Regional Environmental Change

, Volume 18, Issue 1, pp 129–142 | Cite as

Future land use and land cover in Southern Amazonia and resulting greenhouse gas emissions from agricultural soils

  • Jan Göpel
  • Jan Schüngel
  • Rüdiger Schaldach
  • Katharina H. E. Meurer
  • Hermann F. Jungkunst
  • Uwe Franko
  • Jens Boy
  • Robert Strey
  • Simone Strey
  • Georg Guggenberger
  • Anna Hampf
  • Phillip Parker
Original Article


The calculation of robust estimates of future greenhouse gas emissions due to agriculture is essential to support the framing of the Brazilian climate change mitigation policy. Information on the future development of land use and land cover change (LULCC) under the combination of various driving factors operating at different spatial scale levels, e.g., local land use policy and global demands for agricultural commodities, is required. The spatially explicit land use model, LandSHIFT, was applied to calculate a set of high-resolution land use scenarios for Southern Amazonia. The time frame of the analysis was 2010–2030. Based on the generated maps, emission coefficients were applied to calculate annual N2O, CH4, and CO2 emissions from agricultural soils (croplands and pastures). The results indicate that future land use pattern and the resultant greenhouse gas emissions in Southern Amazonia will be strongly determined by global and regional demands for agricultural commodities, as well as by the level of intensification of agriculture and the implementation of conservation policies.


Land use and land cover change Southern Amazonia Scenarios Agriculture Greenhouse gas emissions 


The last decades have seen Southern Amazonia characterized by the conversion of large areas of forest and Cerrado ecosystems to agricultural land (Kohlhepp 2002; Coy 2001), resulting in high CO2 emissions from soils and vegetation in the phase of land clearing (Schmidt et al. 2011; Krogh et al. 2003), as well as substantial losses of biodiversity (Martinelli et al. 2010). In the 1990s and early 2000s, deforestation was one of Brazil’s largest contributors to greenhouse gas emissions (e.g., Lapola et al. 2014). The main drivers of these land use and land cover changes (LULCC) were an accelerated population growth, together with an ongoing trend of urbanization, and the expansion of cropland and pastures due to an increasing global demand for fodder and energy crops, meat, and timber (Godfray et al. 2010a, b). Agricultural land use currently contributes ~ 35% to Brazil’s greenhouse gas emissions (FAO 2014) in the form of N2O and CH4. Consequently, these processes are important drivers of global climate change.

After 2005, increasing agricultural production could be considerably decoupled from further deforestation because of (1) successful initiatives of the Brazilian government and local authorities to protect natural ecosystems (Gibbs et al. 2015a; Nepstad et al. 2014) and (2) the decrease of world market prices for soybean (Hecht 2011; Nepstad et al. 2009), in combination with a further intensification of agricultural systems (Cohn et al. 2014; Macedo et al. 2012). However, several authors expect deforestation and expansion of agricultural land to increase again (e.g., Assunção et al. 2012; Malingreau et al. 2012) as soon as Brazil’s cattle agreement runs out in 2019 (Gibbs et al. 2015b) and world market prices for soy and cotton eventually rise further. Therefore, the analysis of future LULCC trajectories is an essential element for framing policies that aim at reducing land use-related greenhouse gas emissions as an important means of climate change mitigation.

Land use models in combination with scenario techniques are suitable tools to systematically explore future LULCC trajectories (Lambin et al. 2000) because they can capture the interplay between multiple drivers such as agricultural intensification, regional governance and linkages between regional and global markets (Mietzner and Reger 2005; Veldkamp and Lambin 2001). In the last couple of years, different land use models were used to study deforestation processes (Arvor et al. 2013; Aguiar et al. 2007; Soares-Filho et al. 2004, 2006; Walker et al. 2004), direct and indirect LULCC due to biofuel production (Lapola et al. 2010), and the effect of climate change on agricultural expansion (Lapola et al. 2011) within the Amazon region. An overview of these studies is given by Dalla-Nora et al. (2014), who identify two main research needs. Firstly, a more detailed description of the drivers of land use change (e.g., agricultural production) as model input in the form of scenarios and secondly, a more detailed representation of regional land use-related policies (e.g., Soy Moratorium, Cattle Agreement, and Brazilian Forest Code) and management practices (e.g., intensive pasture) within the models. Moreover, while there are several studies investigating greenhouse gas emissions from land use and land cover change such as forest clearing (e.g., Lapola et al. 2010; Fearnside et al. 2009), only few analyses also take into account emissions from agricultural soils (e.g., Galford et al. 2010).

The objectives of our study are twofold. The first is to translate a new set of narrative scenarios (storylines) for Southern Amazonia (Schönenberg et al. 2017) that was developed within the CARBIOCIAL project ( into numerical model input data. The resulting quantitative scenarios go beyond assumptions regarding deforestation and include national and international drivers of land use change (e.g., demands for agricultural commodities and yield increases due to intensification) as well as a detailed representation of regional land use policies (e.g., Brazilian Forest Code). We apply the spatially explicit land use model LandSHIFT (Schaldach et al. 2011) to simulate these quantitative scenarios resulting in a set of land use maps. The second objective is to calculate the resultant greenhouse gas emissions (CO2, N2O, and CH4) from agricultural soils in order to provide insights into the potential role of future agricultural development in Brazil for climate change mitigation.

Material and methods

Study area

The study focuses on the two federal states—Mato Grosso and Pará—in Brazil (Fig. S1), which contain 36 municipalities blacklisted as so-called priority municipalities in terms of monitoring and repressing deforestation through an optimized monitoring system and stricter environmental law enforcement, respectively (MMA 2001). These constitute only 6.6% of the area of all municipalities within the Legal Amazon (IBGE 2014) but accounted for almost 45% of deforestation within the Amazon in 2012. Hotspots of deforestation can be found around the “development highways” (BR-070, BR-158, BR-163).

Mato Grosso has an area of 907,000 km2 and a population of 3.2 million people (IBGE 2013); 69,807 km2 of land is used for soybean cultivation (IBGE 2015), and 1149 km2 was deforested in 2013, which constitutes an increase of 52% in comparison with 2012 (INPE 2013). Another dominant land use sector is cattle ranching, with a total herd size of 28.4 million animals (IBGE 2015). Here, the expansion of area used for soybean cultivation and cattle ranching could be identified as the primary cause of conversion of natural ecosystems to agricultural land (Greenpeace-Brazil 2009; Barona et al. 2010).

Pará has an area of 1.25 million km2 and a population of 8 million people (IBGE 2013). Only 11,969 km2 of the land is used for soybean cultivation (IBGE 2015). In 2013, 2379 km2 was deforested, which shows an increase of 37% in comparison with 2012 (INPE 2013). The dominant land use sector is cattle ranching, with a total herd size of 19.2 million animals (IBGE 2015). The natural vegetation is dominated by dense rainforest (Vieira et al. 2008). A hotspot of LULCC is along the Cuiabá-Santarem highway (BR-163), the most recent of the “development highways” used to acquire the agriculturally underdeveloped northern parts of Brazil for crop cultivation and cattle ranching (Coy and Klingler 2008).

Modeling of land use and land cover change

Model description

Land use and land cover change were simulated with the spatially explicit LandSHIFT model. The model is fully described in Schaldach et al. (2011) and has been tested in different case studies for Brazil (Lapola et al. 2010, 2011). It is based on the concept of land use systems (Turner et al. 2007) and couples components that represent the respective anthropogenic and environmental sub-systems. In our case study, land use change is simulated on a raster with the spatial resolution of 900 m × 900 m that covers the territories of the federal states of Mato Grosso and Pará. LandSHIFT simulates the spatiotemporal dynamics of settlement, cropland, and pasture by regionalizing their state-level drivers to the raster level in 5-year time periods. These drivers include human population, livestock numbers, crop production, and crop yield increases due to technological change. Input on cell-level comprises the state variables “land use type” (Table S1) and “human population density,” as well as a set of parameters that describe its landscape characteristics (e.g., terrain slope), road infrastructure, and zoning regulations.

