Content uploaded by Richard A Houghton
Author content
All content in this area was uploaded by Richard A Houghton on Oct 28, 2015
Content may be subject to copyright.
This article was downloaded by: [186.251.144.58]
On: 20 August 2015, At: 08:58
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick
Place, London, SW1P 1WG
Carbon Management
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tcmt20
The emissions of carbon from deforestation and
degradation in the tropics: past trends and future
potential
Richard A Houghtona
a Woods Hole Research Center, 149 Woods Hole Road, Falmouth, MA 02540, USA.
Published online: 10 Apr 2014.
To cite this article: Richard A Houghton (2013) The emissions of carbon from deforestation and degradation in the
tropics: past trends and future potential, Carbon Management, 4:5, 539-546, DOI: 10.4155/cmt.13.41
To link to this article: http://dx.doi.org/10.4155/cmt.13.41
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose
of the Content. Any opinions and views expressed in this publication are the opinions and views of the
authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not
be relied upon and should be independently verified with primary sources of information. Taylor and Francis
shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and
other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation
to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
The primary purpose of REDD+ is to reduce emis-
sions of carbon and help stabilize the concentration of
CO2 in the atmosphere, thereby limiting the rate and
amount of climatic disruption. When the predecessors
of REDD+ were first proposed in the early 1990s, the
emissions of GHGs from tropical deforestation were
thought to account for 20–25% of global emissions [1] .
That percentage has come down, largely because the
annual emissions of carbon from fossil fuel combus-
tion have increased sharply in recent years (Figure1) [2].
The emissions of carbon from deforestation and forest
degradation in the topics averaged 10–15% of anthro-
pogenic carbon emissions in the decade 2000–2009.
Today they may be less 10%, and the decline raises the
question of whether REDD+ is worth the effort and
expense. This paper suggests that REDD+ is indeed
worth the effort, and that an all-out effort could poten-
tially stabilize atmospheric carbon for a few decades
and, thus, buy time for the development of technolo-
gies that would eventually eliminate dependence on
fossil fuels.
This paper has two parts. Part one provides a brief
review of historic and current emissions of carbon from
deforestation and degradation in the tropics. Reasons for
the variability and uncertainty of emissions estimates
are discussed. Part two explores the implications of
past and current land use (LU) in the tropics for future
reductions in emissions and enhancement of sinks.
A few definitions are in order. LU is defined here
as a use of land that does not change its category, or
cover. For example, harvest of wood, when followed by
regrowth of the forest, is a LU; the forest remains forest.
In contrast, land-cover change (LCC) is defined as the
conversion of one cover type to another, for example,
deforestation changes a forest cover to a largely treeless
cover (a cropland or pasture use). LU and LCC together
(LULCC), refers to the combined effects. Ideally,
LULCC would include all types of land management,
but many are not documented at a global level. It is
important to stress that LULCC refers to changes that
result from direct human activity. Deforestation caused
by fires, storms, insects or disease is not included in
LULCC, unless fires are used intentionally for LCC.
Forest degradation is defined as a reduction in car-
bon density (mgC/ha) of either biomass or soil within
a forest. Thus, it is LU, not LCC. The net effect of
wood harvest lowers, at least temporarily, the carbon
density of forests (although it may increase the storage
Carbon Management (2013) 4(5), 539–546
The emissions of carbon from deforestation and
degradation in the tropics: past trends and future potential
Richard A Houghton*
Land use in the tropics, including both deforestation and forest degradation, is estimated to have emitted
approximately 1.4PgCyr-1 to the atmosphere over the interval 1990–2010 (~15% of anthropogenic carbon
emissions). This net emission is composed of gross emissions of at least 2.6 PgC yr-1 and gross sinks of
1.2PgCyr-1 in forests recovering from wood harvest and in the fallows of shifting cultivation. In contrast
to recent management of tropical forests, future management in the region could be used to stabilize the
concentration of CO2 in the atmosphere, at least temporarily, with the following three measures: a halt to
deforestation and forest degradation, protection of regrowing forests, and the re-establishment of forests
on lands not intensively used now that were forests in the past. Together, these three measures have the
potential to reduce emissions of carbon and increase uptake by as much as 3–5PgCyr-1.
Review
*Woods Hole Research Center, 149 Woods Hole Road, Falmout h, MA 02540, USA
Tel.: +1 508 540 9900; E-mail: rhoughton@whrc.org
Mini Focus: sustainable landscapes in a woRld oF change:
tRopical FoRests, land use and iMpleMentation oF Redd+
future science group 539
ISSN 1758-3004
10.4155/CMT.13.41 © 2013 Future Science Ltd
For reprint orders, please contact reprints@future-science.com
Downloaded by [186.251.144.58] at 08:58 20 August 2015
of carbon in wood products). Given enough time, a
harvested forest may attain its original carbon density
again. While growing, the forest is a net carbon sink.
