The Role of Livestock Production in Carbon and Nitrogen Cycles

Abstract

This review looks at the role of the livestock sector in carbon (C) and nitrogen (N) cycles from a global perspective and considers impacts at the various stages of the commodity chain. With regard to livestock, N and C cycles are closely connected to livestock's role in land use and land-use change. Livestock's land use includes grazing land and cropland dedicated to the production of feed crops and fodder. Considering emissions along the entire commodity chain, livestock currently contribute about 18% to the global warming effect. Livestock contribute about 9% total carbon dioxide (CO2) emissions, but 37% methane (CH4), and 65% nitrous oxide (N2O). The latter will substantially increase over the coming decades, as the pasture land is currently at maximum expanse in most regions; future expansion of the livestock sector will increasingly be crop based. The chapter also reviews mitigation options to reduce C and N emissions from livestock's land use, production, and animal waste.

Full text

ANRV325-EG32-09 ARI 21 September 2007 16:56
The Role of Livestock
Production in Carbon
and Nitrogen Cycles
Henning Steinfeld
1
and Tom Wassenaar
2
1
Animal Production and Health Division, Food and Agriculture Organization
of the United Nations, 00100 Rome, Italy; email: henning.steinfeld@fao.org
2
Joint Research Center of the European Commission, Institute for the Protection
and Security of the Citizen, AgriFish Unit, 21020 Ispra (Varese), Italy;
email: tom.wassenaar@jrc.it
Annu. Rev. Environ. Resour. 2007. 32:271–94
First published online as a Review in Advance on
September 5, 2007
The Annual Review of Environment and Resources
is online at http://environ.annualreviews.org
This article’s doi:
10.1146/annurev.energy.32.041806.143508
Copyright
c
 2007 by Annual Reviews.
All rights reserved
1543-5938/07/1121-0271$20.00
Key Words
air pollution, fossil fuel use, greenhouse gas emissions, land use and
land-use change, terrestrial carbon loss, water depletion
Abstract
This review looks at the role of the livestock sector in carbon (C) and
nitrogen (N) cycles from a global perspective and considers impacts
at the various stages of the commodity chain. With regard to live-
stock, N and C cycles are closely connected to livestock’s role in land
use and land-use change. Livestock’s land use includes grazing land
and cropland dedicated to the production of feed crops and fodder.
Considering emissions along the entire commodity chain, livestock
currently contribute about 18% to the global warming effect. Live-
stock contribute about 9% of total carbon dioxide (CO
2
) emissions,
but 37% of methane (CH
4
), and 65% of nitrous oxide (N
2
O). The
latter will substantially increase over the coming decades, as the pas-
ture land is currently at maximum expanse in most regions; future
expansion of the livestock sector will increasingly be crop based. The
chapter also reviews mitigation options to reduce C and N emissions
from livestock’s land use, production, and animal waste.
271
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
Contents
1. INTRODUCTION .............. 272
2. GLOBAL TRENDS IN
LIVESTOCK PRODUCTION . . 272
3. THE ROLE OF EXTENSIVE
LIVESTOCK PRODUCTION . . 274
3.1. Carbon ...................... 274
3.2. Nitrogen .................... 277
4. THE ROLE OF INTENSIVE
LIVESTOCK PRODUCTION . . 279
4.1. Carbon ...................... 279
4.2. Nitrogen .................... 284
5. SUMMARY OF LIVESTOCK’S
ROLE IN CARBON AND
NITROGEN CYCLES .......... 287
6. TECHNICAL OPTIONS TO
MITIGATE CARBON AND
NITROGEN LOSSES ........... 288
6.1. Sequestering Carbon and
Mitigating Carbon Dioxide
Emissions ..................... 289
6.2. Reducing Methane Emissions
from Enteric Fermentation
Through Improved Efficiency
and Diets ..................... 289
6.3. Mitigating Methane Emissions
Through Improved Manure
Management and Biogas ....... 290
6.4. Mitigating Nitrogen Loss..... 290
1. INTRODUCTION
Livestock are present in most ecosystems of
the planet, and they also shape crop agricul-
ture to a large extent through their demand
for feed. Therefore, livestock activities have
important impacts on virtually all aspects of
the environment, including air and climate
change, land and soil, water and biodiversity.
The impact may be direct, through grazing
for example, or indirect, such as the expan-
sion of soybean production for feed replacing
forests in South America.
Livestock production is technically ex-
tremely diverse. In countries or areas where
there is no strong demand for food products
of animal origin, low-input production
prevails, mainly for subsistence rather than
for commercial purposes. This contrasts
with commercial, high-input production in
areas serving a growing or established high
demand. Such diverse production systems
make extremely diverse claims on resources.
Although intensive livestock production is
booming in many developing countries, there
are still vast areas where extensive livestock
production and its associated livelihoods
maintain their traditional forms. For the
purpose of this review, a distinction is made
between extensive forms of livestock produc-
tion, with limited use of external inputs, and
intensive production. Because feed represents
by far the largest cost item in livestock
production, extensive systems are defined as
depending on low-cost and locally available
feed inputs, whereas intensive systems are
based on marketable high-cost feed items,
such as grains, oil cakes, and cultivated
fodder. This review covers livestock’s role in
carbon (C) and nitrogen (N) cycles from a
global perspective and considers impacts at
the various stages of the commodity chain.
2. GLOBAL TRENDS IN
LIVESTOCK PRODUCTION
The livestock sector is growing and changing
rapidly. Until about 1980, most developing
countries, withthe exception of Latin America
and some Near East countries, had annual per
capita meat consumption of substantially less
than 20 kg. For most people in Africa and
Asia, meat, milk, and eggs were an unafford-
able luxury, consumed only on rare occasions.
A high proportion of the larger livestock in
developing countries was not primarily kept
for food, but for other important functions,
such as providing draught power and manure
as well as serving as an insurance policy and a
capital asset, usually disposed of only in times
of communal feasting or emergency.
Today, the livestock sector is currently
growing faster than the rest of agriculture
in almost all countries. Typically, its share
272 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
in agricultural gross domestic product rises
with income and the level of development
and is above 50% for most Organisation for
Economic Co-operation and Development
countries. The nature of livestock production
is also changing rapidly in many emerging
economies, as well as in developed countries.
Most of this change can be summarized under
the term “industrialization.” Through indus-
trialization, livestock escape most of the envi-
ronmental constraints that have shaped live-
stock production diversely in the wide range
of environments in which it occurs (1).
Growing populations and other demo-
graphic factors, such as age structure and ur-
banization, determine food demand and have
driven the intensification of agriculture for
centuries. Growing economies and individual
incomes have also contributed to growing de-
mand and a shift in diets. These trends have
accelerated over the past two decades in large
parts of Asia, Latin America, and the Near
East, spurring a rapid increase in demand for
animal products and other high-value food-
stuffs such as fish, vegetables, and oils.
Driven by population growth and rising
income in many developing countries, the
global livestock sector has seen a dramatic ex-
pansion over the past decades, although with
considerable differences between developing
and developed countries. In the developing
countries (Table 1), annual per capita con-
sumption of meat has doubled since 1980,
from 14 kg to 28 kg in 2002 (2, 3). Total
meat supply tripled from 45 million tonnes to
134 million tonnes over the same period. De-
velopments have been most dynamic in coun-
tries that have seen rapid economic growth,
notably East Asia, led by China. China alone
accounted for 57% of the increase in to-
tal meat production in developing countries.
For milk, developments are less spectacular
but still remarkable: Total milk production
in developing countries expanded by 118%
between 1980 and 2002; and 23% of that
increase came from one country, India.
There is a great deal of variation in the ex-
tent and character of livestock sector growth.
China and East Asia have experienced the
most impressive growth in consumption and
production, first in meat and more recently
also in dairy. In contrast, India’s livestock sec-
tor continues to be dairy oriented, using tradi-
tional feed resources and crop residues. Brazil,
Argentina, and other Latin American coun-
tries have successfully expanded their domes-
tic feed base, taking advantage of low produc-
tion costs and abundance of land. They have
moved to adding value to feed, rather than
exporting it. They are poised to become the
major meat-exporting region supplying devel-
oped and East Asian countries. Trade in ani-
mal products as well as feed is strongly increas-
ing worldwide, partly driven by the variable
Table 1 Past and projected trends in consumption and production of livestock products
Developing countries Developed countries
1980 1990 2002 2015 2030 1980 1990 2002 2015 2030
Consumption
Annual per capita meat consumption (kg) 14 18 28 33 38 73 80 78 83 88
Annual per capita milk consumption (kg) 34 38 46 57 67 195 200 202 204 211
Total meat consumption (million tonnes) 47 73 137 191 257 86 100 102 113 122
Total milk consumption (million tonnes) 114 152 222 330 449 228 251 265 278 292
Production
Annual per capita meat production (kg) 14 18 28 33 38 75 82 80 85 91
Annual per capita milk production (kg) 35 40 50 62 73 300 301 266 270 280
Total meat production (million tonnes) 45 43 134 190 255 88 103 105 116 126
Total milk production (million tonnes) 112 159 244 359 491 352 378 349 369 387
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 273
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
availability of natural resources. With trade
in these products, natural resources and envi-
ronmental impact are transferred, referred to
as virtual trade in water and nutrients (4).
In the developing countries, livestock pro-
duction is rapidly shifting toward mono-
gastrics, notably poultry and pigs. In fact,
poultry and pigs account for 77% of the ex-
pansion in production. Although total meat
production in developing countries more than
tripled between 1980 and 2004, the growth in
ruminant production (cattle, sheep, and goats)
was 111%, whereas that of monogastrics ex-
panded more than fourfold over the same
period.
These dramatic developments in rapidly
growing developing countries are in stark
contrast with trends in developed countries,
where consumption of livestock products is
growing only slowly or stagnating. With low
or no population growth, markets are satu-
rated in most developed countries. Here, to-
tal meat production increased by only 22%
between 1980 and 2004. Ruminant meat pro-
duction actually declined by 7%, and that of
poultry and pigs increased by 42%. As a result,
the share of production of poultry and pigs has
gone up from 59% to 69% of total meat pro-
duction. Among the monogastrics, poultry is
the commodity with the highest growth rate
across all regions. One reason for this, apart
from very favorable feed conversion, is that
poultry is a meat type acceptable to most re-
ligious and cultural groups.
Today, the livestock sector is a major land
user, spanning more than 3.9 billion hectares,
equivalent to about 30% of the world’s sur-
face land area. The intensity with which the
sector uses land is, however, extremely vari-
able. Of the 3.9 billion hectares, 0.5 are used
for feed crops, generally intensively managed.
Another 1.4 billion hectares are highly pro-
ductive pasture, and the remaining 2 billion
hectares are extensive pastures with relatively
low productivity. This puts the livestock sec-
tor as the largest user of agricultural land, ac-
counting for about 78% of agricultural land,
including as much as 33% of the cropland.
