Alessandro De Pinto*
Jhon Brayan Valencia Garcia**
Alejandro Castro Coca**
Jesus David Martinez**
Jesus David Hoyos**
* International Food Policy Research
** International Center for Tropical
Low Emission Development Strategies for
Agriculture and Other Land Uses: The Case of
Acknowledgements ....................................................................................................................................... 3
1 Introduction .......................................................................................................................................... 4
2 Country Background ............................................................................................................................. 5
2.1 Colombia and Greenhouse Gas Emissions .................................................................................. 5
2.2 Land Use and Land Cover ............................................................................................................ 6
2.2.1 Forest and Deforestation ........................................................................................................ 7
2.2.2 Crop and Livestock Production ............................................................................................. 10
2.2.3 Pasture .................................................................................................................................. 12
2.3 Colombia Climate Change Policies ............................................................................................ 14
2.3.1 National Policies and Plans Associated with Mitigation ....................................................... 14
2.3.2 Emissions reduction in the land use, land use change, agriculture and forestry sectors ..... 15
3 Modeling Framework .......................................................................................................................... 15
3.1 IMPACT Model ........................................................................................................................... 16
3.2 Land Use Model ......................................................................................................................... 19
3.2.1 Data for Upper-Level Model ................................................................................................. 20
3.2.2 Data for Lower-Level Model .................................................................................................. 21
3.3 Crop Emission Model and Other Emissions ............................................................................... 22
3.4 Carbon Stock .............................................................................................................................. 25
4 Economic Trade-Offs, Baseline, and Alternative Scenarios .................................................... 26
5 Baseline Results .................................................................................................................................. 28
6 Policy Simulations ............................................................................................................................... 34
6.1 Pasture ....................................................................................................................................... 34
6.2 Forestry ...................................................................................................................................... 35
6.3 Palm Cultivation......................................................................................................................... 35
6.4 Policy Target 1: Reduction in Area Allocated to Pasture ........................................................... 39
6.5 Policy Target 2: Reduction in the Rate of Deforestation in the Amazon Forest ....................... 40
6.6 Policy Target 3: Increase the Area Allocated to the Cultivation of Palm .................................. 43
7 Discussion and Conclusions .......................................................................................................... 45
8 References .......................................................................................................................................... 48
9 Appendix 1 .......................................................................................................................................... 53
The authors thank our colleagues in the Colombian government who kindly answered our questions and
gave critical advice, without which we could not have developed this study. We are especially grateful for
suggestions and advice from colleagues of the Ministerio de Agricultura (MADR), the Federación Colombiana
de Ganaderos (FEDEGAN) and the Sociedad de Agricultores de Colombia (SAC). We are also grateful to many
colleagues from the Ministry of Environment (MADS), Universidad Javieriana, Universidad Nacional, National
Cereal Growers Federation (FENALCE), the National Potato Growers Association (FEDEPAPA) and the
National Forestry Research and Promotion Institute (CONIF). This work was supported by a grant from the
United States Agency for International Development (USAID) and received support from two Consultative
Group on International Agricultural Research (CGIAR) research programs: Climate Change and Agricultural
Food Security (CCAFS) and Forest, Trees and Agroforestry (FTA). The authors take sole responsibility for the
opinions expressed within this report.
Resource use in many developing countries, from crop production to deforestation, is responsible for
the bulk of greenhouse gasses (GHG) emissions. We also know that there are instances in which the
agricultural and forestry sectors can provide low-cost climate change mitigation opportunities (Golub et
al. 2013; Lubowski and Rose 2013; Smith et al. 2007). From a technical point of view, reducing expected
increases in GHG emissions in agriculture requires the adoption of transformative approaches in the use
of resources. Emphasis has been placed on methods that increase the efficiency in the use of fertilizers,
water, and fossil fuels, as well as waste reduction. A growing body of literature analyzes the effects of
alternative agricultural practices (Antle and Stoorvogel 2008; Diagana et al. 2007; Gilhespy et al. 2014;
Tenningkeit et al. 2012; Schneider and Smith 2008; Smith et al. 2013; Tschakert 2007 ). The livestock
sector has also been the target of research on mitigation opportunities (Golub et al. 2013; Li et al. 2012;
Schils et al. 2007) and the mitigation potential of forests, soil and other biomass, has been amply
analyzed as well (Cacho et al. 2005; Makundi and Sathaye, 2004; Torres et al. 2010; Lubowski and Rose
2013). However, from a policy-making perspective, the design of low emission development strategies is
an example of multi-objective decision making in which policies target the reduction of GHG emissions
while other goals such as increasing agricultural productivity and food security or attaining objectives
such as export goals or economic growth, are preserved. Furthermore, it is important to consider that all
countries are part of a global economic system and therefore it is critical that policies are devised with
full recognition of the role of the international economic environment which, with its effects on
commodity prices, can significantly affect the long-term viability and the budgetary implications of
mitigation policies. The challenge at hand is therefore to reconcile the limited spatial resolution of
macro-level economic models that operate at a subnational or national level with models that function
at a higher spatial resolution, which allow to properly account for changes in carbon stocks and GHG
emissions. To our knowledge there are only a few examples of analyses with similar objectives: Golub et
al (2013) examined the impact on food consumption and income of implementing mitigation policies at
national and regional levels. Schneider et al (2008) estimated mitigation potentials of U.S. agriculture
with regionally disaggregated data and changes in welfare within the agricultural sector. Rutten et al.
(2014) evaluated the effects of select climate change and economic growth scenarios on Vietnam’s
economy. Havlik et al. (2014) estimated the effects of transitioning to more efficient livestock
production system on GHG mitigation and the economy, and Lubowski and Rose (2013) provided a
review of a number of studies that model mitigation potentials of Reducing Emissions from
Deforestation and Forest Degradation (REDD) along with additional role of conservation, sustainable
management of forests and enhancement of forest carbon stocks policies and their impacts on prices
and agricultural production.
In this article we demonstrate that different models, all widely accessible to the public, can be
brought together to help policymakers in their evaluation of trade-offs, opportunities, and repercussions
of alternative mitigation policies in the agricultural sector. While the focus of this work is on Colombia,
the analytical framework can be applied to any country interested in exploring country-wide effects and
economic viability of climate change mitigation policies in agriculture. The approach is based on the use
of public and widely accessible data and we believe that the flexibility and transparency of the approach
proposed in this study can increase decision-makers’ trust in the results. Naturally, additional data and
targeted surveys can increase the accuracy of the results and the framework does not create barriers for
the inclusion of additional input. Nonetheless, it appears clear from our analysis that policy-makers
need substantial support in their decision-making process as the range of options they face can be very
diverse and the effects of their decisions have important, and sometimes unexpected, repercussions.
The effects of the policies we simulated cover the entire spectrum of potential outcomes. We find win-
win policies (reducing land allocated to pasture increase revenues and carbon stock and reduces GHG
emissions), policies with tradeoffs (limiting deforestation in the Amazon increases carbon stock,
decreases emissions, but reduces revenues) and policies that seem to generate clearly inferior results
(increasing the area allocate to oil palm cultivation reduces carbon stock, increases emissions and
reduces revenues). Stakeholders, from government agencies, to producer and consumers’ organizations
to farmers, will benefit from policies devised with the support of solid evidence and the effects of which
can be investigated and evaluated by all the parties affected.
2 Country Background
The background information presented here gives a brief overview of the situation in Colombia with
respect to land use activities associated with greenhouse gas (GHG) emissions. It includes information
on land use and land use change, forests and deforestation, and crop and livestock production. An
overview of policies related to emissions in the agriculture and land use change sectors is presented.
2.1 Colombia and Greenhouse Gas Emissions
Colombia’s unique circumstances make GHG emissions an important topic in the agricultural and forest
sectors. The Amazon biome within the country represents a significantly large carbon stock experiencing
a relatively rapid change. The native savannas of the Eastern Plains of Colombia—the second largest
savanna region in Latin America after the Brazilian Cerrados—has a relatively low carbon stock and
presents some opportunities for carbon sequestration. However, this region is also the focus of much of
the growth within the agricultural sector, with its corresponding emissions from crop production and
cattle ranching. In the rest of the country, with the exception of the high Andes, most lands would
naturally be covered in forests. Nevertheless, because of the importance of agriculture and natural
resource use, these lands provide an important livelihoods function for Colombia.
Although the agriculture sector consists of 7 percent of the gross national product (GNP), the sector
employs 18 percent of the population (CIA 2014). Traditionally, Colombia has had a large number of
smallholder farmers and their agricultural activities have been responsible for a good share of emissions.
There is also a well-established culture of cattle ranching with both small and large livestock keepers.
Urbanization and industrialization have been growing in Colombia, but agricultural and forestry activities
will continue to grow and claim a large share of emissions.
In 2010 Colombia presented its second National Communication on Climate Change to the United
Nations Framework Convention on Climate Change (UNFCCC). The report contains data from the last
National Greenhouse Gases Inventory carried out in 2004. Colombia contributes 0.37 percent (180,010
Gg) of the total worldwide emissions of GHG (49 Gt). This total emission is composed of 50 percent
carbon dioxide (CO2), 30 percent methane (CH4), and 19 percent nitrous oxide (N2O), with the remaining
1 percent of emissions classified as chlorofluorocarbons (CFCs).
The last National Greenhouse Gases Inventory showed that agricultural activities emit 38 percent of
total emissions; energy 37 percent (including fuels and transports), land use, land change, and forestry
14 percent; solid waste 6 percent; and industrial processes 5 percent. More than half of the total GHG
emissions are related to the agricultural sector or encompass land use change, representing 52 percent
of the national total (Figure 2.1; IDEAM 2010). Of the emissions resulting from agricultural activities,
48.5 percent are due to enteric fermentation, 47.5 percent from agricultural soil management, and 2
percent from emissions related to rice cultivation.
Figure 2.1 Participation of GHG by sector in Colombia, 2004
Note: LULUCF signifies Land Use, Land-Use Change, and Forestry
Source: Hydrology, Meteorology, and Environmental Studies Institute (IDEAM 2010).
