Brazil Low Carbon Case study [PDF]

Public Disclosure Authorized

69869

2011

Public Disclosure Authorized

Brazil Low Carbon Case Study Technical Synthesis Report

Public Disclosure Authorized

Public Disclosure Authorized

Land Use, Land-Use Change, and Forestry Coordination Christophe de Gouvello, The World Bank Britaldo S. Soares Filho, CSR-UFMG André Nassar, ICONE

WORLD BANK

Authors Britaldo S. Soares Filho and Letícia Hissa, UFMG André Nassar, Leila Harfuch, Marcelo Melo Ramalho Moreira, Luciane Chiodi Bachion and Laura Barcellos Antoniazzi, ICONE Luis G. Barioni, Geraldo Martha Junior, Roberto D. Sainz, Bruno J. R. Alves, and Magda A. de Lima, EMBRAPA Osvaldo Martins, Magno Castelo Branco, and Renato Toledo, Iniciativa Verde Manoel Regis Lima Verde Leal, CENEA Fábio Marques, Rodrigo Ferreira, Luiz Goulart, and Thiago Mendes, PLANTAR Christophe de Gouvello, Adriana Moreira, Barbara Farinelli, Jennifer Meihuy Chang, and Rogerio Pinto, The World Bank Júlio Hato and Sérgio Pacca, USP Saulo Ribeiro Freitas, Karla Maria Longo and Ricardo Almeida de Siqueira (National Institute for Space Research, INPE)

2011 Brazil Low Carbon Case Study Technical Synthesis Report

Land Use, Land-Use Change, and Forestry Coordination Christophe de Gouvello, The World Bank Britaldo S. Soares Filho, CSR-UFMG André Nassar, ICONE Authors Britaldo S. Soares Filho and Letícia Hissa, UFMG André Nassar, Leila Harfuch, Marcelo Melo Ramalho Moreira, Luciane Chiodi Bachion and Laura Barcellos Antoniazzi, ICONE Luis G. Barioni, Geraldo Martha Junior, Roberto D. Sainz, Bruno J. R. Alves, and Magda A. de Lima, EMBRAPA Osvaldo Martins, Magno Castelo Branco, and Renato Toledo, Iniciativa Verde Manoel Regis Lima Verde Leal, CENEA Fábio Marques, Rodrigo Ferreira, Luiz Goulart, and Thiago Mendes, PLANTAR Christophe de Gouvello, Adriana Moreira, Barbara Farinelli, Jennifer Meihuy Chang, and Rogerio Pinto, The World Bank Júlio Hato and Sérgio Pacca, USP Saulo Ribeiro Freitas, Karla Maria Longo and Ricardo Almeida de Siqueira (National Institute for Space Research, INPE)

© 2011 The International Bank for Reconstruction and Development / The World Bank Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Email: [email protected] All rights reserved

5

This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent.

Rights and Permissions

The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street, NW, Washington, DC, 20433, USA; fax: 202-522-2422; email: [email protected]. The Energy Sector Management Assistance Program (ESMAP) is a global knowledge and technical assistance program administered by the World Bank that assists low- and middle-income countries to increase know how and institutional capacity to achieve environmentally sustainable energy solutions for poverty reduction and economic growth. For more information on the Low Carbon Growth Country Studies Program or about ESMAP’s climate change work, please visit us at www.esmap.org or write to us at:

Energy Sector Management Assistance Program The World Bank 1818 H Street, NW Washington, DC 20433 USA email: [email protected] web: www.esmap.org

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The World Bank does not guarantee the accuracy of the data included in this work and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.

Contents

6

Tables

------------------------------------------------------------------------------- 9

Figures

----------------------------------------------------------------------------- 12

Maps

----------------------------------------------------------------------------- 15

Acronyms and Abbreviations----------------------------------------------------------------------------- 17

1.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

2 .

