
54%

Figure 2 - World textile demand (70 million Mg total fiber use)

(National Cotton Council, 2009)
*Synthetics from fossil organic compounds – polyester, nylon
**Synthetics from plant organic compounds - rayon

Page 6 of 31
 3.0
U.S.
Land required for production of 1 Mg of Fiber (ha)

World
2.5

2.0

1.5

1.0

Less land is required to produce a unit of fiber

0.5 due to improvements in productivity

0.0
1960 1970 1980 1990 2000 2010
Year

Figure 3 - Land required for the production of 1 Mg of fiber (1 Mg = 1 metric ton)

(Meyer et al., 2008)

40.0
35.0
30.0
25.0
Hectares

20.0
15.0
The world’s land devoted to cotton has changed little
10.0 over the past forty years.
5.0
0.0
1960 1970 1980 1990 2000 2010

Year

Figure 4 - Global land area devoted to cotton production (1960 – 2010)

(Meyer et al., 2008)

Page 7 of 31
 2,500
Seed
Fiber
2,000
Yield (kg / ha)

1,500

1,000

500

0
World U.S.
Region
Figure 5 - Average cotton yields for the world and the U.S. from 2005 to 2008
(Meyer et al., 2008)

Water Use
Based on data from the United States Department of Agriculture’s (USDA) Farm and Ranch Irrigation
Survey (USDA, 2004), an average of 4,270 cubic meters of water per hectare (0.36 hectare-meters)
were applied to irrigated cotton in the U.S. While one might imagine that irrigation levels would vary
widely from year to year, the data indicates that over time the national average use of irrigation in
cotton has been following a steady downward trend. The most recent USDA Agricultural Census data
also indicates that in 2007 approximately 1.4 million hectares, representing only 36% of U.S. crop
acres, were irrigated. Distributing U.S. irrigation water use across total US production (all lint
produced non-irrigated and irrigated), the irrigation water for the U.S. is 1800 cubic meters per
Mg(1 Mg = 1 metric ton) of fiber produced or 725 cubic meters of irrigation water per Mg of fiber
and seed produced. From a global perspective, it is estimated about 50% of the world’s cotton
receives irrigation but the situations vary widely. In one place the crop may be completely
dependent on irrigation, and in another, the crop may require only supplemental irrigation with most
water coming from rainfall (Chapagain et al. 2006). Whether by rainfall or through irrigation, cotton’s
share of the global agricultural water footprint is 3% (Hoekstra and Chapagain, 2007), an amount
proportional to cotton’s global land use footprint of 2.5%. And this water produces not only fiber but
valuable cottonseed as well. It is also important to note that cotton is a drought tolerant plant and

Page 8 of 31
can be productive in hostile environments, providing a livelihood for farmers where few other crops
can be produced. For example, in West Texas, research has shown that because of cotton’s lower
water requirements, water use by agricultural overall would actually decrease if more of the area
currently devoted to the production of other crops would be transitioned to cotton (Gowda et al.,
2007).

Energy Use
There is no doubt that energy resources will be limited in the future, and optimizing cotton’s energy
use will continue to be an important priority. The blue bar of Figure 6 represents the average amount
of energy required to produce cotton from planting through the ginning process as reported in four
different studies. The green bar of Figure 6 represents the gross energy content associated with the
cottonseed produced from the lint. Note that even with 60% energy conversion efficiency, the
energy in the seed equals the amount of energy needed to produce the crop.

35
The embedded energy contained in the seed
exceeds the energy used in producing the
30
fiber. Therefore, the cotton production
process is a net energy provider.
Energy (GJ per Mg)

25

20

15

10

0
Energy Used in Production Energy Contained in the Seed

Page 9 of 31
Figure 6 - Energy use estimates for U.S. cotton production and gross energy content of the seed
associated with the fiber
Source: Energy required is average of four studies – see section entitled Energy Use and GHG Emission Data
Sources in text below and Appendix B. Energy content in the seed based on results of samples analyzed by
Minnesota Valley Testing Laboratories, Inc.

Greenhouse Gas (GHG) Emissions

Estimates of GHG emissions produced during cotton production are shown in Figure 7 (kilograms of
carbon dioxide equivalents emitted per kg of cotton fiber produced). The amount of carbon stored in
the fiber and soil (green bar) exceeds the total GHG emissions that occur while growing and ginning
the crop (blue bar). The green bar represents the 42% carbon content of the cotton fiber plus the
carbon that is stored in the soil (Causarano et al., 2006). If credit were given for the amount of
biodiesel that could be produced from the cottonseed oil and the carbon emissions from petroleum
diesel were replaced with biodiesel (~0.6 kilogram CO2e per kilogram of fiber –see Appendix B for
details), then cotton production from the field to the bale actually has higher stored GHG than
portrayed in Figure 7.

Page 10 of 31
 2.5
In the cotton production process,
more CO2e is sequestered in the fiber
and soil than CO2e GHG 's emitted into
kg CO2e per kg of Fiber
2.0
the atmosphere.