Important elements of the scenario storylines (Schönenberg et al. 2017) were related to land use policies and agricultural intensification. In order to reflect these assumptions more accurately, LandSHIFT has been modified by integrating a new agricultural land use type and a new process that describes the further intensification of pasture management. The Legal Intensification Scenario presumes compliance with the new Brazilian forest code (Soares-Filho et al. 2014). For this purpose, a new mosaic land use type (Mosaic Legal Reserve) was implemented. This land use type consists of 20% cropland or pasture and 80% rainforests, which reflect the requirement of the forest code to protect a certain amount of native vegetation on farms. In order to represent the intensification of pasture management described in some of the investigated scenarios, the model includes a new sub-module for calculating the increase of biomass productivity of each pasture cell per time step (specified by the parameter intensification rate) until a defined maximum is reached (parameter maximum intensity).

Model input data

LandSHIFT is initialized with a gridded historic land use map representing the year 2010, which combines MODIS land cover data (Friedl et al. 2010) and census data on the municipality level regarding cropland and pasture area (IBGE 2015). Human population density for each grid cell was derived from the population density data set published by Salvatore et al. (2005). Moreover, the model input comprises spatial datasets in regard to landscape characteristics, road infrastructure, and zoning regulations. Grid level information on terrain slope is based on the SRTM30 data set (Farr and Kobrick 2000). Information on the river network and the road network were derived from the Banco de Nomes Geográficos do Brasil database (IBGE 2012). Spatial data sets with the location of military areas, ecological, and indigenous protected areas were provided by the ZONEAMENTO ecológico-econômico da área de influência da Rodovia BR-163 (Cuiabá-Santarém) (Embrapa Amazônia Oriental 2008) and the Ministério do Meio Ambiente (MMA 2013), respectively. Crop yields and biomass productivity of pasture were calculated with the LPJmL model (Bondeau et al. 2007) on a 0.5° raster for current climate conditions (averaged over the reference period 1971–2000). Data on monthly precipitation, air temperature, cloud cover, and frequency of wet days was taken from the CRU TS 2.1 dataset (Mitchell and Jones 2005). Additional datasets (soil texture, soil moisture, and atmospheric CO2 concentration) were applied according to Sitch et al. (2003). An evaluation of the LPJmL modeling results can be found in Lapola et al. (2009). The simulation results from LPJmL were converted to the 900 × 900-m raster by assigning the respective values to all cells located within each 0.5° cell.

Main input data for the land use simulations from 2010 until 2030 is provided on the state level and includes human population, crop production, crop yield increases due to technological change, and livestock numbers (see Tables S2, S3, and S4). This data is specified as part of the scenarios.

Calculation of N2O and CH4 fluxes from agricultural soils

We used the average N2O emissions reported in Meurer et al.’s (2016) review for different land use types in Brazil. Meurer et al. (2016) showed the non-linear relation between N2O fluxes from soils and pasture age (years since conversion), and hence distinguished between pastures younger and older than 10 years. As we do not have information about the age of the existing pastures in the base year 2010, we consider these pastures to be older than 10 years. For the pastures established during the scenario period between 2010 and 2030, we included the age and applied the corresponding average emissions for the estimation of total annual N2O fluxes. For methane, cropland is reported to be a sink for atmospheric CH4, although positive fluxes from pastures were reported by almost all references included in this study.

Calculation of CO2 fluxes from agricultural soils

CO2 emissions and uptake from agricultural ecosystems were derived from changes in soil organic carbon (SOC) stocks under different land uses. These data were derived in the course of field trials conducted within the Carbiocial project (see Supporting information 3). Soil samples were taken from 29 plots in the study region according to the methods described in the Supporting information. To include the most common soil types of the Amazon region (Quesada et al. 2011), the analysis concentrated on Ferralsols and Acrisols. Additionally, old (> 10 years) and young (≤ 10 years) pastures and croplands were distinguished in order to capture the specific potentials to absorb or emit CO2. The result of the analysis was a set of SOC stocks and SOC stock changes (Table 1) that was applied to determine annual carbon fluxes for the different land use scenarios.
Table 1

Mass-corrected SOC stocks and SOC stock changes for different land use types on Ferralsol and Acrisol for topsoil (0–30 cm), subsoil (30–100 cm), and the complete sampling depth (0–100 cm)



SOC stocks

SOC stock changes

0–30 cm

30–100 cm

0–100 cm

0–30 cm

30–100 cm

0–100 cm

Mg SOC ha−1

SE ±

Mg SOC ha−1

SE ±

Mg SOC ha−1

SE ±

Mg SOC ha−1

SE ±

Mg SOC ha−1

SE ±

Mg SOC ha−1

SE ±


















 Young pasture (≤ 10 years)














 Old pasture (> 10 years)














 Young crop-field (≤ 10 years)














 Old crop-field (> 10 years)























 Young pasture (≤ 10 years)














 Old pasture (> 10 years)














 Young crop-field (≤ 10 years)














Due to mass-correction, the sum of topsoil SOC and subsoil SOC might not be similar to 0–100 cm

SOC, soil organic carbon; n, the amount of individual sampling points; SE, standard error

Scenario storylines and model drivers

An important outcome from the CARBIOCIAL project was a set of four scenarios that describe plausible future development pathways of Southern Amazonia until the year 2030. Each scenario includes a qualitative part (storyline) that provides a short narrative of the future world, and quantitative information that describes the respective main drivers of LULCC (Schönenberg et al. 2017). Input data for the LandSHIFT simulations provided by each scenario include human population (Table S2), livestock numbers (Table S3), crop production and crop yields (Table S4), as well as assumptions regarding intensification of pasture management and environmental conservation.

The storyline of the Trend Scenario describes a growing production of agricultural commodities in the study region. At the same time, further intensification of the agricultural sector leads to increasing crop yields. Natural ecosystems that are not located in protected areas are still converted into cropland and pasture. Migration processes lead to a strong population increase.

The Legal Intensification and the Illegal Intensification Scenario are characterized by a further increase of crop production and livestock numbers due to a growing demand for these agricultural commodities from Asian countries. Crop yields increases are similar to the Trend Scenario. Additionally, the scenarios presume the intensification of cattle ranching as described in “Model description.” In Pará, we assume an intensification rate of 4.5% per time step up to a maximum of 30%. That means that the biomass productivity of any pasture grid cell is increased by 4.5% until biomass productivity is 30% higher than in the base year. As agriculture in Mato Grosso is presumed to be more mechanized, large scale, and world market oriented (Jasinski et al. 2005; Arvor et al. 2012), we assume an intensification rate of 9% up to a maximum value of 50%. These assumptions are based on observed pasture intensification rates in Brazil. According to Wint and Robinson (2007) and Lapola et al. (2014), the stocking density of pastures in Brazil rose continuously from 1990 to 2010, with a total increase of 45% during that period. The two scenarios differ in respect of the assumed enforcement of environmental law. Under Legal Intensification, the conversion of protected areas of any kind is not allowed. In addition, we assume compliance with the Brazilian Forest Code, which implies that cropland and pasture expansion is realized as the new mosaic land use type, leaving 80% of natural land on the newly converted grid cell intact. In contrast, the Illegal Intensification Scenario is characterized by weak law enforcement. Here, areas under ecological conservation status are de facto available for agricultural use. Also, the compliance with the Brazilian Forest Code does not apply.

The Sustainable Development Scenario describes a society that enjoys a social model based on participation, citizenship, an inclusive economic system with clear land titles, and strong law enforcement. Natural resources are well-protected. Due to a global shift towards a more vegetarian diet that is oriented on WHO recommendations (e.g., Srinivasan et al. 2006; Amine et al. 2002), we find a strong decrease in livestock numbers and a significant increase in crop production (soybeans, beans, fruits, and vegetables) for compensating the calorie intake formerly realized by animal products. Due to less immigration from other parts of Brazil, population increase is lower than in the other scenarios.

Modeling protocol

Using the scenario drivers (see Tables S2, S2, and S4) as input, the LandSHIFT model generates maps for 2010 until 2030 in 5-year time steps that depict the resulting land use pattern. For further analysis, we aggregated the 12 crop types into the land use class cropland, the five forest types into the class forest vegetation types (Rainforest), and the two savannah vegetation types into the class savannah vegetation types (Cerrado) according to Table S1. Changes in location and area of the respective land use types were determined by comparing the maps for 2010 and 2030 using GIS software.