Acknowlegments

----------------------------------------------------------------------------- 23

Executive Summary

----------------------------------------------------------------------------- 24

Study Overview

----------------------------------------------------------------------------- 25

Introduction

----------------------------------------------------------------------------- 34

1.1 Context of the Low-carbon Study------------------------------------------------------------------------- 36 1.2 Approach of the LULUCF Summary Report------------------------------------------------------------- 37 Reference Scenario

----------------------------------------------------------------------------- 38

2.1 Emissions from Land Use, Land-use Change, Deforestation, Agriculture and Livestock ------- 38 2.1.1 Effects of Land Use and Land-use Change on Emissions--------------------------------------- 38 2.1.1.1 Deforestation

------------------------------------------------------------------------------- 38

2.1.1.2 Agricultural Production--------------------------------------------------------------------------- 39 2.1.1.3 Livestock Activities ------------------------------------------------------------------------------- 39

2.1.1.4 Forestry-based Carbon Uptake------------------------------------------------------------------ 40

2.1.2 Land Use and Land-use Change Simulation Methodology----------------------------------------- 40

2.1.2.1 Area Available for the Expansion of Productive Activities----------------------------------- 40 2.1.2.2 Economic Land-use, Agriculture and Livestock Modeling: BLUM Model----------------- 43

2.1.2.3 Allocation of Area for Agriculture and Livestock Activities---------------------------------- 49

2.1.3 Land-use Reference Scenario--------------------------------------------------------------------------- 53 2.1.3.1 Division into Geographic Micro-regions-------------------------------------------------------- 63

2.1.3.2 Spatialization of Land-use Change and Deforestation: SIMBRASIL Model--------------- 64

2.1.4 Calculation of Emissions Associated with Land use, Land-use

Change and Deforestation in the Reference Scenario----------------------------------------------------- 68

2.1.4.1 Emissions from Livestock----------------------------------------------------------------------------- 69 2.1.4.1.1 Methodology

------------------------------------------------------------------------------- 70

2.1.4.1.2 Reference Scenario Results--------------------------------------------------------------------- 79

2.1.4.2 Agricultural Emissions ------------------------------------------------------------------------------- 81 2.1.4.2.1 Evaluation of CO2 Emissions from Changes in Soil C Stocks------------------------------ 81

2.1.4.2.2 Greenhouse Gas Production from the Use of Fossil Energy------------------------------- 90

2.1.4.2.3 Synthesis of Emissions from Agricultural Activities--------------------------------------- 91

2.1.5 Emissions from Deforestation------------------------------------------------------------------------- 93 2.2 Carbon Uptake Through Reforestation----------------------------------------------------------------- 98

2.2.1 Methodology

------------------------------------------------------------------------------- 99

2.2.1.1 Details of the Potential Biomass Model--------------------------------------------------------- 99

2.2.1.2 Carbon Uptake Potential through the Restoration of the Legal Reserves---------------- -113 2.2.1.3 Carbon Uptake Potential through the Restoration of Riverside Forests----------------- -114

2.2.1.4 Carbon Uptake Potential through Energy Forest

Plantations in the Cerrado and Atlantic Forest Biomes-------------------------------------------------- 115 2.2.2 Reference Scenario for Forest Restoration------------------------------------------------------ -116

3.

2.2.3 Non-renewable Charcoal and Planted Forests for Renewable Charcoal------------------- -117

7

2.3 Reference-Scenario Emissions Results---------------------------------------------------------------- -120

Mitigation and Carbon Uptake Options------------------------------------------------------------------122 3.1 Mitigation Options in Agriculture: Zero tillage------------------------------------------------------- -122 3.1.1 Emissions Reduction Potential Associated with Zero Tillage-------------------------------- -125 3.1.2 Obstacles Limiting the Expansion of Zero Tillage---------------------------------------------- -127

3.2 Carbon Uptake through the Increase of Planted Forests for Renewable Charcoal-------------- 129 3.2.1 Carbon Uptake Potential Associated with the Increase in Renewable Charcoal Production

------------------------------------------------------------------------------ -129

3.2.2 Obstacles to the Expansion of Production Forests for Renewable Charcoal--------------- -134 3.2.3 Measures for Overcoming Obstacles------------------------------------------------------------- -137

3.3 Carbon Uptake through Native Forest Recovery------------------------------------------------------ -141 3.3.1 Carbon Uptake Potential Resulting from a“Legal Scenario” for Forest Restoration------ 141 3.3.2 Obstacles to Forest Restoration and Ways to Overcome Them------------------------------- -144

3.3.3 Reforestation Support Policies-------------------------------------------------------------------- -146

3.4 Mitigation Options for Livestock Activities------------------------------------------------------------ -152 3.4.1 Main Options Considered for Mitigating Emissions from Livestock------------------------ -152

4.

