NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange 27
B
razil has been unique worldwide in terms of land use. Although
vast areas of forests and savannahs have been converted into
farmland (Fig.1) — placing the country as a leading global
producer of agricultural commodities — it still safeguards the larg-
est tracts of native tropical vegetation on Earth, with extremely
high levels of biodiversity. Patterns of land use change, which until
recently exhibited the highest worldwide absolute rates of tropical
deforestation, largely resulted in low-productivity cattle pastures
2
.
Moreover, climate change issues in Brazil are inextricably related
to land use and land-use change (LUC) as approximately 80% of
the country’s total CO
2
-equivalent (CO
2
e) emissions in 2005 were
sourced from agriculture and LUC
3
.
Demand for farmland is the key immediate driver of LUC
in Brazil, and there is little evidence that agricultural expansion
is grinding to a halt
4–7
. In fact, Brazil holds the greatest potential
for further agricultural expansion in the twenty-rst century
8
.
Understanding recent LUC patterns (Box 1) and visualizing a
sustainable land-use pathway in Brazil have become highly strate-
gic — not only for Brazilians, given that regional and global climate
change, food and energy provision, and biodiversity conservation
are all at stake.
is Review presents an integrated analysis and provides new
insights on recent trends in the Brazilian land-use system. In the
rst two sections we show how Brazil’s agriculture is becoming both
gradually decoupled from deforestation processes and increasingly
intensied and oriented to large-scale farming of trade commodities
throughout the country. Next we explain the economic and political
factors driving those changes. e fourth section reveals the drawbacks
of those changes in aggravating the long history of inequality in land
ownership. We then explore repercussions for climate change, namely
Pervasive transition of the Brazilian land-use system
David M. Lapola
1
*, Luiz A. Martinelli
2
, Carlos A. Peres
3
, Jean P. H. B. Ometto
4
, Manuel E. Ferreira
5
,
Carlos A. Nobre
4
, Ana Paula D. Aguiar
4
, Mercedes M. C. Bustamante
6
, Manoel F. Cardoso
4
,
Marcos H. Costa
7
, Carlos A. Joly
8
, Christiane C. Leite
7
, Paulo Moutinho
9
, Gilvan Sampaio
4
,
Bernardo B. N. Strassburg
10,11
and Ima C. G. Vieira
12
Agriculture, deforestation, greenhouse gas emissions and local/regional climate change have been closely intertwined in Brazil.
Recent studies show that this relationship has been changing since the mid 2000s, with the burgeoning intensification and
commoditization of Brazilian agriculture. On one hand, this accrues considerable environmental dividends including a pronounced
reduction in deforestation (which is becoming decoupled from agricultural production), resulting in a decrease of ~40% in nationwide
greenhouse gas emissions since 2005, and a potential cooling of the climate at the local scale. On the other hand, these changes in
the land-use system further reinforce the long-established inequality in land ownership, contributing to rural–urban migration that
ultimately fuels haphazard expansion of urban areas. We argue that strong enforcement of sector-oriented policies and solving
long-standing land tenure problems, rather than simply waiting for market self-regulation, are key steps to buer the detrimental
eects of agricultural intensification at the forefront of a sustainable pathway for land use in Brazil.
for the country’s greenhouse gas (GHG) emissions balance, and for the
two-way interactions between climate change and land use. Finally, we
discuss the meaning of sustainable land use in Brazil, and suggest how
we can eectively achieve it in the near future.
Decoupling agricultural expansion and deforestation
Although agricultural expansion alone cannot explain the deforesta-
tion rates observed in the past
9
, both processes have long been con-
nected in Brazil
10,11
. is became especially evident in the late 1990s,
with peaks in cropland area and cattle herd size coinciding with peaks
in deforestation in Amazonia and in the Cerrado region (Fig.2).
Since the mid 2000s, annual deforestation trends began to diverge
from uctuations in cropland area and cattle herd size. Although
cropland area and cattle herd continued to increase aer 2004, defor-
estation in all Brazilian biomes plunged to the lowest rates since
monitoring began. e decoupling of agricultural expansion and
deforestation reported for part of the Amazon
12
and elsewhere in the
tropics
13
therefore applies more widely to the whole of Brazil, except
for some subregions such as in the northeast Cerrado, where cropland
expansion is still tied to native vegetation clearing
14,15
.
Nevertheless, the link between agricultural expansion and defor-
estation has weakened rather than disappeared completely, as exem-
plied by the small resurgence in Amazonian deforestation in 2008,
that was driven — in a much weaker way than previously seen— by
increases in cropland area and cattle herd size.