2.2 Land Use and Land Cover
Colombia exhibits a broad range of land covers and land uses in line with a varied physiography of
plains, mountain cordilleras, and intermountain valleys with high overall relief. A latest land use map
generated by the government shows this diversity of land uses and land covers (Figure 2.2). According to
the government statistics ( IGAC, 2013, IAvH et al. 2007), 52 percent of Colombia’s 115 million hectares
(ha) are in natural forests, mostly within the Amazon basin but also including forests on the Pacific coast
and in the northern part of the country. Cultivated pastures and native savanna grasslands make up 26
percent of the land area. These lands are characterized by cattle grazing with low stocking rates and
frequent natural and anthropogenic fires. Cropland is mostly concentrated in the intermountain valleys,
making up about 4 percent of the land surface (Table 2.1).
Figure 2.2 Land cover map of Colombia, 2000
Source: IAvH et al. 2007.
Table 2.1 Land cover for Colombia, 2008
Land cover category
Source: IGAC, 2013, IAvH et al. 2007.
Note: Although the cropland statistics from MADR have a discrepancy from the IGAC’s map, we prioritize MADR’s
data in order to incorporate crop allocation in the study.
2.2.1 Forest and Deforestation
Approximately 65 percent of Colombia is in its natural state, 54 percent in natural forests, and 10
percent in native savannas, which is counted within pasture in Table 2.1. About 30 percent of the
national territory has been cleared of its forest cover. Recently, Colombia has been improving its
capacity to monitor land use change, partly as a result of its participation in different projects related to
reducing emissions from deforestation and degradation. The discussion below focuses on changes in
forest extent based on government reports and two broad-scale remote-sensing analyses available in
the public domain.
In 2011, IDEAM and the Ministry of Environment and Sustainable Development (hereafter MADS
produced a technical report quantifying national deforestation using Landsat and MODIS images for the
period 1990 to 2010, establishing trends in deforestation (Table 2.2). Their analysis shows the total
forest extent decreasing from 64.5 million ha in the 1990–2000 period to 60.2 million ha in the 2005–
2010 period. Most of the reduction took place in the 1990s. Deforestation rates decreased in the first
half of the 2000s before going up again between 2005 and 2010. The average annual deforestation rate
over the entire period is more than 238,000 ha per year. During this entire period most of the
deforestation occurred in the Amazon (approximately 1.7 million ha), followed by the Andean region
(1.4 million ha), the Caribbean (590,000 ha), the Orinoco basin (316,000 ha), and the Pacific region
Table 2.2 Deforestation by regions
Forest 1990 (Ha)
Forest 2000 (Ha)
Forest 2005 (Ha)
Annual average deforestation
Source: IDEAM 2010.
The government report described above generally coincides with two nongovernmental initiatives to
map forest cover. One of these is the recent global study of forest extent carried out using 30 m Landsat
data for the entire world (Hansen et al. 2014). A second initiative called Terra-i monitors forest
vegetation in Latin America using 250 m MODIS data (Reymondin et al. 2012). The first of these studies
has the advantage of greater spatial resolution, and is available at annual increments between 2000 and
2012. The Terra-i dataset monitors vegetation change every 16 days, but is unable to detect small
clearings that are often associated with shifting cultivation. Data from these two studies were extracted
for the period 2005 to 2010 for the national territory of Colombia and compared with the governmental
study described above (Figure 2.3).
The Ministry of Environment and Sustainable Development (MADS) is the lead in management of the
environment and renewable natural resources, responsible for guiding and regulating the environmental planning
and developing policies and regulations for the recovery, conservation, protection, management, planning,
sustainable use, and exploitation of renewable natural resources and the environment in Colombia.
Figure 2.3 Deforestation in hectares for the 5-year period from 2005 to 2010
Source: IDEAM 2011, Hansen et al. 2014, and Reymondin et al. 2012.
Deforestation estimates across the five regions of Colombia are in general agreement among the
three sources of data. The Amazon region is clearly the area where deforestation occurs the most over
the past decade. Deforestation in the Andean and Caribbean regions are relatively higher than in the
Orinoco basin and Pacific region. The higher Andean deforestation reported by IDEAM was not found in
the Terra-i and Hansen datasets. The IDEAM and Terra-i datasets showed higher deforestation estimates
in the Pacific region than did the Hansen dataset.
The main drivers of deforestation in Colombia are pasture expansion, annual and permanent crop
cultivation, illegal drug crop cultivation, infrastructure development, mining, and colonization of frontier
areas. The development of pasturelands is clearly the most important deforestation driver in the
country. Estimates of forest clearing prior to 2000 suggested that two-thirds of this clearing was for
pasturelands and one-third was for cropping (Etter et al. 2006). A more recent analysis has suggested
that 90 percent of forest clearing between 2005 and 2010 was for pastureland development (Nepstad et
Colombia is the fifth largest global producer of palm oil, and the development of plantations is
expected to increase. However, oil palm development in the past mostly occurred on lands that were
already cleared of their forests, a trend that is expected to continue in the future (Castiblanco et al.
2013). Aside from oil palm, cropping as a deforestation driver is mostly associated with smallholder
shifting cultivation. Soybeans—a growing driver of deforestation in the tropics—are relatively
insignificant as a deforestation driver in Colombia (Nepstad et al. 2013). However, illegal drug crop
cultivation is thought to be an important driver of deforestation in Colombia, but its magnitude is not
well known (Davalos et al. 2011, Nepstad et al. 2013).
The principal hotspots of deforestation described above are distributed geographically in four regions
of the country. Figure 2.4 shows this distribution aggregated to municipalities based on the Terra-i
dataset described above. The most important of these hotspots is the Amazon region, with a focus in
the Caqueta department. To the north of this region, the Eastern Plains in the Orinoco basin is a
transition area between native savannas to the north and humid forest to the south. A third focus of
deforestation in Colombia is in the four departments of the Pacific coast, an area with very high rainfall
and abundant tropical forest. Finally, the middle Magdalena Valley—in the center north of the country,
in the departments of Santander, Bolivar, and Antioquia—shows high levels of forest loss between 2004
Figure 2.4 Loss of forest by municipality, 2004–2012
Source: Reymondin et al. 2012.
2.2.2 Crop and Livestock Production
Colombia’s unique geography and varied climates support a large variety of agricultural products. For
example, low elevation areas (< 1000 m) are dominated by banana, plantain, cassava, rice, sugarcane,
maize, palm, tropical fruits, and citrus, among others. At intermediate elevations (between 1000 and
2000 m), the key crops are coffee, bananas, sugarcane, beans, some fruits, and citrus. At higher altitudes
(> 2000 m), corn, wheat, potatoes, and vegetables are the main crops. Figure 2.5 shows the geographic
distribution of the main crop in Colombia presented at the municipality level.
Figure 2.5 Predominant crop by municipality
Source: IGAC, 2013.
There is little indication that there will be large changes in crop area over the coming decades, with
the exception of oil palm. Figure 2.6 presents projections based on historical agricultural area, reported
in municipal agricultural assessments by MADR. Oil palm is expected to increase substantially after
2016, due to its high demand for food products and biofuels. There is also a great deal of interest within
the government to promote the oil palm sector. Palm plantations for oil extraction increased from
18,000 ha in the 1960s to almost 360,000 ha in 2010, mainly in the Meta, Casanare, Cesar, Magdalena,
Bolivar, Cundinamarca, Santander, Norte de Santander, and Nariño departments (Sanchez-Cuervo et al.
Figure 2.6 Area projections for select crops, 2008–2040
Land use intensification shows two factors that are reducing the emissions in the eastern plains of
Colombia: the carbon sequestration due to the increase in oil palm plantations and the decreasing
emissions from savanna burning as a result of increased areas of planted pastures. These processes are
partly compensating for the increasing cattle and irrigated rice emissions (Etter et al. 2010).
Pasturelands and livestock production may change substantially in the coming years. According to the
National Livestock Federation (FEDEGAN), the Colombian livestock inventory totals 23.5 million head of
cattle on 39.2 million ha in pasture. With less than one head of cattle per ha, livestock occupies 26
percent of the total land area of Colombia. Figure 2.7 shows the spatial distribution of managed pasture
and grassland, both of which serve for livestock production. The livestock area has expanded from 14.6
to 38 million ha in the past 50 years, mostly at the expense of tropical forest (MADS, 2012). As seen in
Figure 2.8, the majority of pasture area is located in the eastern plain and Caribbean region. In number
of livestock, the northwestern part of the Amazon region also holds a large share (Figure 2.9).
Figure 2.7 Managed pasture and grasslands cover, 2007
Source: IAvH et al, 2007.
Figure 2.8 Percent of pasture area by municipality
Source: MADR 2011.
Figure 2.7 Number of cattle by municipality
Source: FEDEGAN 2011.
2.3 Colombia Climate Change Policies
2.3.1 National Policies and Plans Associated with Mitigation
Colombia has developed plans and policies for mitigating climate change, establishing priority sectors
with high GHG emission rates. A working group on climate change is led by Ministry of Environment and
Sustainable Development (hereafter MADS), which coordinates Colombian mitigation actions and
strategies for GHG reduction among government ministries with responsibility for the key productive
sectors. The group works closely with other public, private and civil society organizations participating in
emissions reductions research and development. Coordinated by MADS, the key public sector initiative
for reducing emissions is the Colombian Low Carbon Development Strategy (Behrentz et al. 2013).
Launched in 2011 through CONPES 3700
, the ECDBC seeks to decouple the increase of GHG from
national economic growth (DNP 2011). This will be done through the design and implementation of
The National Council on Economic and Social Policy (CONPES) is the highest national planning authority and
serves as an advisory body to the government on all policy related to the economic and social development of
plans, projects and policies aimed at reducing emissions and through simultaneously strengthening
economic growth, in compliance with worldwide standards of efficiency, competitiveness and
environmental performance. The government sectors that are involved in ECDBC include Industry,
Energy, Mining, Transportation, Housing, Waste and Agriculture. One private sector initiative that
supports low emissions development is led by the National Association of the Colombian Businessman
(Herrera et al., 2010). An important civil society initiative was a recent analysis of deforestation drivers
in Colombia and their implications for emissions (Nepstad et al. 2013).
2.3.2 Emissions reduction in the land use, land use change, agriculture and forestry
Priority areas for low emissions development in the agriculture, forestry and land use sectors of
Colombia include reducing emissions from deforestation and degradation (REDD), oil palm, livestock,
forestry and fertilizers. Efforts to reduce emissions from deforestation are currently centered on a
proposed results-based financing initiative that could bring several hundred million dollars to efforts to
reduce deforestation in the Colombian Amazon. The joint effort proposed by Colombia, Norway and
Germany – which aims at zero deforestation by 2020 – has been announced by the governments and
has just recently begun the implementation process. In the oil palm sector, the government has
proposed to increase the area in oil palm to three million ha in order to satisfy market demand,
including needs for biofuels to be mixed with gasoline (Castiblanco et al. 2013). The major initiative in
the livestock sector is to reduce the pasture area by ten million ha while at the same time increasing the
cattle herd (FEDEGAN 2006). Other initiatives related to forestry development, fertilizer use and others
have been discussed with little or no implementation to date.