Other definitions will be given in context below.
Past & current emissions of carbon from
deforestation & degradation of tropical forests
A recent review of 13 estimates of global carbon emis-
sions from LULCC found average emissions of 1.1 ± 0.5
for the period 1990–2010 [3]. Although the estimates
were of global emissions, they are thought to represent
tropical emissions because, first, estimates for the trop-
ics, alone are similar to global estimates [4 –7] and, sec-
ond, because extra-tropical regions are nearly neutral
with respect to carbon emissions from LULCC. That is,
the sources of carbon from decay of logging slash, burn-
ing and oxidation of wood products are approximately
balanced by the sinks of carbon in forests regrowing
from earlier harvests and agricultural abandonment.
However, the estimate of 1.1 PgC yr-1 for the tropics
includes degradation but does not include peat swamps.
The emissions of carbon from the draining and burning
of peat forests in southeastern Asia, principally Indone-
sia and Malaysia, are estimated to have released another
0.3 PgC yr-1 in recent years [8], and thus emissions of
carbon from tropical deforestation and degradation
are likely to equal approximately 1.4 ± 0.5 PgC yr-1
at present.
The 13 estimates of carbon emis-
sions were calculated with two types
of models. Most of the analyses
used Dynamic Global Vegetation
Models, which are based on physi-
ological processes such as photosyn-
thesis, respiration, growth, decay,
allocation, and so on. The models
simulate the uptake and release of
carbon as controlled by environmen-
tal factors: the effects of light, water,
CO2, temperature and, sometimes,
nitrogen on plant growth and on
decomposition. The fluxes of car-
bon vary through time as a result
of changes in these environmental
parameters. All analyses of LULCC
included the conversion of lands to
crops and pastures (LCC) based
on historic datasets [9, 10]; some also
included the effects of wood harvest
(LU) [11]. The models are most often
run with and without LULCC, the
difference being an approximation
of net emissions from LULCC.
A few of the analyses used what
has come to be known as a bookkeeping approach. In
this approach, all of the carbon on lands experiencing
LULCC is redistributed when management first takes
place [1 2] . For example, when a forest is converted to a
cropland, the living biomass is assigned to the atmosphere
(representing burning), to organic matter pools left on
site to decay, and to wood product pools that decay at
rates simulating different wood products. Soil organic
carbon is lost in the years after f irst cultivation and recov-
ers again if the cropland returns to its natural state. Aban-
doned croplands and pastures accumulate carbon as the
natural ecosystems that replace them recover.
It is worth emphasizing that the net and gross emis-
sions calculated with the bookkeeping approach are only
those resulting from direct human activity. The sources
and sinks of carbon from unmanaged lands are not
counted. Furthermore, the gross f luxes of carbon are based
on changes in carbon storage, not on changes in metabo-
lism. Photosynthesis and respiration are not a part of the
accounting. What is counted are changes in the pools of
biomass, soils, litter and wood products as a direct con-
sequence of LULCC. Unlike analyses based on Dynamic
Global Vegetation Models, the bookkeeping approach
does not include the effects of environmental change (e.g.,
increasing CO2, warming temperatures, changing mois-
ture regimes and nitrogen deposition). Evergreen forests
in North America, for example, grow at the same rate
whether in 1850 or 2000 in the bookkeeping calculations.
Figure1. Annual emissions of carbon from combustion of fossil fuels and net emissions from
land use and land-cover change. Emissions from peat swamps not included.
1850
PgC yr-1
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0
1
2
3
4
5
6
7
8
9
10
Land use
Fossil fuel and cement
Year
Carbon Management (2013 ) 4(5) future science group
540
Review Houghton
Downloaded by [186.251.144.58] at 08:58 20 August 2015
Differences among the models used to estimate the
flux of carbon from LULCC account for some of the
differences among estimates. Other differences result
from the activities included (or not) in different analy-
ses, but the standard deviation for 13 model estimates
was only approximately ±0.25 PgC yr-1, suggesting that
many errors were offsetting. The primary uncertain-
ties result from different rates of LULCC and different
estimates of biomass density. Current rates of deforesta-
tion reported by the latest Forest Resource Assessment
of the FAO, for example, are based on country reports
often lacking up-to-date inventories [13 ]. An indepen-
dent satellite-based assessment [14] found lower rates of
forest loss in 1990–2005 than reported by FAO [13], and
an increasing trend rather than the decreasing trend. A
systematic, tropics-wide monitoring of forests would
reduce the uncertainties of emissions considerably [15] .