1.5

1.0

0.5

0.0
GHG Emitted During Production GHG Stored in the Fiber and Soil

Figure 7 - Greenhouse gas emissions from U.S. cotton production, and GHG reductions due to
carbon stored in the fiber and soil
Source: GHG emissions based on average of data from four studies see section entitled Energy Use and GHG
Emission Data Sources in text below and Appendix B.

Energy Use and GHG Emission Data Sources

Three recently published studies and Cotton Incorporated’s 2008 grower survey examines U.S.
cotton’s impact on energy and air resources. A study by Nelson et al. (2009) evaluated the energy
use and greenhouse gas emissions of different tillage systems for different crops in the U.S., including
cotton. This study is published in a peer reviewed scientific journal and was conducted without input
or funding from Cotton Incorporated (it was funded largely by the U.S. Department of Energy).

A second study was released in 2009 by a collaborative group of agricultural producers, food and
retail companies, conservation organizations and agribusinesses. This 35 member group called Field
to Market: The Keystone Alliance for Sustainable Agriculture, developed natural resource indicators
(land, water, energy use, soil loss, GHG emission) to examine U.S. agriculture’s impact on the

Page 11 of 31
environment over the last 20 years from the production of corn, soybeans, wheat and cotton
(Keystone, 2009). The development of the natural resource indicators and the study’s results are
based on publically available USDA data. Cotton Incorporated is a member of the Field to Market
Alliance.

Cotton Incorporated funded engineers and scientists at the Center for Agricultural and Rural
Sustainability at the University of Arkansas to evaluate U.S. cotton’s energy and GHG footprints to
help guide future research and enable U.S. cotton producers to continue making strides in reducing
cotton production’s impact on natural resources. Work is currently underway to prepare a
manuscript of the Arkansas study for submission into a peer-reviewed scientific journal (Matlock et
al., 2008; Matlock et al., 2009).

Finally, data on 2007 agricultural chemical use in cotton (USDA, 2008b) were used in combination
with energy and agricultural GHG emission factors from West and Marland (2002), and Cotton
Incorporated’s Natural Resource Survey (Reed, et al., 2009) to develop an additional evaluation of
energy use and GHG emissions using the most current available data. Details on the methods used to
combine these data sources are given in Appendix C.

Average results from these four studies are used in Figures 6 and 7, and specific values from the
individual studies are reported in Table B.1 of Appendix B to this report.

Ecosystem Services
Quantifying cotton’s impact on ecosystem services (benefits that people obtain from ecosystems
such as freshwater, timber, climate regulation, erosion control, and recreation) is a difficult metric to
develop because of the numerous groups monitoring and collecting data and the myriad
methodologies being utilized. Cotton Incorporated, along with many others, is evaluating several
different approaches to quantify ecosystem impact. Measuring agriculture’s impact on water quality
is one such approach. The data source under consideration to measure impact on water quality is the
U.S. Geological Survey (USGS) National Water Quality Assessment (NAWQA) program (Toccalino,
2007). This program regularly monitors water quality and tests for agricultural compounds in surface
Page 12 of 31
and ground waters. Based on USGS reports that monitor the number of nitrate and pesticides found
to be above human health levels in surface and ground water, incidences of contamination of surface
and ground water by fertilizers or pesticides are rare in the U.S. [see Table 2.3 of Keystone (2009)].

Thriving ecosystems are essential for the health of habitat upon which wildlife depends. U.S. cotton
producers concern for ecosystems and wildlife habitat is evidenced by growers’ enrollment in a
number of USDA wildlife and environmental conservation programs such as the Conservation Reserve
Program (CRP), the Environmental Quality Incentives Program (EQIP), Wildlife Habitat Incentives
Program (WHIP), Highly Erodible Land and Wetland Conservation (HELC-WC), Conservation Security
Program (CSP) or other regional and local conservation programs such as Ducks Unlimited or Delta
Wildlife. According to Cotton Incorporated’s Natural Resource Survey (Reed, et al.,2009), 39% of
growers participate in EQUIP and 34% of growers participate in CRP. These programs make a
difference as 62% of cotton producers reported an increase in wildlife in and around their cotton
fields over the last 10 years (Reed et al., 2009).

Improved water quality and wildlife habitats are in large part the result of the steady decline in the
number of insecticide applications and overall use of pesticides since the mid-1990’s (Figure 8). In
2007, 34% of U.S. cotton acres did not require a single insecticide application (USDA, 2008b). This is
due to the success of the boll weevil eradication program and an increase in the adoption of Bt
cotton.

Page 13 of 31
 6
Insecticide Applications
5 Made to Cotton

0
1986 - 1990 1991 - 1995 1996 - 2000 2001 - 2005 2006 - 2008

Figure 8 - Number of insecticide applications made to U.S. cotton

(National Cotton Council, 2009).

There are examples where modern cotton production systems are actually restoring ecosystem
services. In studies of the Beasley Lake watershed in the Mississippi delta region of the U.S., Locke et
al. (2008) report: “Applying combinations of conservation practices can significantly reduce nonpoint
source pollution. For example, converting row crops to reduced tillage and transgenic herbicide-
resistant crops in BLW reduced suspended sediment (70% reduction), total phosphorus (41%
reduction), and pesticide concentrations in lake water. Corresponding increases in Secchi visibility
(97%) and chlorophyll a (a primary productivity indicator) likely contributed to improved fish
productivity (e.g., fish weight increase comparing 1998 and 2004: Micropterus salmoides 87%,
Lepomis macrochirus, 65%) during this period.”