3 Modeling Framework
We combine and reconcile economic and biophysical models to provide estimations of temporal shifts in
GHG emissions, carbon stock, and economic revenues. This approach includes the use of the following
The International Model for Policy Analysis of Agricultural Commodities and Trade
(IMPACT; Rosegrant et al. 2012) model, a global partial equilibrium agriculture model that
allows for policy and agricultural productivity investment simulations;
A spatially explicit model of land use choices to determine the possible effects of future
changes in the drivers of land use choices; and
DeNitrification–DeComposition crop model (DNDC; Li 2007) that estimates spatially
explicit profiles of GHG emissions from cropland with varying crop genetic productivity
shifters, management systems, and climate scenarios.
The use of this modeling environment provides detailed country-level results embedded in a
framework consistent with global outcomes. By using the crop model, which incorporates the most
updated knowledge on GHG emissions generated by crop production, it is possible to simulate the
effects on GHG emissions of current and alternative agricultural management practices (Figure 3.1).
We choose to work, in a somewhat arbitrary manner but with the goal of making the results of our
analysis as relevant as possible to policy-makers, with a relatively short time horizon, about 20 years
(from 2008 – 2030).
Figure 3.1 Workflow of the modeling approach
3.1 IMPACT Model
For the country of Colombia, which opens its market to the global economy, it will be crucial to estimate
the policy effectiveness and its consequence under the influence of global market forces.
IFPRI’s IMPACT model uses a system of linear and nonlinear equations to approximate the underlying
production and demand relationships of world agriculture. It uses country-level elasticity supply and
demand estimates (Rosegrant et al. 2012). The world’s food production and consumption is
disaggregated into 115 countries and regional groupings, with a further disaggregation in many regions
to the river basin level and with the basic unit of analysis being the food production unit (FPU, see Figure
3.2). The model includes 64 commodities (Table 3.1), including all major cereals, soybeans, roots and
tubers, meats, milk, eggs, oils, oilcakes and meals, vegetables, fruits, sugarcane and beets, and cotton.
IMPACT models the behavior of a competitive world agricultural market for crops and livestock, and is
specified as a set of country or regional submodels, within each of which supply, demand, and prices for
agricultural commodities are determined. The country and regional agricultural submodels are linked
through trade so that the interactions among country-level production, consumption, and commodity
prices are captured through net trade flows in global agricultural markets. Demand is a function of
prices, income, and population growth. Growth in crop production in each country is determined by
crop prices and the rate of productivity growth, from agricultural research and development,
agricultural extension and education, markets, infrastructure, and irrigation.
Figure 3.2 IMPACT’s global partitioning
Source: Rosegrant et al. 2012.
Table 3.1 IMPACT commodities
Groundnuts for oil
Sheep and goats
Rapeseed for oil
Soybeans for oil
Cassava and other roots &
Sunflower seeds for oil
Oil palm fruit
Other roots & tubers
Palm kernel oil
Palm kernel meal
Total other oilseeds
Total other oilseeds for oil
Total other oils
Total other oilseed meal
Source: Rosegrant et al. 2012.
IMPACT is a partial equilibrium agricultural model in which output from several other models are
imputed. The models that provide output include Global Circulation Models (GCM), which simulate
climate scenarios under global climate change; Decision Support System for Agrotechnology Transfer
(DSSAT) crop model suite, which estimates yields with varying crop genetic productivity shifters;
management systems and climate change scenarios; Spatial Production Allocation Model (SPAM), which
provides crop allocation and management techniques such as rainfed/irrigated (Rosengrant et al. 2012);
and some IMPACT submodels, such as IMPACT Water Simulation Model (IWSM) and IMPACT Global
Hydrologic Model (IGHM). The resolution of IMPACT model’s final output is at a subnational or national
level, while some of the connecting models use higher spatial resolution.
3.2 Land Use Model
The utilization of a land use model is the most prominent addition to the IMPACT core component. It is
essential, given the importance that land use change plays in changes in GHG emissions and in carbon
stocks, and its impact on the economy. Furthermore, the effectiveness of most agricultural practices
that could mitigate GHG emissions is strongly dependent on the specific characteristics of the land
where these efforts are taking place. It is only by giving proper consideration to the specific
characteristics of the location where mitigation efforts are taking place that a higher degree of accuracy
in the assessment of GHG emissions is achieved.
We use a model of land use choices based on a Maximum Likelihood method (Li et al. 2014). This is a
spatially explicit model of land use choices that captures the main drivers of land use change. The model
estimates changes in land use during two time periods. Required data are the land use maps and crop
allocation datasets, variables that are likely to affect profit levels from land uses, and other location-
specific characteristics consisting of biophysical and socioeconomic variables (see Table 3.2).
Table 3.2 Summary of the variables used for the land use model
Choice variable: crop shares within
provinces (2008 and 2030)
Crop area, crop suitability, commodity producer price,
elevation, terrain slope, soil pH, annual precipitation,
annual mean temperature
Choice variable: land cover (2008 and
elevation, terrain slope, soil pH, annual precipitation,
annual mean temperature, population density, travel
time to major cities, conserved areas, indigenous
The basic logic of this modeling approach is that each landholder will choose a land use that
maximizes the stream of benefit. As shown in Figure 3.3, the model structure consists of two levels of
land use choices. The upper level estimates choices among large aggregation categories, such as
cropland, pasture, forest, and other uses. The upper choices are determined by biophysical and
socioeconomic variables. The lower level estimates crop choices of landholders within cropland and is
estimated using cross-sectional data. Subsequently, using projections for the future values of key drivers
of land use choices, the land use choices in a future period are determined. A detailed description of the
conceptual framework, model specification, and estimation method is provided in Appendix 1.
Figure 3.3 Structure of the land use model
The explanatory variables used in the model are listed in Table 3.2. For the upper-level model, we
use a wide variety of biophysical and socioeconomic variables that influence land use choice, while for
the lower-level section we employ biophysical variables related to crop production, as well as some
economic variables such as crop prices, which we consider to be key determinants of land allocation
choices among crops. To simulate land use choice for the year 2030, we use projections of commodity
prices and the climate scenario MIROC General Circulation Model in combination with socioeconomic
assumptions: AR5-SSP2 scenario (Shared Socioeconomic Pathway 2—middle of the road scenario).
3.2.1 Data for Upper-Level Model
As mentioned in Chapter 1.4, we use the Colombian government statistics for cropland area (IGAC,
2013) as well as for forest and for pasture (IAvH et al. 2007). Based on the data available, we classify the
land cover using the categories of cropland, forest, pasture, and other land uses. Other land uses
includes shrub and secondary vegetation.
Population density data were gathered for the year 2000 from Gridded Population of the World,
Version 3, with a spatial resolution of ~1 km (CIESIN/Columbia University/CIAT 2005). We use a lagged-
population density value to mitigate the endogeneity issue that the use of this variable might cause in
the land use model.
The land use model uses area and price projections from IMPACT. IMPACT generates scenarios based on four
different Global Circulation Models (GCMs): Hadley, IP SL, MIROC, and GFDL. While the choice of MIROC is
arbitrary, the average difference among IMPACT projections for crop areas and prices for the year 2030 for
Colombia using the four GCMs is rather low, about 0.5 percent (with some larger differences for palm and cocoa
areas using GFDL: 3 percent and 2 percent, respectively). MIROC seems to provide a “middle of the road” scenario
for Colombia and for the year 2030. We acknowledge that additional exploration of alternative GCM-generated
scenarios is necessary to obtain a more robust set of results.
Data on market accessibility, measured by travel time to major cities (cities of 50,000 or more people
in year 2000), were generated from a global map of accessibility with a resolution of 30 arc-second (~1
km) (Nelson 2008). This variable was computed using a cost-distance algorithm that computes the cost
of time required to travel across pixels in the map. Factors affecting the travel cost include transport
network and environmental and political factors (see Nelson 2000, 2008 for details).
Elevation and slope data were generated from a resampled Shuttle Radar Topography Mission
(SRTM) digital elevation data with a 1-km spatial resolution, produced by the U.S. National Aeronautics
and Space Administration (NASA) and the U.S. National Geospatial-Intelligence Agency (NGA). The
original SRTM data were available with a 90-meter resolution. The original elevation data contain “no-
data” observations where water or heavy shadow prevented the quantification of elevation. These
observations were further processed by Jarvis et al. (2008) to fill in the no-data voids.
Climate data, measured by mean precipitation and mean annual temperature, are generated from
WorldClim. Using MIROC Global Climate Model−AR5 SSP2, we apply rates of change to extrapolate the
climate condition for the year 2030. Soil pH was collected from Harmonized World Soil Database
(FAO/IIASA/ISRIC/ISS-CAS/JRC 2012). Conservation practice and indigenous reserves are defined as a
dummy variable, extracted using the Colombian government’s map (SINAP).
3.2.2 Data for Lower-Level Model
In contrast to the upper-level model, the lower-level model combines local agricultural statistics with
geo-referenced data. We used a dataset for cropland physical area at municipality level (in total 1,121
municipalities) for the year 2008 (MADR 2008). Based on these data, we identified the five major annual
crops in terms of areas of cultivation: cassava, maize, potato, rice, and sugarcane, and four major
perennial crops: cacao, coffee, palm, and plantain. To determine the crop choice that returns the
highest benefit at each location, we introduce the domestic producer price for each crop, adjusted by
the effect of market accessibility for each location.
Domestic commodity producer prices were generated for the period 2008−2030 from IMPACT. We
use the 2008 data to estimate the model and use the 2008−2030 data as a baseline for policy simulation.
The use of country-level commodity prices calculated from IMPACT substantially reduces concerns
about endogeneity between municipal crop choices and local producer prices.
We assume that domestic commodity producer prices can be observed at the major markets located
in cities with populations greater than 50,000 and that farm-level producer prices move synchronously
with domestic prices. With these assumptions, we estimate the spatially explicit prices, , for each
location and each crop using a distance decay function:
where represents the domestic producer price of commodity j and is travel time from
each location n to its nearest major market. The specification of distance decay function is arbitrary.