3. THE ROLE OF EXTENSIVE
LIVESTOCK PRODUCTION
Extensive livestock production affects natural
resources through its land use and associated
land-use change, which lead to land degrada-
tion, affecting natural resource cycles in a va-
riety of ways. Extensive livestock production
systems and their corresponding land use are
strongly associated with ruminant livestock
species. The digestive system of ruminants
also directly contributes to modifications of
the C and N cycles, whereas modifications of
the latter affect the water cycle.
3.1. Carbon
Livestock’s role in global C cycles is mainly
determined by the extensive current land
use through pastures, their expansion at
the expense of forests, and their degrada-
tion. Furthermore, livestock release impor-
tant amounts of methane (CH
4
).
Savannah burning. Savannah burning is
often associated with extensive livestock
farming and has important effects on the
global biological C cycle. Burning is common
worldwide in establishing and managing pas-
tures. Fire removes ungrazed grass, straw, and
litter; stimulates fresh growth; and can control
the density of woody plants (trees and shrubs).
As many grass species are more fire tolerant
than tree species (especially seedlings and
saplings), burning can determine the balance
between grass cover and ligneous vegetation.
Fire stimulates the growth of perennial
grasses in savannahs and provides nutritious
regrowth for livestock. Controlled burning
prevents uncontrolled and possibly more de-
structive fires and consumes the combustible
lower layer at an appropriate humidity stage.
The environmental consequences of
rangeland and grassland fires depend on the
environmental context and conditions of
burning. In tropical savannah areas, burning
has significant environmental impact because
of the large area affected and the relatively
low level of control. In 2000, burning affected
274 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
some 4 million km
2
. More than two thirds of
this occurred in the tropics and subtropics (5).
Globally about three quarters of this burning
took place outside forests. Savannah burning
represented some 85% of the area burned
in Latin American fires, 60% in Africa, and
nearly 80% in Australia (5).
Usually, savannah burning is not consid-
ered to result in net CO
2
emissions, because
emitted amounts of carbon dioxide (CO
2
) re-
leased in burning are recaptured in grass re-
growth. Apart from CO
2
, biomass burning
releases important amounts of other globally
relevant trace gases (NO
x
, CO, and CH
4
) and
aerosols (6, 7). Many of the emitted elements
lead to the production of tropospheric ozone
(8, 9), which is another important greenhouse
gas influencing the atmosphere’s oxidizing
capacity, and bromine, also released in signif-
icant amounts from savannah fires, decreases
stratospheric ozone (8).
Smoke plumes may be redistributed
locally, transported throughout the lower tro-
posphere, or entrained in large-scale circu-
lation patterns in the mid and upper tropo-
sphere. Often, fires in convection areas take
the elements high into the atmosphere, cre-
ating increased potential for climate change.
Satellite observations have found large ar-
eas with high O
3
and CO levels over Africa,
South America, and the tropical Atlantic and
Indian Oceans (10). Aerosols produced by the
burning of pasture biomass dominate the at-
mospheric concentration of aerosols over the
Amazon Basin and tropical Africa (7, 11).
Concentrations of aerosol particles are highly
seasonal. An obvious peak occurs in the dry
(burning) season, which contributes to cool-
ing both through increasing atmospheric scat-
tering of incoming light and the supply of
cloud condensation nuclei. High concentra-
tions of cloud condensation nuclei from the
burning of biomass stimulate rainfall produc-
tion and affect large-scale climate dynamics
(12).
Desertification of pastures. In contrast to
pasture burning, desertification of grazing
land causes a net loss of C to the atmosphere.
Desertification reduces productivity and veg-
etation cover and also changes C and nutri-
ent cycles and stocks. Although changes in
aboveground biomass and C stocks are of-
ten small, total soil C usually declines. Asner
et al. (13) found in Argentina that desertifica-
tion resulted in little change in woody cover,
but soil organic C declined by 25% to 80%
in areas with long-term grazing. Soil erosion
accounts for part of this loss, but the majority
stems from the nonrenewal of decaying or-
ganic matter stocks, leading to a significant
net emission of CO
2
.
Lal (14) estimated the C loss resulting from
desertification. Assuming a loss of 8–12 tonnes
of soil C per hectare (15) on a desertified land
area of 1 billion hectares (16), the total historic
loss would amount to 8–12 billion tonnes of
soil C. Similarly, degradation of aboveground
vegetation has led to an estimated C loss of
10–16 tonnes per hectare—a historic total of
10–16 billion tonnes. Thus, the total C loss as
a consequence of desertification may be 18–
28 billion tonnes of C (17).
Livestock occupy about two thirds of
the global dry land area, and the rate of
desertification has been estimated to be
higher for grazing land than for other land
uses [3.2 million hectares per year against
2.5 million hectares per year for cropland
(16)]. Considering only soil C loss (i.e., about
10 tonnes of C per hectare), CO
2
emissions
from the desertification of pastures amount
to 100 million tonnes of CO
2
per year.
Another, largely unknown, influence on
the fate of soil C is the feedback effect of
climate change. In higher-latitude cropland
zones, global warming is expected to increase
yields by virtue of longer growing seasons
and CO
2
fertilization. At the same time,
however, global warming may also accelerate
decomposition of C already stored in soils
(18, 19). Although there are large uncer-
tainties, van Ginkel et al. (20) estimate the
magnitude of this effect, at current rates
of increase of CO
2
in the atmosphere, at
a net absorption of 0.036 tonnes of C per
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 275
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
hectare per year in temperate grassland, after
deducting the effect of rising temperature
on decomposition. Recent research indicates
that the magnitude of the temperature rise
on the acceleration of decay may be stronger,
with already very significant net losses over
the past decades in temperate regions (21, 22).
Pasture expansion into forest. Livestock’s
role in the deforestation process is of partic-
ular importance in Latin America where the
the largest net loss of forests and resulting
C fluxes occur. Latin America is the world
region where expansion of pasture and arable
land for feedcrops is strongest, mostly at the
expense of forest area. Wassenaar et al. (23)
showed that most of the cleared area ends
up as pasture and identified large areas where
livestock ranching is probably a primary
motive for clearing. The conversion of forest
into pasture releases considerable amounts of
C into the atmosphere, particularly when the
area is not logged but simply burned. Cleared
patches may go through several changes of
land use. Over the 2000–2010 period, the
pasture areas in Latin America are projected
to expand into forest by an annual average
of 2.4 million hectares—equivalent to some
65% of expected deforestation.
Forest clearing produces a complex pat-
tern of net fluxes that change direction over
time; the calculation of related C fluxes is
the most complex of the emissions inventory
components. Responses of biological systems
vary over time. The Intergovernmental Panel
on Climate Change (IPCC) (24) estimated
the average annual flux owing to tropical de-
forestation for the decade 1980 to 1989 at
1.6 ± 1.0 billion tonnes C as CO
2
. Only about
50% to 60% of the C released from forest
conversion in any one year was a result of
the conversion and subsequent biomass burn-
ing in that year. The remainder were delayed
emissions resulting from oxidation of biomass
harvested in previous years (25).
On the basis of the assumption that forests
are completely converted into climatically
equivalent grasslands and croplands (24,
p. 192) and combining changes in C density
of both vegetation and soil in the year of
change, emissions from conversion of forests
to pastures can be estimated at approximately
1.7 billion tonnes of CO
2
per year. Although
it takes more than a year to reach this
new status because of inherited or delayed
emissions, the resulting emission estimate
would not change much as the transformation
process is continuous.
In addition to producing CO
2
emissions,
land conversion also results in emissions of
other gases. For example, Mosier et al. (26)
noted that upon conversion of forest to graz-
ing land, CH
4
oxidation by soil microorgan-
isms is typically greatly reduced, and grazing
lands may even become net sources in situa-
tions where soil compaction from cattle traffic
limits gas diffusion.
Enteric fermentation. On the basis of ani-
mal numbers and liveweights, the total live-
stock biomass amounts to some 0.7 billion
tonnes (2). According to the function estab-
lished by Muller & Schneider (27; cited by
28), applied to standing stocks per country and
species (with country-specific liveweight), the
CO
2
from the respiratory process of livestock
amounts to some 3 billion tonnes of CO
2
.
Emissions from livestock respiration are part
of a rapidly cycling biological system, where
the plant matter consumed was itself created
through the conversion of atmospheric CO
2
into organic compounds. Livestock respira-
tion is therefore not considered to be a net
source.
However, the specific characteristics of ru-
minant physiology lead to net gaseous emis-
sions of C as CH
4
. Livestock are globally
the most important source of anthropogenic
CH
4
emissions. However, there are signif-
icant spatial variations in CH
4
emissions
from enteric fermentation. In Brazil, CH
4
emissions from enteric fermentation totaled
9.4 million tonnes in 1994–93% of agricul-
tural emissions and 72% of the country’s total
emissions of CH
4
. Over 80% of this origi-
nated from beef cattle (29), and the Brazilian
276 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
cattle population has increased with some
25% since (2). In the United States, CH
4
from enteric fermentation totaled 5.5 million
tonnes in 2002, again overwhelmingly orig-
inating from beef and dairy cattle. This was
71% of all agricultural emissions and 19% of
the country’s total emissions (30).
Levels of CH
4
emissions are affected by
energy intake and several other animal and
diet factors including quantity and quality of
feed, animal body weight, age and amount of
exercise. Therefore, assessing CH
4
emission
from enteric fermentation in any particular
country requires a detailed description of the
livestock population (species, age, and pro-
ductivity categories), combined with informa-
tion on the daily feed intake and the feed’s
CH
4
conversion rate. As many countries do
not possess such detailed information, an ap-
proach that is based on standard emission fac-
tors is generally used in emission reporting.
CH
4
emissions from enteric fermentation
change as production systems intensify and
move toward higher feed use and increased
productivity. We have attempted a global es-
timate of total CH
4
emissions from enteric
fermentation in the livestock sector using re-
gional and production system-specific emis-
sion factors [as outlined by Steinfeld et al.
(31)]. Total global emissions of CH
4
from en-
teric fermentation are estimated at 86 million
tonnes CH
4
annually, which is roughly in line
with the global estimate by the United States
Environmental Protection Agency (32) of
about 80 million tonnes of CH
4
annually. The
distribution across regions, species, and pro-
duction systems of such CH
4
emission is given
in Table 2 (31). The relative global impor-
tance of mixed systems compared to grazing
systems reflects the fact that about two thirds
of all ruminants are held in mixed systems.