3.4.2 Obstacles and Proposals for Overcoming Them------------------------------------------------ -155

3.5 Reduction of Emissions from Deforestation---------------------------------------------------------- -156

Low-Carbon Land-Use Scenario in Brazil---------------------------------------------------------------159

4.1 Additional Needs for Land for Carbon Uptake Activities and Biofuel Export-------------------- -159

4.2 Toward a New Pattern of Productivity for the Livestock Industry--------------------------------- -160

4.3 Mitigation Potential of Direct Emissions from Livestock in the Low-carbon Scenario--------- 164

4.4 A New Land-use Scenario for the Low-carbon Scenario--------------------------------------------- -169 4.5 Reduction of Deforestation in the Low-carbon Scenario-------------------------------------------- -176

4.6 Additional Measures for Protecting the Forest from Deforestation------------------------------- -179

4.7 Balance of emissions from land use and land-use change in the Low-carbon Scenario-------- -186

4.8 Key Uncertainties for Emissions Estimates----------------------------------------------------------- -191 4.9 Benefits Related to Reducing Aerosol Emissions Resulting from Deforestation by Burning- -193 4.9.1 Methodology: Numerical Modeling with CCATT-BRAMS------------------------------------- -194 4.9.1.1. Calculation of Aerosol Emissions-------------------------------------------------------------- -196

4.9.1.2 Aerosol Emissions in the Reference and Low-carbon Scenarios-------------------------- -199

4.9.2 Results

4.9.2.1 Precipitation

4.9.9.2 Temperature

------------------------------------------------------------------------------ 202

------------------------------------------------------------------------------ -202

------------------------------------------------------------------------------ -206

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

3.1.3 Proposals for Overcoming Obstacles------------------------------------------------------------- -128

4.9.3 Summary of the Reduction of Impacts on Rainfall and Temperature

Regimes in the Low-carbon Scenario----------------------------------------------------------------------- -208

5. Analysis of Transition Costs from the Reference Scenario to the Low-carbon Scenario----211 5.1 Costs of Reducing Emissions from Deforestation---------------------------------------------------- -218 5.1.1 Improving Livestock Productivity---------------------------------------------------------------- -218

8

5.1.2 Forest Protection

------------------------------------------------------------------------------ -220

5.2 Forest Recovery: Legal Forest Reserves---------------------------------------------------------------- -226 6.

5.3 Renewable Charcoal

------------------------------------------------------------------------------ 229

5.4 Emissions Abatement with Zero Tillage--------------------------------------------------------------- -233 Conclusion

------------------------------------------------------------------------------241

7. Annex A: Analysis of Low-carbon Scenarios------------------------------------------------------------244

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

7. 1 Herd Optimization Scenario----------------------------------------------------------------------------- -247

8.

7.2 Production Forest Scenario ------------------------------------------------------------------------------ 250 7.3 Ethanol Scenario and Production Forests------------------------------------------------------------- -251 7.4 Legal Scenario (Reforestation of the Legal Reserve)------------------------------------------------- -253

7.5 Aggregate Scenario: Herd, Production Forests, Ethanol, Forest Restoration-------------------- -256 References

------------------------------------------------------------------------------264

Tables Table 1: Summary of additional land needs in the reference and Low-carbon Scenarios.-----------------30 Table 2: Comparison of emissions distribution among sectors in the reference and

Low-carbon Scenarios, 2008-30------------------------------------------------------------------------------------31