The emissions of carbon from degradation are even
more uncertain. Although forest degradation is wide-
spread, it is also highly variable geographically and
the estimate in Ta ble 1, in which degradation might
account for approximately 15% of net LULCC emis-
sions, is likely an underestimate. Overall, the error for
net emissions of carbon from LULCC is estimated to
be approximately ±0.5 PgC yr-1 and could be higher if
modeled estimates include the effects of environmental
changes on LULCC emissions [3]. The results from one
bookkeeping model are used here to elaborate some of
the details of carbon emissions. One detail is the long-
term trend in annual emissions since 1850 (Figure2).
Early on, the net emissions from LULCC were almost
entirely from the mid-latitudes as agriculture expanded
rapidly in North America, Australia and later in China
and the Former Soviet Union. Emissions in the trop-
ics grew significantly early in the twentieth century in
Latin America, followed by tropical Asia by the mid cen-
tury. The combined emissions from tropical America,
Asia and Africa grew almost continuously up to the
1990s and seem to have declined a little since then.
The growing emissions from the tropics, combined with
the declining emissions (since the late 1950s) outside
the tropics have kept global emissions from LULCC
between 1 and 1.5 PgC yr-1 (if peat swamps are included)
since approximately 1960.
In contrast to the net emissions from LULCC,
gross emissions and gross rates of uptake have gener-
ally increased (Figure3). The gross emissions include
the losses of carbon to the atmosphere from burning
and decay, including the decay of soil organic matter as
a result of cultivation. Gross uptake includes the accu-
mulation of carbon in regrowing forests and in soils
recovering from cultivation. Currently, gross emissions
are approximately 4 PgC yr-1 globally and approximately
2.6 PgC yr-1 in the tropics. Gross rates exceed net rates
as a result of rotational LUs; for
example, the harvest of wood with
subsequent forest recovery, and
the repeated clearing and aban-
donment of shifting cultivation.
The net fluxes of carbon for these
rotational LUs are small relative to
the offsetting gross emissions and
sinks (Ta bl e 1). By comparison, the
gross and net emissions from defor-
estation are the same; there are no
sinks when forests are converted to
permanent agricultural lands.
Comparison of two
recent studies
A recent study by Harris et al. [16 ]
reported gross emissions of car-
bon that were only approximately
35% of those reported by Baccini
et al. [7] and another recent sum-
mary of the world’s forests [17] (0.8
vs 2.3 PgC yr-1). The low estimate
by Harris et al. [16 ] was surpris-
ing because the ana lysis appeared
to be similar in many ways to the
ana lysis by Baccini et al. [7]. Both
estimated the emissions of carbon
from deforestation in the tropics,
and both used spatial data derived
from satellites. If state-of-the-art
estimates of carbon emissions from tropical deforesta-
tion vary by a factor of three, there is little hope for the
implementation of REDD+.
However, the term ‘deforestation’ was used differently
by the two studies. Harris et al. used the term in the
strict sense, consistent with the IPCC Good Practice
Guidelines category of ‘forestland converted to other
land,’ or LCC as used here [1 6] . Baccini et al. used the
term to encompass the broader set of emissions from
LULCC, consistent with much of the published litera-
ture, including the IPCC 4th Assessment Report [7]. In
the parlance of the IPCC’s Good Practice Guidelines,
Baccini et al. included emissions from activities occur-
ring on ‘forestland remaining forestland’. To minimize
confusion and misinterpretation, future analyses should
make clear the definitions used and the implications of
those definitions for broader discussions. Here the term
LULCC is defined to include not only deforestation but
forest degradation as well.
The gross emissions from deforestation alone
(0.81 PgC yr-1 [16 ] ) are not the same as the gross emis-
sions from LULCC (2.3 PgC yr-1 [7]), which include
emissions from wood harvest and (rotational) shifting
Key terms
Management: Direct human activities;
for example, clearing of forests for
croplands and harvest of wood. It is
synonymous with land use and land-
cover change in this article. In contrast
to management, there are two other
processes that may influence changes
in terrestrial carbon storage: indirect
effects (i.e., environmental factors, such
as CO2, climate and nitrogen deposition)
and natural effects. Distinguishing
among these three processes is often
difficult. Changes in carbon storage in
unmanaged or natural ecosystems are
not included in estimates of the land
use and land-cover change flux; and the
indirect and natural effects of
environmental change on managed
lands are also excluded to the ex tent
possible.