3.2. Nitrogen
Diatomic nitrogen (N
2
) in the atmosphere
is a large and stable pool of N. In con-
trast, the modest capability of natural ecosys-
tems to drive the N cycle constituted a major
Table 2 Global methane emissions from enteric fermentation in 2004
Emissions (million tonnes CH
4
per year by source)
Region/country Dairy cattle Other cattle Buffalo Sheep and goats Pigs Total
Sub-Saharan Africa 2.30 7.47 0.00 1.82 0.02 11.61
Asia
a
0.84 3.83 2.40 0.88 0.07 8.02
India 1.70 3.94 5.25 0.91 0.01 11.82
China 0.49 5.12 1.25 1.51 0.48 8.85
Central and South America 3.36 17.09 0.06 0.58 0.08 21.17
West Asia and North Africa 0.98 1.16 0.24 1.20 0.00 3.58
North America 1.02 3.85 0.00 0.06 0.11 5.05
Western Europe 2.19 2.31 0.01 0.98 0.20 5.70
Oceania and Japan 0.71 1.80 0.00 0.73 0.02 3.26
Eastern Europe and the Commonwealth
of Independent States
1.99 2.96 0.02 0.59 0.10 5.66
Other developed countries 0.11 0.62 0.00 0.18 0.00 0.91
Total 15.69 50.16 9.23 9.44 1.11 85.63
Livestock production system
Grazing 4.73 21.89 0.00 2.95 0.00 29.58
Mixed 10.96 27.53 9.23 6.50 0.80 55.02
Industrial 0.00 0.73 0.00 0.00 0.30 1.04
a
Excludes China and India.
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 277
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
hurdle in satisfying the food needs of growing
populations (33). Extensive livestock, and par-
ticularly ruminant, production has tradition-
ally contributed to alleviating this problem by
making reactive N fixed in grass and other
fodder available to humans through animal
products. But in doing so, it often impover-
ishes natural ecosystems, particularly because
livestock use the scarce resource with a low ef-
ficiency. In fact, the majority of N they ingest
enters the so-called N cascade (34) by which
N is transported downstream or downwind
in different forms to a series of temporary
reservoirs.
N assimilation efficiency varies consid-
erably among different animal species and
products. According to estimates by van der
Hoek (35), average N assimilation efficiency
is around 20% for pigs and 34% for poultry.
For the United States, Smil (36) calculated the
protein conversion efficiency of intensively
produced dairy products at 40%, whereas that
of beef cattle is only 5%. The low N assimila-
tion efficiency of extensively held ruminants,
in particular cattle, is partly inherent to large
animals with long gestation periods and a high
basal metabolic rate. But the global cattle herd
also comprises a large draught animal popula-
tion whose task is to provide energy, not pro-
tein. For example, in 1995, cattle and horses
still accounted for 25% of China’s agricultural
energy supply (37). In addition, in many areas
of the world, grazing animals are fed at bare
maintenance level, consuming without pro-
ducing much.
As a result, a huge amount of N is re-
turned to the environment through animal
excretions. When directly deposited on pas-
ture or crop fields, some of the reactive N
reenters the plant production cycle. A large
share leaves the system through gaseous emis-
sion, and volatilization, and, to a lesser ex-
tent, leaching and erosion. We focus here on
the two dominant reactive forms in which N
is lost, nitrous oxide (N
2
O) and ammonia,
both associated with an important negative
environmental impact, contributing to global
warming and air pollution, respectively.
N lost to the atmosphere following
deposition of manure. Excreta directly de-
posited on land have high N loss rates
through substantial ammonia volatilization.
Wide variations in the quality of forages con-
sumed by ruminants and in environmental
conditions make N emissions from manure
directly deposited on land difficult to quan-
tify. The Food and Agriculture Organization
of the United Nations (FAO) & International
Fertilizer Industry Association (IFA) (38) es-
timate the N loss via NH
3
volatilization from
animal manure, after application, to be 23%
worldwide. Smil (39) estimates this loss to be
at least 15% to 20%.
The IPCC proposes a standard N loss
fraction from ammonia volatilization of 20%,
without differentiating between applied and
directly deposited manure. Considering the
substantial N loss from volatilization during
storage (see the section on intensive produc-
tion), total ammonia volatilization following
excretion can be estimated at ∼40%. This
rate can be applied to directly deposited ma-
nure, assuming that the lower share of N in
urine in tropical land-based systems is com-
pensated by higher temperature. We estimate
that in the mid-1990s ∼30 million tonnes of
N were directly deposited on land by an-
imals in the more extensive systems, pro-
ducing an NH
3
volatilization loss of some
12 million tonnes N. In addition, the postap-
plication loss of managed animal manure is
about 8 million tonnes N (38), resulting in a
total ammonia volatilization N loss from an-
imal manure on land of ∼20 million tonnes
N. These figures continue to grow. Even
taking the very conservative IPCC ammo-
nia volatilization loss fraction of 20% and
subtracting manure used for fuel results in
an estimated NH
3
volatilization loss follow-
ing manure application/deposition of some
25 million tonnes N in 2004.
With regard to N
2
O, the soil emissions
originating from the remaining external N
input, which is after subtraction of ammo-
nia volatilization, depend on a variety of fac-
tors, particularly soil water-filled pore space,
278 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
organic C availability, pH, soil temperature,
plant/crop uptake rate, and rainfall character-
istics (26). However, because of the complex
interaction and the highly uncertain result-
ing N
2
O flux, the revised IPCC guidelines are
based on N inputs only and do not consider
soil characteristics. Despite this uncertainty,
manure-induced soil emissions are clearly the
largest livestock source of N
2
O worldwide.
Emission fluxes from animal grazing (unman-
aged waste, direct emission) and from the use
of animal waste as crop fertilizer are of a
comparable magnitude. The grazing-derived
N
2
O emissions are in the range of 0.002–
0.098 kg N
2
O-N/kg N excreted, whereas the
default emission factor used for fertilizer use
is set at 0.0125 kg N
2
O-N/kg N. Nearly all
data pertain to temperate areas and to inten-
sively managed grasslands. Here, the N con-
tent of dung, and especially urine, is higher
than from less intensively managed grasslands
in the tropics or subtropics. It is not known
to what extent this compensates for the en-
hanced emissions in the more phosphorus-
limited tropical ecosystems.
Emissions from applied manure need to
be calculated separately. The FAO/IFA study
(38) estimates the N
2
O loss rate from applied
manure at 0.6%, i.e., lower than most mineral
N fertilizers, resulting in an animal manure
soil N
2
O loss in the mid-1990s of 0.2 million
tonnes N. Following the IPCC methodology
would increase this to 0.3 million tonnes N.
In the mid-1990s, animal excreta directly
deposited on pastures loaded approximately
30 million tonnes N on land in the more ex-
tensive systems. Applying the IPCC “overall
reasonable average emission factor” (0.02 kg
N
2
O-N/kg of N excreted) to this total re-
sults in an animal manure soil N
2
O loss of
0.6 million tonnes of N. This gives a total
N
2
O emission of about 0.9 million tonnes N
in the mid-1990s.
Applying the IPCC methodology to the
current estimates of the livestock production
system and animal numbers (31) results in an
overall loss of 1.7 million tonnes N per year in
the form of direct animal manure soil N
2
O.
Of this, 0.6 million tonnes derive from ex-
tensive grazing systems, 1.0 million tonnes
from mixed systems (of which roughly 0.9 mil-
lion tonnes originate from the more extensive
mixed systems), and 0.1 million tonnes from
industrial production systems.
4. THE ROLE OF INTENSIVE
LIVESTOCK PRODUCTION
The use of marketable and higher quality feed
items in the fast-growing intensive livestock
sector reduces land requirement per unit of
output because the feed items themselves are
intensively produced too. But in order to sus-
tain a high quality, intensive animal produc-
tion, this smaller land area needs to be sup-
plied with large amounts of external inputs
like nutrients, water, and energy. The mo-
bilization of these resources and the losses
associated with their intensive use affect the
respective natural cycles. Because of the in-
herent inefficiency of animals in transform-
ing even the resulting high-grade products,
subsequent losses occur, which are not only
important in absolute terms but also because
of their spatially concentrated nature.
4.1. Carbon
Intensive livestock production systems have
higher requirements for fossil fuel at all stages
of the production process. However, com-
pared to extensive modes of production, they
have less impact on land-use changes, except
through the expansion of crop land for feed
production.
Fossil fuel. Significant additions of CO
2
to
the atmosphere result from fossil fuel use
at the different stages of intensive animal
production.
Use of fossil fuel to manufacture fertilizer.
A large share of the world’s crop produc-
tion is fed to animals. This includes about
670 million tons of cereals and large amounts
of agro-industrial by-products such as brans
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 279
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
Table 3 Chemical fertilizer nitrogen used for feed and
pastures in selected countries
Country
Share of total N
consumption (%)
Absolute amount
(1000 tonnes/year)
United States 51 4697
China 16 2998
France
a
52 1317
Germany
a
62 1247
Canada 55 897
United Kingdom
a
70 887
Brazil 40 678
Spain 42 491
Mexico 20 263
Turkey 17 262
Argentina 29 126
a
Countries with a considerable amount of N fertilized grassland.
and oil cakes (2). Mineral N fertilizer is ap-
plied to much of the corresponding cropland,
and about 97% of N fertilizers are derived
from synthetically produced ammonia via the
energy-intensive Haber-Bosch process.
Combining fertilizer use by crop for the
year 1997 (40) with the fraction of these crops
used for feed in major N fertilizer consuming
countries (41) shows that animal production
accounts for a very substantial share of fertil-
izer consumption. Table 3 (40, 41) gives ex-
amples for selected countries. Except for the
Western European countries, production and
consumption of mineral fertilizer are increas-
ing in these countries. This high proportion
of N fertilizer for feed crops is largely owing
to maize, which not only covers large areas in
temperate and tropical climates but also de-
mands high doses of N fertilizer. More than
half of total maize production is used as feed.
Use of N fertilizer for maize and other ani-
mal feed crops is especially high in N-deficit
areas such as North America, Southeast Asia,
and Western Europe. In fact in 18 of the
66 maize producing countries analyzed, maize
is the crop with the highest N fertilizer con-
sumption (40). In 41 of these 66 countries,
maize is among the first three crops in terms
of N fertilizer consumption. The projected
production of maize in these countries shows
that its area generally expands at a rate inferior
to that of production, suggesting an enhanced
yield, partially owing to an increase in fertil-
izer consumption (41).
Other feed crops such as barley and
sorghum are also important consumers of
chemical N fertilizer. Despite the fact that
some oil crops are associated with N-fixing
organisms themselves, their intensive produc-
tion also often makes use of N fertilizer. Such
crops predominantly used as animal feed, in-
cluding rapeseed, soybean and sunflower, gar-
ner considerable amounts of N fertilizer: 20%
of Argentina’s total N fertilizer consump-
tion is applied to production of such crops,
110,000 tonnes of N fertilizer (for soybean
alone) in Brazil, and over 1.3 million tonnes
in China. In addition, pastures receive a con-
siderable amount of N fertilizer in a number of
countries.
The countries of Table 3 together repre-
sent the vast majority of the world’s N fertil-
izer use for feed production and spread about
14 million tonnes of N fertilizer per year
for the animal food chain. Adding Common-
wealth of Independent States and Oceania, the
total amounts to ∼20% of the annual 80 mil-
lion tonnes of N fertilizer consumed world-
wide. Furthermore, considering fertilizer use
that can be attributed to by-products other
than oil cakes, in particular brans, may well
take the total up to some 25%.