While this is only one example for the U.S., such improvements are occurring in cotton production
systems around the globe as recently summarized by Naranjo (2009). For example, Naranjo (2009)
reports reductions of insecticide applications in the range of 40 to 60% in China, Australia, South
Africa, Argentina and Mexico due to the adoption of insect resistant cottons. Since the adoption of Bt
cotton in 1997 the volume of insecticide use in China has decreased 34.5% (Brookes, 2009). And
James (2008) reports that according to the Center for Chinese Agricultural Policy (CCAP), small
farmers adopting Bt cotton increased yield by 9.6% and reduced insecticide use by 60% (James,
Page 14 of 31
2008). Bt cotton has also had a substantial positive impact on pesticide use in India. In 1998, Indian
cotton pesticides represented 30% of the total pesticide market, whereas in 2006, that number had
fallen to only 18% (James, 2008). Similarly, cotton insecticides represented 42% of the total Indian
insecticide market in 1998 compared to 28% in 2006 (James, 2008). Further, in the five years from
2001-2006 (years of highest rate of Bt cotton adoption), pesticide sales in India decreased by 22%
(James, 2008).

Globally, the 23% reduction in insecticide use attributed to the use of Bt cotton is greater than that
recorded for any other Bt crop (Brookes, 2008). Importantly, of the associated increase in
environmental benefits from Bt adoption, over half occurred in developing countries (Brookes, 2008).
Despite these gains, there are situations where crop protection products must be used. When
pesticides are used, regulations established by the USDA, U.S. Environmental Protection Agency
(USEPA) and U.S. Food and Drug Administration (FDA) ensure that they are used in a manner that
minimizes environmental impact. The USEPA requires an extremely cautious approach to use of plant
protection products. For regulatory approval, every product and its use undergoes rigorous testing to
assure safety to humans, animals, and the environment. In fact, in 2007 the USEPA completed its 10
year review of pesticides as mandated by Congress' Food Quality Protection Act. The law required a
re-review of every plant protection product and imposed additional factors to mitigate any
uncertainties as to their safety. Where products failed to meet updated and stricter precautionary
environmental and health standards, registration and use of such products in the U.S. was not
allowed.

While developing a clear metric to measure ecological services is an ongoing challenge, current out-
come based indicators support a trend of lower impact of cotton production on ecosystem services.

Ongoing improvements
The primary goal of Cotton Incorporated’s use of LCA techniques is to evaluate cotton’s
environmental footprint and identify areas for improvement. As part of that process we have
evaluated the relative contribution of different inputs and components of the cotton production
process to identify where our largest resource use occurs. For example, Table 1 shows the total and
Page 15 of 31
relative amount of energy use by production activity for three different production scenarios, each
representing different environments in the U.S. Details on the methodology used to make these
estimates are included in Appendix C. The first column represents a non-irrigated production system
that is representative of many areas of the mid-south and southeast with typical rainfall patterns.
The second column represents the same geographic area where supplemental irrigation is used,
while the third column represents a more irrigation intensive production system that would be
representative of West Texas.

In all three production systems, particularly the first two, fertilizers represent a large portion of
cotton’s energy footprint, largely due to energy associated with nitrogen. While Cotton Incorporated
has a long history of funding nitrogen management research, upon seeing that nitrogen use has the
greatest impact on energy and GHG emissions (GHG emissions closely parallel energy), greater efforts
were directed towards improving nitrogen use efficiency. Researchers are now: 1) evaluating on-the-
go sensors that measure and adjust the crop’s nitrogen requirement in real-time; 2) conducting
nitrogen management recommendation validation trials in 10 states; 3) searching for nitrogen fixing
cover crops adapted to a wide range of environments; 4) exploring traditional breeding techniques in
hope of finding a cotton plant that requires less nitrogen.

A second area that will also require ongoing attention is the impact of irrigation on cotton’s overall
environmental footprint. Not only are water resources becoming limited in many areas of the U.S.
and the world, the energy associated with pumping irrigation water can quickly dominate cotton’s
energy footprint. For example, the second column of Table 1 shows the divisions of energy use in an
irrigated system where supplemental irrigation is applied from a shallow well. In this scenario,
irrigation is a small part of cotton’s energy inputs (column 2, Table 1). However, when 250 mm of
irrigation water is applied from a pumping depth of 90 meters the energy requirement of irrigation
increases nearly 50% relative to a non-irrigated system (12.68 Gj/Mg to 19.22 Gj/Mg). Therefore, in
areas where pumping from deep wells is required, investment in renewable energies (wind or solar)
could be a feasible way for producers to manage future energy costs as well as to lower their carbon
emissions. Note that some cotton producers obtain irrigation water from reservoirs rather than
wells. In this situation the energy partition from irrigation would be very low since drawing water

Page 16 of 31
from a reservoir requires less pumping than from wells. In some areas, irrigation will continue to be
an important part of an efficient production system. In arid areas, the highest yields are achieved by
proper irrigation ensuring a consistent level of production and maximum land use efficiency. In more
humid regions, a small amount of irrigation at the right time can prevent significant yield loss and
provide a stable supply of quality fiber and cottonseed from year to year.