Equation (1) has desirable properties that the producer price decays at a moderate speed and the
spatially explicit price is between 0.368 ( 0.368) and .
Data on biophysical suitability are derived from the global Agro-ecological Zones (GAEZ, v1.0)
assessment (Fischer et al. 2001), which was developed by the International Institute for Applied Systems
Analysis (IIASA) and the Food and Agriculture Organization of the United Nations (FAO). Suitability data
are available at a resolution of approximately 9 km at the equator. These data are used as a proxy for
the maximum attainable yield in Colombia.
3.3 Crop Emission Model and Other Emissions
To estimate GHG emissions from cropland in Colombia, we employed the biogeochemical model—
DeNitrification and DeComposition (DNDC) (Li et al. 1992)—along with high-resolution remote sensing
data (5 arc min grid) and information about local agricultural practices. Carbon dioxide (CO2), methane
(CH4), and nitrous oxide (N2O) emitted from soil in cropland are all accounted for. Figure 3.4 illustrates
data required to run the DNDC model and Figure 3.5 shows the process followed for the computation of
Figure 3.4 Data requirement for the use of the DNDC model
Figure 3.5 Simulation flow of GHG emissions in DNDC model
We first identified cropland and mosaic cropland land use areas categorized by National Aeronautics
and Space Administration’s Moderate Resolution Imaging Spectroradiometer (MODIS) and further
divided them into a 5-minute cell grid (about 10 by 10 km cells). For each cell, we applied the DNDC
model to estimate crop productivity and GHG emissions. To derive the crop and location-specific data
required for simulations in DNDC, information about local cropping systems and nitrogen fertilization
rates were obtained through interviews and literature reviews while remote sensing data were used to
characterize crop calendars (that is, planting and harvesting dates) (Sacks et al. 2010) and soil physical
and chemical properties (such as texture, soil C content, pH, and bulk density) (FAO/IIASA/ISRIC/ISS-
CAS/JRC 2012). To derive future weather series corresponding to the MIROC climatic scenario and
ensure consistency with the IMPACT scenario, we utilize climatological data developed by the CGIAR
Research Program on Climate Change, Agriculture and Food Security and the MarkSim weather
generator accompanied by the data (www.ccafs-climate.org/pattern_scaling). Note that this is an
essential step given that GHG emissions and changes in soil organic C vary with soil types and other
location-specific biophysical determinants.
Finally, DNDC simulations were run to make forward projections of GHG emissions from four annual
crops—maize, rice, cassava, and potato—and two perennial crops—plantain and palm—for the period
of 2008 to 2030 (Figure 3.6). We assumed current crop management practices and input levels stay
constant. We further computed global warming potentials (GWP) by using fluxes in GHG and changes in
soil organic C content. In order to smooth out the effects of single weather event fluctuations, such as a
year with abnormal precipitation or temperatures, we computed the average per hectare GWP for each
crop and for each pixel over the period 2008 to 2030. Given that total emissions are computed by
multiplying per hectare emissions by crop area estimations, changes in total emission reflect only the
effect of change in crop allocation decisions.
Figure 3.6 Calculation of greenhouse gas warming potential in cropland emission with DNDC
For the remaining major crops cultivated in Colombia—coffee, cacao, and sugarcane—we used
emission factors available in the literature (Palm et al. 2002, Hergoualc’h et al. 2008, de Figueiredo
2010). In addition to cropland emissions, we considered GHG emissions from land allocated to pasture,
such as methane emission from enteric fermentation and manure management, as well as N2O
emissions from manure left on the ground. To generate location-specific estimations, we used national
statistics for GHG emissions from livestock (FAO 2014) and IMPACT values for the country herd size to
generate per-head emission estimates.
Using total herd size and total hectares of pasture, we also
estimate the number of animals per hectares and therefore emissions per hectare of pasture. With
respect to GHG emissions from forest and other land use, due to lack of data, we disregard emissions
from other sources, such as biomass burning and decomposition.
The herd size is computed using IMPACT model’s output on number of slaughtered animals and by assuming a
constant ratio of slaughtered animal and herd size.
3.4 Carbon Stock
We estimate the following pools of carbon stocks: above- and belowground biomass as well as soil
organic carbon (SOC). Changes in stock are estimated in proportion to area changes at the municipality
level for each land cover type (Figure 3.7). We make the simplifying assumption that, for uses other than
cropland, carbon stock does not change through time if land remains in the same land use.
Figure 3.7 Calculation of carbon pools and GHG fluxes
We provide an estimation of above -and belowground biomass in forest, grassland, other land use,
and major perennial crops. For forest, grassland, and other land use, we use a spatial aboveground
biomass dataset (Saatchi et al. 2011, Anaya et al. 2009) to calculate average biomass for each land use.
According to the country’s land cover map (IAvH et al. 2007), nearly 50 percent of “other land uses”
category in Colombia consists of shrub and secondary vegetation. Biomass for this land cover category is
also taken into account. For the estimation of belowground biomass, root-shoot ratio for each land
cover type is adopted from the available literature (IPCC 2006, Saatchi et al. 2011). For major perennial
crop, we used biomass figures from Dossa et al. 2007, Koskela et al., 2000, and Henson et al. 2012.
Municipal average SOC per land use type was elaborated using a global SOC dataset (Hiederer and
Köchy 2011), land cover map (IAvH et al. 2007), and soil type map (Niels 2010). Average SOC is
calculated for extracting existing combinations of soil type and land cover in each municipality by
overlaying each map and weighing each combination according to the corresponding area size within
the municipality. Table 3.3 provides an overview of all carbon pools and GHG fluxes accounted for in the
Table 3.3 Carbon pools and GHG fluxes by land use category
Other land uses
a Figures are only for perennial crops. b Figures are only for areas with shrub and secondary vegetation.
4 Economic Trade-Offs, Baseline, and Alternative Scenarios
Policies that aim at reducing emissions can generate trade-offs by reducing or increasing area allocated
to crops, pasture, or forests or by affecting yields or production costs. Due to lack of data, we don’t
consider possible changes in costs of production or input saving effects and we only account for changes
in revenues brought about by projected changes in yields, prices, and land allocations. Forest economy
wasn’t included due to lack of data. Yields and producer prices from national statistics (FAO 2014) are
used for the year 2008 while growth rates for yields and prices from the IMPACT model are used for
Figure 4.1 summarizes the methodology used to compare alternative policy scenarios with the baseline.
The modeling framework we use for the analysis of low emission development strategies generates
estimations and projections for the base year (2008) and the end-period (2030). Therefore, the annual
rates of change must be extrapolated from the starting and ending values. The annual rate of change is
essential to compute the total amount of GHG emissions generated under a particular scenario.
Figure 4.1 Computation of changes in GHG emissions and carbon stock for baseline and alternative policy
The stylized example provided in Figure 4.1 is a case where both GHG emission and carbon stock
increase in 2030 compared to 2008. Total change in emissions is obtained by the extrapolated yearly
changes throughout the time period considered (area triangle ABC). Similarly, the computation of
changes in GHG emissions deriving from the implementation of a particular policy would be obtained by
measuring the area of triangle ACD. Note that the effect of each policy is computed against the changes
in emissions and carbon stock projected for the baseline.
For the baseline, we assume that all field activities and agronomic practices stay constant, and
therefore our computation of GHG emissions is based only on changes in land use due to changes in
commodity prices and other underlying assumptions in IMPACT regarding changes in GDP, population
growth, and climatic conditions.
Carbon stocks present in the soil and above- and below-ground biomass in each land use are also
assumed constant, thus carbon stock accounting also is based only on land use change. Changes in
revenues are similarly computed as cumulative effects over the 21 years under consideration.
5 Baseline Results
IMPACT model results, reported in Table 5.1, indicate that global market forces determine significant
increases in area for oil palm, plantain, and sugar cane (20%, 13%, and 31% respectively). Decreases in
area are projected for cacao, maize, and rice. The interpretation of these results is rather complex as
these changes are dependent on global changes in incomes and diets (the decrease in demand for
economically inferior goods such as rice can be attributed to that). A more detailed interpretation of
these trends can be found in Msangi and Rosegrant 2011.
Table 5.1 IMPACT projected changes in prices and areas for crop and livestock commodities, 2008–2030
change in price
Source: Authors. * Area for pasture.
These changes in crop areas are used as an input in the land use model. The technical details on how
the project changes are used in the land use model are provided in Appendix 1. We rely heavily on the
land use model for the spatial allocation of the changes in areas and therefore we need to evaluate how
correctly the model predicts land use choices.
Table 5.2 and Table 5.3 present an assessment of the predictive power of the upper and lower levels
of the land use model, respectively. In each table, columns 1 and 2 correspond to the observed land use
derived from the data sample; columns 3 and 4 correspond to the predicted land use estimated from
the land use model; the last two columns present the discrepancies by subtracting entries in columns 1
and 2 from entries in columns 3 and 4. We find that the model performs well in predicting most land
uses. As far as the upper level is concerned, the model provides generally accurate in-sample
predictions. The highest and lowest discrepancies, expressed in percentage terms, are -0.2 percent for
perennial cropland and -2.3 percent for forests. Regarding the lower-level component, the model tends
to overestimate perennial and annual croplands by a total of 0.2 percent and 0.4 percent, respectively.
These discrepancies arise mainly from overestimating areas allocated to palm, cacao, maize, and rice.
Even so, the predicted errors of these four crops in terms of percentage are in a reasonable range, from
2.3 percent to 3.2 percent. The model performs well in the prediction of the remaining crops in terms of
both magnitude and percentage.
Table 5.2 Assessment of the predictive power of the upper-level model
Table 5.3 Assessment of the predictive power of the lower-level model
Table 5.4 reports the results for the upper level of the land use model and shows how changes in
agricultural product prices are projected to affect large land use categories. The most significant change
is in forested area: a decrease of 3.4 million hectares, an average loss of some 150,000 hectares per
year. Most of the change is caused by an increase in pastureland, 2.6 million hectares, and in lower
amounts by growth in perennial and annual crop areas.
Table 5.4 Predicted land use change, 2008–2030
Land use category
Other land uses
Source: Authors. Both 2008 and 2030 figures are model predictions.