On the basis of these figures, the corre-
sponding emission of CO
2
can be estimated.
Energy requirement in modern natural gas–
based systems varies between 33 and 44 gi-
gajoules (GJ) per tonne of ammonia. Taking
into consideration additional energy use in
packaging, transport, and application of fer-
tilizer [estimated to represent an additional
cost of at least 10% (42)], an upper limit of
40 GJ per tonne has been applied here. En-
ergy use in the case of China is considered
to be some 25% higher, i.e., 50 GJ per tonne
of ammonia, because not only its N fertilizer
production is based on coal, but it is mostly
produced in small- and medium-sized, rela-
tively energy-inefficient, plants (43). Taking
280 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
the IPCC emission factors for coal in China
[26 tonnes of C per terajoule (TJ)] and for
natural gas elsewhere (17 tonnes C/TJ), and
estimating C 100% oxidized, results in an es-
timated annual emission of CO
2
of more than
40 million tonnes at this initial stage of the
animal food chain.
On-farm fossil fuel use. In intensive pro-
duction systems, the bulk of the energy is
spent on production of feed, either forage for
ruminants or concentrated feed for poultry
or pigs. Like the energy used for fertilizer,
important amounts of energy are spent on
seeds, herbicides and pesticides, diesel for
machinery, and electricity. On the basis
of data presented by Ryan & Tiffany (44)
on Minnesota, a major U.S. state in terms
of agricultural production with a focus on
intensive livestock production, the bulk of
Minnesota’s on-farm CO
2
emissions from
energy use is related to feed production and
exceeds the emissions associated with N fer-
tilizer use. Using the average maize fertilizer
application (150 kg N per hectare for maize
in the United States) results in emissions for
Minnesota maize of about one million tonnes
of CO
2
, compared with 1.26 million tonnes
of CO
2
from on-farm energy use for corn
production. These estimates are obtained
from combining data by Ryan & Tiffany (44)
with efficiency and emission factors from the
United States’ Common Reporting Format
report submitted to the United Nations
Framework Climate Change Convention
(UNFCCC) in 2005. At least half the CO
2
emissions of the two dominant commodities
and CO
2
sources in Minnesota (maize and
soybean) can be attributed to the (intensive)
livestock sector. Taken together, feed pro-
duction and pig and dairy operations make
the livestock sector by far the largest source
of agricultural CO
2
emissions in Minnesota.
In the absence of similar estimates repre-
sentative of other world regions, it is not pos-
sible to provide a reliable quantification of the
global CO
2
emissions that can be attributed to
on-farm fossil fuel use by the intensive live-
stock sector. The energy intensity of produc-
tion as well as the source of this energy vary
widely. A rough indication of the fossil fuel
use–related emissions from intensive systems
can, nevertheless, be obtained by considering
that the expected lower energy need for feed
production at lower latitudes (lower energy
need for corn drying, for example) and the of-
ten lower level of mechanization, which are
compensated overall by a lower energy use
efficiency and a lower share of relatively low
CO
2
-emitting sources (natural gas and elec-
tricity). Minnesota figures can then be com-
bined with global feed production and live-
stock populations in intensive systems. As a
conservative estimate, CO
2
emissions induced
by on-farm fossil fuel use for feed produc-
tion may be 50% higher than that from feed-
dedicated N fertilizer production, i.e., some
60 million tonnes CO
2
globally. To this, farm
emissions related directly to livestock rearing
need to be added, which can be estimated at
roughly 30 million tonnes of CO
2
.
Fossil fuel for processing and transport. A
number of studies assess the energy expendi-
ture for processing animals for meat and other
products (45). The wide variation among en-
terprises makes generalizations difficult. Es-
timating related emissions is still more haz-
ardous as it is highly uncertain what the
source of this energy is and how this varies
throughout the world. Among livestock prod-
ucts, large amounts of energy are used to
pasteurize milk and transform it into cheese
and dried milk. In the above cited case of
Minnesota, this makes the dairy sector re-
sponsible for the second highest CO
2
emis-
sions from food processing. The largest emis-
sions result from soybean processing, caused
by energy-intensive physical and chemical oil
extraction. Considering the value fractions
(46) of these two commodities two thirds
of these soy-processing emissions can be at-
tributed to the livestock sector. Extrapolat-
ing to country level results in a total animal
product and feed processing-related emission
of the United States in the order of a few
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 281
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
million tonnes CO
2
. Therefore, the proba-
ble order of magnitude for the emission level
related to global animal-product processing
would be several tens of million tonnes of
CO
2
.
Although rough indications, these figures
show that emissions from energy use by
processing are much higher than that from
transport in the animal food chain. Transport
occurs mainly at two key stages: that of
feed to animal production sites and that of
animal products to consumer markets. Large
amounts of bulky raw ingredients for concen-
trate feed are shipped around the world. One
of the most notable long-distance feed trade
flows is for soybean, which is also the largest
traded volume among feed ingredients,
as well as the one showing the strongest
increase. Among soybean trade flows, the
one from Brazil to Europe is of a particularly
important volume. Cederberg & Flysj
¨
o (47)
studied the energy cost of shipping soybean
cake from the Mato Grosso to Swedish dairy
farms. Applying the resulting energy need
to the annual soybean cake shipped from
Brazil to Europe, combined with the IPCC
emission factor for ocean vessels, results in an
annual emission of approximately 32 thou-
sand tonnes of CO
2
. Similarly, we combined
traded volumes of pig, poultry, and bovine
meat (2, accessed December 2005) with
respective distances, vessel capacities and
speeds, fuel use of main engine and auxiliary
power generators for refrigeration, and their
respective emission factors (48). The selected
flows represent some 60% of international
meat trade. Annually, they produce some
500 thousand tonnes of CO
2
. This represents
more than 60% of total CO
2
emissions in-
duced by meat-related sea transport, because
the trade flow selection is biased toward the
long-distance exchange. By contrast, surface
transport to and from the harbor has not been
considered. Assuming, for simplicity, that
the latter two effects compensate each other,
the total annual meat transport-induced
CO
2
emission would be in the order of 800–
850 thousand tonnes of CO
2
.
Cropland expansion into forest. Driven by
the rising demand for animal products, the
fast-growing intensive livestock sector drives,
in turn, important increases in feed produc-
tion. This rise in the demand for feed is not
only met by intensification, but also by ex-
pansion of the corresponding cropland. Large
losses of C to the atmosphere occur where
this expansion occurs at the expense of forest.
Wassenaar et al. (23) showed that, although
less than pasture, large areas of cropland too
will replace forest in the Neotropics. Much
of this land will produce soybean primar-
ily destined to feed use (49, 50). If we con-
servatively assume that half of the projected
cropland expansion into tropical forests in
Bolivia and Brazil can be attributed to provid-
ing feed for the livestock sector; this results
in an additional annual deforestation for feed
of over 0.5 million hectares per year. Simi-
lar changes in temperate forest in Argentina
are not considered. Applying the same proce-
dure as for pasture expansion into forest (see
Section 3, above) results in an estimated cor-
responding emission of 0.7 billion tonnes of
CO
2
per year.
Carbon loss from soils cultivated to
produce feed. Most of the organic C loss
from soil occurs at the original conversion of
natural cover into managed land. However,
soils are the largest reservoir of the terres-
trial C cycle and continue to release signif-
icant amounts of C for a very long time af-
ter conversion. It is impossible to evaluate the
shares attributable to inherited emissions and
to management practices at a broad scale be-
cause land use and management change much
faster. Under appropriate management prac-
tices (such as zero tillage), agricultural soils
can conserve a large part of the original C
content. Losses beyond this level can be con-
sidered to be a result of management prac-
tices. Excluding emissions originating from
crop residues, Sauv
´
e et al. (51) found annual
loss rates of about 100 kg CO
2
per hectare on
permanently cultivated temperate brown soils
under conventional cultivation practices. The
282 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
large area in temperate regions that produces
coarse grains and oil crops for feed is under
large-scale intensive management, which is
still dominated by conventional tillage prac-
tices. Using the above figure as an approxi-
mation of the average loss rate for temper-
ate climate soils with moderate organic matter
content, the approximately 1.8 million km
2
of
arable land cultivated with maize, wheat, and
soybean for feed adds an annual CO
2
flux in
the order of 18 million tonnes to the livestock
balance.
Tropical soils have lower average C
content (24, p. 192), and therefore lower emis-
sions. By contrast, the considerable expan-
sion of large-scale cropping of feeds, not only
into uncultivated areas, but also into previ-
ous pastureland or subsistence cropping, may
increase CO
2
emission. These, however, are
impossible to gauge.
In addition, practices such as soil liming
contribute to net emissions. Soil liming is a
common practice in more intensively culti-
vated tropical areas because of soil acidity.
Brazil, for example, estimated its CO
2
emis-
sions owing to soil liming at 8.99 million
tonnes in 1994 (52), and these have most
probably increased since then. To the ex-
tent that these emissions concern cropland
for feed production, they can be attributed
to the livestock sector. Comparing reported
emissions from liming from national com-
munications of various tropical countries to
the UNFCCC with the importance of feed
production in those countries adds another
10 million tonnes CO
2
.
Methane released from animal manure.
As discussed above, the contribution of in-
tensive production systems to CH
4
emissions
from enteric fermentation is comparatively
low. For CH
4
released from animal manure,
the situation is the opposite; manure directly
deposited on land, or otherwise handled in
a dry form, does not produce significant
amounts of CH
4
. But substantial amounts are
released from the anaerobic decomposition
of organic material in livestock manure. This
occurs mostly when manure is managed in liq-
uid form, such as in lagoons or holding tanks.
Lagoon systems are typical for most large-
scale pig operations over most of the world
(except in Europe). These systems are also
used in large dairy operations in North
America and in some developing countries,
for example, Brazil.
CH
4
emissions from livestock manure are
influenced by a number of factors that af-
fect the growth of the bacteria responsible for
CH
4
formation, including ambient tempera-
ture, moisture, and storage time. The amount
of CH
4
produced also depends on the energy
content of manure, which is determined to
a large extent by livestock diet. Higher en-
ergy feed produces larger amounts of manure
with more volatile solids, increasing the sub-
strate from which CH
4
is produced. How-
ever, to a certain extent this is offset by higher
digestibility of feed, and thus less wasted
energy.
The default emission factors currently
used in country reporting to the UNFCCC
do not reflect the strong changes in the global
livestock sector. For example, Brazil’s country
report to the UNFCCC (52) mentions a CH
4
emission from manure of 0.38 million tonnes
in 1994, mainly from dairy and beef cattle.
However, Brazil also has a very strong indus-
trial pig production sector, where an estimated
95% of manure is held in open tanks for sev-
eral months before application (EMBRAPA,
personal communication). Hence, Steinfeld
et al. (31) reassessed and applied emission fac-
tors to the animal population figures specific
to each production system. The total annual
global emission of CH
4
from manure decom-
position is estimated at 17.5 million tonnes of
CH
4
; this is substantially higher than previous
estimates.