In the second step of the analysis, we first classified cropland and pasture in the 2030 map as “new” and “old” and additionally assigned each cell in the 2010 and 2030 maps the Ferrasol and Acrisol soil type. Then, the annual N2O and CH4 emissions as well as CO2 emissions (derived from SOC stock changes) were calculated for each cropland and pasture cell in both maps using the emission coefficients described in “Calculation of N2O and CH4 fluxes from agricultural soils” and “ Calculation of CO2 fluxes from agricultural soils.” In order to compare the values of emission of the greenhouse gases CO2, N2O, and CH4, the emitted amounts were converted into global warming potential (GWP) [CO2e] according to Myhre et al. (2013) in relation to a time horizon of 20 years (GWPCH4 = 86, GWPN2O = 268).

In all of our scenarios, we assume that increases of crop yields can be achieved until 2030 by technological improvements and a more intensive management. Potential negative effects of climate change on crop yields as discussed, e.g., by Lapola et al. (2011), were not considered in the analysis. In order to investigate if climate change might hinder the further increase of crop yields, we have conducted an additional simulation study with the crop model MONICA (Nendel et al. 2011). For detailed information regarding the applied method and results regarding climate-driven crop yield changes refer to Supporting information 4.


Model output

The main model output comprises the time-series of grid maps showing land use type, as well as population and livestock densities. Figure 1 shows the simulated land use maps for the base year 2010 and the scenarios in 2030. Based on these maps, aggregated information on the state level is produced, including area quantities of each land use type. Figure 2 shows the land use and land cover change for each of the scenarios between 2010 and 2030 and the resultant changes of annual GHG emissions. Figure 3 presents total global warming potential (GWP) [CO2e] for the greenhouse gases CO2, N2O, and CH4 in 2030.
Fig. 1

Simulation results from LandSHIFT. Land use maps of Southern Amazonia in the year 2010 and for the four Carbiocial scenarios in 2030. For a description of the land use types and their aggregation, see “Model description” and Table S1

Fig. 2

Land use and land cover change in Pará and Mato Grosso between 2010 and 2030 and resultant annual GHG (CO2, N2O, CH4) emissions in 2030. (a Land use and land cover change Pará 2010–2030. b Land use and land cover change Mato Grosso 2010–2030. c Annual GHG emissions Pará 2030 d Annual GHG emissions Mato Grosso 2030)

Fig. 3

Annual emissions of CO2, N2O, and CH4 in total global warming potential (GWP) [CO2e] in Pará and Mato Grosso in 2030

Trend Scenario

In Pará, the loss of tropical rainforests amounts to 113,370 km2 (− 11.5%), while 12,879 km2 of Cerrado vegetation is converted into urban and agricultural land. The majority of deforestation can be found in close proximity to the newly established “development highways” (BR-163, BR-230) and along the eastern border of the state. The largest fraction of the converted land is used for pasture (102,271 km2), which almost doubles in comparison with 2010. Cropland expands by 16.4%, from 147,960 to 172,190 km2, despite the assumed yield improvements due to technological change (i.e., more efficient crop varieties; improved agricultural management). Urban area expands from 599 to 640 km2 by 6.9%.

In Mato Grosso, 34,360 km2 (− 20.1%) of Cerrado are converted, followed by rainforest with 30,136 km2 (− 8.4%) and grassland with 2143 km2 (− 11.1%). Most of the loss of natural vegetation can be witnessed along the BR-163 (central north-south axis) with ongoing expansion to the east and west from this starting point, along the east-west axis in southern Mato Grosso (BR-070), and along the eastern north-south axis (BR-158). The area in central southern Mato Grosso (Pantanal) is not affected by LULCC as it is defined as a nature conservation area. Pasture area expands from 168,198 to 252,786 km2 (+ 50.3%). Cropland decreases by 8.3%.

In Pará, annual N2O fluxes more than double due to the expansion of pasture area and cropland. In Mato Grosso, total annual emissions from pasture soils almost double between 2010 and 2030. Eighty-three percent of pastures in 2030 are older than 10 years and account for 63% (0.02 Mt) of the total annual N2O fluxes from pasture. The slight decline in cropland leads to an emission decrease. The emission patterns are the same for methane, with the difference that most of the fluxes are negative and thus, soils are a CH4 sink. The only exceptions are pastures, since the emission coefficient assumed accounts for CH4 fluxes from the soil to the atmosphere. In Pará, annual CO2 emissions from agricultural soils increase from 70.37 Mt. in 2010 to 215.91 Mt. in 2030. During the same period, in Mato Grosso, annual CO2 emissions rise from 38.42 to 224.43 Mt. The main contributor in both states is old cropland (> 10 years), followed by old pasture. Annual uptake by young cropland in 2030 amounts to 0.70 Mt. in Pará and 3.51 Mt. in Mato Grosso, respectively. Total annual CO2, N2O, and CH4 emissions in 2030 add up to 463.9 Mt. CO2e.

Legal Intensification Scenario

In Pará, 57,339 km2 of rainforest (− 5.8%) and 14,721 km2 of Cerrado vegetation (− 59.7%) are converted. Urban area increases by 6.2%, from 599 to 636 km2. Cropland increases by 50.5%, from 134,641 to 222,677 km2. In contrast, pasture is slightly decreasing. The results for Mato Grosso indicate a loss of 8937 km2 of Cerrado and grassland ecosystems. The converted area is utilized mainly for crop cultivation, with cropland increasing from 221,389 to 240,987 km2 (+ 8.9%), while pasture decreases from 168,198 to 157,536 km2 (− 6.3%).

In both states, the decrease of pasture can be attributed to the intensification of grazing management (see “Model input data”), which is sufficient to absorb the additional pressure by increasing livestock numbers and therefore, seems to be a suitable tool to substantially reduce LULCC.

Due to the compliance to the Brazilian Forest Code (see “Scenario storylines and model drivers”), the resulting land use pattern has a very different characteristic compared with the Trend Scenario (Fig. 1). As we have previously described, the newly allocated cropland cells have a mosaic land cover consisting of 20% cropland or pasture and 80% of the original natural land cover type (see “Model description”). This might have positive effects, for example, on biodiversity compared with larger agricultural entities (Wright et al. 2012), but de facto means that human activities affect a larger area. Another negative side effect is that cells with potentially high crop yields can only partly be used for crop production. As a result, the production that could have been generated on this land has to be realized on additionally converted land with lower crop yields, which will lead to an over-proportionally expansion of cropland.

Total annual N2O fluxes slightly increase in Pará and Mato Grosso mainly due to an expansion of cropland. In contrast, annual N2O fluxes from pastures are slightly lower (~ 5%) in 2030 as compared with 2010. This decline is more than compensated by the mentioned increase of N2O fluxes from cropland. At the same time, the soils’ uptake of atmospheric CH4 increases in both states, also because of an expansion of cropland. In Pará, annual CO2-emissions from agricultural soils increase from 70.37 Mt. in 2010 to 244.14 Mt. in 2030. In Mato Grosso, we find an increase to 242.04 Mt. Similar to the Trend Scenario, old cropland and old pasture are the main sources. Annual CO2 uptake by young cropland amounts to 0.24 Mt. in Pará and 4.60 Mt. in Mato Grosso. Total annual CO2, N2O, and CH4 emissions from agricultural soils in 2030 amount to 499.97 Mt. CO2e.

Illegal Intensification Scenario

In Pará, 99,377 km2 of tropical rainforest is converted. Cerrado decreases by 62.2% from 24,648 to 9327 km2. Grassland and Cerrado diminish almost completely. Urban area spreads by 6.9%, from 599 to 640 km2. Cropland increases by 26.5%. The scenario assumes the possibility to convert land that is under conservation (e.g., nature reserves), thus opening up spaces that are not allowed for conversion in all other scenarios. Such areas are mainly located in north-western and in north-eastern Pará on an east-west axis on the Ilha de Marajó. At the same time, we see that pressure is relieved from regions less favorable for crop cultivation (e.g., due to lower potential crop yields). Examples can be found in north-eastern Pará, east of Baía de Marajó, and in south-eastern Pará, north of the state border to Mato Grosso. Pasture expands by 74.4% mainly in the regions of south-western Pará (Parque Nacional do Rio Novo) and west of the BR-163 that are favorable for conversion due to their proximity to roads and urban centers. Further areas are acquired in the north-western region of Pará close to Bahia and to the north of Bahia. Generally speaking, the opening up of regions formerly protected due to their ecological richness leads to an increased destruction of rainforest and other natural vegetation cover.