9

Table 3: Comparison between total pasture area and area of residual vegetation convertible into farmland/forests in the regions of the BLUM Model (1000 ha)----------------------------------------------------43 Table 4: Brazil - Area allocated and production of products covered by the BLUM Model-----------------46

Table 5: Data sources--------------------------------------------------------------------------------------------------47 Table 6: Macroeconomic projections-------------------------------------------------------------------------------49 Table 8: Projection of areas occupied by production forests----------------------------------------------------55 Table 9: Productive land use (crops, pasture and forests) in the different regions of Brazil-------------- 55

Table 10: Land use (1000 ha) in the six regions of the model for the Reference Scenario------------------56 Table 11: Dairy cattle herd - Reference Scenario-----------------------------------------------------------------57

Table 12: Land use for Brazil - Reference Scenario---------------------------------------------------------------59 Table 13: Description of the base developed for the implementation of SIMBRASIL-----------------------65

Table 14: Categories of animals considered in the analysis of livestock emissions-------------------------72 Table 15: Zootechnical coefficients considered for each productive system---------------------------------77 Table 16: Greenhouse gas emissions per animal and per carcass equivalent in kg in different

production systems----------------------------------------------------------------------------------------------------78

Table 17: Estimates of area, herd, proportion of the herd in productive systems and emissions for the Reference Scenario----------------------------------------------------------------------------------------------------80 Table 18: Areas under different uses and total area in 1990, by state-----------------------------------------82 Table 19: Soil C stock under native vegetation for each region of the Blum model--------------------------86

Table 20: Change factors for soil C Stock for each type of land use---------------------------------------------88

Table 21: CO2, CH4, and N2O accumulated emissions from emissions from agriculture from 20102030, expressed in CO2eq for the reference scenario----------------------------------------------------------- 91 Table 22: Land use in Brazil between 1990 and 2005------------------------------------------------------------94

Table 23Risk of extinction of arboreal forest species in Brazil in 2000----------------------------------------95

Table 24: Points of the different IBGE fertility classes---------------------------------------------------------- 104

Table 25: Entries in the environmental database.-------------------------------------------------------------- 107

Table 26: Projection of CO2 uptake in the Reference Scenario (use of coal and/or non-renewable/renewable charcoal) - 2010 to 2030--------------------------------------------------------------------------------- 120

Table 27: Methane emissions from convertional planting and zero tillage in irigated rice areas in and a different location, comparison of emissions reductions between the two uses---------------------------123 Table 28: Cumulative costs and revenue in the reference and Low-carbon Scenarios with the adoption of zero tillage from 2010 to 2030---------------------------------------------------------------------------------- 125

Table 29: Greenhouse gases produced in the Low-carbon Scenario: adoption of zero tillage in 100% of the agricultural area from 2015 to 2030------------------------------------------------------------------------- 125

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 7: Land competition matrix in Brazil------------------------------------------------------------------------50

Table 30: CO2 uptake from forest plantations for renewable charcoal scenario 1------------------------ 132

Table 31: CO2 uptake from forest plantations for renewable charcoal scenario 2------------------------ 133 Table 32: Measures proposed to surmount obstacles----------------------------------------------------------138

10

Table 33: Area needed for reforestation under Brazil’s Legal Reserve Law, by state----------------------143

Table 34: Average productivity of selected crops in different countries, 2008---------------------------- 157

Table 35: Mitigation and carbon uptake options for a Low-carbon Scenario and associated needs for additional land--------------------------------------------------------------------------------------------------------160 Table 36: Comparison of land-use results for the reference and Low-carbon Scenarios

(millions of ha)-------------------------------------------------------------------------------------------------------- 170 Table 37: Snapshot of protected areas in the Amazon Biome and ARPA participation-------------------180

Table 38: Resources of the INPE for monitoring the Amazon by satellite----------------------------------- 182

Table 39: Implementation of the Public Forest Management Systems: Benefits and Losses------------ 184 Table 40: Summary of expenditure anticipated for Public Forest Management services in 2009------185