China has the largest country-level CH
4
emission from manure in the world, mainly
from pigs. At a global level, emissions from
pig manure represent almost half of total live-
stock manure emissions. Just over a quarter of
the total CH
4
emission from managed manure
originates from industrial systems, nearly
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 283
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
exclusively from pigs (85%). About 70%
originates from mixed systems, mainly from
intensive large ruminant operations in Europe
and North America and from small- and
middle-scale pig farms in China.
4.2. Nitrogen
Although livestock-induced atmospheric N
emissions in extensive systems lead to a re-
duction of an essential nutrient in often al-
ready fragile ecosystems, the situation is again
quite the opposite on the intensive side. Large
additions of N, disrupting ecosystem equi-
librium, result from chemical N fertilizer
use and animal concentrations. As a result
not only the air is loaded with often pol-
luting forms of N, but also soil and water
resources.
N emissions from feed-related fertilizer.
It has been estimated that humans have al-
ready doubled the natural rate of N enter-
ing the land-based N cycle, and this rate is
continuing to grow (53). Synthetic fertilizers
now provide about 40% of all the N taken up
by crops (54). Although less so than animals,
crop production too uses the additional re-
source at a rather low efficiency of about 50%.
The fate of the remainder in the N cascade
(34) is often hard to predict. Chemical pro-
cesses involving nitrous oxides are particularly
complex (26). Soil N
2
O emissions are, among
others, determined by temperature and soil
moisture. They also depend on the chemical
balance of the soil; whereas terrestrial ecosys-
tems in the Northern Hemisphere are lim-
ited by N, tropical ecosystems, currently an
important source of N
2
O and NO, are often
limited by phosphorus. N fertilizer inputs into
these phosphorus-limited ecosystems gener-
ate NO and N
2
O fluxes that are 10 to 100
times greater than the same fertilizer addition
to N-limited ecosystems (55). Chemical form,
mode, and timing of application are also im-
portant variables.
Average N losses as ammonia from syn-
thetic fertilizer use are more than twice as high
(18%) in developing countries than in devel-
oped and transition countries (7%). Most of
this difference in loss rates is caused by higher
temperatures and the dominant use of urea
and ammonium bicarbonate in the developing
world. In developing countries, about 50% of
the N fertilizer used is in the form of urea
(38). Bouwman et al. (56) estimate that NH
3
emission losses from urea may be 25% in trop-
ical regions and 15% in temperate climates. In
China, 40% to 50% of the N fertilizer used is
in the form of ammonium bicarbonate, which
is highly volatile. The NH
3
loss from ammo-
nium bicarbonate may be 30% on average in
the tropics and 20% in temperate zones. By
contrast, the NH
3
loss from injected anhy-
drous ammonia, widely used in the United
States, is only 4% (56).
In Section 4.1, we estimated that 20% to
25% of mineral fertilizer use (about 20 million
tonnes N) can be ascribed to feed production
for the intensive livestock sector. Assuming
an average mineral fertilizer NH
3
volatiliza-
tion loss rate of 14% (38), livestock produc-
tion causes a global NH
3
volatilization from
mineral fertilizer of 3.1 million tonnes NH
3
-
N per year.
N
2
O emissions for major world regions
can be estimated using the FAO/IFA (38)
model. N
2
O emissions amount to 1.25% ±
1% of the N applied. This estimate is the av-
erage for all fertilizer types, as proposed by
Bouwman (57) and adopted by IPCC (48).
Under the same assumptions as for NH
3
above, livestock production can be consid-
ered responsible for a global N
2
O emission
from mineral fertilizer of 0.2 million tonne
N
2
O-N per year.
There is also N
2
O emission from legu-
minous feedcrops, even though they do not
generally receive N fertilizer because the rhi-
zobia in their root nodules fix N that can
be used by the plant. Studies have demon-
strated that such crops show N
2
O emissions
of the same level as those of fertilized non-
leguminous crops. On the basis of the to-
tal area of soybean and pulses, and the share
of production used for feed, a total of about
284 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
75 million hectares in 2002 (2) released an-
other 0.2 million tonnes of N
2
O-N. Adding
alfalfa and clovers would probably double this
figure, although there are no global estimates
of their cultivated areas. Russelle & Birr (58),
for example, show that soybean and alfalfa to-
gether harvest some 2.9 million tonnes of fixed
N in the Mississippi River Basin, with the N
2
fixation rate of alfalfa being nearly twice as
high as that of soybean (see also the review in
Reference 39). It is therefore plausible that
intensive forms of livestock production are
linked to more than 0.5 million tonnes per
year of total N
2
O-N emission from soils un-
der leguminous crops.
Mineral fertilizer N loss to aquatic sources
and subsequent emissions. N uptake rate in
global crop production is estimated at about
50% on average (35, 39). Smil (39) attempted
to derive a global estimate of N losses from
fertilized cropland. He estimates that glob-
ally, in the mid-1990s, about 37 million tonnes
N were exported from cropland through ni-
trate leaching (17 million tonnes N) and soil
erosion (20 tonnes N). In addition, a frac-
tion of the volatilized ammonia from mineral
fertilizer N (11 million tonnes N per year)
finally also reaches surface waters after depo-
sition (3 million tonnes N per year). This N
is gradually denitrified in subsequent reser-
voirs of the N cascade (34). The resulting en-
richment of aquatic ecosystems with reactive
N causes emissions not only of N
2
, but also
of N
2
O. Galloway et al. (33) estimate the to-
tal anthropogenic N
2
O emission from aquatic
reservoirs at about 1.5 million tonnes N, orig-
inating from a total of some 59 million tonnes
N transported to inland waters and coastal ar-
eas. Feed and forage production induces a loss
of N to aquatic sources of some 8 to 10 million
tonnes per year, assuming that such losses are
in line with N fertilization shares of feed and
forage production (some 20% to 25% of the
world total, see the carbon section). Applying
the overall rate of anthropogenic aquatic N
2
O
emissions (1.5/59) to the livestock, induced
mineral fertilizer N loss to aquatic reservoirs
results in livestock-induced emissions from
aquatic sources of around 0.2 million tonnes
NN
2
O.
N lost to the atmosphere from stored
manure. Because of the low N assimilation
efficiency of livestock, large amounts of N are
lost in the intensive production units. The N
harvested in feed over a wide area is used and
largely lost in a spatially highly concentrated
manner by highly intensive or industrial pro-
duction units with very large animal numbers.
This concentration is often aggravated by ex-
cessively N-rich diets. The environmental ef-
fect of these enormous concentrations of N
depends on the fate of the manure, which is
highly variable: lost from stored manure, ap-
plied in excess to nearby land or in a more
balanced manner over a wider area, sold as
fertilizer, or simply discharged into surface
waters.
Although of variable duration and mode,
storage of manure occurs in all industrial pro-
duction units. For the most part, excreted
N compounds mineralize rapidly during this
first phase of manure management. In urine,
typically over 70% of the N is present as
urea (48). Uric acid is the dominant N com-
pound in poultry excretions. The hydrolysis of
both urea and uric acid to NH
3
/NH
4
+
is very
rapid in urine patches. The magnitude of N
2
O
emissions depends on environmental condi-
tions. N
2
O emissions occur when waste is first
handled aerobically, allowing ammonia or or-
ganic N to be converted to nitrates and nitrites
(nitrification). If an anaerobic stage follows,
nitrates and nitrites are reduced to N
2
, with
intermediate production of N
2
O and nitric
oxide (NO) (denitrification). These emissions
are most likely to occur in dry waste-handling
systems, which have aerobic conditions but
contain pockets of anaerobic conditions ow-
ing to saturation. The amount of N
2
O re-
leased during storage and treatment of animal
wastes also depends on temperature. Unfor-
tunately, there is not enough quantitative data
to establish a relationship between the degree
of aeration and N
2
O emission from slurry
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 285
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
during storage and treatment. When ex-
pressed in N
2
O-N/kg N in the waste (i.e., the
share of N in waste emitted to the atmosphere
as N
2
O), losses from animal waste during stor-
age range from less than 0.0001 kg N
2
O-N/kg
N for slurries to more than 0.15 kg N
2
O-N/kg
N in the pig waste of deep-litter stables. Any
estimation of global manure emission needs
to consider these uncertainties. Expert judge-
ment, on the basis of existing manure manage-
ment in different systems and world regions,
combined with default IPCC emission factors
(31), suggests N
2
O emissions from stored ma-
nure equivalent to 0.7 million tonnes N per
year.
With regard to ammonia, rapid degrada-
tion of urea and uric acid to ammonium leads
to very significant N losses through volatiliza-
tion during storage and treatment of manure.
Although actual emissions are subject to many
factors, particularly the manure management
system and ambient temperature, most of the
NH
3
N volatilizes during storage, before ap-
plication or discharge. On the basis of the ani-
mal population in industrial systems (31), and
their estimated manure production (48), the
current amount of N in the corresponding
animal waste can be estimated at 10 million
tonnes and the corresponding NH
3
volatiliza-
tion from stored manure at 2 million tonnes
N. Estimated volatilization losses from stored
manure in mixed systems are much higher,
though lower on a per animal basis: We esti-
mate that currently over 16 million tonnes N
of ammonia volatilizes annually from mixed
systems; approximately 5 of these originate
from more intensive production systems such
as western dairy operations and medium-size
pig holdings in China. On the one hand,
this N loss reduces emissions from manure
once applied to fields; on the other, it gives
rise to N
2
O emissions further down the N
cascade.
Aquatic N
2
O emissions after manure N
loss. N loss to the atmosphere after ap-
plication on soil was discussed in conjunc-
tion with those following direct deposition in
Section 3.2, the former being clearly much
lower than the latter. Contrary to the situa-
tion in extensive systems, a very large amount
of N leaves the local system through surface
water flows, either resulting from leaching of
excess amounts applied to land or from direct
discharge.
In the mid-1990s, about 25 million tonnes
of N from animal manure remained available
per year for plant uptake in the world’s crop-
lands and in intensively used grasslands af-
ter losses to the atmosphere during storage
and following application and direct deposi-
tion. Uptake depends on the ground cover:
Legume/grass mixtures can take up large
amount of added N, whereas loss from row
crops is generally substantial, and losses from
bare soil are much higher still. We suppose
that N losses from grazing land, through
leaching and erosion, are negligible. Apply-
ing the crop N-use efficiency of 40% to the
remainder of animal manure N applied to
cropland, about 9 or 10 million tonnes N en-
tered the N cascade mostly through water in
the mid-1990s. Applying the N
2
O loss rate
for subsequent N
2
O emission (see the above
section on subsequent emissions from min-
eral fertilizer N loss), an additional 0.2 mil-
lion tonnes N N
2
O are emitted through this
channel. N
2
O emissions of similar size can
be expected to have resulted from the rede-
posited fraction of the volatilized NH
3
from
manure that reached the aquatic reservoirs in
the mid-1990s. Total N
2
O emissions follow-
ing N losses would, therefore, have been in
the order of 0.4 million tonnes N N
2
O per
year in that period.