Gross fluxes of carbon: Gross fluxes of
carbon from land use and land-cover
change (LULCC) include the emissions
of carbon from fires associated with
LULCC, decay of woody debris,
oxidation of wood products from decay
of soil organic matter. For many types of
LULCC, these gross emissions are largely
offset by the uptake of carbon in
growing forests and recovering soils.
Gross and net emissions from
deforestation are equal. Gross and net
emissions are very dif ferent for wood
harvest with forest recovery and for the
rotational aspects of shifting cultivation.
The emissions of carbon from deforestation & degradation in the tropics Review
future science group www.future-science.com 541
Downloaded by [186.251.144.58] at 08:58 20 August 2015
cultivation, as well as from deforestation ( Tabl e1). Defor-
estation (in the strict sense) accounts for only 42% of
gross emissions in the ana lysis by Baccini et al., but for
86% of net emissions [7]. Forest degradation, by contrast,
accounts for 58% of gross emissions, but only 14% of
net emissions. The gross emissions from wood harvest
and shifting cultivation are large and nearly balanced
by the gross uptake of carbon in recovering forests
and fallows. Because these rotational aspects of LU do
not change forest cover, reductions in biomass density
represent forest degradation rather than deforestation.
As mentioned above, the estimate that forest degrada-
tion accounts for only 14% of net LULCC emissions is
probably an underestimate.
When the (net or gross) emissions from deforesta-
tion are compared, the estimate by Harris et al. is only
16% lower than the estimate by Baccini et al. (0.81 vs
0.96 PgC yr-1, respectively) [7, 16] . In fact, if the emissions of
carbon from cultivated soils are omitted from the estimate
by Baccini et al ., as they were in Harris et al., the two inde-
pendent estimates are identical (0.81 PgC yr-1) (Figure4)
[16 ,7]. Neither ana lysis included the emissions of carbon
from draining and burning of peatlands. Furthermore,
the estimate of 0.81 PgC yr-1 is for deforestation alone,
and not forest degradation.
However, the agreement of 0.81 PgC yr-1 for deforesta-
tion is fortuitous for at least two reasons. First, a region-
by-region comparison showed a reasonable agreement
only in Latin America (Baccini et al. 0.47 PgC yr-1; Harris
et al. 0.44 PgC yr-1). For tropical Asia and sub-Saharan
Africa the differences were nearly a factor of two, and off-
setting (Baccini et al. 0.11 PgC yr-1 and 0.23 PgC yr-1, for
Asia and Africa, respectively; Harris et al. 0.26 PgC yr-1
and 0.11 PgC yr-1, respectively).
The major reason for different
estimates of emissions in tropical
Asia and sub-Saharan Africa was the
rates of deforestation used by the two
analyses. Baccini et al. used rates of
deforestation (2000–2010) reported
in the 2010 Forest Resources
Assessment of the FAO [13]. The
rates were obtained from country
surveys. The surveys are not nec-
essarily based on current data, and
revisions are often reported in sub-
sequent years. The study by Harris
et al. [16 ], used estimates of gross for-
est loss obtained from satellite [18].
While the satellite data are more up
to date than those reported in the
Forest Resource A ssessment, they are
not necessarily a measure of defores-
tation [13 ]. The satellite data capture
deforestation, but they also capture
a number of other processes that
would not be included in the FAO’s
definition of deforestation. Natural
disturbances and clear-cut logging,
for example, may both be followed
by forest recovery, in which case they
Table1. Gross and net emissions of carbon (PgCyr-1) from land-use and
land-cover change activities in the tropics for the period 2000–2005.
Type of management Gross emissions Gross uptake Net emissions
Change in forest area
Deforestation 0.960 00.960
Afforestation - 0.015 -0. 015
Degradation
Wood harvest (industrial) 0.450†-0.446 0.004‡
Fuelwood harvest 0.230 - 0.146 0.084§
Shifting cultivation cycle¶0.640#- 0.558 0.082††
Subtotal for degradation 1.32 0 -1.150 0.170
Total 2.280 -1.165 1 .11 5
Negative valu es indicate carbon removed f rom the atmosphere.
†Emissions from lo gging debris and wood produ cts.
‡Emissions from lo gging debris and wood produ cts, and uptake by recovering forests.
§Both emissions and uptake by recoverin g forests.
¶The first time a forest is cleared for shifting cultivation that clearing is deforestation. Subsequent
clearing of fallows is counted here as fo rest degradation.
#Emissions from the re clearing of fallows.
††Net emissions from the reclearing and regrowth o f fallows.