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 41: Comparison of cumulative emissions distribution among sectors in the reference and Low-carbon Scenarios, 2010-30.---------------------------------------------------------------------------------------189

Table 42: Emissions factors (g/kg) for different biomes for CO2 and aerosols (particulate matter with a diameter less than 2.5 micrometers - PM2.5)------------------------------------------------------------------- 197

Table 43: Total annual aerosol emissions (tons per hectare and per year) throughout the country for the reference (REF) and low carbon (LC) scenarios. Also shown are the figures of absolute differences (LC-REF) and differences in percentages (LC-REF (%)) between emissions for the two scenarios--- 200 Table 44: Mitigation potential and marginal abatement cost of various alternatives, based on three discount rates------------------------------------------------------------------------------------------------------------- 214

Table 45: Comparison of sector benchmark IRRs and break-even carbon prices for various mitigation options----------------------------------------------------------------------------------------------------------------- 216 Table 46: Volume of incentive required (undiscounted) in order to achieve the emissions reductions considered in the Low-carbon Scenario from 2010 to 2030-------------------------------------------------- 217

Table 47: Investments and expenditures for prototypical livestock systems (2009-30)---------------- 219 Table 48: Economic and financial performance of prototypical livestock systems (2009-2030)------ 219 Table 49: Investment and expenses in the reference and Low-carbon scenarios------------------------- 220

Table 50: Comparable economic and financial performance in the livestock sector.--------------------- 220 Table 51: Projection of costs for forest protection in areas where deforestation is illegal

(in millions of US$).-------------------------------------------------------------------------------------------------- 222

Table 52: Livestock-sector investments and expenses to release land to absorb additional lands needed in the reference and Low-carbon Scenarios (2010-30)--------------------------------------------------------226 Table 53: Marginal abatement cost-------------------------------------------------------------------------------- 230 Table 54: Summary of economic parameters for the 2010-2030 period----------------------------------- 231

Table 55: Investments in the additional use of renewable charcoal compared with total mitigation measures in the Brazilian industrial sector, considering the adjusted potential-------------------------------231 Table 56: Hypotheses of the technical-economic analysis---------------------------------------------------- 232

Table 57: Discrimination of costs considered in the study---------------------------------------------------- 239

Table 58: Emissions reduction potential in tons of CO2eq, average abatement cost during the period and price to be paid per ton of C to compensate the implementation of zero tillage---------------------- 240

Table 59: Relationship of the Low-carbon Scenarios developed for this study.---------------------------- 245

Table 60: Area necessary for the reforestation of the legal reserve by state in Brazil (hectares)------- 247

Table 61: Balance of supply and demand for selected products, herd optimization scenario----------- 248 Table 62: Land use in Brazil, herd optimization scenario (1000 ha)-----------------------------------------249

Table 63: Regional allocation of pasture areas, reference and herd optimization scenario (thousand hectares).--------------------------------------------------------------------------------------------------------------249

11

Table 64: Regional distribution of cattle herd, reference scenario and herd optimization scenario

(100 head)-------------------------------------------------------------------------------------------------------------250

Table 65: Regional distribution of the production forest in the reference and production forest scenarios (thousand hectares)--------------------------------------------------------------------------------------------251 Table 67: Regional sugar cane distribution in the reference scenario, the herd optimization scenario and the ethanol scenario (in thousand hectares)---------------------------------------------------------------253 Table 68: Reforestation needs in order to comply with the Legal Reserve in the regions of the model (1000 ha)--------------------------------------------------------------------------------------------------------------254

Table 69: Pasture area in the regions of the model in 2009 and 2030 (in 1000 ha), in the Reforestation Scenario of the LR.---------------------------------------------------------------------------------------------------- 254 Table 70: Presentation of quantitative results by state for the Atlantic Forest and Cerrado-------------256

Table 71: Comparison of land use results in all scenarios for Brazil.-----------------------------------------258

Table 72: Comparison of results for pasture area in all scenarios for Brazil and regions.----------------259