Figure 5.2 shows projected change for the two crops with the largest growth in area. Among the
annual crops considered, sugarcane shows the greatest projected change in area with a gain of 107,000
hectares, and area allocated to oil palm production increases by 71,000 hectares. Rice, cacao, and maize
areas are projected to decrease by some 7,000, 3,000, and 2,000 hectares respectively. Projected
changes in land use translate into changes in carbon stocks, and these are reported in Table 5.5. The
model results show a net decrease in carbon stock, a total of 364 Tg C equivalent to 1,335 Tg CO2 eq.
This large reduction in the stock of carbon is due mostly to the decrease in forested area.
Figure 5.1 Projected changes in forest and pasture areas at municipality level, 2008–2030
Figure 5.2 Projected changes in sugarcane and palm area at municipality level, 2008–2030
Table 5.5 Change in carbon stocks, 2008–2030
Table 5.6 reports the estimated changes in GHG emissions from cropland and pastureland. By far, the
greatest change in emissions is caused by pastureland expansion as well as changes in livestock
intensity. These are driven by increases in demand for livestock products as projected by IMPACT for the
year 2030. The next significant increase in GHG emissions comes from oil palm cultivation. This is
projected to generate a total increase equivalent to 2.1 Tg of CO2 eq. Note that these changes are driven
completely by changes in land use, and we assume that land management practices, as well as livestock
management practices, remain unchanged for the entire period considered. This means, for instance,
that the emission reduction in plantain and cassava is due to a shift of cultivated area into locations
where the same and management practices coupled with different land characteristics generate lower
per-hectare emissions. Revenues deriving from the agricultural activities considered increase by a total
US$44 billion during the period 2008 - 2030.
Table 5.6 GHG emissions and revenues in cropland and pasture, 2008–2030
Change in area
per ha GHG
per ha GHG
Total change in
Total change in
Source: Authors calculations.
** Difference between accumulated GHG during 2008–2030 and 21 years of GHG emissions by keeping
the annual GHG emission constant at the beginning of year.
*** Difference between accumulated revenue during 2008–2030 and the total of 21 years of revenue by
keeping the annual revenue constant at the beginning of year.
6 Policy Simulations
Colombian public and private sectors are invested in finding viable strategies to reduce emissions from
the agriculture sector and from land use change. Our goal is to provide research results that reflect the
interests and priorities of a wide range of stakeholders interested or involved in the designing of Low
Emission Strategies. For this, we held a series of workshops designed to solicit opinions and ideas on
high-priority policy objectives from government officials, private-sector representatives, and
Sectoral workshops for forestry, livestock, and food crops were held and preliminary research results
were presented to facilitate targeted discussions. Following this first series of workshops, another series
of meetings was held in which stakeholders from all the sectors were brought together and policy
targets were identified. Once consensus among stakeholders was reached concerning the importance of
certain scenarios, those scenarios were then simulated in our modeling framework.
The following are the policy objectives chosen:
1. Reduce the area allocated to pasture.
2. Reduce the rate of deforestation in the Amazon forest.
3. Increase the area allocated to the cultivation of palm.
Upon completing the final workshop, additional information on each policy was gathered and
numerical targets were defined. Given that some of the policy targets are clearly aspirational in nature,
we simulated two stages of accomplishment, (1) scenarios in which a goal is fully accomplished and (2)
scenarios in which only half of the target is achieved. It is essential to note that the goals are simulated
with respect to the baseline, which represents our best prediction of what the landscape will look like in
2030. All these scenarios are simulated in the land use model by altering the benefits deriving from the
targeted land use, in comparison to alternative uses, until the desired amount of hectares is achieved.
The Colombian livestock sector association, Federación Colombiana de Ganaderos (FEDEGAN), proposed
a strategy called “Plan Estratégico de la Ganadería 2019 (PEGA)” (FEDEGAN 2006). This strategy
proposes an increase of the total number of heads up to 48 million, while reducing pasture area down to
24 million hectare (equivalent to a 10-million-hectare reduction in pastureland) by 2019. The Ministry of
Agriculture, meanwhile, proposes reducing land allocated to pasture to a total of 13 million hectares.
FEDEGAN proposes to return the 10 million hectares to natural vegetation through promoting
reforestation and silvopasture systems (FEDEGAN 2006). Such changes in land use can lead to significant
changes in carbon stocks and GHG emissions. In this study, we simulated the effects of a reduction in
pastureland of 10 million hectares, as well as a “middle of the road” scenario in which only 5 million
hectares are removed from pasture.
It is estimated that during the period 2005–2010, 0.2–0.4 million hectares of forested area were lost in
the Amazon, corresponding to an annual deforestation rate of 0.5–1 percent for the region (see Table
2.2). Prevention of deforestation in the Amazon forest is considered a critical objective for climate
change mitigation as well as biodiversity and cultural conservation. “The Colombian Low Carbon
Development Strategy” proposed by the government sets as a goal a complete halt to deforestation in
the Amazon region, while lowering the rate of deforestation in the rest of the country.
The implementation mechanisms considered range from promoting sustainable forest production, to
agroforestry, to zoning and regulation, to community capacitation, to name a few. Challenges associated
with achieving these goals are the market forces of infrastructure and mining development; livestock
and cropland expansion, including biofuel crop; illegal crop production; sustainability of policy
implementation; and effects of climate change on ecosystems.
We simulated two scenarios: a complete stop to deforestation in the Amazon forest and a reduction
by half of the deforestation in the Amazon predicted for 2030.
6.3 Palm Cultivation
The Colombian government is planning to incentivize the cultivation of strategic crops. The list of
strategic crops includes palm, sugarcane, and soybeans, the expansion of which have been driven by
both biofuel and dietary demands. The expectation is that the majority of the area expansion would
occur in pastureland, which is widely considered underutilized land. Because each crop has different
GHG emission profiles as well as different rates of carbon sequestration (in the case of perennial crops)
that varies from location to location, any crop expansion policy will have implications for a GHG
We simulated two policy objectives, (1) an expansion in area allocated to palm cultivation by a total
of 1.5 million hectares
and (2) a scenario in which 50 percent of this goal is achieved. The land use
model simulates the case where the utility of palm production is increased at the country level by
implementing certain policies (Table 6.1).
The government plans indicate a goal of over three million hectares of oil palm plantations by 2020. However, as
other authors have indicated (Castiblanco et al 2012) it highly unlikely that this goal will be met. Our IMPACT
simulations also indicate that if such increase in production were to be achieved, world prices would be depressed
by about five percent, which makes attaining the goal even more difficult.
Table 6.1 Summary of policy simulations
Policy target fully met
5 million ha
10 million ha reduction of
deforestation in the
A complete stop to
deforestation in the
Total land allocated
to palm production
Total land allocated to
palm production increased
by 1.5 million ha
We examined the outcome of such goals in terms of their total potential for GHG emission reduction.
Furthermore, we assess the effects of policies on total agricultural output and total revenue so that we
can characterize the various options not only using the potential for emission reduction but also their
economic trade-offs. Table 6.2 shows the changes in land use induced by the implementation of these
objectives. Table 6.3 shows the effects on carbon stock, GHG emissions, and total revenue from
Table 6.2 Land use change under alternative policy scenarios
Policy targets partially met
Policy targets fully met
Other land uses
Table 6.3 Summary of predicted total changes in carbon stock (C stock), GHG emissions, and revenues, 2008–
Change in C stock
Total change in GHG
Total change in revenues
Reduce pasture by
10 million hectares
Reduce pasture by
5 million hectares
Zero deforestation in the
Reduce deforestation in
the Amazon by 50%
Increase area allocated
to oil palm by 1.5 million
Increase area allocated
to oil palm by 750,000
6.4 Policy Target 1: Reduction in Area Allocated to Pasture
Baseline results predict a 2.6-million-hectare expansion to a total of 31.8 million hectares of
pastureland, equivalent to a 9 percent growth during the period 2010 – 2030. Results of the policy
simulation (Table 6.2) should be interpreted as assessing the changes in land use brought about by a
policy that limits expansion and reduces pasture area by a total 10 million hectares compared to the
baseline. As can be observed in Table 6.2 , the effects of a 10-million-hectare reduction in pasture area
causes growths in cropland (1.1 million hectares), forest (4.1 million hectares), and other land uses (4.7
million hectares). These changes correspond to an estimated increase in carbon stock of some 402 Tg C
compared to the baseline (Table 6.3). GHG emissions are also affected. Emissions from pastureland and
the associated livestock are expected to decrease while emissions from increased crop production rise.
Given our assumptions on livestock density (heads 0.92 in 2008 and 0.99 in 2030), the reduction in
pasture area leads to a reduction in revenue compared to the baseline, which is estimated to be about
$8 billion over the 21 years considered. However, the $18 billion increase in revenue deriving from
cropland expansion is expected to compensate for the loss and generate a net gain of some $10 billion.
We also estimate that in order to offset the economic loss in the livestock sector (maintain livestock
revenue at baseline level), it would be necessary to raise the livestock intensity by 34 percent from the
2008 level (livestock density: 1.44). The intermediate scenario, 50 percent achievement of the policy
goal, also returns the double benefit of a reduction of emissions and economic gains. Carbon stock is
projected to increase by 196 Tg C while GHG emissions from crop production and pastureland are
reduced by a total of 76 Tg CO2eq. The expected net change in revenues is estimated at $5 billion.
Figure 6.1 Projected effects of a policy that reduces pastureland by 10 million hectares, 2008–2030
6.5 Policy Target 2: Reduction in the Rate of Deforestation in the Amazon
Results for the baseline scenario indicate a decrease in forested area of 3.4 million hectares during the
period 2008–2030, equivalent to an annual average loss of some 155,000 hectares. Results also indicate
that the loss in carbon stock due to deforestation is substantial: a total of 774 Tg of carbon for the
period under consideration. Deforestation occurs in other areas besides the Amazon forest (Figure 6.2
however, given its importance, the government puts an emphasis on the Amazon. The implementation
of a policy that stops deforestation in the Amazon would result in an additional 1.3 million hectares of
forest compared to the baseline (Table 6.2 ). However, this would be at the expense of land that would
otherwise transition into pasture, 947,000 hectares less compared to the baseline, and into about
17,000 hectares of land allocated to crop production. Considered all together, the changes in land use
lead to 111 Tg C of net increase in carbon stock compared to the baseline and a total reduction in GHG
emissions quantified to be about 17 Tg CO2eq, compared to the baseline. However, the reduction in
pasture and cropland expansion into the forests causes a total loss of $1 billion in revenue that would
have derived from the sale of livestock products and food crops.