Efforts to improve nitrogen and water use efficiency make good business sense as well. All sources of
energy will be very expensive in the future and minimizing cotton’s energy footprint will increase the
cotton producers' profitability.

Page 17 of 31
Table 1 - Total energy requirements and percent energy requirement of major processing events for
three U.S. cotton production scenarios

Production Scenario

Irrigation: 50 mm; 9 Irrigation: 250 mm; 90

Non-irrigated meter pumping depth meter pumping depth
Total Energy Input (GJ per
Mg fiber produced) 12.7 13.8 19.2
Yield (Mg per ha) 0.8 1.1 1.7

Source: Percent of Total Energy Usage

Planting Seed Production 3% 2% 1%
Tillage 10% 8% 4%
Fertilizers* 48% 50% 35%
Crop Protection Products* 18% 17% 11%
Harvest 10% 7% 3%
Ginning 12% 13% 9%
Irrigation 0% 3% 37%
Total 100% 100% 100%

* Includes energy embedded in the compounds, and energy required to apply to the field.

Page 18 of 31
References
1. Brookes, G. & Barfoot, P. (2008). Global Impact of Biotech crops: Socio-economic and
Environmental Effects, 1996-2006. AgBioForum, 11 (1):21-38.
2. Brookes, G. & Barfoot, P. (2009). GM Crops: Global Socio-economic and Environmental
Impacts 1996-2007. PG Economics 2009
http://www.pgeconomics.co.uk/pdf/2009globalimpactstudy.pdf.

3. Causarano, H. J., A. J. Franzluebbers, D. W. Reeves, and J. N. Shaw. 2006. Soil Organic Carbon
Sequestration in Cotton Production Systems of the Southeastern United States: A review. J.
Env. Qual. 35:1374-1383
4. Chapagain, A.K., A.Y. Hoekstra, H.H.G. Savenije, R. Gautam. 2006. The Water Footprint of
Cotton Consumption: An assessment of the Impact of Worldwide Consumption of Cotton
Products on the Water Resources in the Cotton Producing Countries. Ecol. Econ. 60:186-203.
5. FAO, 2005. 2005 harvested acres from the United Nations Food and Agricultural Organization
6. Gowda, P.H., R. Louis Baumhardt, Allison M. Esparza, Thomas H. Marek, and Terry A. Howell.
2007. Suitability of Cotton as an Alternative Crop in the Ogallala Aquifer Region. Agronomy
Journal 99:1397-1403.
7. Hoekstra, A.Y. and A.K. Chapagain. 2007. Water Resource Manage 21:35–48.
8. James, Clive. 2008. Global Status of Commercialized Biotech/GM Crops: 2008. ISAA Brief No.
39. ISAAA: Ithaca, NY
9. Keystone, 2009. Environmental Resource Indicators for Measuring Outcomes of On-Farm
Agricultural Production in the United States. First Report, January 2009
10. Locke, M.A., S.S. Knight, S. Smith Jr., R.F. Cullum, R.M. Zablotowicz, Y. Yuan, and R.L.
Bingner.2008. Environmental Quality Research in the Beasley Lake Watershed, 1995 to 2007:
Succession from Conventional to Conservation Practices. J. of Soil and Water Conservation
63(6):430-442.
11. Matlock, M., Greg Thomas, Darin Nutter, Thomas Costello. 2008. Energy Use Life Cycle
Assessment for Global Cotton Production Practices. March 15 Final report to Cotton
Incorporated by the Center for Agricultural and Rural Sustainability, University of Arkansas
Division of Agriculture, Fayetteville, AR.

Page 19 of 31
12. Matlock, M., L. Nalley, and Z. Clayton-Niederman. 2009. Carbon Life Cycle Assessment of
United States Cotton: 'A View of Cotton Production Practices and their Associated Carbon
Emissions for Counties in 16 Cotton Producing States. Feb. 5 2009 Final report to Cotton
Incorporated by the Center for Agricultural and Rural Sustainability, University of Arkansas
Division of Agriculture, Fayetteville, AR
13. Meyer, Leslie , Stephen MacDonald, James Kiawu, COTTON AND WOOL SITUATION AND
OUTLOOK YEARBOOK. Washington, D.C.: Economic Research Service, U.S. Department of
Agriculture, November 2008.
(http://usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1228)
14. National Cotton Council Beltwide Cotton Conference Proceedings. 1987-2009. 2009. Cotton
Insect Losses Reports
15. National Cotton Council. 2009. Personal communication; calculation based on PCI Redbook,
World Synthetic Supply/Demand http://www.pcifibres.com/
16. Nelson, Richard G., Chad M. Hellwinckel, Craig C. Brandt and Tristram O. West, Daniel G. De La
Torre Ugarte, and Gregg Marland. 2009. Energy Use and Carbon Dioxide Emissions from
Cropland Production in the United States, 1990–2004. J. Environ. Qual. 38:418–425.
17. Naranjo, Steve. 2009. Impacts of Bt Crops on Non-Target Invertebrates and Insecticide Use
Patterns. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural
Resources 2009 4, No. 011.
18. Reed, J.N., E.M. Barnes and K.D. Hake. 2009. Technical Information Section. US Cotton
Growers Respond to Natural Resource Survey. International Cotton Advisory Committee, THE
ICAC RECORDER, Vol. XXVII. No. 2, June 2009.
19. Toccalino, Pl., J.E. Norma, N.L. Booth, and J.S. Zogorski. 2007. Health-based Screening Levels:
A Tool for Evaluating What Water-Quality Data May Mean to Human Health. USGS, NAWQA.
20. USDA, 2004. Farm and ranch irrigation survey (2003). Vol. 3, Special Studies, Part 1. Issued
November 2004. United States Department of Agriculture, National Agricultural Statistics
Service.
21. USDA, 2008a. Oil Crops Yearbook (Update 03/08) Stock #89002 Economic Research Service,
United States Department of Agriculture
http://usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1290