Carbon Management (2013 ) 4(5) future science group
542
Review Houghton
-0.25
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
0.25
0.50
0.75
PgC yr-1
1.00
1.25
1.50
1.75
Global
Tropics
Temperate
0.00
Year
Figure2. Annual net emissions of carbon from land use and land-cover change in tropical
regions, temperate and boreal regions, and globally. Emissions from peat swamps not
included.
Downloaded by [186.251.144.58] at 08:58 20 August 2015
are not deforestation as defined by
FAO. The reclearing of fallows for
shifting cultivation may appear from
space as forest loss, but the fallows
are not considered forests by the FAO
until they are at least 5 years old.
Aside from the offsetting differ-
ences in tropical Asia and sub-Saha-
ran Africa, a second reason why the
agreement for pan-tropical defores-
tation emissions may be fortuitous
is that the methods used by the two
groups were very different. Baccini
et al. included the emissions from
soils [7]; Harris et al. did not [16 ]. Har-
ris et al. considered the years 2000–
2005; Baccini et al., 2000–2010.
Perhaps most importantly, Harris
et al. calculated ‘committed’ emis-
sions. Committed emissions assume
that all of the carbon lost as a result
of deforestation is lost at the time
of deforestation. The estimate is
obtained by multiplying the area
deforested by the carbon density of
the forest lost. In contrast, Baccini
et al. used a bookkeeping model that
accounts for lags in emissions and
uptake, thus enabling an estimate
of net emissions. If the emissions
today from LU activities in the past
are equal to the emissions delayed
to future years, then the committed
emissions and the actual emissions
will be the same. When rates of LU
are changing from year to year, this
equivalence is unlikely.
Thus, despite the remarkable
agreement at the pan-tropical scale,
that agreement does not define the
precision or accuracy available for
monitoring changes in terrestrial
carbon storage. Although the efforts
of Harris et al. [16 ] and Baccini et al.
[7] are both notable steps forward,
neither provides emissions estimates
with the level of resolution or degree
of certainty needed to support per-
formance-based mechanisms, such
as REDD+. Although the resolution
for accounting is national or subna-
tional, the resolution required for
measurement may be tens of meters.
The limitation of coarse spatial
-1.5
-0.5
0.5
1.0
Carbon emission (+) or uptake (-) (PgC yr-1)
1.5
2.5
0.64
0.23
0.45
0.81
0.004
0.082
0.084
0.81
0.81
0.15
-0.015
0.15
-0.446
-0.146
-0.558
-0.015
2.0
0
Gross fluxes Net fluxes Deforestation flux
Harris et al. [16]
Baccini et al. [7] Baccini et al. [7]
-1.0
Shifting cultivation
Fuelwood harvest
Industrial logging
Deforestation
Soils
Afforestation
Figure4. Average annual gross emissions (+) and accumulations (-) of carbon from
deforestation and other land-use and land-cover change activities in the tropics.
Emissions from peat swamps not included.
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
-2.0
-1.5
-1.0
-0.5
0
0.5
Gross carbon flux (PgC yr-1)
1.0
1.5
2.0
2.5
3.0
Non-tropics emissions
Non-tropics uptake
Tropical emissions
Tropical uptake
Year
Figure3. Annual gross emissions and gross uptake of carbon from tropical regions and from
temperate and boreal regions. Emissions from peat swamps not included.
The emissions of carbon from deforestation & degradation in the tropics Review
future science group www.future-science.com 543
Downloaded by [186.251.144.58] at 08:58 20 August 2015
resolution is that it fails to recognize the heterogeneity
of forest biomass and fails to account for any bias in the
biomass density of deforestation. That is, deforestation
may not be distributed randomly over a landscape of
varying carbon densities but may occur in forests with
systematically lower or higher biomass density than the
average. To capture such bias, co-location of data on
deforestation and degradation with data on carbon den-
sities at the spatial and temporal resolutions of deforesta-
tion and degradation would be ideal, although perhaps
not achievable with satellite data.
The conclusions from a comparison of these two
studies are not that a ‘consensus’ exists [19 ], and not
that the two groups are so different that uncertainties
in measurement, reporting, and verification preclude
the adoption of REDD+. Rather, the data and meth-
ods to monitor, record and verify carbon emissions for
REDD+ in individual countries exist, but were not used
in calculating the pan-tropical estimates. The commu-
nity can monitor REDD+ at the country scale, but nei-
ther Baccini et al. nor Harris et al. did it with the spatial
resolution that is now available. The data requirements
are, first, high-resolution deforestation maps (30 m),
second, high-resolution biomass maps (250 m or less)
and, third, co-location of the two maps so that the bio-
mass of the forests actually deforested can be obtained.