Towards a commoditization of the land
From 1990 to 2011 the land area used for cropping in Brazil grew
from ~530,000 to ~680,000km
2
. Large-scale farming of commod-
ity crops (namely soybean, sugarcane and maize) accounted for all
1
Laboratório de Ciência do Sistema Terrestre, Departamento de Ecologia, Universidade Estadual Paulista, Rio Claro, São Paulo 13506-900, Brazil,
2
Centro
de Energia Nuclear na Agricultura, Universidade de São Paulo, Piracicaba, São Paulo 13400-000, Brazil,
3
School of Environmental Sciences, University
of East Anglia, Norwich NR4 7TJ, UK,
4
Centro de Ciência do Sistema Terrestre, Instituto Nacional de Pesquisas Espaciais, São José dos Campos, São
Paulo 12227-010, Brazil,
5
Instituto de Estudos Sócio-Ambientais, Laboratório de Processamento de Imagens e Geoprocessamento, Universidade Federal
de Goiás, Goiânia 74001-970, Brazil
6
Departamento de Ecologia, Universidade de Brasília, Brasília, Distrito Federal 70919-970, Brazil,
7
Departamento
de Engenharia Agrícola, Universidade Federal de Viçosa, Viçosa, Minas Gerais 36570-000, Brazil,
8
Departamento de Biologia Vegetal, Universidade
Estadual de Campinas, Campinas, São Paulo 13081-970, Brazil,
9
Instituto de Pesquisa Ambiental da Amazônia, Brasília, Distrito Federal 71503-505, Brazil,
10
International Institute for Sustainability, Rio de Janeiro 22460-320, Brazil,
11
Departamento de Geografia, Pontifícia Universidade Católica do Rio de
Janeiro, Rio de Janeiro 22451-900, Brazil,
12
Museu Paraense Emílio Goeldi, Belém, Pará 66040-170, Brazil.*e-mail: dmlapola@rc.unesp.br
REVIEW ARTICLE
PUBLISHED ONLINE: 20 DECEMBER 2013 | DOI: 10.1038/NCLIMATE2056
© 2014 Macmillan Publishers Limited. All rights reserved
28 NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange
of that increase (Fig. 2b). By 1990, the area occupied by these
monocultures represented 53% of all cultivated area, and 21years
later this proportion increased to 70%. Large-scale commodity
farming represented 83% (2011 US$84 billion) of the country’s
gross crop production value in 2011
16
. Contrastingly, although rep-
resenting less than 20% of the gross value of crop production and
occupying only 24% of the country’s farmland, smallholder agri-
culture is responsible for a large fraction of the production of staple
foods in Brazil
17
. e area cultivated with traditional staple crops
such as rice, beans and cassava has contracted by ~30,000 km
2
(−25%). at shrinking, however, has been largely compensated by
production intensication
18
. In fact all Brazilian crops, and cattle
ranching in particular, were subject to a pronounced intensication,
well above world averages
1,6,10,16,19
(Supplementary Fig.1). Although
still inecient in many regions (mean cattle density ≈1 head per
hectare), the stocking density of Brazilian pastures increased ~45%
during the 1990–2011 period
2,18
(Fig.2c). Increased exports of beef
and soybean by 720% and 530%, respectively
1,20
, and the high share
of genetically modied crops in the agricultural matrix (Fig.2b) are
other strong indicators that Brazilian agriculture is turning to an
export-oriented large-scale commodity farming pathway.
Intensication of cattle ranching has been widely shown to be
the central pivot of the land-use transition to more environmentally
friendly agriculture in Brazil
5,6,21–24
, resulting in land being spared
for other uses. is option has to be carefully evaluated, however,
given that — in light of land rent theory
13,25
— agricultural intensi-
cation and its increased economic attractiveness leads to expan-
sion, rather than contraction of cultivated and grazing land
21,26–28
.
So far Brazil is experiencing the opposite eect, with cattle ranch-
ing intensication leading to a reduction, or at least no expansion
of total pasture area, along with (governance-driven) declining
deforestation rates.
Economic and political conjunctures
Increases in both domestic and international demand for beef, feed
and renewable energy, induced by greater purchase power achieved
in Brazil in the past ~15 years coupled with market liberaliza-
tion in Russia and China
5,10
,
have created new agricultural market
opportunities that Brazil and other countries have taken advantage of.