Page 20 of 31
22. USDA, 2008b. Agricultural Chemical Usage 2007 Field Crops Summary. May 2008. United
States Department of Agriculture, National Agricultural Statistics Service. Ag Ch 1 (08).
23. Valco, T.D., J.K. Green, R.A. Isom, D.S. Findley, T.L. Price, and H. Ashley. 2009. The Cost of
Ginning Cotton – 2007 survey results. 2009 Beltwide Cotton Conferences, NCC
24. West, T,O., and Gregg Marland. 2002. A Synthesis of Carbon Sequestration, Carbon Emissions,
and Net carbon Flux in Agriculture: Comparing Tillage Practices in the United States.
Agriculture, Ecosystems and Environment 91:217–232.

Page 21 of 31
Appendix A. Summary of units of measure and U.S. cotton data sources

One of the challenges in evaluating LCI data is that it has not always been reported in a common
format. Table A.1. is a summary of the key metrics used throughout the main report using the
reference data below and provides a comparison of values for different units of measure.

Table A.1. Summary of metrics used for U.S. cotton land, energy, GHG, and water use calculations
Land Units: kg/ha lb/acre ha/kg ha/Mg*
Fiber: 933 833 0.0011 1.072
Seed: 1,403 1,250 0.0007 0.713
Fiber + Seed: 2,336 2,083 0.0004 0.428

cubic meters
Water Units per Mg L / kg lb/acre-in gal/lb

Irrigation and fiber produced 1811 1811 125 217

Irrigation and (fiber + seed produced) 725 725 313 87

Energy Units (per mass of fiber produced): GJ/Mg BTU/lb MJ/kg MJ/lb
To produce fiber: 20.4 8,767 20.4 9.2
Stored in seed: 32.4 13,924 32.4 14.7

Air Units (per mass of fiber produced) kg CO2e/kg lb CO2e / lb kg Ce/kg kg CO2e/lb
GHG Emissions: 1.8 1.8 0.49 3.97
Stored in lint & soil: 2.2 2.2 0.59 4.78

A summary of the references for Table A.1.:

• Land – Fiber data from Meyer et al. (2008); Seed data from USDA (2008a).
• Energy – Mean of the data reported in: Keystone (2008); Matlock et al. (2008); Nelson et al.
(2009); West and Marland (2002) and Reed et al. (2009)
• GHG – Mean GHG emissions from averages reported in: Keystone (2008); Matlock et al.
(2009); Nelson et al. (2009); West and Marland (2002) coefficients combined with Reed, et al.
(2009)
• Water - USDA (2004).

Additional details on assumptions needed to summarize these data are provided in Appendix B.

Page 22 of 31
Appendix B. Data sources used to derive energy use and GHG emissions
estimates for U.S. cotton production

Values for energy use and GHG emissions during U.S. cotton production were derived by averaging
data from four individual studies. The studies and data are listed in Table B.1. below and are the
basis for Figures 6 and 7 in the main report.

Table B.1. References used to estimate the energy and greenhouse gas emissions associated with
the production of one Mg of cotton fiber in the U.S.
Energy GHG emissions
Study (GJ/Mg) (kg CO2e kg-1)
Nelson et al. (2009) 19.4 1.3
Matlock et al. (2008; 2009) 26.6 2.2
Keystone (2009) 15.6 1.5
Appendix C* 20.0 2.2
Average 20.4 1.8
*Applying coefficients from West and Marland (2002) to data from the USDA (2008b) and Cotton
Incorporated’s 2008 Natural Resource Survey (Reed, et al., 2009). See Appendix C for details. Other
details specific to each study are noted in the text below.

None of the studies listed in Table B.1 included the energy use or greenhouse gas emissions from
transport of cotton to the gin or the ginning process itself. Therefore, data on U.S. ginning energy use
from Valco et al. (2009) were used to add an additional energy use of 1.8 GJ per Mg of fiber and 0.2
kg of CO2e per kg of fiber to all of the studies estimates to accurately portray the total footprint to a
bale ready for delivery to a textile mill (Table B.1 includes these additions). More detail on how these
values were derived is provided in Appendix C.