In Mato Grosso, rainforest cover only slightly decreases by 0.2%. The largest share of land conversion is at the expanse of Cerrado vegetation, as it diminishes by 6.8%. Most of this conversion is located in southern Mato Grosso (Pantanal). Large areas are also converted along a north-south axis in central Mato Grosso and along the courses of the rivers Rio Juruena and Rio São Manuel; 9.9% of grassland vegetation is converted. Urban area is estimated to expand by 0.3%. Most of the natural vegetation cover (27,939 km2) is converted to cropland, corresponding to an expansion by 12.6%. At the same time, pasture area decreases by 10.5%. Due to expansion of pasture taking place mainly in central southern Mato Grosso, large areas of rainforest can be spared from deforestation, e.g., along the courses of the rivers Rio Juruena and Rio São Manuel and west of the Rio Xingú along a north-south axis. This again shows the suitability of intensification measures to preserve rainforest, but also the necessity to legally protect rare and ecologically valuable zones from destruction.

In Pará, total annual N2O fluxes from agricultural soils increase by 68% from 2010 to 2030 due to an expansion of cropland and pasture. In Mato Grosso, total annual N2O fluxes amount to 0.034 Mt. for 2030. This is an increase of 2.4% compared with 2010; 1.4% of total pasture emissions are derive from young pastures. Uptake of atmospheric methane is reduced in Pará but increased in Mato Grosso. While the uptake reduction in Pará is mainly driven by an expansion of pasture area, in Mato Grosso, the loss of pasture area and expansion of cropland lead to decreasing annual CH4 emissions to the atmosphere and an increasing CH4 uptake. Annual CO2 emissions from agricultural soils increase to 227.14 Mt. in Pará and to 239.74 Mt. in Mato Grosso. Main sources are old cropland and old pasture. As a consequence of the increasing pasture area in Pará, annual emissions from young pasture also play a significant role (10.31 Mt). Again, young pastures act as a carbon sink in both states, with an annual uptake of 1.25 Mt. CO2 in Pará and of 4.33 Mt. CO2 in Mato Grosso. Total annual CO2, N2O, and CH4 emissions are equal to 483.73 Mt. CO2e.

Sustainability scenario

In Pará, 3766 km2 of natural vegetation cover are converted into croplands. As rainforest is fully protected (see “Scenario storylines and model drivers”) the majority of the converted area (98.5%) is Cerrado vegetation. Caused by the lower meat consumption (see “Scenario storylines and model drivers”), pasture areas considerably decrease until 2030. In total, 89,038 km2 (− 86.7%) of the original pasture (2010) can be released. In contrast, cropland increases by 53.9%, from 134,641 to 227,636 km2. Most of the newly established cropland area is found in areas that were formerly used for grazing (characterized by relatively high crop yields) and could be released due to the declining demand for meat. Similar to the Trend Scenario, new cropland is located in regions west of Rio Tocatins, along the “development highways” (BR-163, BR-230) and around the shores of the Amazonas in western Pará (close to Santarém).

In Mato Grosso, 36,731 km2 (− 21.5%) of Cerrado is lost. Also, grassland vegetation diminishes considerably, by 5761 km2 (− 30.1%). Similar to Pará, rainforest area is protected (see “Scenario storylines and model drivers”) and remains constant. As expected, most of the converted area is used for crop cultivation. Consequently, cropland expands from 212,389 to 266,481 km2 (+ 20.4%). Triggered by the shift in diets, pasture area is slightly decreasing.

In total, annula N2O fluxes are reduced by 2.9% in Pará but increase by 10.3% in Mato Grosso. In Pará, the highest emission reduction is caused by the decrease of pasture area, whereas cropland expansion increases annual N2O fluxes. The increase of cropland in Mato Grosso and thus, the increase of N2O emissions, surmount the decrease due to a reduction of pasture area, and lead a total annual N2O emission increase of 10%. The total annual uptake of atmospheric methane increases by 0.022 Mt. in Pará and by 0.006 Mt. in Mato Grosso. The main driver in Pará is the reduced pasture area, which leads to 0.01 Mt. less annual CH4 fluxes to the atmosphere. The decrease of annual CH4 emission in Mato Grosso is mainly resulting from an expansion of cropland that functions as a sink of atmospheric CH4. Due to the strong expansion of cropping area, young cropland forms a significant carbon sink. Annual CO2 uptake amounts to 29.99 Mt. in Pará and 24.92 Mt. in Mato Grosso, respectively. Nevertheless, in Pará, annual emissions increase to 196.50 Mt. and in Mato Grosso to 224.86 Mt. in 2030. The main contributor in both cases is old cropland, while emissions from pasture are lower than in the other scenarios due to the decline of pasture area. In the case of the Sustainable Development Scenario, we calculated total annual CO2, N2O, and CH4 emissions from agricultural soils in 2030 to be 432.70 Mt. CO2e.


Differences between land use and land cover change scenarios

The largest reduction of rainforest was simulated for the Trend Scenario. The main driver is the expansion of pasture. This is typical for the dynamics of pioneer frontier development in this region of Brazil, where newly deforested area is first converted into pastures, and after being used for several years, converted into cropland (e.g., Coy and Klingler 2008; Pacheco 2012). The loss of rainforest could be considerably reduced in the Legal Intensification Scenario (compliance with the Brazilian Forest Code), indicating that, especially in Pará, effective governance and conservation of natural habitats play an important role in reducing deforestation (Arima et al. 2014). The compliance with conservation policies leads to a reduction of deforestation by 35% in the case of the Legal Intensification Scenario in comparison with the Illegal Intensification Scenario. These results are reinforced by Soares-Filho et al. (2010), who argues that 37% of deforestation could be halted by the establishment of new protected areas. However, especially in Pará, at the frontline of the agricultural frontier, land holders often do not acquire large parcels of land and split them into 80%/20% shares, but rather, acquire small parcels of rainforest connected to one of the development corridors. A split of these small parcels according to the regulations of the Brazilian Forest Code imposes a high risk of habitat fragmentation, which is an important factor for the loss of species diversity (Chaplin-Kramer et al. 2015). Additionally, this development does not occur from the edge of the rainforest inward, but rather, from the inside. This might lead to higher carbon losses due to the higher amount of carbon stored in the central parts of the forest compared with the forest edges (Chaplin-Kramer et al. 2015).

For both states, the highest increase of pasture area occurs in the Trend Scenario. In this particular case, we see a strong increase of meat demand, while there is no intensification of pasture management. Consequently, the additional feed demand has to be fulfilled by further area expansion alone.

The highest increase of cropland area is projected for the Sustainable Development Scenario for both states, which can be explained by the soaring demands for crop products due to a shift towards a healthy diet. Parts of this additional demand are fulfilled by crops (e.g., soybean) formerly used as feedstock. It is important to note that a substantial share of the newly acquired cropland is located on former pasture. Interestingly, in Pará, under the Illegal Intensification Scenario, the expansion of cropland into protected areas leads to a considerable reduction of the total increase of cropland area. In these regions, areas with higher crop yields could be converted, thus sparing land from conversion into cropland.

The highest reduction of pasture in Pará could be achieved for the Sustainable Development Scenario due to the lower demand for meat products. Pasture expansion under the Legal Intensification Scenario is substantially lower than under Illegal Intensification. This underlines the important role of governance and for reducing area expansion, while fulfilling the growing demand for agricultural commodities, as emphasized by Soares-Filho et al. (2010). In contrast, the reduction of pasture in Mato Grosso is highest in both intensification scenarios, indicating that in this particular case, the effect of a more intensive pasture management outweighs the effect of the reduction of meat production.