Table 73: Results from the cattle herd in the Reference Scenarios and herd optimization scenarios and aggregate (1000 head).----------------------------------------------------------------------------------------------260 Table 74: Results for land use abd herd for selected products inhte aggregates scenario---------------261

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 66: Land use in Brazil, ethanol scenario (in 1000 hectares)-------------------------------------------252

Figures

12

Figure 1: GHG mitigation wedges in the Low-carbon Scenario, 2008-30.------------------------------------32

Figure 2: Calculation of available area for the expansion of productive activities.--------------------------41 Figure 3: Land use by class, excluding the Pampa, Caatinga and Pantanal biomes.-------------------------42 Figure 4: Methodological land-use diagram-----------------------------------------------------------------------53

Figure 5: Evolution of the demand for land in Brazil by crop in the Reference Scenario - 2006-30 (million ha).-------------------------------------------------------------------------------------------------------------59

Figure 6: Architecture of the LULUCF Study, with an emphasis on the components that include the deforestation factor---------------------------------------------------------------------------------------------------------64 Figure 7: Example of the database prepared for simulations of land-use change and cover.--------------65

Figure 8: First part of the spatially explicit model for land-use change and soil cover - land allocation.-67 Figure 9: Spatially explicit land-use change and soil cover model - simulation of land-use change.-----68

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 10: Information flow in the analytical model -------------------------------------------------------------71

Figure 11: Variation i the pasture area occupied by type of productive system

in the reference scenario----------------------------------------------------------------------------------------------80

Figure 12: Area of the country occupied by agriculture, pasture, planted forests and complementary area in the form of native vegetation and other uses, from 1990-2030.---------------------------------------83 Figure 13: Fictitious land-use change scheme for three crops (A, B, and C).----------------------------------85

Figure 14: CO2, N2O and CH4 emissions from agriculture during the 2010-2030 period, expressed in CO2 equivalents in the Reference Scenario.------------------------------------------------------------------------92 Figure 15: Deforestation dynamic in the three main biomes in Brazil in the reference scenario (km2/ year)----------------------------------------------------------------------------------------------------------------------96 Figure 16: Emissions from land use in the Reference Scenario.------------------------------------------------98

Figure 17: Diagram of potential carbon removals by reforestation for the Cerrado and Atlantic Forest biomes.----------------------------------------------------------------------------------------------------------------- 100 Figure 18: Points attributed to the WCMI values in the model. Modified by Iverson et al. (1994).----- 102 Figure 19: Points attributed to the amount of rainfall in the model,

according to Iverson et al. (1994)--------------------------------------------------------------------------------- 102 Figure 20: Points attributed to altitude classes in the model, modified by Iverson et al. (1994).------- 103 Figure 21: Points Attributed to the degree of incline of the land, modified by Iverson et al. (1994).--- 104

Figure 22: Graph of the box-plot where the distribution of the amounts for altitude, rainfall and months of hydrous deficit may be observed for the Embrapa and WorldClim databases.------------------------- 106

Figure 23: Logistical function of biomass uptake using local biomass potential and age of vegetation as parameters.----------------------------------------------------------------------------------------------------------- 115

Figure 24: Reference Scenario for charcoal with a low level of legal restrictions; participation of thermo-reduction agents in the Brazilian iron and steel market. ------------------------------------------------- 118

Figure 25: Reference Scenario for charcoal with a high level of legal restrictions: participation of thermo-reduction agents in the Brazilian iron and steel producing market ------------------------------------ 119 Figure 26: Projection of CO2 Emissions for the Reference Scenario (charcoal)--------------------------- 120 Figure 27: Reference Scenario results, emissions from land use and land-use change, 2009-30.------ 121 Figure 28: Percentage of reduction of soil and water losses from zero tillage (ZT) compared to conven-

tional planting (CP).-------------------------------------------------------------------------------------------------- 124 Figure 29: CO2e stock from forest plantations for renewable charcoal in Scenario 1-------------------- 132 Figure 30: CO2e stock in forest plantations for renewable charcoal in Scenario 2.----------------------- 133