We have updated these figures for the cur-
rent livestock production system estimates,
using the IPCC methodology for indirect
emissions. The current overall “indirect”
animal manure N
2
O emission following
volatilization and leaching then totals around
1.3 million tonnes N per year. However, this
methodology is beset with high uncertainties.
The majority of N
2
O emissions, or about
0.9 million tonnes N, originates from mixed
systems.
286 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
Table 4 Summary of current impacts on the carbon cycle
Process Impact on C cycle
Contribution from
extensive systems
a
Contribution from
intensive systems
a
N fertilizer production Addition of atmospheric CO
2
— 0.04
On-farm fuel use Addition of atmospheric CO
2
— ∼0.09
Savannah burning Changing carbon distribution in vegetation
Contribution to climate change
Majority of burned
area worldwide
—
Pasture desertification Soil carbon loss
Addition of atmospheric CO
2
∼0.1 —
Deforestation Soil and vegetation carbon loss
Addition of atmospheric CO
2
Changing local carbon cycle
∼1.7 ∼0.7
Soil tillage Soil carbon loss
Addition of atmospheric CO
2
— ∼0.02
Soil liming Addition of atmospheric CO
2
— ∼0.01
Enteric fermentation Addition of atmospheric CH
4
1.6 0.20
Methane from manure Addition of atmospheric CH
4
0.17 0.20
Processing Addition of atmospheric CO
2
— 0.01–0.05
Transport Addition of atmospheric CO
2
— ∼0.001
a
Quantified contributions concern additions to and removals from the atmospheric pool and all are expressed in billion tonnes CO
2
equivalent.
5. SUMMARY OF LIVESTOCK’S
ROLE IN CARBON AND
NITROGEN CYCLES
Tables 4 and 5 summarize the range of
afore-described impacts of the livestock sec-
tor on the C and N cycles. Not all impacts are
quantifiable, either because of their complex
nature, as in the case of Savannah burning,
or because of a lack of information, as in the
case of eutrophication of surface water from
manure ammonia. Figures corresponding to
processes leading to N
2
O emissions comprise
direct as well as indirect emissions. The
figures should be considered as indicative
values, particularly because the role of mixed
Table 5 Summary of current impacts on the nitrogen cycle
Process Impact on N cycle
Estimated
contribution from
extensive systems
a
Estimated contribution
from intensive
systems
a
Mineral fertilizer application Eutrophication of aquatic systems — 8–10
Addition of atmospheric N
2
O — 0.4 (0.2)
Volatilization/deposition of NH
3
— 3.1
Leguminous feed cropping Addition of atmospheric N
2
O — 0.5 (0.2)
Extensive grazing N loss from local terrestrial pools 18 —
Addition of atmospheric N
2
O 1.8 (0.8) —
Volatilization/deposition of NH
3
6 —
Manure management Addition of atmospheric N
2
O 1.3 (0.6) 0.5 (0.2)
Volatilization/deposition of NH
3
11 7
Eutrophication of aquatic systems More than 10
b
a
Quantified contributions are expressed in million tonnes N per year, except for additions of atmospheric N
2
O, also expressed in billion tonnes
CO
2
equivalent (between parentheses).
b
Overall estimate based on mid-1990s figures but lacks information on the importance of direct discharge of manure to water.
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 287
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
crop-livestock systems is often large. Despite
the importance of this category of production
systems, little information is available on
its characteristics in different parts of the
world. In addition, and as a consequence,
estimates for the parts of mixed systems to
be considered as respectively extensive and
intensive production remain largely based on
expert knowledge.
Livestock-related emissions of CO
2
are a
huge component of the global C budget when
deforestation for pasture and feed crop land
as well as pasture degradation are taken into
account. Although small by comparison, the
livestock food chain is becoming more fossil
fuel intensive, as the shift from traditional lo-
cal feed resource-based ruminant production
to intensive monogastrics fundamentally en-
tails a concomitant shift away from solar en-
ergy to fossil fuel.
The leading role of livestock in CH
4
emis-
sions has long been a well-established fact.
With the decline of ruminant livestock in rel-
ative terms, and the overall trends toward
higher productivities also in ruminant pro-
duction, the importance of enteric fermenta-
tion will likely not grow much more. While
much lower in absolute terms, CH
4
emis-
sions from animal manure are considerable
and rapidly growing. Together they represent
some 80% of agricultural emissions and about
35% to 40% of the total anthropogenic CH
4
emissions.
Livestock activities contribute in a princi-
pal way to the emission of nitrous oxides, the
most aggressive greenhouse gas. Their con-
tribution to the global budget is as high as
65%, and 75% to 80% to agricultural emis-
sions; current trends suggest that these lev-
els will substantially increase over the coming
decades.
Global anthropogenic atmospheric emis-
sion of ammonia has recently been estimated
at some 47 million tonnes N (33), with 94%
produced by the agricultural sector. The live-
stock sector contributes about 68% to this
share, mainly from deposited and applied ma-
nure. The resulting air and environmental
pollution is a local or regional more than a
global environmental problem. Similar levels
of N depositions can indeed have substantially
different environmental effects according to
the type of ecosystem they affect.
It is evident from the above that, although
it is certain that the livestock sector plays a
major role in C and N cycles worldwide, ma-
jor knowledge gaps exist. In particular, these
relate to feedback mechanisms, such as the ef-
fect of climate change and its impact on ter-
restrial C pools. Furthermore, impacts of the
N cycle depend to a large extent on soil con-
ditions, but knowledge is still insufficient to
allow for precise and reliable estimates of the
related, important fluxes. Equally important,
the limited information available on varying
forms of waste management in different world
regions constitutes another important obsta-
cle toward more accurate estimates. However,
currently available information would already
warrant a revision of the IPCC default emis-
sion factors. Finally, there is a need to translate
information on environmental impacts into
more accessible material for use by consumers
and decision makers, such as comparative life
cycle analyses among production systems, lo-
cations, and food sectors.
6. TECHNICAL OPTIONS TO
MITIGATE CARBON AND
NITROGEN LOSSES
Even though there are important knowledge
gaps in assessing livestock’s role in C and N
cycles, sufficient evidence points to a large
and growing contribution. Just as the livestock
sector has large and multiple impacts on both
cycles, so there are multiple and effective op-
tions for mitigation. Much can be done, but
to get beyond a “business-as-usual” scenario
requires a strong involvement of public pol-
icy. Most of the options are not cost neutral,
and simply enhancing awareness will not lead
to widespread adoption. While the policy and
institutional aspects are criticial, we do not
examine these here but only present the main
technical options.
288 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
6.1. Sequestering Carbon and
Mitigating Carbon Dioxide
Emissions
Compared to the amounts of C released from
changes in land use and land degradation,
emissions from the food chain are small. For
C, the environmental focus needs therefore
to be on addressing issues of land-use change
and land degradation. Here, the livestock sec-
tor offers significant potential for C seques-
tration, particularly in the form of improved
pastures.
In addressing land-use change, the chal-
lenge lies in slowing and eventually halting
and reversing deforestation. Vlek et al. (59)
consider that the only available option to free
up the land necessary for C sequestration
would be intensification of agricultural
production on some of the better lands, for
example, by increased fertilizer inputs. They
demonstrate that the increased CO
2
emis-
sions related to the extra fertilizer production
would be far outweighed by the sequestered
or avoided emissions of organic C related to
deforestation. Apart from improved fertilizer
use, other options for intensification include
the use of higher-yielding, better-adapted
varieties and improved land and water
management. Although rationally attractive,
the “sequestration through intensification”
paradigm may not be effective in all sociopo-
litical contexts and requires a functioning
regulatory framework.
A huge potential exists for net sequestra-
tion of C in cultivated soils. The C sink ca-
pacity of the world’s agricultural and degraded
soils is 50% to 66% of the historic C loss from
soils of 42 to 78 gigatonnes of C (60). There
are proven new practices that can improve
soil quality and raise soil organic C levels
(e.g., conservation tillage and organic farm-
ing), which achieve yields comparable to con-
ventional intensive systems. The full potential
for terrestrial soil C sequestration is uncertain,
because of insufficient data and understanding
of soil organic C dynamics at all levels, includ-
ing molecular, landscape, regional, and global
scales (61). According to the IPCC (62), im-
proved practices typically allow soil C to in-
crease at a rate of about 0.3 tonnes of C per
hectare per year. It is unclear if this rate is sus-
tainable: Research shows a relatively rapid in-
crease in C sequestration for a period of about
25 years and a gradual leveling thereafter (63).
Improved grassland management is an-
other major area where soil C losses can
be reversed, leading to net sequestration, by
the use of trees, improved pasture species,
fertilization, and other measures. Because pas-
ture is the largest anthropogenic land use,
improved pasture management could poten-
tially sequester more C than any other terres-
trial sink (62, Table 4). There would also be
additional benefits, particularly preserving or
restoring biodiversity in many ecosystems.
In the humid tropics, silvo-pastoral sys-
tems are one approach to C sequestration and
pasture improvement. In dryland pastures,
some aspects of dryland soils may help in C
sequestration. Dry soils are less likely to lose
C than wet soils, as lack of water limits soil
mineralization and therefore the flux of C to
the atmosphere. Consequently, the residence
time of C in dryland soils is sometimes even
longer than in forest soils. Although the rate at
which C can be sequestered in these regions is
low, it may be cost-effective, particularly tak-
ing into account all the side benefits for soil
improvement and restoration (17).
6.2. Reducing Methane Emissions
from Enteric Fermentation Through
Improved Efficiency and Diets
The most promising approach for reduc-
ing CH
4
emissions from livestock is by im-
proving the productivity and efficiency of
livestock production, through better nutri-
tion, genetics, animal health, and general hus-
bandry practices. Greater efficiency means
that a larger portion of the energy in the an-
imals’ feed is directed toward the creation of
useful products, so that CH
4
emissions per
unit product are reduced. The trends to-
ward high-performing animals and toward
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 289
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
monogastrics and poultry, in particular, are
valuable in this context because they reduce
CH
4
per unit of product.
A number of technologies exist to reduce
CH
4
release from enteric fermentation. The
basic principle is to increase the digestibility
of feedstuff, either by modifying feed or by
manipulating the digestive process. Examples
of improvements in fibrous diets are the
use of feed additives or supplements and
the increased level of starch or rapidly fer-
mentable carbohydrates in the diet (so as to
reduce excess hydrogen and subsequent CH
4
formation). In many instances, such improve-
ments may not be profitable at farm level.
However, national planning strategies in
large countries could potentially bring about
such changes. For example, Eckard et al. (64)
suggest that concentrating dairy production
in the temperate zones of Australia could
potentially decrease CH
4
emissions because
temperate pastures are likely to be higher in
soluble carbohydrates and easily digestible
cell wall components.