Pasture intensification in Pará only leads to a reduction of pasture expansion in the case of the Legal Intensification Scenario. The assumed intensification rate of 4.5% is too low to halt pasture expansion for the case of the Illegal Intensification Scenario, suggesting that a higher intensification rate is necessary to completely fulfill the demand for meat products without further conversion of natural ecosystems. For Mato Grosso, a higher intensification rate was assumed, thus, expansion of pastures could be halted in the case of both intensification scenarios, indicating the need for pasture intensification as a means of habitat conservation, as has been discussed by Galford et al. (2013). We found that an increase of pasture productivity leads to a reduction of pasture area by close to 44% when comparing Trend and Legal Intensification Scenarios focusing on a Brazilian hotspot of cattle ranching. This result is supported by Cardoso et al. (2016), who used a life cycle analysis to compare five different scenarios for beef production in Brazil, with each scenario representing a higher degree of pasture intensification. The authors found that the introduction of a forage legume on Brazilian pastures, thereby increasing the digestible biomass productivity on pastures, led to a reduction of pasture area by 36% in Brazil.

Roads to a more sustainable use of land resources in Southern Amazonia

In our scenarios, we investigated three main mechanisms that can be part of strategies aiming for a sustainable use of land resources: land conservation policies, agricultural intensification, and changing human consumption pattern.

The outstanding role of land conservation policies becomes obvious in all scenarios, as suggested by Arima et al. (2014). As the climate and soil conditions of natural ecosystems in Southern Amazonia are often very suitable for agriculture (e.g., Lambin et al. 2013; Lambin and Meyfroidt 2011), without effective conservation measures in our simulation runs, these are being converted into cropland and pasture when agricultural production increase cannot be compensated by intensification. In the scenarios, land conservation is realized either as a land use mosaic under the Brazilian Forest Code, or by the strict protection of specific ecosystems. An analysis of the effects of these land conservation approaches on biodiversity was beyond the scope of this paper.

In “Differences between land use and land cover change scenarios,” we elaborated on the intensification of pasture management. It plays a crucial role in reducing the area that is needed for cattle grazing. Especially in Mato Grosso, intensification was identified as a powerful instrument to stop further expansion of pasture, even under increasing livestock production (Galford et al. 2013). In the case of Pará, intensification measures were less relevant for reducing pasture expansion. Here, the compliance with environmental policies (e.g., Brazilian Forest Code) and the conservation status of natural habitats, in combination with a shift of human consumption patterns towards a more crop-based diet leading to decreasing meat consumption and decreasing livestock numbers, had the strongest impact in terms of a reduction of pasture expansion and rainforest loss.

Also, the expansion of cropland was strongly influenced by the demand for crop products and crop yield increases due to technological change and intensification of agricultural management. In both states, the largest expansion of cropland can be found in the Sustainable Development Scenario, where it could be compensated by the drastic decline of pasture area. Interestingly, we can see a reduction of cropland area in the Trend Scenario in Mato Grosso. This effect can be traced back to the decoupling of agricultural production and area expansion that could be witnessed over the latter half of the first decade of this century and can be explained by agricultural intensification (e.g., Gollnow and Lakes 2014; Macedo et al. 2012). However, if we assume further increase of crop production, e.g., in the Legal and Illegal Intensification Scenarios, this effect is canceled out. Furthermore, the results from the model runs with the crop model MONICA (Nendel et al. 2011) indicate that negative climate effects under an A1B emission scenario can be compensated by technological improvements. Compared with these results, the assumed yield increases in the scenarios are in a plausible range, and it is likely that climate change, at least until 2030, will not prevent further agricultural intensification. It is important to note that this situation might change by the mid- or end of this century, when changes in temperature and precipitation are projected to become more intense (e.g. Marengo et al. 2012), with potentially stronger negative impacts on crop yields (e.g. Rosenzweig et al. 2014), thus putting additional pressure on natural land resources.

Our findings are supported by Bringezu et al. (2012) and Stehfest et al. (2009), who also see changes of human consumption behavior in combination with more intensive land use as a crucial element of avoiding further LULCC.

Agriculture and greenhouse gas emissions

In contrast to the comprehensive analysis by Galford et al. (2010) on the effect of alternative deforestation futures on greenhouse gas emissions in Mato Grosso, in this article, we focus on cropland and pasture and their role as sources or sinks of CO2, N2O, and CH4. But compared with Galford et al. (2010), our study investigates a wider range of scenario assumptions and covers a larger geographic region by also incorporating the state of Pará. As our scenario analysis shows (Fig. 2), the Sustainable Development Scenario produces the lowest annual GHG emissions compared with the other scenarios. None of the other scenarios shows a reduction of total annual GHG emissions in 2030 compared to the Trend Scenario. This finding again underlines the potential of a change of human consumption patterns to decrease GHG emissions from agricultural soils while, at the same time, satisfying the growing global demand for agricultural products.

In Mato Grosso, a decrease of annual N2O emissions from cropland is only calculated for the Trend Scenario due to the reduction of cropland area. The Illegal Intensification Scenario shows the lowest amount of annual CO2 emissions from pasture as it is characterized by the lowest extend of pasture area of all scenarios due to the strong intensification of grazing management. Also, pasture expands into protected areas (mainly the Pantanal) that are characterized by higher biomass productivity, thus reducing the net area demand. Compared with the base year, we find decreasing annual CO2 emissions from pasture for all but the Trend Scenario. This can be explained by the increasing proportion of old pastures to young pastures, and the fact that young pastures (≤ 10 years) on Acrisols tend to emit three times more CO2 than older pastures. Consequently, it would be a good measure to reactivate older degraded pastures instead of transforming natural vegetation in order to reduce CO2 emissions from agricultural soils. This measure is a part of the strategies implemented in the Brazilian national Low-Carbon Agriculture (ABC) program aiming at reducing agriculture-related CO2 emissions, while increasing agricultural productivity and assisting forest restoration (MAPA 2012).

As expected, in Pará, the largest reduction of annual N2O and CH4 emissions from pasture is achieved for the Sustainable Development Scenario. The highest increase of annual N2O and CH4 emissions from pasture is calculated for the Trend Scenario, as we assume high rates of livestock production increases, while simultaneously restricting the possibility to realize this production by means of pasture intensification. The highest increase of annual CH4 uptake by cropland in Pará was achieved in the case of the Sustainable Development Scenario. In Pará, the highest reduction of annual CO2 emissions from pasture is calculated for the Sustainable Development Scenario (− 98.2%), where it can be attributed to the strong decline of area used for meat production. The highest increase of annual CO2 emissions from pasture is calculated for the Illegal Intensification Scenario, closely followed by the Trend Scenario. Here, the additional emissions from an expansion of pasture area cannot be compensated by the shift of proportion of old pastures to young pastures.

For both states, a substantial increase of annual CO2 emissions from cropland is calculated for all scenarios due to a decline of young cropland that acts as a carbon sink in favor of old cropland that acts as a source of CO2 emissions.

An important part of our analysis is the consideration of the age of these land use types which influences their emission behavior. This approach is comparable with the study of Galford et al. (2010), who divided pastures into young (0–3 years), middle (4–5 years), and old (≥ 6 years). However, they only use this information in regard to N2O emissions; in our study, we additionally focus on age-related CO2 emissions from pastures and croplands. Furthermore, our study is based on a very broad data basis since it combines both our own observations from the specific study areas, and information from an extensive literature research.

Our results clearly indicate that the way agriculture in Southern Amazonia will develop in the coming decades not only affects the loss of natural ecosystems, but also the amount of greenhouse gas emissions from agricultural soils. Therefore, the Brazilian efforts for avoiding deforestation should be accompanied by policies aiming at a more climate-friendly agriculture.