However, some factors specic to Brazil made it feasible to intensify
agriculture without increasing deforestation.
A convergence of conditions such as internal market
regulations, creation of more protected areas, command-and-
control crack-down on illegal deforestation and credit barriers
imposed by the federal government on municipal counties in
deforestation frontiers were largely responsible for the decoupling
of deforestation and agricultural expansion in the Amazon
12,29–32
.
As a consequence of increased enforceable restrictions on illegal
deforestation, it is reasonable to assume that land will become a
scarce resource in frontier areas
21
, such as in southern Amazonia,
which may lead to a positive feedback for agricultural intensi-
cation in those regions, a process now observed in highly
consolidated agricultural areas of southern Brazil
26,33
. In such
consolidated rural areas, strong intensication of cattle ranching
and steps taken by the sugarcane agro-industry to comply with
European market requirements
34
, for instance, have been push-
ing agriculture to set new environmental standards (for exam-
ple, prevention of illegal deforestation) in large-scale commodity
farming.
One of the strongest factors driving the above changes in Brazilian
agriculture was the political power exerted by the large-scale agri-
business sector at the National Congress, and oen at the state and
municipal levels, mainly in the Amazon, Cerrado and Caatinga
regions. e rural caucus is the largest elected ‘interest group’ in the
Brazilian Congress at present, having historically held 20–50% of the
voting power in the Congress lower house since the onset of the re-
democratization process in late 1980s
35,36
. ese legislators have, for
instance, strongly opposed land redistribution and agrarian reform
policies but also inuenced the growth of rural credit availability from
US$15billion in 1990 to US$75billion in 2009 (2009US$)
37
. at
countrywide credit boost allowed heavy investments on agrochemi-
cals for soil improvement in the Cerrado and Amazonian farmland,
and the genetic development of cultivars adapted to tropical cli-
mates
22
, for example
.
It is undisputable that such subsidies to the Brazilian agro-indus-
try contributed to the economic surge in Brazil, as this sector has
accounted for 25% of the country’s gross domestic product (GDP)
over the past two decades. However, it is worth questioning what
societal sectors have beneted from this commoditization process, as
Brazil continues to exhibit — along with a few other countries — the
worst inequalities anywhere on Earth in terms of income and land
ownership distribution
38,39
.
a b
100
5
100
5
Amazon
Pantanal
Atlantic Forest
Pampas
Caatinga
Cerrado
Grid cell
coverage
(%)
Grid cell
coverage
(%)
Figure 1 | Spatial distribution of agriculture in Brazilian biomes in 2000. a, Pastures. b, Croplands. Data from ref. 88. Grid cell size is 5x5 arcmin.
REVIEW ARTICLE
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2056
© 2014 Macmillan Publishers Limited. All rights reserved
NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange 29
Collateral eects on land distribution
Land distribution is a long-term chronic problem in Brazil. e 2006
national census revealed that nearly 75% of all agricultural land area
(2.3millionkm
2
) is in the hands of large-scale commodity-oriented
farmers, who own only 10% of all farm land titles in the country.
Conversely, the remaining 25% of farmland is occupied by small-
holders, who represent 90% of all Brazilian rural properties (Fig.4).
Despite the eventual environmental and socioeconomic gains
achieved during the past decade with the technological improve-
ments and policy regulations of Brazil’s agricultural sector
40
, the
above changes in the land-use system reinforced the historical
inequality in land ownership. In the 1986–2006 period there was
an increase both in the number of large farms and the total area
they occupied. e area occupied by farms larger than 1000 ha
increased 1.6% in this period, representing an extra ~170,000km
2
of largeholdings, namely in the Cerrado and Amazonian agricul-
tural frontiers. Moreover, there has been a decrease in the number
of small farms and the total area occupied by small-scale farming,
particularly in the Caatinga, probably owing to the hardship faced
by smallholders in competing with large-scale commodity farm-
ing
41
and voluntary changes in lifestyle. Some consequences of these
changes in land distribution are discussed below.
Over 80% of the expansion in cropland in Brazil from 1990 to
2011 occurred in the Amazon and Cerrado regions
18
. Amazonia
and northern portions of the Cerrado are also the only regions
where pasture area has increased (at the expense of native veg-
etation) over the past 20years
18,91
. Nevertheless, pasture area evo-
lution over that period for the entire country is debatable, with
Brazil’s ocial data accounting for a ~13% reduction (1.78 to
1.53 millionkm
2
)
18
, whereas Food and Agriculture Organization
(FAO) statistics indicate a ~6% increase (1.84 to 1.96millionkm
2
)
restricted to the 1990s
1
(Fig.2c).