More advanced technologies are also be-
ing studied but are not yet widely ap-
plied. These include reduction of hydrogen
production by stimulating acetogenic bacte-
ria, defaunation (eliminating certain protozoa
from the rumen), and vaccination (to reduce
methanogens). These options have the advan-
tage of being applicable to free-ranging rumi-
nants as well, although the latter option may
encounter resistance from consumers (65).
Defaunation has been proven to result in a
20% reduction in CH
4
emissions on average
(66), but regular dosing with the defaunating
agent remains a challenge.
6.3. Mitigating Methane Emissions
Through Improved Manure
Management and Biogas
CH
4
emissions from anaerobic manure man-
agement can be readily reduced with exist-
ing technologies. Such emissions originate
from intensive mixed and industrial systems;
these commercially oriented holdings usu-
ally have the capacity to invest in such tech-
nologies. The potential for emission abate-
ment from manure management is consider-
able, and multiple options exist. A first ob-
vious option to consider is balanced feeding
(increased C to N ratios) because it also in-
fluences other emissions. Additional measures
include anaerobic digestion (producing biogas
as an extra benefit), flaring/burning (chemi-
cal oxidation, burning), special biofilters (bi-
ological oxidation) (65, 67), composting, and
aerobic treatment. It is assumed that biogas
can achieve a 50% reduction in emissions in
cool climates, and higher in warmer climates,
for manures that otherwise would be stored as
liquid slurry. Various systems exist to exploit
this huge potential, such as covered lagoons,
pits, tanks, and other liquid storage struc-
tures. These are suitable for large- or small-
scale biogas systems, with a wide range of
technological options and different degrees of
sophistication. Additionally, covered lagoons
and biogas systems produce a slurry that can
be applied to rice fields instead of untreated
dung, leading to reduced CH
4
emissions
(68).
6.4. Mitigating Nitrogen Loss
An important mitigation pathway lies in rais-
ing low animal N assimilation efficiency
through more balanced feeding by optimizing
proteins or amino acids to match the exact re-
quirements. Improved feeding practices also
include grouping animals by gender and phase
of production and by improving the feed con-
version ratio through tailoring feed to physi-
ological requirements.
But even with these measures, manure still
contains large quantities of N. The use of an
enclosed tank can nearly eliminate N loss dur-
ing storage and offers an important synergy
with respect to mitigating CH
4
emissions and
production of biogas; N
2
O emissions from the
subsequent spread of (digested) slurry can also
be reduced.
The key to reducing N loss resulting
from the application/deposition of manure
is the fine-tuning of waste application to
land with regard to environmental conditions,
290 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
including timing as well as amounts and form
of application in response to crop physiol-
ogy and climate. Another technological op-
tion is the use of nitrification inhibitors (NIs)
that can be added to urea or ammonium
compounds. Monteny et al. (65) cite exam-
ples of substantially reduced emissions. Some
of these substances can potentially be used
on pastures where they act upon urinary N,
an approach being adopted in New Zealand
(69). Costs of NIs may be offset by increased
crop/pasture N uptake efficiency. The degree
of adoption of NIs may depend on public per-
ception of introducing yet another chemical
into the environment (65).
Livestock play an important role in both
the global C and N cycle. The contribution
to the C cycle mainly stems from livestock’s
land use and role in land-use change, in par-
ticular deforestation and pasture degradation.
Livestock’s role in the N cycle is mainly deter-
mined by their demand for concentrate feed
and by livestock waste storage and disposal. As
the scope for pasture expansion and intensifi-
cation is limited, extensive livestock is stagnat-
ing, but industrial livestock is growing rapidly.
Subsequently, there is an ongoing shift toward
a growing role of livestock in the N cycle and
a stagnating or declining role in the C cycle,
albeit from a very high level.
FUTURE ISSUES
There is a major uncertainty in the quantification of N
2
O emissions, in particular with
regard to different livestock production systems as well as their feeding and waste man-
agement practices, which need to be underpinned by more accurate modeling of soil
N
2
O fluxes. Furthermore, a major knowledge gap is the feedback mechanisms resulting
from climate change. On the biophysical side, quantifications of C loss from agricultural
soils and from land degradation (LU) are still unreliable, and better quantification is also
needed of the effects on above- and belowground C pools when there are changes in land
use. Such quantifications are extremely difficult, which is why they are often neglected.
DISCLOSURE STATEMENT
The authors are not aware of any biases that might be perceived as affecting the objectivity of
this review.
LITERATURE CITED
1. Naylor R, Steinfeld H, Falcon W, Galloway J, Smil V, et al. 2005. Losing the links between
livestock and land. Science 310:1621–22
2. Food Agric. Organ. (FAO). 2006. FAO Statistical Databases. Rome: FAO
3. Food Agric. Organ. (FAO). 2006. World agriculture: towards 2030/2050. Interim Rep. Food
Agric. Organ., Rome
4. Galloway JN, Burke M, Bradford EG, Naylor R, Falcon W, et al. 2007. International trade
in meat: the tip of the pork chop. Ambio. In press
5. Tansey K, Gr
´
egoire J-M, Stroppiana D, Sousa A, Silva J, et al. 2004. Vegetation burning in
the year 2000: global burned area estimates from SPOT VEGETATION data. J. Geophys.
Res. 109:D14S03
6. Crutzen PJ, Andreae MO. 1990. Biomass burning in the tropics: impact on atmospheric
chemistry and biogeochemical cycles. Science 250(4988):1669–78
7. Scholes M, Andreae MO. 2000. Biogenic and pyrogenic emissions from Africa and their
impact on the global atmosphere. Ambio 29:23–29
8. Vet R. 1995. GCOS observation programme for atmospheric constituents: background,
status and action plan. Glob. Clim. Observ. Syst. Rep. 20. WMO, Geneva
9. Crutzen PJ, Goldammer JG, eds. 1993. Fire in the environment: the ecological,
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 291
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
atmospheric, and climatic importance of vegetation fires. Environ. Sci. Res. Rep. 13. Wiley,
Chichester. 400 pp.
10. Thompson AM, Witte JC, Hudson RD, Guo H, Herman JR, Fujiwara M. 2001. Tropical
tropospheric ozone and biomass burning. Science 291:2128–32
11. Artaxo P, Martins JV, Yamasoe MA, Proc
´
opio AS, Pauliquevis TM, et al. 2002. Physical
and chemical properties of aerosols in the wet and dry seasons in Rondonia, Amazonia. J.
Geophys. Res. 107(D20):8081–95
12. Andreae MO, Crutzen PJ. 1997. Atmospheric aerosols: biogeochemical sources and roles
in atmospheric chemistry. Science 276:1052–57
13. Asner GP, Borghi CE, Ojeda RA. 2003. Desertification in central Argentina: changes in
ecosystem carbon and nitrogen from imaging spectroscopy. Ecol. Appl. 13(3):629–48
14. Lal R. 2001. The potential of soils of the tropics to sequester carbon and mitigate the
greenhouse effect. Adv. Agron. 76:1–30
15. Swift MJ, Seward PD, Frost PGH, Qureshi JN, Muchena FN. 1994. Long-term exper-
iments in Africa: developing a database for sustainable land use under global change. In
Long-Term Experiments in Agricultural and Ecological Sciences, ed. RA Leigh, AE Johnston,
pp. 229–51. Wallingford: CABI Publ.
16. UN Environ. Program. (UNEP). 1991. Status of Desertification and Implementation of the
United Nations Plan of Action to Combat Desertification. Nairobi: UNEP. 79 pp.
17. Food Agric. Organ. (FAO). 2004. Carbon sequestration in dryland soils. World Soil Resour.
Rep. 102, FAO, Rome
18. Niklinska M, Maryanski M, Laskowski R. 1999. Effect of temperature on humus respiration
rate and nitrogen mineralization: implications for global climate change. Biogeochemistry
44:239–57
19. Scholes RJ, Schulze ED, Pitelka LF, Hall DO. 1999. Biogeochemistry of terrestrial ecosys-
tems. In The Terrestrial Biosphere and Global Change: Implications for Natural and Managed
Ecosystems, ed. B Walker, W Steffen, J Canadell, J Ingram, pp. 271–303. Cambridge, UK:
Univ. Cambridge Press
20. van Ginkel JH, Whitmore AP, Gorissen A. 1999. Lolium perene grasslands may function
as a sink for atmospheric carbon dioxide. J. Environ. Qual. 28:1580–84
21. Bellamy PH, Loveland PJ, Bradley RI, Lark RM, Kirk GJD. 2005. Carbon losses from all
soils across England and Wales 1978–2003. Nature 437:245–48
22. Schulze DE, Freibauer A. 2005. Carbon unlocked from soils. Nature 437:205–6
23. Wassenaar T, Gerber P, Verburg PH, Rosales M, Ibrahim M, Steinfeld H. 2007. Projecting
land use changes in the Neotropics: the geography of pasture expansion into forest. Glob.
Environ. Change 17(1):86–104
24. IPCC. 2001. Climate Change 2001: The Scientific Basis, ed. JT Houghton, Y Ding, DJ
Griggs, M Noguer, PJ van der Linden, et al. Cambridge, UK/New York: Cambridge
Univ. Press. 881 pp.
25. Houghton RA. 1991. Tropical deforestation and atmospheric carbon dioxide. Clim. Change
19(1–2):99–118
26. Mosier A, Wassmann R, Verchot L, King J, Palm C. 2004. Methane and nitrogen oxide
fluxes in tropical agricultural soils: sources, sinks and mechanisms. Environ. Dev. Sustain.
6(1–2):11–49
27. Muller W, Schneider B. 1985. Heat, water vapour and CO
2
production in dairy cattle
and pig housing. Part 1. Provisional planning data for the use of heat exchangers and heat
pumps in livestock housing. Tieraztl. Umsch. 40:274–80 (In German)
28. Ni J-Q, Hendriks J, Coenegrachts J, Vinckier C. 1999. Production of carbon dioxide in a
292 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
fattening pig house under field conditions. 1. Exhalation by pigs. Atmos. Environ. 33:3691–
96
29. Minist. Ci
ˆ
enc. Tecnol. 2002. Primeiro Invent´ario Brasileiro de Emiss˜oes Antr´opicas de Gases de
Efeito Estufa: Emiss˜oes de Metano da Pecu´aria. Bras
´
ılia: Empresa Brasileira Pesqui. Agropecu.
(EMBRAPA)
30. US Environ. Prot. Agency. 2004. US Emissions Inventory 2004: Inventory of U.S. Greenhouse
Gas Emissions and Sinks: 1990–2002. EPA-430-R-04-003, Washington, DC: US EPA
31. Steinfeld H, Gerber P, Wassenaar T, Castel V, Rosales M, de Haan C. 2006. Livestock’s
Long Shadow: Environmental Issues and Options. Rome: FAO/LEAD. 390 pp.