Unfortunately the Ice, Cloud and land Elevation Satel-
lite providing the LiDAR data used by both studies to
determine biomass is no longer in orbit.
Three mechanisms for helping stabilize the
concentration of CO2 in the atmosphere
To limit climate change and prevent further climatic
disruption requires stabilization of the concentration of
CO2 and other GHGs in the atmosphere. Such a stabili-
zation will require >90% reductions in the emissions of
carbon from fossil fuels, the primary source of carbon
to the atmosphere. Nevertheless, management policies
affecting the area and carbon density of forests can help
in the short term (~50 years), while alternatives to fossil
fuels are developed, and can be beneficial for a host of
other reasons as well – reasons that include sustainable
development, biodiversity, energy and water resources,
and other ecosystem services.
Stabilization of the CO2 concentration in the
atmosphere immediately would require a reduction
in emissions of approximately 4 PgC yr-1 ( Tabl e 2).
Approximately half of the carbon emitted each year
from fossil fuel combustion (7.8 PgC yr-1) and LU
change (1.0 PgC yr-1) accumulates in the atmosphere
(4.0 PgC yr-1). The rest is taken up by the ocean and
land. The carbon sinks in the ocean and on land are
responding to the concentration of CO2 in the atmo-
sphere. If emissions were reduced by approximately
4 PgC yr-1, the concentration in the atmosphere would
remain what it had been the year before because the
ocean and land sinks would still be responding to
the same (air–sea or air–land) gradients. That is, the
atmospheric increase would drop to zero, while the
ocean and land sinks would continue as before. Their
uptake would not continue at the same rate for long,
however, because the gradients with the atmosphere
would decline as their pools increased. Thus, emissions
would have to be reduced by more than 4 PgC yr-1 in
subsequent years to keep the atmosphere at a constant
concentration (i.e., stabilized). Nevertheless, the CO2
concentration would be stabilized with an immediate
reduction of 4 PgC yr-1 in emissions.
Can such a reduction be achieved through man-
agement of tropical forests? Potentially, yes, through
three mechanisms: a halt to deforestation and forest
degradation; protection of regrowing forests; and the
re-establishment of forests on lands formerly forested
but not intensively used now.
The first two of these mechanisms are quantified
by the analyses reviewed in the first part of this paper.
Stopping tropical deforestation and degradation would
reduce emissions by 1.4 PgC yr-1. Second, allowing
secondary forests and the fallows of shifting cultiva-
tion to continue growing (no further harvesting or
clearing) would take another 1–3 PgC yr-1 out of the
atmosphere and store it in growing forests.
The third mechanism would be to re-establish for-
ests on lands that once supported forests, but that are
now without forests. An area of 500 million ha would
provide a global sink of approximately 1 PgC yr-1 if the
annual accumulation of carbon in trees and soil were
a modest 2 MgC ha-1 yr-1. The area required would be
less if rates of accumulation were higher. The area is
large, approximately half the area of the USA or half
the area of China. It is not clear that this area is avail-
able in the tropics. Globally, however, the land area in
crops is three-times larger, and the land area in pas-
ture and rangelands is five-times larger, so the area for
new forest is not unthinkable. It is not clear, however,
Table2. Mean global carbon budget 2000–2009.
Source or sink of carbon PgCyr-1
Emissions
Fossil fuel combustion and
cement production
7.8±0.4
Land-use change 1.0±0.5
Accumulations
Atmospheric growth rate 4.0±0.1
Ocean sink 2.4±0.5
Residual terrestrial sink 2.4±0.8
Data taken from [2].
Carbon Management (2013 ) 4(5) future science group
544
Review Houghton
Downloaded by [186.251.144.58] at 08:58 20 August 2015
where these lands are, or whether they would support
forests again, even if they supported forests in the past.
Many of the available lands may now be impoverished
or infertile as a result of overuse. Even if they did not
immediately support native forests, however, the lands
might support plantations of certain species.
In terms of carbon, reduced emissions or increased
uptake are equivalent. If implemented all at once,
the reduced emissions from these three measures
(3–5 PgC yr-1) are enough to stabilize the concentration
of CO2 in the atmosphere.
Each of these measures has costs. The financial
costs are beyond the scope of this calculation, but they
would be large, as would the benefits. Allowing all
fallows within the shifting cultivation cycle to regrow
would require that shifting cultivation become perma-
nent. Allowing all secondary forests to regrow would
effectively preclude wood harvest for either timber
or fuel wood. It might be better, in terms of carbon
emissions, to use land for energy production (biofuels)
other than for carbon storage [20–22]. Furthermore,
the carbon sink in growing forests would diminish as
forests aged. The sink might continue for 40–50 years
before gradually diminishing. It offers a window dur-
ing which the world might make the transition out
of fossil fuels.