Figure 31: Comparison of CO2e stock in Scenarios 1 and 2 and the Reference Scenario.----------------- 134 Figure 32: Carbon uptake potential of forest recovery activities and production forests.--------------- 143

13

Figure 33: Mitigating measures for the construction of the Low-carbon Scenario------------------------ 161 Figure 34: Change in pasture area occupied according to the type of productive system

-(million hectares)---------------------------------------------------------------------------------------------------- 162 Figure 35: Variation in number of head of cattle in productive systems, 2009-30------------------------ 163

Figure 36: Projection of Brazilian herd productivity between 2009 and 2030 for the reference and Low-carbon Scenarios.--------------------------------------------------------------------------------------------------- 164 Figure 39: Comparison of methane emissions per unit of meat (kg CO2e per kg), 2008–30.----------- 167 Figure 40: Evolution of Brazil’s demand for land, by crop, 2006-30 (millions of ha).--------------------- 171 Figure 41: Evolution of deforestation in the Low-carbon Scenario (curve) (km2 per year).------------ 178

Figure 42: Evolution of deforestation in the Low-carbon Scenario (LCS) and Reference Scenarios (RS) (thousands of ha per year).----------------------------------------------------------------------------------------- 178

Figure 43: Identification of forest degradation patterns in the Amazon within the framework of the DEGRAD program. Source: INPE, 2009.------------------------------------------------------------------------------ 182

Figure 44: Reference Scenario results: emissions from land use and land-use change, 2009–3.------- 187

Figure 45: Emissions from land use and land-use change under the new land-use dynamic in the Low-carbon Scenario.----------------------------------------------------------------------------------------------------- 188 Figure 46: Comparisons of gross emissions distribution among sectors in the reference and Low-carbon Scenarios, 2008–30.-------------------------------------------------------------------------------------------- 190

Figure 47: Transport processes simulated by the CCATT-BRAMS, including plume rise, deep and shallow convective transport by cumulus, diffusion in the PBL, dry and wet deposition.--------------- 195

Figure 48: Estimate of total annual aerosol emissions in Brazil for the reference and Low-carbon Scenarios (Table 43).-------------------------------------------------------------------------------------------------------- 199

Figure 49: Average monthly precipitation in the 6 regions analyzed in the reference and Low-carbon Scenarios from 2007 to 2008 compared to data obtained from the Agência Nacional de Águas (National Water Agency - ANA), which corresponds to monthly precipitation during the period from 1982 to 2005. The margins of error represent the standard deviation for each month.--------------------------- 203

Figure 50: Average monthly precipitation in the 6 regions analyzed in the reference and Low-carbon Scenarios from 2007 to 2030 (bar graph left axis). Also shown is the difference between the reference and Low-carbon Scenarios (line graph right axis).------------------------------------------------------------- 204 Figure 51: Difference in precipitation (mm) between the reference and Low-carbon Scenarios considering the years 2007 to 2030 during the February, March and April (A), May, April and June (B), August, September and October (C), and November, December and January (D) trimesters.-------------------- 205

Figure 52: Average monthly temperature in the 6 regions analyzed in the reference and Low-carbon Scenarios in 2007 and 2008 compared to data obtained from the National Meteorology Institute (Instituto Nacional de Meteorologia - INMET), which corresponds to the monthly climatology of temperature during the period from 1977 to 2000. The margins of error represent the standard deviation for each month.------------------------------------------------------------------------------------------------------------------ 206

Figure 53: Difference in temperature (Celsius) between the reference and Low-carbon Scenarios for the years 2007-2030 for the February, March and April (A), April, May and June (B), August, September and October (C), and November, December and January (D) trimesters.----------------------------------- 208

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 37: projection of pasture area in Brazil from 2009 to 2030 (Low Carbon Scenario)------------- 165

Figure 38: Comparison of methane emissions from beef-cattle raising (MtCO2e per year), 2008-30-166

Figure 54: Difference in accumulated rainfall between the reference and Low-carbon Scenarios, using the 2007-2030 average. The color scale refers to amounts in millimeters of rainfall per year.--------- 209 14