The Nelson et al. (2009) data used was reported in their Table 1 for the 2004 crop year. As data were
reported on a per hectare basis, the average U.S. fiber yield of 944 kg per hectare in 2004 (Meyer et
al., 2008) was used to convert to a units of GHG and energy use per unit of fiber produced.

Page 23 of 31
Matlock et al. (2008) reported energy required per unit of seed and fiber produced. A 40% lint
percentage was used to convert to energy per unit of fiber. This corresponds to 1.5 units of seed per
unit of lint produced. The 1.5 value was computed by calculating the ratio of total U.S. seed
production as reported by the USDA (2008a) to total fiber produced as reported in Meyer et al.
(2008) for the 2004 to 2008 crop years and using the average ratio calculated for that time period.

Soil carbon data used for Figure 7: Causarano et al. (2006) provided an average amount of carbon
stored in the soil in no-tillage cotton fields from numerous published studies in the southeastern U.S.
as 0.67 Mg C per ha per year (2.6 CO2e per ha per year). Data from Cotton Incorporated’s 2008
Natural Resource survey indicate that about one-third of cotton producers are using no-till, so the
value used was divided by 3 (=0.63 CO2e per ha per year).

Carbon off-sets due to cottonseed: The estimates of GHG emissions for cotton production include
emissions associated with the production of both the fiber and the cottonseed since no additional
production steps are required to obtain cottonseed; during the production of a kg of fiber, 1.5 kg of
cottonseed is produced. Cottonseed has an oil content of 14% and is highly valued as a feed
ingredient. Therefore, it is reasonable to claim a credit for the energy embodied in cottonseed. One
approach to take in estimating the credit is to calculate the amount of biodiesel that could be
produced from the oil in the seed, and then calculate the amount of CO2e that would be eliminated if
that amount of petroleum-based diesel were replaced with the biodiesel. Using the previously noted
production of 1.5 kg of seed for every kg of fiber produced, and 14% oil content in the seed,
approximately 0.24 liters of cottonseed-derived biodiesel could be produced for every kg of fiber
produced (1.5 kg x 0.14 = 0.21 kg of oil x 1.14 liter of oil per kg oil = 0.24 liters of fuel). This has the
impact of preventing the emission of 0.64 kg of C02e due to petroleum diesel savings [using the diesel
CO2e emission factor at point of combustion from West and Marland (2002) and an assumed diesel
energy content of 0.0387 GJ per liter]. That is, if a cotton producer used biodiesel from his
cottonseed, his GHG emissions would be reduced by 0.64 kg of C02e.

Page 24 of 31
Appendix C. Methods used to derive energy use and GHG emissions estimates
for U.S. cotton production

West and Marland (2002) provided an extensive list of energy use and greenhouse gas emission
factors for agricultural operations in the U.S. These factors were used in combination with USDA
reported information and data from the Cotton Incorporated’s Natural Resource Survey (Reed, et al.,
2009) of U.S. cotton producers, conducted by OrgWide Services on behalf of the U.S. cotton industry
in 2008 to provide a current estimate of U.S. cotton’s energy and GHG footprint. The following
narrative describes estimates of energy and GHG emissions in sequential order of cotton production
(from planting the seed through harvest and ginning). The resulting energy and GHG values from the
combined information are summarized in Table C.1.

Land Preparation and Planting

Data from the Natural Resource Survey indicates that on average, U.S. cotton producers conduct
three tillage operations per year, and that only 46% conduct any deep tillage operations. This is also
consistent with the Natural Resource Survey statistic that two-thirds of U.S. cotton producers report
they are using some form of conservation tillage practices. Therefore, to characterize tillage
operations, one operation was assumed to be a deep tillage operation but only at a rate of every
other year. This accounts for the fact that only 46% of producers are using deep tillage, and from a
case study perspective, many producers who do use deep tillage often only do so every other year.
An estimate of deep tillage fuel use was taken from Raper and Bergtold (2007) as this operation was
not included in West and Marland (2002). An estimate of carbon emissions was developed by using
the ratio of carbon emissions to energy use for a moldboard plow of West and Marland (2002) and
multiplying that ratio by the energy use associated with deep ripping [(26.75 kg Ce/ha/(1122+102MJ)
= 0.0219 kg Ce / (ha MJ)]. The final two tillage operations were represented by one pass with a disk
and one cultivation. The energy values used include both the diesel fuel use of the operation and the
energy embodied in manufacturing, transportation and repair of the machinery associated with the
operation (MTR).

West and Marland (2002) provide estimates of 33 MJ of energy and 0.65 kg of Ce emissions per kg of
cotton planting seed used. An optimal planting rate of 100,000 seed per ha is typical of university

Page 25 of 31
Cooperative Extension Service recommendations in the U.S. (e.g., NCSU 2009) and current cotton
seed varieties tend to average 11,000 seed per kg (e.g.,
http://www.deltaandpine.com/dp/products/srg), resulting in a planting seed rate of 9.0 kg per ha.

Within-Season Management
In order to represent a “typical” cotton farm, the total irrigation water used for cotton was as
reported in USDA (2004) and then distributed across all cotton acres, resulting in an average depth of
irrigation water applied of 15 cm. An average pumping dept of 63 m at an operation pressure of 69-
kPa was based on responses from Reed et al. (2009) for irrigated fields. Table C.2 provides additional
details on how the total irrigation energy and GHG values were derived for Table C.1.