Finding as many as 500 million ha of degraded lands in
the tropics, or globally for that matter, will be a challenge,
not only because the area is large, but, more importantly,
because those same lands will be sought after by those
seeking to increase agricultural production for a growing
number of people. Indeed, the competition for land for
food, fiber, feed, fuel, carbon storage, biodiversity and
other ecosystem services is going to intensify even in the
absence of climatic change. Nevertheless, informed deci-
sions about how much land to use, for what purpose, and
where, are fundamental to sustainable uses of resources.
Future perspective
Understanding how past LULCC activities affect today’s
sources and sinks of carbon on land provides perspective
and knowledge for planning how lands might be used
henceforth to reduce carbon emissions and increase car-
bon sinks. The obvious conclusion is that current tropics-
wide rates of deforestation and forest degradation have
to be reduced, and the good news is that those rates are
declining in some cases, locally and nationally. Outside
the tropics, the trends were reversed decades ago, and in
Brazil rates of deforestation have been declining steadily
over the last 9 years [10 1]. Forests are accumulating car-
bon nearly everywhere [1 7] , and large areas are suitable for
reforestation [23]. The challenge is to provide incentives to
make such carbon accumulation continue. REDD+ offers
that incentive. The technical capacity for measurement,
reporting, and verification will continue to improve,
but it is already adequate for REDD+. The adoption of
REDD+ is largely a question of political will and econom-
ics at this point.
Acknowledgements
The author thanks S Goetz and two anonymous reviewers for helpful
comments on an early draft.
The emissions of carbon from deforestation & degradation in the tropics Review
future science group www.future-science.com 545
Executive summary
Past sources & sinks of carbon from land use & land-cover change in the tropics
13 recent estimates of carbon emissions from land use and land-cover change (LULCC) indicate mean net emissions of 1.1PgCyr-1 for the
period 1990–2010 (~15% of total anthropogenic carbon emissions). Most of these emissions are from tropical regions, although none of the
studies included emissions from tropical peat swamps.
Net emissions from the tropics are likely to equal approximately 1.4PgCyr-1 when the draining and burning of southeast Asian peatlands
are included.
Deforestation accounts for most of these annual net emissions. Forest degradation is estimated to account for 15–35% of LULCC emissions,
but the uncertainty is large.
The large difference in emissions reported by two recent studies was mostly the result of different definitions and measurement of
deforestation. With similar definitions the two studies agreed within ±0.15PgCyr-1 in all regions.
A combination of field measurements and satellite data is capable right now of meeting measurement, reporting and verification
requirements for REDD+. Improvements will come from co-locating rates of deforestation and degradation with aboveground carbon
densities at temporal and spatial resolutions similar to those of LULCC activities (tens of meters).
Future sources & sinks of carbon from LULCC
Although combustion of fossil fuels accounts for approximately 90% of current carbon emissions worldwide, the concentration of carbon
dioxide in the atmosphere could be stabilized immediately through massive efforts of forest management in the tropics.
A halt to deforestation and forest degradation would reduce emissions by 1.4PgCyr-1.
Allowing secondar y forests to regrow could remove 1–3PgCyr-1 from the atmosphere.
Reforesting 500millionha of forest on lands that were once forested and not intensively used for food production could remove another
1PgCyr-1.
Together, these management efforts could reduce current emissions by 3–5PgCyr-1 and stabilize the carbon content of the atmosphere
long enough (~50years) for low-carbon energy technologies to replace fossil fuels.
Downloaded by [186.251.144.58] at 08:58 20 August 2015
Financial & competing interests disclosure
Support for the preparation of this paper was generously provided by the
Woods Hole Research Center (MA, USA). The author has no other
relevant affiliations or financial involvement with any organization or
entity with a financial interest in or f inancial conflict with the subject
matter o r materials disc ussed in the manuscr ipt apart from those d isclosed.
No writing assistance was utilized in the production of this
manuscript.
References
Papers of special note have been highlighted as:
n of interest
nn of considerable interest
1 Houghton RA. The role of the world’s
forests in global warming. In: World Forests
for the Future. Their Use and Conservation.
Ramakrishna K, Woodwell GM (Eds).
Yale University Press, CT, USA, 21–58
(199 3).
2 Le Quéré C, A ndres RJ, Boden T et al. The
global carbon budget 1959–2011. Earth Syst.
Sci. Data Discuss. 5, 165–185 (2013).
nn The most comprehensive and up-to-date
summa ry of t he annual global carbon
budget.