The scenario in which we simulate an
intermediate achievement of the policy goal (deforestation is reduced by half) returns similar trade-offs
between mitigation and economic returns. The increase in carbon stock, which amounts to 78 Tg C, and
the 8 Tg CO2eq of reduction in cropland and pasture emissions, come at the expenses of agricultural
revenue, which is reduced by about $0.2 billion.
It is assumed that forest does not generate a revenue stream. Although this is likely not true in all cases, lack of
data prevents us from accounting for the contribution of forest products to the economy.
Figure 6.2 Projected effects of a policy that halts deforestation in the Amazon, 2008–2030
6.6 Policy Target 3: Increase the Area Allocated to the Cultivation of Palm
The baseline scenario projects a growth in area allocated to oil palm cultivation of about 71,000 by
2030. In order to reach the policy target of 1.5 million hectares, land allocated to oil palm has to
increase by more than 1 million hectares. Our simulation reveals that this policy could have significant
unintended consequences if not implemented judiciously. Model results (Table 6.2 ) indicate that the
expansion of oil palm production occurs at the expenses of area allocated to perennial crops, causing
decreases in plantain area (339,000 hectares), coffee area (243,000 hectares), cacao area (126,000
hectares), and other perennials (131,000 hectares). This affects other land uses as well; total changes in
land use results in 63 Tg CO2eq of net effect of GHG emission growth, consisting of a decrease in carbon
stock of 11 Tg C and an increase in GHG emissions of 24 Tg CO2eq, compared to the baseline. It is
important to note that this policy objective comes with heavy economic costs. In contradiction to
conventional wisdom, the expansion of oil palm cultivation only partially occurs on pastureland. The net
effect of this expansion is a loss in revenue equivalent to $18 billion for the period 2008–2030,
compared to the baseline.
The intermediate-goal scenario (50 percent of the policy goal) has instead a climate change
mitigation effect. This is due to complex land use transitions, the net effects of which are an increase in
land uses such as forest and pastureland, with a consequent increase in carbon stock quantified at 10 Tg
C and a total 8 TgCO2eq of emission growth. Meanwhile, agricultural revenue is still negatively affected,
with a reduction of $10 billion. This is due to the reduction of other perennial crops.
Figure 6.3 Projected effects of a policy that expands palm cultivated area to a total of 1.5 million hectares, 2008–
Figure 6.4 Trade-off between revenues and GHG emission reduction
7 Discussion and Conclusions
Results reveal the importance of considering the full scope of interactions and changes in the various
land uses when planning for GHG reduction policies. The carbon stock stored in forests often
overwhelms the possible increases in GHG emissions generated by food crop production, as other
studies have noted (Burney et al. 2010, Gockowski and Sonwa 2011, Li et al 2014). Overall shifts in land
uses determined by changes area allocated to agricultural production can have great effects on existing
carbon stock which might be more significant than the resulting changes in GHG emission from crop
cultivation. Our results indicate that one additional hectare allocated to agriculture, on average,
increases GHG emissions by some 2.5 Mg CO2eq per year while one hectare of forest lost in the Amazon
results in a loss of carbon stock (above- and below-ground biomass) equivalent to some 367 Mg CO2eq.
This means that it would take about 146 years for an additional hectare of agricultural land to cause the
same damage of one lost hectare of Amazon forest. The adoption of alternative agricultural practices in
crop production was not indicated as a priority by the stakeholders consulted and therefore none of the
possible GHG-emission reducing alternatives was simulated. However, to provide some perspective on
the possible contribution of mitigation activities applied to crop production, we simulated a hypothetical
reduction of GHG emissions by 50 percent across all crops. By the end of 2030, this would generate a
total reduction in GHG emissions of about 120 Tg CO2eq. Our results for the baseline scenario project a
total loss of carbon stock equivalent to 1,335 Tg CO2eq. Therefore, even a drastic reduction in GHG
emissions represents less than 10 percent of the change in carbon stock induced by changes in land
uses. This is not to say that reduction of emissions in crop production is not important, but rather to
emphasize the importance of considering the totality of effects of policies that induce changes in the
land use. Figure 6.4 provides an overview of the outcomes of the simulated policies with respect to the
baseline. For comparison purposes, we express changes in carbon stock in CO2eq and add them to
changes in GHG gasses. This allows to compare performances for the period under consideration. Of
course, we need to keep in mind that changes in carbon stock are a one-time event while changes in
emissions are yearly recurrences. “Win-win” policies are those represented in the upper left quadrant of
the graph (decreased GHG emissions and increased revenues). The policy that targets a reduction of
land allocated to pasture fairs better than all others and returns significant mitigation and economic
gains. It is particularly important that, even though reducing pasture in our simulation is essentially
equivalent to a reduction in the total number of animals and a consequent decrease in revenue, the
increment in revenues from other agricultural products offsets the loss and generates a net gain. The
outlook of this policy appears particularly favorable considering that increases in animal density and in
pasture productivity are thought to be within reach. The policy that targets deforestation in the Amazon
forest is a typical example of a policy that generates trade-offs. The significant gains in terms of carbon
stock and reduction in GHG emissions generate projected loss in total revenues of $1 billion. Expressed
in dollars per ton of GHG abated, these results indicate costs in the order of $2 to $3 per Mg CO2eq.
However, even if these costs are not prohibitive, this policy ranks relatively low compared to reducing
pastureland: the gain in carbon stock is about one-fourth of what would be achieved with the pasture
reduction policy. This is because a policy that targets pasture both avoids deforestation and induces a
more productive use of the land. The choice between the two policies therefore becomes an issue of
political feasibility and implementation challenges.
Plans that target increasing land allocated to oil palm cultivation need to be considered with extreme
caution and investigated further. Model results suggest that past trend of oil palm cultivation taking
over underproductive pastureland might not continue into the future and that, beyond certain levels of
increase in oil palm area, a detrimental reduction in land allocated to other perennial crops could
actually take place. This is not necessarily because oil palm cultivation replaces, for example, coffee, but
rather because oil palm might displace annual crop production and this in turn displace perennial crops.
Land use transitions can be complex. The model indicates that other important factors besides economic
profit might be at play and these impede the desired land use transition. This suggests that additional
research on this issue is warranted and that particular attention should be given to the instruments and
incentives that should be put in place to achieve the desired outcome.
Some general considerations can be made. Given the complexity of low emissions development
strategies, modeling approaches, frameworks, and tools should be adaptable, open, and transparent.
Modeling frameworks should be adaptable so that policy makers can explore the consequences of using
different data sets and incorporate new information as it becomes available. Modeling frameworks and
tools should be open to the inclusion of input from different models so that the robustness of the results
can be assessed. In the case presented in this article, we have chosen models with which the authors
were most familiar. Each one of models chosen comes with its own set of strengths and weaknesses. For
example, the use of a partial equilibrium model like IMPACT limits the scope of the analysis to the
agricultural sector and the use of a parametric model to determine land use choices does not consider
land uses that are not already present at the time of the analysis. These limitations become more or less
severe according to the specific situation and country analyzed. However, the advantage of the
approach proposed is that each one of the models does have alternatives and researchers have the
option to replace them or integrate them with other models. We believe that this level of transparency
and openness is essential to generate trust in the results among end users.
Anaya, J. A., E. Chuvieco, and A. Palacios-Orueta. 2009. “Aboveground Biomass Assessment in Colombia:
A Remote Sensing Approach.” Forest Ecology and Management 257 (4): 1237–46.
Antle, J. M., & Stoorvogel, J. J. 2008 “Agricultural carbon sequestration, poverty, and
sustainability.“ Environment and Development Economics 13(03): 327-352.
Beheydt, D., Boeckx, P., Sluetel, S., Li, C., van Cleemput, O., 2007. “Validation of DNDC for 22 long-term
nitrous oxide field emission measurements.” Atmos. Environ. 41: 6196–6211.
Behrentz, E., Á. Cadena, H. Mutis, J. F. Pérez, R. Rosales, R. Delgado, et al. 2013. Estrategia Colombiana
de Desarrollo Bajo en Carbono. Bogota: Universidad de los Andes.
Burney, J., Davis, S., Lobell, D. 2010 “Greenhouse gas mitigation by agricultural intensification.”
Proceedings of the National Academy of Sciences of the United States of America 107(26):12052–12057.
Cacho, O. J., Marshall, G. R., & Milne, M. 2005. “Transaction and abatement costs of carbon-sink
projects in developing countries.” Environment and Development Economics 10(05): 597-614.
Cai, Z., T. Sawamoto, C. Li, G. Kang, J. Boonjawat, A. Mosier, R. Wassmann, and H. Tsuruta. 2003 “Field
validation of the DNDC model for greenhouse gas emissions in East Asian cropping systems.” Global
Biogeochem. Cycles 17(4): 1107 doi: 10.1029/2003GB002046.
Castiblanco, C., A. Etter, and T. M. Aide. 2013. “Oil Palm Plantations in Colombia: A Model of Future
Expansion.” Environmental Science and Policy 27: 172–83.
CIA (Central Intelligence Agency). 2014. “CIA Factbook.” Accessed November 14, 2014. www.cia.gov.
CIESIN (Center for International Earth Science Information Network)/Columbia University/CIAT (Centro
Internacional de Agricultura Tropical). 2005. Gridded Population of the World, Version 3 (GPWv3):
Population Density Grid, Future Estimates. Palisades, NY: NASA Socioeconomic Data and Applications
Center (SEDAC). Accessed March 16, 2012. http://sedac.ciesin.columbia.edu/data/set/gpw-v3-
Chomitz, K, Gray, D. 1996. “Roads, land use, and deforestation: a spatial model applied to Belize.” World
Bank Economic Review 10:487–512.
De Figueiredo, E. B., A. R. Panosso, R. Romão, and N. La Scala. 2010. “Greenhouse Gas Emission
Associated with Sugar Production in Southern Brazil.” Carbon Balance and Management 5 (3): 1–7.
Deng, J., Zhou, Z., Zhu, B., Zheng, X., Li, C., Wang, X., Jian, Z., 2011. “Modelling nitrogen loading in a small
watershed in southwest China using a DNDC model with hydrological enhancements.” Biogeosciences 8,
De Pinto, A. and Nelson, G. 2009. “Land use change with spatially explicit data: a dynamic approach.”