Amazon Region. Since the early 1990s, Brazilian Amazonia
entered a renewed phase of colonization and land use, in which tax
incentives played a lesser role and prots from logging and large-
scale agriculture and cattle ranching, as well as low land prices,
drove much of the frontier expansion
94
. is process has been
supported by government and bilateral investment programs in
infrastructure, such as transport facilities and energy provision
7
.
Despite this pressure, there has been a prominent decline in overall
deforestation since 2005: from an annual average of ~18,000km
2
in the 1990–2004 period to ~10,500km
2
in 2005–2012, with the
lowest rate ever of 4,571 km
2
in 2012 (Fig. 2a) (drivers of this
declining deforestation are addressed in the ‘Economic and politi-
cal junctures’ section). Pastures for beef production remain the
dominant land use, occupying 60% to 80% of deforested land
92
(Fig.1), with regional cattle numbers reaching more than 50mil-
lion
18
heads since 2004.
Cerrado Region. Agriculture now occupies nearly 1millionkm
2
of the Cerrado, or ~50% of the biome’s original extent
18,88
. Cattle
ranching is also by far the dominant land use, but a fraction of
these pastures has been replaced recently by advancing large-scale
mechanized cropping of soybean and sugarcane, for example
18,19,95
.
In fact the Cerrado is Brazil’s most important beef producing
region, hosting the largest extent of pasturelands and ~50% of
the national herd (Fig. 1a). e pronounced conversion of the
Cerrado into soybean monoculture over the past two decades
was one of the main contributors to the expansion in total crop-
land area in Brazil (Fig.2b). However, as in the Amazon, annual
deforestation rates are falling from a mean of ~16,000km
2
in early
2000s to ~6,500km
2
in 2010 (Fig.2a). Yet the high suitability of
the Cerrado topography and soils for mechanized agriculture, the
reduced number and total extent of protected areas
81
(Fig.3b), the
lack of a well-established and routinized deforestation surveillance
program, and potential leakage pressure resulting from declining
deforestation in Amazonia all indicate that the Cerrado will con-
tinue to be a principal region of LUC in Brazil
14,29,53
.
Atlantic Forest Region. e Atlantic Forest biome, an extremely
threatened biodiversity hotspot, hosts most of Brazil’s croplands
(Fig. 1b), and is inhabited by ~125 million people, including
several major metropolitan areas such as São Paulo and Rio de
Janeiro. Only 12% (~160,000 km
2
) of the original vegetation
remains, less than 50% of which is located in protected areas
89,91
.
However, the area of secondary forest has been increasing in
some regions, as predicted by forest transition theory for con-
solidated agricultural frontiers
33
, owing to the widespread transi-
tion to mechanized agriculture (which does not operate in steep
areas), and to market-driven enforcement of environmental laws
(for example, the Forest Code bill). Dominant land uses in the
region are large-scale sugarcane farming and cattle ranching
19,88
with relatively high cattle stocking rates (~2 head per hectare)
spread throughout the biome’s southwest region
2
. In the state of
São Paulo alone, sugarcane cropland increased from ~18,000km
2
in 1990 to ~52,000km
2
in 2011
18
. Although most of this recent
sugarcane expansion is occurring in previous pasture lands
19,44,95
,
it can be argued that the livestock demand once met by these pas-
tures has been at least partly relocated to the Amazon and Cerrado
regions (where pastures have expanded at expense of native veg-
etation)
21,44,96
, even though methods to objectively detect these
indirect LUC have yet to be developed
97
.
Caatinga Region. e semi-arid polygon of northeast Brazil
known as Caatinga comprises 970,000 km² of predominantly
thorn-scrub vegetation. From a total human population of
~21million, 44% live in the rural areas, relying heavily on small-
holder and seasonal agriculture, goat husbandry (the dominant
land use along with subsistence cropping) and rewood harvest-
ing (the major driver of deforestation in the region)
98
. Recent irri-
gation projects have prioritized export-oriented fruit production.
However, major impacts due to poor land management, timber
extraction, poorly planned irrigation projects and increasing fre-
quency of severe droughts are contributing to the expansion of
desertication, with degraded areas accounting for 40,000km
2
,
leading to consequent loss of biodiversity, carbon stocks and soil
structural and chemical properties
99
. Available estimates indicate
that annual deforestation occurred at rates of ~5,900 km
2
yr
–1
over the 1994–2002 period, decreasing aer that to ~1,900km
2
in
2009
15
(Fig.2a).