32. US Environ. Prot. Agency. 2005. Global Warming—Methane. Washington, DC: US EPA
33. Galloway JN, Dentener FJ, Capone DG, Boyer EW, Howarth RW, et al. 2004. Nitrogen
cycles: past, present, and future. Biogeochemistry 70:153–226
34. Galloway JN, Aber JD, Erisman JW, Seitzinger SP, Howarth RW, et al. 2003. The nitrogen
cascade. BioScience 53(4):341–56
35. van der Hoek KW. 1998. Nitrogen efficiency in global animal production. Environ. Pollut.
102:127–32
36. Smil V. 2002. Nitrogen and food production: proteins for human diets. Ambio 31(2):126–31
37. Mengjie W, Yi D. 1996. The importance of work animals in rural China. World Anim.
Rev.(FAO) 86:65–67
38. Food Agric. Organ., Int. Fertil. Ind. Assoc. (FAO/IFA). 2001. Global Estimates of Gaseous
Emissions of NH
3
, NO and N
2
O from Agricultural Land. Rome: FAO. 106 pp.
39. Smil V. 1999. Nitrogen in crop production: an account of global flows. Glob. Biogeochem.
Cycles 13(2):647–62
40. Food Agric. Organ. (FAO). 2002. Fertilizer use by crop. Jt. Rep. FAO, IFA, IFDC, IPI, PPI.
Rome. 45 pp. 5th ed.
41. Food Agric. Organ. (FAO). 2003. World Agriculture: Towards 2015/30, An FAO Perspective,
ed. J. Bruisnsma. Rome/London: FAO/Earthscan
42. Helsel ZR. 1992. Energy and alternatives for fertilizer and pesticide use. In Energy in Farm
Production. Vol. 6: Energy in World Agriculture, ed. RC Fluck, pp. 177–201. New York:
Elsevier
43. Price L, Worrell E, Martin N, Lehman B, Sinton J. 2000. China’s industrial sector in
an international context. Rep. LBNL 46273, Environ. Energy Technol. Div., Lawrence
Berkeley Natl. Lab., Berkeley, CA
44. Ryan B, Tiffany DG. 1998. Energy use in Minnesota agriculture. Minn. Agric. Econ. No.
693 (Fall), Minn. Extension Service, Univ. Minn.
45. Sainz R. 2003. Framework for calculating fossil fuel use in livestock systems. Livest. Environ.
Dev. Init. Rep.
46. Chapagain AK, Hoekstra AY. 2004. Water footprints of nations. Vol. 1: Main report. Value
of Water Res. Rep. Ser. 16. UNESCO-IHE. Delft, Neth. 76 pp.
47. Cederberg C, Flysj
¨
o A. 2004. Life cycle inventory of 23 dairy farms in south-western
Sweden. Rep. 728. Swedish Inst. Food Biotechnol. (SIK), G
¨
oteborg. 59 pp.
48. Intergov. Panel Clim. Change (IPCC). 1997. Revised 1996 IPCC Guidelines for National
Greenhouse Gas Inventories. Reference Manual. Vol. 3. http://www.ipcc-nggip.iges.or.jp/
public/gl/invs6.htm
49. Morton D, DeFries R, Shimabukuro Y, Anderson L, Arai E, et al. 2006. Cropland expan-
sion changes deforestation dynamics in the southern Brazilian Amazon. Proc. Natl. Acad.
Sci. USA 103(39):14637–41
50. Nepstad DC, Stickler CM, Almeida OT. 2006. Globalization of the Amazon soy and beef
industries: opportunities for conservation. Conserv. Biol. 20(6):1595–603
www.annualreviews.org
•
Livestock in Carbon and Nitrogen Cycles 293
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

ANRV325-EG32-09 ARI 21 September 2007 16:56
51. Sauv
´
e JL, Goddard TW, Cannon KR. 2000. A preliminary assessment of carbon dioxide emis-
sions from agricultural soils. Presented at Alberta Soil Sci. Workshop, 37th, Medicine Hat,
Alberta
52. Minist. Sci. Technol. 2004. Brazil’s Initial National Communication to the United Nations
Framework Convention on Climate Change. Brazil: Minis. Sci. Technol., Gen. Coord. Glob.
Clim. Change, Brasilia. 271 pp.
53. Vitousek PM, Aber JD, Howarth RW, Likens GE, Matson PA, et al. 1997. Human alter-
ation of the global nitrogen cycle: sources and consequences. Ecol. Appl. 7(3):737–50
54. Smil V. 2001. Enriching the Earth: Fritz Haber, Carl Bosch, and the Transformation of World
Food Production. Cambridge, MA: MIT Press. 411 pp.
55. Hall SJ, Matson PA. 1999. Nitrogen oxide emissions after nitrogen additions in tropical
forests. Nature 400(6740):152–55
56. Bouwman AF, Lee DS, Asman WAH, Dentener FJ, van der Hoek KW, Olivier JGJ.
1997. A global high-resolution emission inventory for ammonia. Glob. Biogeochem. Cycles
11(4):561–87
57. Bouwman AF. 1995. Compilation of a global inventory of emissions of nitrous oxide. PhD thesis.
Wageningen Univ., Wageningen, Neth. 143 pp.
58. Russelle MP, Birr AS. 2004. Large-scale assessment of symbiotic dinitrogen fixation by
crops: soybean and alfalfa in the Mississippi River basin. Agron. J. 96:1754–60
59. Vlek PLG, Rodr
´
ıguez-Kuhl G, Sommer R. 2004. Energy use and CO
2
production in
tropical agriculture and means and strategies for reduction or mitigation. Environ. Dev.
Sustain. 6(1–2):213–33
60. Lal R. 2004. Soil carbon sequestration impacts on global climate change and food security.
Science 304(5677):1623–27
61. Metting F, Smith J, Amthor J. 1999. Science needs and new technology for soil car-
bon sequestration. In Carbon Sequestration in Soils: Science, Monitoring and Beyond, ed. N
Rosenberg, R Izaurralde, E Malone, pp. 1–34. Proc. St. Michaels Workshop. Columbus,
OH: Battelle
62. Intergov. Panel Clim. Change (IPCC). 2000. Land Use, Land Use Change and Forestry. A
Special Report of the IPCC. Cambridge, UK: Cambridge Univ. Press
63. Lal R, Kimble J, Follett R, Cole CV. 1998. Potential of U.S. Cropland to Sequester Carbon
and the Greenhouse Effect Mitigation. Chelsea, MI: Sleeping Bear. 128 pp.
64. Eckard R, Dalley D, Crawford M. 2000. Impacts of potential management changes on
greenhouse gas emissions and sequestration from dairy production systems in Australia. In
Management Options for Carbon Sequestration in Forest, Agricultural and Rangeland Ecosystems,
ed. R Keenan, AL Bugg, H Ainslie, pp. 58–72. Canberra: CRC Greenh. Acc.
65. Monteny GJ, Bannink A, Chadwick D. 2006. Greenhouse gas abatement strategies for
animal husbandry. Agric. Ecosyst. Environ. 112:163–70
66. Hegarty RS. 1998. Reducing methane emissions through elimination of rumen proto-
zoa. Meeting the Kyoto target. In Implications for the Australian Livestock Industries, ed. PJ
Reyenga, SM Howden, pp. 55–61. Canberra: Bur. Rural Sci.
67. Melse RW, van der Werf AW. 2005. Biofiltration for mitigation of methane emissions
from animal husbandry. Environ. Sci. Technol. 39(14):5460–68
68. Mendis M, Openshaw K. 2004. The clean development mechanism: making it operational.
Environ. Dev. Sustain. 6(1–2):183–211
69. Di HJ, Cameron KC. 2003. Mitigation of nitrous oxide emissions in spray-irrigated grazed
grassland by treating the soil with dicyandiamide, a nitrification inhibitor. Soil Use Manag.
19(4):284–90
294 Steinfeld
·
Wassenaar
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

AR325-FM ARI 21 September 2007 16:39
Annual Review of
Environment
and Resources
Volume 32, 2007
Contents
I. Earth’s Life Support Systems
Feedbacks of Terrestrial Ecosystems to Climate Change
Christopher B. Field, David B. Lobell, Halton A. Peters, and Nona R. Chiariello pppppp1
Carbon and Climate System Coupling on Timescales from the
Precambrian to the Anthropocene
Scott C. Doney and David S. Schimel pppppppppppppppppppppppppppppppppppppppppppppppppppp31
The Nature and Value of Ecosystem Services: An Overview
Highlighting Hydrologic Services
Kate A. Brauman, Gretchen C. Daily, T. Ka’eo Duarte, and Harold A. Mooney ppppp 67
Soils: A Contemporary Perspective
Cheryl Palm, Pedro Sanchez, Sonya Ahamed, and Alex Awiti ppppppppppppppppppppppppp 99
II. Human Use of Environment and Resources
Bioenergy and Sustainable Development?
Ambuj D. Sagar and Sivan Kartha ppppppppppppppppppppppppppppppppppppppppppppppppppppp131
Models of Decision Making and Residential Energy Use
Charlie Wilson and Hadi Dowlatabadi ppppppppppppppppppppppppppppppppppppppppppppppppp169
Renewable Energy Futures: Targets, Scenarios, and Pathways
Eric Martinot, Carmen Dienst, Liu Weiliang, and Chai Qimin pppppppppppppppppppppp205
Shared Waters: Conflict and Cooperation
Aaron T. Wolf ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp241
The Role of Livestock Production in Carbon and Nitrogen Cycles
Henning Steinfeld and Tom Wassenaar ppppppppppppppppppppppppppppppppppppppppppppppppp271
Global Environmental Standards for Industry
David P. Angel, Trina Hamilton, and Matthew T. Huber ppppppppppppppppppppppppppppp295
Industry, Environmental Policy, and Environmental Outcomes
Daniel Press ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp317
vii
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.

AR325-FM ARI 21 September 2007 16:39
Population and Environment
Alex de Sherbinin, David Carr, Susan Cassels, and Leiwen Jiang pppppppppppppppppppp345
III. Management, Guidance, and Governance of Resources and Environment
Carbon Trading: A Review of the Kyoto Mechanisms
Cameron Hepburn pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp375
Adaptation to Environmental Change: Contributions
of a Resilience Framework
Donald R. Nelson, W. Neil Adger, and Katrina Brown pppppppppppppppppppppppppppppppp395
IV. Integrative Themes
Women, Water, and Development
Isha Ray ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp421
Indexes
Cumulative Index of Contributing Authors, Volumes 23–32 pppppppppppppppppppppppp451
Cumulative Index of Chapter Titles, Volumes 23–32 ppppppppppppppppppppppppppppppppp455
Errata
An online log of corrections to Annual Review of Environment and Resources articles
may be found at http://environ.annualreviews.org
viii Contents
Annu. Rev. Environ. Resourc. 2007.32:271-294. Downloaded from arjournals.annualreviews.org
by Instituto de Eclogia, A.C, Consorcio CONACyT on 02/27/08. For personal use only.