Environmental and Resource Economics 43:209–229.
Diagana, B., Antle, J., Stoorvogel, J., & Gray, K. 2007. “Economic potential for soil carbon sequestration
in the Nioro region of Senegal’s Peanut Basin.” Agricultural Systems 94(1): 26-37.
Dossa, E. L., E. C. M. Fernandes, W. S. Reid, and K. Ezui. 2007. “Above- and Below-ground Biomass,
Nutrient and Carbon Stocks Contrasting an Open-Grown and a Shaded Coffee Plantation.” Agroforestry
Systems 72: 103–15.
Etter, A., C. McAlpine, K. Wilson, S. Phinn, and H. Possingham. 2006. “Regional Patterns of Agricultural
Land Use in Deforestation in Colombia.” Agriculture, Ecosystems and Environment 114 (2–4): 369–86.
FAO (Food and Agriculture Organization of the United Nations). 2014. “FAOSTAT.” Accessed June 1,
FAO/IIASA/ISRIC/ISS-CAS/JRC. 2012. “Harmonized World Soil Database Version 1.2.” Rome and
FEDEGAN (Federación Colombiana de Ganaderos). 2006. “Plan Estratégico de la Ganadería Colombiana
Fischer, G., M. Shah, H. van Velthuizen, and F. Nachtergaele. 2001. “Global Agro-ecological Assessment
for Agriculture in the 21st Century.” IIASA Research Report 02-02. Laxenburg, Austria: International
Institute for Applied Systems Analysis.
Gilhespy, S. L., Anthony, S., Cardenas, L., Chadwick, D., del Prado, A., Li, C., Yeluripati, J. B. 2014. “First 20
years of DNDC (DeNitrification DeComposition): Model evolution.” Ecological Modelling 292: 51–62.
Giltrap, D.L., Li, C.S., Saggar, S., 2010. DNDC: a process based model of greenhouse gas fluxes from
agricultural soils. Agric. Ecosyst. Environ. 136: 292–300.
Gockowski, J. & Sonwa, D. 2011. “Cocoa intensification scenarios and their predicted impact on CO2
emissions, biodiversity conservation, and rural livelihoods in the Guinea rain forest of West Africa.”
Environ Management 48:307–321.
Golub, A. A, Henderson, B. B., Hertel, T. W., Gerber, P. J., Rose, S. K., & Sohngen, B. 2013. “Global climate
policy impacts on livestock, land use, livelihoods, and food security.” Proceedings of the National
Academy of Sciences of the United States of America 110(52): 20894–9. doi:10.1073/pnas.1108772109
Grote, R, Lehmanna, E., Brümmer, C., Brüggemanna, N, Szarzynski, J. Kunstmanna, H, Karlsruhe, F., 2009.
“Modelling and observation of biosphere–atmosphere interactions in natural savannah in Burkina Faso,
West Africa.” Physics and Chemistry of the Earth 34: 251–260.
Havlík, P., Valin, H., Herrero, M., Obersteiner, M., Schmid, E., Rufino, M. C., Notenbaert, A. 2014.
“Climate change mitigation through livestock system transitions.” Proceedings of the National Academy
of Sciences of the United States of America 111(10): 3709–14. doi:10.1073/pnas.1308044111
Henson, I. E., R. Ruiz, and H. M. Romero. 2012. “The Greenhouse Gas Balance of the Oil Palm Industry in
Colombia: A Preliminary Analysis. I. Carbon Sequestration and Carbon Offsets.” Agronomia Colombiana
30 (3): 359–69.
Hergoualc’h, K., U. Skiba, J. M. Harmand, and C. Hénault. 2008. “Fluxes of Greenhouse Gases from
Andosols Under Coffee in Monoculture or Shaded by Inga Densiflora in Costa Rica.” Biogeochemistry 89
Hiederer, R., and M. Köchy. 2011. “Global Soil Organic Carbon Estimates and the Harmonized World Soil
Database”. European Union. Italy.
Hijmans, Robert J., et al. 2005. "Very high resolution interpolated climate surfaces for global land
areas." International journal of climatology 25(15): 1965-1978.
IAvH (Instituto de Investigación de Recursos Biológicos Alexander Von Humboldt), IDEAM (Instituto de
Hidrología, Meteorología y Estudios Ambientales), IIAP (Instituto de Investigaciones Ambientales del
Pacífico Jhon Von Newman), INVEMAR (Instituto de Investigaciones Marinas y Costeras), and SINCHI
(Instituto Amazónico de Investigaciones Científicas). 2007. “Informe del Estado del Medio Ambiente y de
los Recursos Naturales Renovables.” Bogotá: Instituto de Hidrología, Meteorología y Estudios
IDEAM (Instituto Nacional de Hidrología Meteorología y Estudios Ambientales de Colombia). 2010. “2nd
Comunicación Nacional ante la Convención Marco de las Naciones Unidas sobre Cambio Climático.”
IGAC (Instituto Geografico Agustin Codazzi). 2013. “Sistema de Informacion Geogrfica para la Planeacion
y el Ordenamiento Territorial.” Accessed November 18, 2013. http://sigotn.igac.gov.co/sigotn/.
IPCC (Intergovernmental Panel on Climate Change). 2006. “2006 IPCC Guidelines for National
Greenhouse Gas Inventories.” Prepared by the National Greenhouse Gas Inventories Programme,
Eggleston H.S., Buendia L., Miwa K., Ngara T. and Tanabe K. (eds). Hayama, Japan: Institute for Global
Jarvis, A., H. I. Reuter, A. Nelson, and E. Guevara. 2008. “Hole-Filled SRTM for the Globe Version 4.”
Accessed August 5, 2011. www.cgiar-csi.org/data/srtm-90m-digital-elevation-database-v4-1.
Li, C., S. Frolking, and T. A. Frolking. 1992. “A model of nitrous oxide evolution from soil driven by rainfall
events: 1. Model structure and sensitivity.” J. Geophys. Res. 97: 9759 – 9776.
Li, C. 2007. “Quantifying Greenhouse Gas Emissions from Soils: Scientific Basis and Modeling Approach.”
Soil Science and Plant Nutrition 53: 344–52.
Li, C., Salas, W., Zhang, R., Krauter, C., Rotz, A., & Mitloehner, F. 2012. “Manure-DNDC: a biogeochemical
process model for quantifying greenhouse gas and ammonia emissions from livestock manure systems.”
Nutrient Cycling in Agroecosystems 93(2): 163–200. doi:10.1007/s10705-012-9507-z
Li, M., A. De Pinto, J. Ulimwengo, L. You, and R. Robertson. 2014. “Modeling Land-use Allocation with
Mixed-level Data: An Econometric Analysis for the Democratic Republic of the Congo.” Environment and
Resource Economics. DOI 10.1007/s10640-014-9775-y.
Lubowski, R. N., A. J. Plantinga, and R. N. Stavins. 2006. “Land-Use Change and Carbon Sinks:
Econometric Estimation of the Carbon Sequestration Supply Function.” Journal of Environmental
Economics and Management 51 (2): 135–52.
Lubowski, R. N., & Rose, S. K. 2013. “The Potential for REDD+: Key Economic Modeling Insights and
Issues.” Review of Environmental Economics and Policy 7(1): 67–90. doi:10.1093/reep/res024
MADR (Ministry of Agriculture and Rural Development). 2011. “Statistical Data of Agriculture and Rural
Development 2001−2010.” Hanoi: Agricultural Publishing House.
MADS (Ministry of Environment and Sustainable Development). 2012. “Estrategia Colombiana de
Desarrollo Bajo en Carbono.” Bogotá.
Makundi, W. R., & Sathaye, J. A. 2004. “GHG Mitigation Potential and Cost in Tropical Forestry – Relative
Role for Agroforestry.” Environment, Development and Sustainability 6: 235-260.
McFadden, D. L. 1977. “Modeling the Choice of Residential Location.” New Haven, CT: Yale University
Cowles Foundation for Research in Economics.
Msangi, S. & M. Rosegrant. 2011. “Feeding the Future’s Changing Diets: Implications for Agriculture
Markets, Nutrition, and Policy.” 2020 Conference: Leveraging Agriculture for Improving Nutrition and
Health. New Delhi, India.
Nelson, A. 2000. “Accessibility, Transport and Travel Time Information.” CIAT Hillsides Project Report.
Cali, Colombia: International Center for Tropical Agriculture.
Nelson G., Hellerstein D. 1997. “Do roads cause deforestation? Using satellite images in econometric
analysis of land use.” American Journal of Agricultural Economics 79:80–88.
Nelson G.C., De Pinto A., Harris V., Stone S. 2004. “Land use and road improvements: a spatial
perspective.” International Regional Science Review 27:297–325.
Nelson, G. C., Valin, H., Sands, R. D., Havlik, P., Ahammad, H., Deryng, D., Elliott, J., Fujimori, S., Heyhoe,
E., Kyle, P., von Lampe, M., Lotze-Campen, H., Mason d’Croz, D., van Meijl, H., van der Mensbrugghe, D.,
Müller, C., Popp, A., Robertson, R., Robinson, S., Schmid, E., Schmitz, C., Tabeau, A., Willenbockel, D.
2014. “Climate change effects on agriculture: Economic responses to biophysical shocks.” Proceedings of
the National Academy of Sciences of the United States of America 111 (9): 3274-3279.
Nepstad, D. T. Bezerra, D. Tepper, K. McCann, C. Stickler, D. G. McGrath, et al. 2013. “Addressing
Agricultural Drivers of Deforestation in Colombia: Increasing Land-Based Production While Reducing
Deforestation, Forest Degradation, Greenhouse Gas Emissions and Rural Poverty.” Report to the United
Kingdom, Foreign and Commonwealth Office and Department of Energy Climate Change, Forests and
Climate Change Programme. Earth Innovation Institute.
Niels, H. B. 2010. “IPCC Default Soil Classes Derived from the Harmonized World Soil Database. Version
1.1.” Retrieved from: http://www.isric.org/data/ipcc-default-soil-classes-derived-harmonized-world-soil-
data-base-ver-11. Accessed May 27, 2014.
Palm, C. A., J. C. Alegre, L. Arevalo, P. K. Mutuo, A. R. Mosier, and R. Coe. 2002. “Nitrous Oxide and
Methane Fluxes in Six Different Land Use Systems in the Peruvian Amazon.” Global Biogeochemical
Cycles 16 (4): 21-1–21-13.