Pantanal Region. Despite being the most intact biome in Brazil
(only 15% of its original extension has been converted to anthro-
pogenic uses, mostly for cattle ranching)
91
, forestry, the construc-
tion of hydroelectric dams and navigation are building pressure
for LUC in the seasonally ooded Pantanal region. Nevertheless,
the region has also experienced a decline in deforestation over
the past decade
15
(Fig.2a). Marked environmental concerns over
the runaway expansion of sugarcane plantations in neighbouring
biomes motivated a legal ban prohibiting sugarcane monoculture
in the Pantanal (and Amazonia)
100
.
Box 1 | LUC context within major Brazilian biomes.
REVIEW ARTICLE
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2056
© 2014 Macmillan Publishers Limited. All rights reserved
30 NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange
Food security. Despite the steady shrinkage in overall crop-
land area allocated to staples such as rice and beans (Fig.2b),
food crop production has increased owing to yield gains
(Supplementary Fig.1), which dismisses scaremongering in pro-
jections of food scarcity. However, there could be potential eects
on the physical access to food. Although the 4.7% reduction in
the number of smallholdings (Fig.4a) involved an arable area of
only ~12,000 km
2
, this included over 470,000 individual land-
holdings. It can therefore be argued that the livelihoods of at least
470,000 households have probably changed with this transition,
especially in terms of their nancial and physical access to safe
food. Nevertheless, increases in per capita income (mainly in
the Caatinga owing to short-term welfare policies)
40,42
suggests
that access to food may not have been aected, pending more in-
depth research.
Rural migration and urbanization. Brazil has become highly
urbanized in the past decades, with only 15% of the entire pop-
ulation now living in rural areas (Supplementary Table1) as a
consequence of both rural exodus driven by capital input into
Brazilian agriculture and escalating urban job supply
43
. In fact,
this creates a positive feedback in which smaller rural popula-
tions lead to lower reliance on human labour in farming
22,44,45
.
is is particularly favourable to large-scale mechanized farming,
but aggravates the burgeoning population pressure of Brazilian
urban areas
46
, which have grown by approximately 400,000km
2
(+135%) in the 1992–2010 period (Supplementary Table 1).
Urban (instead of rural) population growth was strongly associ-
ated with tropical deforestation in the early 2000s
47
. Heavily com-
moditized regions, such as the sugarcane belt in the state of São
Paulo, have up to 98% of the population living in urban areas
18
.
Such unplanned urban growth has caused severe environmental
and public health problems
48
. is is especially alarming consid-
ering that over 11million Brazilians live in slums with even worse
sanitation conditions, and that these people are amongst the most
vulnerable to climate change in Brazil
49
.
Interactions with climate change
While on the one hand land use in Brazil has been reported to be
subject and susceptible to global climate change, on the other hand it
is also a driver of climatic changes at the local and regional scales. In
this section we explore these two-way interactions between land use
and climate in Brazil, as well as related changes in GHG emissions.
Coping with the eects of global climate change. Agriculture
in Brazil is frequently exposed to the eects of climate extremes.
For instance, the 2005 drought in western Amazonia impacted
agricultural production and food security
50
; the 2010 oods in
southern Brazil destroyed one-seventh of the rice production
in the state of Rio Grande do Sul; and climatic adversities in
2010–2011 inuenced a sugarcane shortfall that forced the sec-
tor to make large imports of ethanol to meet overall demand.
Both large-scale farmers and smallholders are vulnerable to these
extremes in Brazil today, although the impacts on livelihoods are
only undisputable for smallholders
51,52
.