Uncertainties and limitations of the study

Uncertainties of the simulation results can be separated into uncertainties related to the input data and uncertainties related to the structure of the model. Regarding input data quality, the disparity between different input data sources has to be mentioned. For instance, the underestimation of IBGE-based crop production data in comparison to crop area derived from MODIS satellite data. A reason for this underestimation might be the illegal agricultural activity in Brazil. A new extensive study suggests that close to 90% of Brazil’s deforestation from 2000 to 2012 was illegal (Lawson et al. 2014). This illegally cleared area was used mainly for crop production or cattle ranching (Klingler et al. 2017). Data on crop production on these illegally cleared areas and the areas themselves are not included in the IBGE agricultural survey. Yet, MODIS satellite images capture all agricultural areas. This leads to a discrepancy of observed cropland to agricultural production concerning the land use change model and the step of base-map generation. MODIS satellite images suggest 35% more cropland in the study area than is discernible from IBGE crop production and area statistics. This mismatch is further reinforced through the following process. In the base-map generation step, agricultural production numbers are allocated to observed cropland. If the production of agricultural commodities is underestimated, some cropland areas are left without crop production and are therefore classified as fallow land (set-aside) cells. In the next modeled time-step, these areas are used first for agricultural or pasture expansion, thus sparing areas classified as rainforest or Cerrado from deforestation. An example for model uncertainties is the simplification of agricultural management. For instance, we neglect the information that double cropping has been adapted by close to 60% of the farmers in Mato Grosso (e.g., Lapola et al. 2014). If this management practice was integrated into the model, we could expect a significantly lower pressure on land resources, as one single plot of cropland could satisfy the demand for two different crop types (e.g., soy and maize) each year. The inclusion of these processes into our land use model will play an important part in upcoming studies.

Furthermore, as described earlier (see “Scenario storylines and model drivers”), our estimates of annual CO2, N2O, and CH4 only consider emissions caused after clearing due to persisting land use. Information on emissions caused during the conversion process, burning of biomass, or changes in biomass carbon content were neglected. So far, there is little knowledge regarding the emission of N2O and CH4 during the conversion process; here, further research is needed.


We have successfully translated a new set of narrative socioeconomic scenarios for Southern Amazonia (Schönenberg et al. 2017) into numerical model input data. The quantitative scenarios have been simulated resulting in spatially explicit land use scenarios and a new inventory of the related greenhouse gas emissions from agricultural soils. The generated maps have a higher spatial resolution than previous efforts with the LandSHIFT model (Lapola et al. 2010; Lapola et al. 2011) and hence, can also contribute to a further refinement of studies, for example, related to carbon emissions from deforestation and the loss of biodiversity due to land use change (e.g., Chaplin-Kramer et al. 2015).

Since the representation of drivers of land use change and their interplay as well as land use policies, both in the scenarios and the applied land use model, is more refined than in most other simulation studies available for the Amazon region, we believe that our results can provide valuable information to scientists and policymakers alike (1) regarding the effects of particular combinations of driving factors of land use change on greenhouse gas emissions from agricultural soils, and (2) for the development of climate change mitigation strategies and a more sustainable use of land resources.

In the light of the described limitations, the model-based scenario analysis should not be misunderstood as a method to predict concrete future events. Rather, it provides a powerful tool to systematically explore plausible constellations of social and economic drivers and the emerging trajectories and dynamics of LULCC, together with its related environmental consequences.



This study has been conducted as part of the Carbiocial project (funding reference number 01LL0902A-01LL0902N) commissioned by the German Federal Ministry of Education and Research. We would like to thank the entire project team for their contribution to this research.

Supplementary material

10113_2017_1235_MOESM1_ESM.docx (611 kb)
ESM 1 (DOCX 610 kb)