All chemical use rates are based on data from the USDA (2008). For pesticide use and other
chemicals, an average of energy and GHG data of the herbicide, insecticide and fungicide values
reported in Table 4 of West and Marland (2002) was used. An exact figure for lime use by cotton
producers in the U.S. is not reported by the USDA; however, the USDA Economic Resource Service
reports that only 34% of cotton acres were ever treated with lime (2003 survey data from
http://www.ers.usda.gov/Data/ARMS/app/Crop.aspx) and this value was in strong agreement with
the fact only 37% of producers in the Natural Resource Survey survey (Reed et al., 2009) report lime
applications. In West and McBride (2005), they suggest approximately 30 Tg of lime was applied to
all U.S. agricultural crops in 2001. Assuming approximately 125 million ha of U.S. crop land, this
corresponds to an average annual lime use of 238 kg of lime per ha. Such a rate is reasonable – for
example, assume a typical field that does require lime receives 2420 kg per ha every three years or
807 kg per ha per year. Noting 34% of cotton acres are treated with lime, a “typical” value of 274 kg
per ha per year was used in this analysis. The previous values discussed did not include applications
for the fertilizers and other chemicals. The Natural Resource Survey data indicate that on average,
producers made a total of 6 chemical applications per year (pesticides, growth regulators and harvest
aids) and values taken directly from Table 7 in West and Marland (2002) were used in the calculations
shown in Table C.1.

Harvest and Ginning

Cotton harvest data were not reported by West and Marland (2002); therefore, fuel use data from
Willcutt and Barnes (2008) were used to estimate harvest energy use with the MTR for harvest with a
Page 26 of 31
combine from Table 7 of West and Marland (2002). Transportation energy to take the cotton from
the field to the gin is from Hamann et al. (2008). Ginning energy use was based on data reported in
Valco et al. (2009).

Soil Nitrous Oxide Emissions

The Intergovernmental Panel on Climate Change (IPCC) assumes in their 2001 Third Assessment
Report that 1% of nitrogen applied to soil is emitted as N2O, although this amount varies significantly
based on soil type, soil conditions and climate, as well as amount and timing of application. Snyder,
et al. (2007) applied some of these variables to the IPCC value of 1% and determined that 1% was
equivalent to 1.27 kg Ce per kg N. Based on a nitrogen use rate in cotton of 94.4 kg/ha, the total
carbon equivalent emissions for nitrogen use is 119.8 kg Ce/ha, a value that represents 21% of the
total GHG emissions during cotton production (Table C.1.). However, a high level of uncertainty
exists as to the magnitude of N loss as N20 (see review by Snyder et al., 2007 and Rochette and Eriksen-
Hamel, 2008) and it is likely that the default IPCC factor could be less than 1% (1.27 kg Ce per kg N) for
U.S. cotton production. Indeed, both USDA ERS Resource Management Survey (ARMS, see
http://ers.usda.gov/Data/ARMS/app/Crop.aspx, data for the 2003 cotton season) and Natural
Resource Survey data support:
• Cotton producers are very aware of the potential losses associated with nitrogen application
and a majority, 73% (ERS) apply nitrogen after planting to minimize losses,
• a majority, 80% (ERS), incorporate nitrogen fertilizers into the soil to minimize losses due to
nitrogen volatilization,
• Cotton producers use soil testing extensively (86% of Natural Resource Survey respondents) to
determine fertilizer rates to insure the proper amount is applied.

All of these factors contribute to increased nitrogen use efficiency and would likely lower any nitrous
oxide emissions that may occur during U.S. cotton production. Note that approximately the same
value for nitrous oxide emissions was assumed for the GHG estimates from the Keystone (2009) and
Matlock et al. (2009) studies, but not in the values reported by Nelson et al. (2009).

Relationship between per area and per mass basis

In order to translate data between a per hectare basis and per mass of fiber basis, a yield of 933 kg
cotton per ha was used (2004 to 2008 average U.S. fiber yield from Meyer et al., 2008). Using this
Page 27 of 31
yield, the per hectare values of Table B.1 corresponds to 20 MJ of energy and 2.2 kg CO2e equivalent
emissions on the basis of kg of cotton fiber produced.

Note GHG emissions values are expressed in units of Ce that in Appendix C, according to
nomenclature of West and Marland (2002); however, the term carbon dioxide equivalents (CO2e) was
used in the main body of this report as recommended by Bowman et al. (2009), who advocates using
C02e to reference all GHG emissions.