3 Houghton RA, House JI, Pongratz J et al.
Carbon emissions from land use and land-
cover change. Biogeosciences 9, 5125 –5142
(2012).
4 DeFries RS, Houghton R A, Hansen MC,
Field CB, Skole D, Townshend J. Carbon
emissions from tropical deforestation and
regrowth based on satellite observations for
the 1980s and 90s. Proc. Natl Acad. Sci. US A
99, 14256 –14261 (20 02).
5 Achard F, Eva HD, Mayau x P, Stibig H-J,
Belward A. Improved estimates of net
carbon emissions from land cover change in
the tropics for the 1990s. Global Biogeochem.
Cycles 18(2), GB2008 (20 04).
6 van der Werf GR, Morton DC, DeFries RS
et al. CO2 emissions from forest loss. Nat.
Geosci. 2, 737–738 (2009).
7 Baccini A , Goetz SJ, Walker WS et al.
Estimated ca rbon dioxide emissions from
tropical deforestation improved by carbon-
density maps. Nat. Clim. Chang. 2, 182 –185
(2012).
n Discusses net carbon emissions from both
tropical deforestation and forest
degradation.
8 Hooijer A, Page S, Canadell JG et al. Current
and future CO2 emissions from drained
peatlands in southeast Asia. Biogeosciences 7,
150 5 –1514 (2 010).
9 K lein Goldewijk K. E stimating global land
use change over the past 30 0 years: the
HYDE database. Glob. Biogeochem. Cycles 15,
417–433 (2001).
10 Ramankutty N, Foley JA. Estimating
historical changes in global land cover:
croplands from 1700 to 1992. Glob.
Biogeochem. Cycles 13, 997–1027 (1999).
11 Hurtt GC, Frolking S, Fearon MG et al. The
underpinnings of land-use histor y: three
centuries of globa l gridded land-use
transitions, wood harvest activity, and
resulting secondary lands. Global Change
Biolog y 12, 1–22 (2006).
12 Houghton RA, Hobbie JE, Melillo JM et al.
Changes in the carbon content of terrestrial
biota and soils between 1860 and 1980: a net
release of CO2 to the atmosphere. Ecol.
Monogr. 53, 235–262 (1983).
13 FAO. Global Forest Resources Assessment 2010.
FAO, Rome, Italy (2010).
14 Lindquist EJ, D’Annunzio R, Gerrand A et al.
Global Forest Land-Use Change From 1990 to
2005. FAO, Rome, Italy (2012).
15 Grainger A. Difficulties in tracking the long-
term globa l trend of tropical forest area. Proc.
Natl Acad. Sci. USA 105(2), 818–823 (2008).
16 Harris NL, Brown S, Hagen SC et al.
Baseline map of carbon emissions from
deforestation in tropical regions. Science 336,
1573 –1576 (2012).
n Carbon emissions calculated from gross rates
of forest loss.
17 Pan Y, Birdsey RA, Fang J et al. A large and
persistent carbon sink in the world’s forests.
Science 333, 988– 993 (2 011).
n A summa ry of t he carbon balance of the
world’s forests, 1990–2007.
18 Hansen MC, Stehman SV, Potapov PV.
Quantification of global gross forest cover
loss. Proc. Natl Acad. Sci. USA 107,
8650–8655 (2010).
19 Harris NL, Brown S, Hagen SC, Baccini A,
Houghton R A. Progress Toward a Consensus
On Carbon Emissions From Tropical
Deforestation. The Meridian Institute,
Washington, DC, USA (2012).
20 Ellison D, Lundblad M, Petersson H. Carbon
accounting and t he climate politics of
forestr y. Enviro. Sci. Pol. 14, 10 62 –1078
(2 011).
21 Schulze E-D, Körner C, Law BE, Haberl H,
Luyssaert S. Large-scale bioenergy from
additiona l harvest of forest biomass is neither
sustainable nor greenhouse gas neutral. GCB
Bioenergy 4(6), 611– 616 (2012).
22 Vanhala P, Repo A, Liski J. Forest bioenergy
at the cost of carbon sequestration? Curr.
Opin. Environ. Sustain. 5(1), 41– 46 (2 012).
23 Lamb D, Erskine PD, Parrotta JA.
Restoration of degraded tropical forest
landscapes. Science 310, 162 8 –1632 (20 05).
Website
101 Foreign Affairs. A light in the forest.
www.foreignaffairs.com/articles/138849/
jeff-tollefson/a-light-in-the-forest
Carbon Management (2013 ) 4(5) future science group
546
Review Houghton
Downloaded by [186.251.144.58] at 08:58 20 August 2015