  1. Aguiar APD, Câmara G, Escada MIS (2007) Spatial statistical analysis of land use determinants in the Brazilian Amazonia: exploring intra-regional heterogeneity. Ecol Model 209(2):169–188. CrossRefGoogle Scholar
  2. Amine E, Baba N, Belhadj M, Deurenbery-Yap M, Djazayery A, Forrester T, Galuska D, Herman S, James W, M’Buyamba J, Katan M, Key T, Kumanyika S, Mann J, Moynihan P, Musaiger A, Prentice A, Reddy K, Schatzkin A, Seidell J, Simpopoulos A, Srinujata S, Steyn N, Swinburn B, Uauy R, Wahlqvist M, Zhao-su W, Yoshiike N (2002) Diet, nutrition and the prevention of chronic diseases: report of a joint WHO/FAO expert consultation. World Health Organization, Geneva ISBN: 924120916XGoogle Scholar
  3. Arima EY, Barreto P, Araújo E, Soares-Filho B (2014) Public policies can reduce tropical deforestation: lessons and challenges from Brazil. Land Use Policy 41:465–473. CrossRefGoogle Scholar
  4. Arvor D, Meirelles M, Dubreuil V, Begue A, Shimabukuro YE (2012) Analyzing the agricultural transition in Mato Grosso, Brazil, using satellite-derived indices. Appl Geogr 32(2):702–713. CrossRefGoogle Scholar
  5. Arvor D, Dubreuil V, Simões M, Bégué A (2013) Mapping and spatial analysis of the soybean agricultural frontier in Mato Grosso, Brazil, using remote sensing data. GeoJournal 78(5):833–850. CrossRefGoogle Scholar
  6. Assunção J, Gandour CC, Rocha R (2012) Deforestation slowdown in the legal Amazon: prices or policies. Clim Pol Initiat 1:03–37. Google Scholar
  7. Banco de Nomes Geográficos do Brasil (IBGE), 2012. Accessed Apr 2012
  8. Barona E, Ramankutty N, Hyman G, Coomes OT (2010) The role of pasture and soybean in deforestation of the Brazilian Amazon. Environ Res Lett 5(2):024002. CrossRefGoogle Scholar
  9. Bondeau A, Smith PC, Zaehle S, Schaphoff S, Lucht W, Cramer W, Smith B (2007) Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Glob Chang Biol 13(3):679–706. CrossRefGoogle Scholar
  10. Bringezu S, O’Brien M, Schütz H (2012) Beyond biofuels: assessing global land use for domestic consumption of biomass: a conceptual and empirical contribution to sustainable management of global resources. Land Use Policy 29(1):224–232. CrossRefGoogle Scholar
  11. Cardoso AS, Berndt A, Leytem A, Alves BJ, de Carvalho IDN, de Barros Soares LH, Boddey RM (2016) Impact of the intensification of beef production in Brazil on greenhouse gas emissions and land use. Agric Syst 143:86–96. CrossRefGoogle Scholar
  12. Chaplin-Kramer R, Sharp RP, Mandle L, Sim S, Johnson J, Butnar I, Kareiva PM (2015) Spatial patterns of agricultural expansion determine impacts on biodiversity and carbon storage. Proc Natl Acad Sci 112(24):7402–7407. CrossRefGoogle Scholar
  13. Cohn AS, Mosnier A, Havlík P, Valin H, Herrero M, Schmid E, Obersteiner M (2014) Cattle ranching intensification in Brazil can reduce global greenhouse gas emissions by sparing land from deforestation. Proc Natl Acad Sci 111(20):7236–7241. CrossRefGoogle Scholar
  14. Coy M (2001) Globalisierung in Brasilien: Raumwirksamkeit und Reaktionen. Beispiele aus städtischen und ländlichen Regionen. In: Krömer G, Borsdorf A, Parnreiter C (eds.) Lateinamerika im Umbruch. Geistige Strömungen im Globalisierungsstress. Innsbrucker Geographische Studien Band 32. Geographie Innsbruck, Innsbruck. ISBN: 978-3-901182-32-7120916XGoogle Scholar
  15. Coy M, Klingler M (2008 ) Pionierfronten im brasilianischen Amazonien zwischen alten Problemen und neuen Dynamiken. Das Beispiel des „Entwicklungskorridors “Cuiabá (Mato Grosso)–Santarém (Pará). Innsbrucker Geographische Gesellschaft (Hg.) Jahresbericht 2008-2010:109-129. Geographie Innsbruck, Innsbruck ISBN: 3901182837Google Scholar
  16. Dalla-Nora EL, de Aguiar APD, Lapola DM, Woltjer G (2014) Why have land use change models for the Amazon failed to capture the amount of deforestation over the last decade? Land Use Policy 39:403–411. CrossRefGoogle Scholar
  17. Embrapa Amazônia Oriental: Agência de Desenvolvimento da Amazônia (2008) ZONAMENTO ecológico-econômico da área de influência da Rodovia BR-163 (Cuiabá-Santarém). Belem, PA. Accessed June 2013
  18. FAO (2014) FAOSTAT Emissions Database. Accessed 11 Dec 2014
  19. Farr TG, Kobrick M (2000) Shuttle radar topography mission produces a wealth of data. Eos, Trans Am Geophys Union 81(48):583–585. CrossRefGoogle Scholar
  20. Fearnside PM, Righi CA, de Alencastro Graça PML, Keizer EW, Cerri CC, Nogueira EM, Barbosa RI (2009) Biomass and greenhouse-gas emissions from land use change in Brazil’s Amazonian “arc of deforestation”: the states of Mato Grosso and Rondônia. For Ecol Manag 258(9):1968–1978. CrossRefGoogle Scholar
  21. Friedl MA, Sulla-Menashe D, Tan B, Schneider A, Ramankutty N, Sibley A, Huang X (2010) MODIS collection 5 global land cover: algorithms and characterization of new datasets. Remote Sens Environ 115:168–182. CrossRefGoogle Scholar
  22. Galford GL, Melillo JM, Kicklighter DW, Cronin TW, Cerri CE, Mustard JF, Cerri CC (2010) Greenhouse gas emissions from alternative futures of deforestation and agricultural management in the southern Amazon. Proc Natl Acad Sci 107(46):19649–19654. CrossRefGoogle Scholar
  23. Galford GL, Soares-Filho B, Cerri CE (2013) Prospects for land use sustainability on the agricultural frontier of the Brazilian Amazon. Philos Trans Royal Soc Lond B: Biol Sci 368(1619). doi:
  24. Gibbs HK, Rausch L, Munger J, Schelly I, Morton DC, Noojipady P, Walker NF (2015b) Brazil’s Soy Moratorium. Science 347(6220):377–378. CrossRefGoogle Scholar
  25. Gibbs HK, Munger J, L’Roe J, Barreto P, Pereira R, Christie M, Amaral T, Walker NF (2015a) Did ranchers and slaughterhouses respond to zero-deforestation agreements in the Brazilian Amazon? Conserv Lett 9(1):32–42. CrossRefGoogle Scholar
  26. Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, Muir JF, Toulmin C (2010a) Food security: the challenge of feeding 9 billion people. Science 327(5967):812–818. CrossRefGoogle Scholar
  27. Godfray HCJ, Crute IR, Haddad L, Lawrence D, Muir JF, Nisbett N, Whiteley R (2010b) The future of the global food system. Philos Trans Royal Soc B: Biol Scis 365(1554):2769–2777. CrossRefGoogle Scholar
  28. Gollnow F, Lakes T (2014) Policy change, land use, and agriculture: the case of soy production and cattle ranching in Brazil, 2001–2012. Appl Geogr 55:203–211. CrossRefGoogle Scholar
  29. Greenpeace-Brazil (2009) Amazon cattle footprint. Mato Grosso: state of destruction. Accessed June 2015
  30. Hecht SB (2011) From eco-catastrophe to zero deforestation? Interdisciplinarities, politics, environmentalism and reduced clearing in Amazonia. Environ Conserv 39(1):4–19. CrossRefGoogle Scholar
  31. Instituto Brasileiro de Geografia e Estatística (IBGE) (2013) Indicadores IBGE. IBGE http://wwwibgegovbr/estadosat/indexphp. Accessed August 2013
  32. Instituto Brasileiro de Geografia e Estatística (IBGE) (2014) Amazonia Legal 2014. IBGE. Accessed June 2015
  33. Instituto Brasileiro de Geografia e Estatística (IBGE) (2015) Indicadores IBGE Estatística da Produção Agrícola. Accessed Aug 2015
  34. INPE (2013) PRODES. Accessed May 2013
  35. Jasinski E, Morton D, DeFries R, Shimabukuro Y, Anderson L, Hansen M (2005) Physical landscape correlates of the expansion of mechanized agriculture in Mato Grosso, Brazil. Earth Interact 9(16):1–18. CrossRefGoogle Scholar
  36. Klingler M, Richards PD, Ossner R (2017) Cattle vaccination records question the impact of recent zero-deforestation agreements in the Amazon. Regional Environmental Change, this issueGoogle Scholar
  37. Kohlhepp G (2002) Regionalentwicklung im Amazonasgebiet Brasiliens. In: Donato H, Kutschat RSG, Tiemann J (eds.): Institut Martius-Staden. Jahrbuch 2001–2002 (49). Nova Bandeira, São Paulo. ISSN 1677-051XGoogle Scholar
  38. Krogh L, Noergaard A, Hermansen M, Greve M, Balstroem T, Breuning-Madsen H (2003) Preliminary estimates of contemporary soil organic carbon stocks in Denmark using multiple datasets and four scaling-up methods. AGEE 96:19–28. Google Scholar
  39. Lambin EF, Rounsevell MDA, Geist HJ (2000) Are agricultural land use models able to predict changes in land use intensity? Agric Ecosyst Environ 82(1):321–331. CrossRefGoogle Scholar
  40. Lambin EF, Meyfroidt P (2011) Global land use change, economic globalization, and the looming land scarcity. Proc Natl Acad Sci 108(9):3465–3472. CrossRefGoogle Scholar
  41. Lambin EF, Gibbs HK, Ferreira L, Grau R, Mayaux P, Meyfroidt P, Munger J (2013) Estimating the world's potentially available cropland using a bottom-up approach. Glob Environ Chang 23(5):892–901. CrossRefGoogle Scholar
  42. Lapola DM, Priess JA, Bondeau A (2009) Modeling the land requirements and potential productivity of sugarcane and jatropha in Brazil and India using the LPJmL dynamic global vegetation model. Biomass Bioenergy 33(8):1087–1095. CrossRefGoogle Scholar
  43. Lapola DM, Schaldach R, Alcamo J, Bondeau A, Koch J, Koelking C, Priess JA (2010) Indirect land use changes can overcome carbon savings from biofuels in Brazil. Proc Natl Acad Sci 107(8):3388–3393. CrossRefGoogle Scholar