Pathak, H., Li, C., Wassmann, R., 2005. Greenhouse gas emissions from Indian rice fields: calibration and
upscaling the DNDC model. Biogeosciences 2: 113–123.
Rosegrant, M.W., Zhu, T., Msangi, S., Sulster, T. 2008. “Global Scenarios for Biofuels: Impacts and
Implications.” Applied Economic Perspectives and Policy 30 (3): 495-505.
Rosegrant, M. W., and the IMPACT Development Team. 2012. “International Model for Policy Analysis of
Agricultural Commodities and Trade (IMPACT): Model Description.” Washington, DC: International Food
Policy Research Institute. Accessed July 2013. www.ifpri.org/ourwork/program/impact-model.
Rutten, M., van Dijk, M., van Rooij, W., & Hilderink, H. 2014. “Land Use Dynamics, Climate Change, and
Food Security in Vietnam: A Global-to-local Modeling Approach.” World Development 59: 29–46.
Saatchi, S. S., N. L. Harris, S. Brown, M. Lefsky, E. T. A. Mitchard, W. Salas, et al. 2011. “Benchmark Map
of Forest Carbon Stocks in Tropical Regions Across Three Continents.” Proceedings of the National
Academy of Sciences of the United States of America 108 (24): 9899–904.
Sacks, W. J., D. Deryng, J. A. Foley, and N. Ramankutty. 2010. “Crop Planting Dates: An Analysis of Global
Patterns.” Global Ecology and Biogeography 19: 607–20.
Schils, R. L. M., Olesen, J. E., del Prado, a., & Soussana, J. F. 2007. “A review of farm level modelling
approaches for mitigating greenhouse gas emissions from ruminant livestock systems.” Livestock
Science, 112(3), 240–251.
Schneider, U. A., & Smith, P. 2008. “Energy intensities and greenhouse gas emission mitigation in global
agriculture.” Energy Efficiency 2(2): 195–206. doi:10.1007/s12053-008-9035-5
SINAP (Sistema Nacional de Áreas Protegidas). 2012. “GISDATA: Parques Nacionales Naturales de
Colombia, Reservas Naturales de la Sociedad Civil, Areas protegidas de las categorías regionales, y
Reservas forestales protectoras nacionales. Versión 2.” Data provided by Grupo de Sistemas de
Información y Radiocomunicaciones de Parques Nacionales Naturales de Colombia.
Smith, P., Martino, D., Cai, Z., Gwary, D., Janzen, H., Kumar, P., Rice, C. 2007. “Policy and technological
constraints to implementation of greenhouse gas mitigation options in agriculture.” Agriculture,
Ecosystems & Environment 118(1-4): 6–28.
Smith, P., Haberl, H., Popp, A., Erb, K.-H., Lauk, C., Harper, R., Rose, S. 2013. “How much land-based
greenhouse gas mitigation can be achieved without compromising food security and environmental
goals?” Global Change Biology 19(8): 2285–302. doi:10.1111/gcb.12160
Torres, A. B., Marchant, R., Lovett, J. C., Smart, J. C. R., & Tipper, R. 2010. “Analysis of the carbon
sequestration costs of afforestation and reforestation agroforestry practices and the use of cost curves
to evaluate their potential for implementation of climate change mitigation.” Ecological Economics,
Tschakert, P. 2007. “Environmental services and poverty reduction: Options for smallholders in the
Sahel.” Agricultural Systems 94(1): 75-86.
Tenningkeit, T., Kahrl, F., Wolcke, J., & Newcombe, K. 2012. “Economics of Agricultural Carbon
Sequestration in Sub-Saharan Africa.” In Ev. Wollenberg, A. Nihart, M.-L. Tapio-Bistrom, & M. Grieg-Gran
(Eds.), Climate Change Mitigation and Agriculture (pp. 144-158). New York and London: Earthscan.
Train, K. 2003. “Discrete Choice Methods with Simulation.” New York: Cambridge University Press.
Uchida, H. and Nelson, A. Agglomeration Index: Towards a New Measure of Urban Concentration. 2010.
WIDER Working Paper 2010/29. UNU-WIDER.
Werner, C., Butterbach-Bahl, K., Haas, E., Hickler, T., & Kiese, R. 2007. “A global inventory of N 2 O
emissions from tropical rainforest soils using a detailed biogeochemical model.” Global Biogeochemical
Cycles 21(3) GB3010.
9 Appendix 1
Conceptual Framework of the Land Use Model
Consider how a decision-maker allocates land use in a municipal,
indexed by n; n = 1, …, N. Suppose the
decision-maker is a risk-neutral landowner; s/he chooses land uses to maximize the present discounted
value of the stream of the expected net benefits from the land. The land grid could be allocated to K
alternative major uses, indexed by k; k = 1, …, K. Among these uses, perennial cropland and annual
cropland are two nests of crop, within which the decision maker is allowed to select different crops,
indexed by j; j = 1, …, JK. Our data for Colombia includes four major perennial crops (cacao, coffee, oil
palm, plantain) and five major annual crops (maize, potato, rice, sugarcane, cassava); the remaining
crops are lumped into other perennial crops and other annual crops, respectively. Under these
simplifying assumptions, the steady-state decision rule that emerges from the related dynamic
optimization problem is to put a land parcel to the use generating the greatest present discounted value
of net benefit (Lubowski et al. 2006). That is, allocate a parcel of land to use j if
, and if , (1)
where is the one-period expected net benefit from allocating land parcel to use j ( ). The
potential value of depends on attributes of the parcel, such as land quality, weather conditions,
locational characteristics, and economic conditions in the surrounding area, as well as attributes
associated with alternative choices, such as price and yield.
A standard practice in the land use modeling literature is to decompose into a deterministic
and a random error term .
represents the expected average net benefit from
allocating land use in a grid; represents the deviation from the average net benefit and is often
assumed to follow a normal, logit, or type-I extreme value distribution. We further decompose
two deterministic components:
depends on variables that describe the nest k; it differs over nests but not over alternatives within
each nest. depends on variables that describe nested use j ( ); it varies over choices within the
Under the assumption analogous to the nested logit model about , for example, that is
correlated within nest k but uncorrelated across nests, the probability of grid n allocated to alternative j
( ) can be derived as a product of two multinomial logit probabilities (McFadden 1977, Train 2003):
We choose municipal as the unit of analysis in order to keep the units of upper- and lower-level models
consistent. We apply the coefficient of variation of elevation to control potential heterogeneity within a municipal.
where the parameter is a measure of the degree of independence in among the alternatives in
each nest and
, often called inclusive value of nest k. A higher value of
implies less correlation and indicates complete independence.
Equation (3) defines marginal probability of choosing any alternative in nest k and equation (4)
defines conditional probability of choosing alternative j given that any alternative in nest k is chosen. We
refer to the marginal probability as an upper-level model and to the conditional probability as a lower-
level model, reflecting their relative positions in the hierarchy structure. In equation (4), is treated as
a scale parameter that scales coefficient parameters implicitly defined in . For those nests where
there are no alternative choices inside, the conditional probability given in (4) equals 1 and the inclusive
value is reduced to
. Insert into (3) and will be canceled. Probabilities (3) and (4) are
fundamental equations in the nested logit model.
Land Use Model Specification
The average expected net benefits from allocating every grid n to nest k, , is specified as
, where is a vector of location-specific variables describing population density, market
accessibility, conservation practice, topography, soil pH level, and weather conditions; is vector of
coefficients on . Within any nest k, the average expected net benefits from allocating each location n
to alternative j, , is specified as , where is a vector of crop-specific variables,
including the crop prices, land suitability, topography, soil pH level, and weather conditions, and an
inertia variable (a lagged crop share) that captures land use conversion costs; is a vector of
coefficient parameters specific to crop.
Note that if the choice-specific variables perfectly captured the average expected net revenues for
each crop at the level of the individual grid, then should simply reflect the marginal net benefit and
would not be expected to differ over crops. We allow this parameter varying across crops in our
specification because both crop-specific variables are originally measured at relatively coarse resolutions
and hence cannot perfectly capture the average expected net benefits for each crop at each location. To
eliminate from the lower-level model, we introduce by dividing by such that
This scaling facilitates the estimation.
We estimate the model in a “bottom-up” sequential fashion, starting from the estimation of the
lower-level model and using the estimated coefficients to calculate the inclusive values, which enter the
upper-level model as explanatory variables. This approach exploits the fact that the choice probabilities
can be decomposed into marginal and conditional probabilities that are logit functions.
1. The growth rate of crop area between the base year and the projected year is consistent with
that generated from IMPACT baseline projection.
2. The specification of the nested land use model is correct; the parameter estimates of the land
use model are consistent.
Allocating Projected Crop Area Based on the Nested Land Use Model
Let n be municipal index, j be crop index, and t be time period index. At the base year t = 0, share of crop
j can be expressed as
where is the estimated crop probability from land use model and is the error between the
agricultural census (H) and the land use model estimation (P). Rearranging equation (9), we have
Hence, for any n, the adjusted share of crop j is
Let’s now turn to the projected year t = 1. Let gj be the growth rate of crop area between time 0 and
1. Then the area of crop j at time 1 can be extrapolated as
Rearranging equation (13), we have
For any municipal, the adjusted share of crop j at time 1 is
Therefore, changes in crop share at municipal can be derived as
Land Use Model Performance Assessment
Error! Reference source not found. and Error! Reference source not found. present an assessment
of the predictive power of the upper and lower levels of the land use model, respectively. In each table,
columns 1 and 2 correspond to the observed land use derived from the data sample; columns 3 and 4
correspond to the predicted land use estimated from the land use model; the last two columns present
the discrepancies by subtracting entries in columns 1 and 2 from entries in columns 3 and 4. We find
that the model performs well in predicting most land uses. As far as the upper level is concerned, the
model provides generally accurate in-sample predictions. The highest and lowest discrepancies,
expressed in percentage terms, are -0.2 percent for perennial cropland and -2.3 percent for forests.
Regarding the lower-level component, the model tends to overestimate perennial and annual croplands
by a total of 0.2 percent and 0.4 percent, respectively. These discrepancies arise mainly from
overestimating areas allocated to oil palm, cacao, maize, and rice. Even so, the predicted errors of these
four crops in terms of percentage are in a reasonable range, from 2.3 percent to 3.2 percent. The model
performs well in the prediction of the remaining crops in terms of both magnitude and percentage.