For the future, the government’s outlook on agricultural growth
singles out global climate change as a large source of uncertainty
that will steer the magnitude of production growth in the next
decades
4
. Some studies on the impact of future climate change on
the yields of crops that are widely cultivated in Brazil
consistently
point out substantial losses in productivity if no adaptation meas-
ures are taken (especially for soybean crops)
53,54–57
. e exception
is sugarcane, the productivity of which is projected to increase
throughout the country. However, fewer studies
53,56
attempted
to quantify how these yield changes could inuence land use. In
fact, the results of these studies were more relevant to elucidat-
ing cause–eect relationships within the Brazilian land system,
rather than predicting future land-use patterns. e geography
of Brazilian agriculture under future scenarios of climate change
therefore remains largely undetermined, given uncertainties
regarding the CO
2
fertilization eect on crop yields
58
, highly-var-
iable projections of rainfall
59
and the evolution of both prices and
0
100
200
300
400
500
600
700
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Crop area (×10
3
km
2
)
b
Other crops Rice
Beans Maize
GM maize Sugarcane
Soybean GM soybean
120
140
160
180
200
220
1,200
1,400
1,600
1,800
2,000
2,200
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Cattle herd (million heads)
c
IBGE pasture area
FAO pasture area
Cattle herd
0
0.3
0.6
0.9
0
5
10
15
20
25
30
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Deforestation
(×10
3
km
2
yr
–1
)
a
Amazon
Cerrado
Caatinga
Atlantic Forest
Pantanal
Pampas
Pasture area (×10
3
km
2
)
Year
Year
Year
Figure 2 | Trends in land-use change and agricultural expansion in
Brazil during the 1990–2012 period. a, Deforestation rates in all Brazilian
biomes
3,15,89,90
. Remaining native primary vegetation area in each biome:
Amazon: 80%; Cerrado: 51%; Caatinga: 54%; Atlantic Forest: 12%;
Pantanal: 85%; Pampa: 46%
89,91,92
. b, Area under each crop type
18
. GM,
genetically modified. c, Total pasture area and cattle herd size
1,18
; asterisks
represent the values provided by Brazil’s ocial census data and the red
dashed line indicates the trends between these values.
REVIEW ARTICLE
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2056
© 2014 Macmillan Publishers Limited. All rights reserved
NATURE CLIMATE CHANGE | VOL 4 | JANUARY 2014 | www.nature.com/natureclimatechange 31
investments in agricultural adaptation to climate change. As a
rst-order estimate, 2011 US$480–570million should be invested
yearly until 2050 to adapt Brazilian agriculture to the ravages
of climate change
56
. In light of that, smallholder and subsist-
ence farmers, who are far less economically and institutionally
supported than commodity farmers, will certainly be heavily
impacted by climate change
51,52
.
Eects on local and regional climate. Continuing changes in the
Brazilian land-use system may also imply alterations in the local
and regional climate. Although past studies revealed a local and
regional warming originating from forest and savannah conversion
into pastures
60–62
, now the encroachment of pastures by commodity
crops leads to divergent biosphere–atmosphere relationships. In the
Cerrado biome, for example, changes in albedo and evapotranspi-
ration cause an average regional warming of ~1.6°C driven by the
replacement of the natural vegetation by cropland or pasture, and a
cooling of ~0.9°C when those pastures are converted to sugarcane
elds
63,64
. Conversely, modelling studies suggest that the large-scale
substitution of Amazonian pastures by soybean elds will lead to
local warming and reduced precipitation compared with pastures.
is is due to the marked increase in albedo caused by a decrease
in leaf area index, which consequently reduces evapotranspiration in
soybean elds between growing seasons
65,66
. However, the net eect is
likely to be a cooling of surface temperatures, considering maize culti-
vation in the soybean o-season, a prevalent double-cropping system
in most Brazilian soybean farms.
Although it is still uncertain whether these changes in land surface
properties will lead to regional climatic change, there are indications
that they will result in important local and seasonal eects
64
. And
importantly, the reduction in deforestation and concomitant mainte-
nance of regional forest–atmosphere heat and moisture uxes gradu-
ally moves the country away from the worst regional climate changes
projected from large-scale substitution of Amazonian forests and
Cerrado savannahs by pastures and croplands
60,65–67
or from degra-
dation of the Caatinga vegetation
68
.
Despite the countrywide reduction in deforestation in the past
decade, the number of re spots detected yearly over the same
Fire scars
a
b
c
d
1992
Biodiversity conservation
areas
Biome limits
2010
Indigenous people
reservation
Biome limits
Biome limits
Biome limits
Roads
Roads
0 1,000 km500
In operation
Planned
In implementation
Figure 3 | Biofuels, roads, protected areas and fire in Brazilian biomes. a, Bioenergy (ethanol) plants and road infrastructure
5
. b, Protected areas.
c, Urban areas in 1992 and 2010 (as detected from nightlight glow)
93
. d, Fire spots detected in the period 2002–2012
75
. The scale bar in a applies to all panels.
REVIEW ARTICLE
NATURE CLIMATE CHANGE DOI: 10.1038/NCLIMATE2056
© 2014 Macmillan Publishers Limited. All rights reserved