  44. Lapola DM, Schaldach R, Alcamo J, Bondeau A, Msangi S, Priess JA, Silvestrini R, Soares-Filho BS (2011) Impacts of climate change and the end of deforestation on land use in the Brazilian Amazon. Earth Interact 15:1–29. CrossRefGoogle Scholar
  45. Lapola DM, Martinelli LA, Peres CA, Ometto JP, Ferreira ME, Nobre CA, Vieira IC (2014) Pervasive transition of the Brazilian land use system. Nat Clim Chang 4(1):27–35. CrossRefGoogle Scholar
  46. Lawson S, Blundell A, Cabarle B, Basik N, Jenkins M, Canby K (2014) Consumer goods and deforestation: an analysis of the extent and nature of illegality in Forest conversion for agriculture and timber plantations. Forest Trend Report Series, Washington DC. Accessed May 2014
  47. Macedo MN, DeFries RS, Morton DC, Stickler CM, Galford GL, Shimabukuro YE (2012) Decoupling of deforestation and soy production in the southern Amazon during the late 2000s. Proc Natl Acad Sci 109(4):1341–1346. CrossRefGoogle Scholar
  48. Malingreau JP, Eva HD, De Miranda EE (2012) Brazilian Amazon: a significant five year drop in deforestation rates but figures are on the rise again. Ambio 41(3):309–314. CrossRefGoogle Scholar
  49. Marengo JA, Chou SC, Kay G, Alves LM, Pesquero JF, Soares WR, Tavares P (2012) Development of regional future climate change scenarios in South America using the eta CPTEC/HadCM3 climate change projections: climatology and regional analyses for the Amazon, São Francisco and the Paraná River basins. Clim Dyn 38(9–10):1829–1848. CrossRefGoogle Scholar
  50. Martinelli LA, Naylor R, Vitousek PM, Moutinho P (2010) Agriculture in Brazil: impacts, costs, and opportunities for a sustainable future. Curr Opin Environ Sustain 2(5):431–438. CrossRefGoogle Scholar
  51. Meurer KHE, Franko U, Stange CF, Dalla Rosa J, Madari BE, Jungkunst HF (2016) Direct nitrous oxide (N2O) fluxes from soils under different land use in Brazil—a critical review. Environ Res Lett 11:023001. CrossRefGoogle Scholar
  52. Mietzner D, Reger G (2005) Advantages and disadvantages of scenario approaches for strategic foresight. Int J Technol Intell Plan 1(2):220–239. Google Scholar
  53. Ministério da Agricultura, Pecuária e Abastecimento (MAPA) (2012) Plano Setorial de Mitigação e de Adaptação às Mudanças Climáticas para a Consolidação de uma Economia de baixa Emissão de Carbono na Agricultura. Plano ABC (Agricultura de Baixa Emissão de Carbono). Coordenado por Casa Civil da Presidência da República, Ministério da Agricultura, Pecuária e Abastecimento (MAPA) e Ministério do Desenvolvimento Agrário (MDA). Versão final-13/01. ISBN 978-85-7991-062-0Google Scholar
  54. Ministério do Meio Ambiente (MMA) (2001) Causas e dinamica do desmatamento na Amazonia. MMA, Brasilia DF. ISBN: 8587166271Google Scholar
  55. Ministério do Meio Ambiente (MMA) (2013) Indigenous Areas Mato Grosso and Pará. Available at: Accessed Mar 2013
  56. Mitchell TD, Jones PD (2005) An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int J Climatol 25(6):693–712. CrossRefGoogle Scholar
  57. Myhre G, Shindell D, Bréon FM, Collins W, Fuglestvedt J, Huang J, Koch D, Lamarque J-F, Lee D, Mendoza B, Nakajima T (2013) Anthropogenic and natural radiative forcing. In: Stocker TF, Qin D, Plattner GK, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of hte Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Google Scholar
  58. Nendel C, Berg M, Kersebaum KC, Mirschel W, Specka X, Wegehenkel M, Wieland R (2011) The MONICA model: testing predictability for crop growth, soil moisture and nitrogen dynamics. Ecol Model 222(9):1614–1625. CrossRefGoogle Scholar
  59. Nepstad D, Soares-Filho BS, Merry F, Lima A, Moutinho P, Carter J, Stella O (2009) The end of deforestation in the Brazilian Amazon. Science 326(5958):1350–1351. CrossRefGoogle Scholar
  60. Nepstad D, McGrath D, Stickler C, Alencar A, Azevedo A, Swette B, Hess L (2014) Slowing Amazon deforestation through public policy and interventions in beef and soy supply chains. Science 344(6188):1118–1123. CrossRefGoogle Scholar
  61. Pacheco P (2012) Actor and frontier types in the Brazilian Amazon: assessing interactions and outcomes associated with frontier expansion. Geoforum 43(4):864–874. CrossRefGoogle Scholar
  62. Quesada CA, Lloyd J, Anderson LO, Fyllas NM, Schwarz M, Czimczik CI (2011) Soils of Amazonia with particular reference to the RAINFOR sites. Biogeosciences 8(6):1415–1440.
  63. Rosenzweig C, Elliott J, Deryng D, Ruane AC, Müller C, Arneth A, Jones JW (2014) Assessing agricultural risks of climate change in the 21st century in a global gridded crop model intercomparison. Proc Natl Acad Sci 111(9):3268–3273. CrossRefGoogle Scholar
  64. Salvatore M, Pozzi F, Ataman E, Huddleston B, Bloise M, Balk D, Yetman G (2005) Mapping global urban and rural population distributions. FAO, Rome. Accessed Apr 2013
  65. Schaldach R, Alcamo J, Koch J, Kölking C, Lapola DM, Schüngel J, Priess JA (2011) An integrated approach to modelling land use change on continental and global scales. Environ Model Softw 26(8):1041–1051. CrossRefGoogle Scholar
  66. Schmidt MWI, Torn MS, Abiven S, Dittmar T, Guggenberger G, Janssens IA, Kleber M, Kögel-Knabner I, Lehmann J, Manning DAC, Nannupieri P, Rasse DP, Weiner S, Trumbore SE (2011) Persistence of soil organic matter as an ecosystem property. Nature 478:49–56. CrossRefGoogle Scholar
  67. Schönenberg R, Schaldach R, Lakes T, Göpel J, Gollnow F (2017) Inter- and transdisciplinary scenario construction to explore future land-use options in Southern Amazonia. Ecol Soc 22(3).
  68. Sitch S, Smith B, Prentice IC, Arneth A, Bondeau A, Cramer W, Thonicke K (2003) Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Glob Chang Biol 9(2):161–185. CrossRefGoogle Scholar
  69. Soares-Filho B, Alencar A, Nepstad D, Cerqueira G, Diaz V, del Carmen M, Voll E (2004) Simulating the response of land-cover changes to road paving and governance along a major Amazon highway: the Santarém–Cuiabá corridor. Glob Chang Biol 10(5):745–764. CrossRefGoogle Scholar
  70. Soares-Filho B, Nepstad DC, Curran LM, Cerqueira GC, Garcia RA, Ramos CA, Schlesinger P (2006) Modelling conservation in the Amazon basin. Nature 440(7083):520–523. CrossRefGoogle Scholar
  71. Soares-Filho B, Moutinho P, Nepstad D, Anderson A, Rodrigues H, Garcia R, Silvestrini R (2010) Role of Brazilian Amazon protected areas in climate change mitigation. Proc Natl Acad Sci 107(24):10821–10826. CrossRefGoogle Scholar
  72. Soares-Filho B, Rajão R, Macedo M, Carneiro A, Costa W, Coe M, Alencar A (2014) Cracking Brazil’s forest code. Science 344(6182):363–364. CrossRefGoogle Scholar
  73. Srinivasan CS, Irz X, Shankar B (2006) An assessment of the potential consumption impacts of WHO dietary norms in OECD countries. Food Policy 31(1):53–77. CrossRefGoogle Scholar
  74. Stehfest E, Bouwman L, van Vuuren DP, den Elzen MG, Eickhout B, Kabat P (2009) Climate benefits of changing diet. Clim Chang 95(1–2):83–102. CrossRefGoogle Scholar
  75. Turner BL, Lambin EF, Reenberg A (2007) The emergence of land change science for global environmental change and sustainability. Proc Natl Acad Sci 104(52):20666–20671. CrossRefGoogle Scholar
  76. Veldkamp A, Lambin EF (2001) Predicting land use change. Agric Ecosyst Environ 85:1):1–1):6. CrossRefGoogle Scholar
  77. Vieira ICG, Toledo PM, Silva JMC, Higuchi H (2008) Deforestation and threats to the biodiversity of Amazonia. Braz J Biol 68(4, Suppl):949–956. CrossRefGoogle Scholar
  78. Walker R, Drzyzga SA, Li Y, Qi J, Caldas M, Arima E, Vergara D (2004) A behavioral model of landscape change in the Amazon basin: the colonist case. Ecol Appl 14(sp4):299–312. CrossRefGoogle Scholar
  79. Wint W, Robinson T (2007) Gridded livestock of the world 2007. FAO, Rome ISBN 978-92-5-106791-9Google Scholar
  80. Wright HL, Lake IR, Dolman PM (2012) Agriculture—a key element for conservation in the developing world. Conserv Lett 5(1):11–19. CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany 2017

Authors and Affiliations

  • Jan Göpel
    • 1
  • Jan Schüngel
    • 1
  • Rüdiger Schaldach
    • 1
  • Katharina H. E. Meurer
    • 2
  • Hermann F. Jungkunst
    • 3
  • Uwe Franko
    • 4
  • Jens Boy
    • 5
  • Robert Strey
    • 5
  • Simone Strey
    • 5
  • Georg Guggenberger
    • 5
  • Anna Hampf
    • 6
  • Phillip Parker
    • 6
  1. 1.Center for Environmental Systems Research (CESR)University of KasselKasselGermany
  2. 2.Department of EcologySwedish University of Agricultural Sciences - SLUUppsalaSweden
  3. 3.Institute for Environmental SciencesUniversity of Koblenz-LandauLandauGermany
  4. 4.Department of Soil Physics, Helmholtz Centre for Environmental Research—UFZHalle (Saale)Germany
  5. 5.Institute of Soil ScienceLeibniz Universität HannoverHannoverGermany
  6. 6.Leibniz-Zentrum für Agrarlandschaftsforschung (ZALF) e. VMünchebergGermany

Personalised recommendations