Page 28 of 31
Table C.1. Estimates of energy use and carbon emissions from a typical U.S. cotton field

Total
Energy Total Ce
Energy Energy Carbon Use Rate (MJ per (kg Ce
Input Value Units value Carbon Units Use Rate Units ha) per ha)

Tillage Operations & Planting:

Deep tillage (46%) 604 MJ/ha 13.29 kg Ce / ha 0.5 per year 302 6.6
Disk 400 MJ/ha 8.72 kg Ce / ha 1 per year 400 8.7
Cultivate 210 MJ/ha 4.57 kg Ce / ha 1 per year 210 4.6
Plant 312 MJ/ha 6.79 kg Ce / ha 1 per year 312 6.8
Planting seed 33 MJ/kg 0.65 kg Ce / kg seed 9.0 kg / ha 296 5.8

Within-season management
Irrigation 5124 MJ/ha 175.06 kg Ce / ha See Table C.2 5124 175.1
Total chemical use
(excludes fertilizers) 280 MJ/kg 4.94 kg Ce / kg AI 5.72 kg AI/ha 1601 28.2
N 57 MJ/kg 0.86 kg Ce/ kg N 94.4 kg /ha 5422 80.9
P 20 5 7 MJ/kg 0.17 kg Ce / kg P205 32.7 kg /ha 229 5.4
K 20 7 MJ/kg 0.12 kg Ce / kg K20 41.1 kg /ha 282 4.9

CaCO3 2 MJ/kg 0.04 kg Ce/ kg CaCO3 238.0 kg/ha 407 8.5

Fertilizer application 566 MJ/ha 12.35 kg Ce / ha 1.0 per year 566 12.4
Other chemical
applications 119 MJ/ha 2.54 kg Ce / ha 6.0 per year 714 15.2

Harvest and Ginning

Harvest (pick and
module) 1128 MJ/ha 24.44 kg Ce / ha 1.0 per year 1128 24.4
Transport to gin 113 MJ/ha 2.48 kg Ce / ha 1.0 per year 113 2.5
Ginning 1572 MJ/ha 47.47 kg Ce / ha 1.0 per year 1572 47.5

Soil emissions
Nitrous Oxide na 1.27 kg Ce / kg N 94.4 kg /ha na 119.8

Totals 18,679 557

Page 29 of 31
Table C.2. - Calculation of weighted average of irrigation energy and GHG emissions for a typical
U.S. cotton farm.

Energy Source 1 Energy Required 2 (MJ GHG Emission 3 (kg

Energy Source (%) per ha) C per ha)
Electricity 4 55% 4,720 236
Natural Gas 21% 6,498 94
Diesel 24% 4,850 106
Weighted Average 5124 175

1
Based on Table 5 of West and Marland (2002)
2
Based on energy needed to pump 15 cm of water per ha from an average dept of 61 m at
an operating pressure of 69 kPa using equations 19.1 and 19.2 of Hoffman et al. (1992)
3
Using the same ratio of kg C to GJ energy of Table 5 of West and Marland (2002)
4
Actual energy used at pump = 449 kWh. The U.S. Deparment of Energy reports actual
energy use values of 10.5 MJ per kWh for the U.S. in 2004
(http://www1.eere.energy.gov/ba/pba/intensityindicators/delivered_electricity.html)

Page 30 of 31
Additional References for Appendix C
1. Bowman, T.E., E. Maibach, M.E. Mann, S.C. Moser, and R.C.J. Somerville. 2009. Creating a
common climate language. Science 324:36-37.
2. Hamann, M., C.B. Parnell, S.L. Simpson, S.A. Emsoff, and C. Hammock. 2008. Economic
analysis of transporting seed cotton with semi-tractor trailers. 2008 Beltwide Cotton
Conferences, Nashville, Tennessee, January 8-11, 2008 pg. 788 - 795
3. Hoffman, G.J., T.A. Howell, and K.H. Solomon. 1992. Management of Farm Irrigation Systems,
ASAE Monograph Number 9. St. Joseph MI
4. NCSU, 2009. 2009 Cotton Information. North Carolina Cooperative Extension Service, College
of Agriculture and Life Sciences, North Carolina State University. 209 pp.
5. Raper and Bergtold. 2007. In-row subsoiling: A review and suggestions for reducing cost of
this conservation tillage operation. Applied engineering in Agriculture 23(4):463-471.
6. Rochette, P. and N.S. Eriksen-Hamel. 2008. Chamber Measurements of Soil Nitrous Oxide
Flux: Are Absolute Values Reliable? Soil Sci. Soc. Am. J. 72(2):331-342
7. Snyder, C.S., T.W. Bruulsema, and T.L. Jensen. 2007. Greenhouse gas emissions from cropping
systems and the influence of fertilizer management—a literature review. International Plant
Nutrition Institute, Norcross, Georgia, U.S.A.
8. Valco, T.D., J.K. Green, R.A. Isom, D.S. Findley, T.L. Price, and H. Ashley. 2009. The cost of
ginning cotton – 2007 survey results. 2009 Beltwide Cotton Conferences, NCC.
9. West, T.O. and A.C. McBride. 2005. The contribution of agricultural lime to carbon dioxide
emissions in the United States: dissolution, transport, and net emissions. Agriculture,
Ecosystems and Environment 108:145–154.
10. West, Tristram, O., and Gregg Marland. 2002. A synthesis of carbon sequestration, carbon
emissions, and net carbon flux in agriculture: comparing tillage practices in the United States.
Agriculture, Ecosystems and Environment 91:217–232.
11. Willcut, M.H., and E.M. Barnes. 2008. Fuel consumption in spindle picker cotton harvesting
systems. 2008 Beltwide Cotton Conferences, Nashville, Tennessee, January 8-11, 2008.

Page 31 of 31