Figure 55: Difference between the reference and low-carbon scenarios in average air temperature, taken between 2007 and 203. The color scale reflects the amount in celsius------------------------------ 210

Figure 56: Marginal Abatement Cost (8-percent social discount rate) and break-even carbon price (considering an IRR of 12%) for deforestation avoidance measures.--------------------------------------- 225 Figure 57: Variations in forest restoration costs by intervention scenario--------------------------------- 228

Figure 58: MAC and equilibrium price of carbon for CO2 uptake through legal forest restoration.---- 229

Figure 59: Percentages of distribution of investments by group of measures. ---------------------------- 232

Figure 60: Cost of land in the state of São Paulo between 1995 and 2008----------------------------------- 234

Figure 61: Variation in crops prices used in the present study----------------------------------------------- 236

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 62: Results of the Reference Scenario (Up) and aggregate Low-carbon scenario (down)------ 263

Maps Map 1: Map of the main regions of the land-use model. ----------------------------------------------------------44

Map 2: Dynamics of areas where sugar cane and cotton are grown in the Reference Scenario (20102030)------------------------------------------------------------------------------------------------------------------- 60

15

Map 3: Dynamics of areas where rice and beans are grown for the Reference Scenario (2010-2030)-61

Map 4: Dynamics of areas where corn and soybean are cultivated for the Reference Scenario (20102030).------------------------------------------------------------------------------------------------------------------ 62 Map 5: Dynamic of planted forest and pasture areas for the Reference Scenario (2010-2030)--------- 63 Map 6: Simplification of the soils map for Brazil with six soil categories ----------------------------------- 86 Map 8: Deforestation in the Reference Scenario (2010-2030)----------------------------------------------- 96 Map 9: Carbon stock mosaic---------------------------------------------------------------------------------------- 97

Map 10: Boundaries of the Cerrado and the Atlantic Forest, extracted from the Map of

Brazilian Biomes----------------------------------------------------------------------------------------------------- 108 Map 11: Altimetry based on the Digital SRTM Land Model. Original data offer an average altitude

with 3” resolution. The model presented was re-modeled for 30”------------------------------------------ 108

Map 12: Declivity in percentages based on the digital land model------------------------------------------- 109

Map 13: Average annual precipitation in millimetres---------------------------------------------------------- 109

Map 14: Length of the growing season indicated by the sum of months with higher precipitation than 50 mm------------------------------------------------------------------------------------------------------------------ 110

Map 15: Average temperature of the hottest month of the year---------------------------------------------- 111 Map 16: Soil fertility map for Brazil------------------------------------------------------------------------------- 111

Map 17: Map fo plant cover in Brazil for 2000------------------------------------------------------------------- 112 Map18: Map of Brazilian Ecosystems, IBGE, representing an estimate of the dsitribution of “original” plant formations with a simplified legend indicating areas of transition----------------------------------- 113

Map 19: Map of carbon uptake potential through the forest restoration of the Legal Reserve in the Cerrado and Atlantic Forest in tCO2/ha------------------------------------------------------------------------------- 114

Map 20: Map of carbon uptake potential through the restoration of riverside forests in the Cerrado and Atlantic Forest biomes in tCO2/ha--------------------------------------------------------------------------------- 114 Map 21: Forest productivity (tCO2/ha/year) for the Cerrado and Atlantic Forest biomes-------------- 116

Map 22: Mitigation by crop, 2010 to 2030------------------------------------------------------------------------ 126 Map 23: Total emissions from agriculture, 2010 to 2030------------------------------------------------------ 127

Map 24: CO2 uptake potential through forest restoration by 2030 and total CO2 uptake potential---- 144 Map 25: Number of heads of cattle-------------------------------------------------------------------------------- 168

Map 26: Total cumulative emissions from livestock, 2010-2030-------------------------------------------- 169

Map 27: Dynamics of sugar cane cultivation and cotton in the Low-carbon Scenario-------------------- 172

Map 28: Dynamics of the rice and bean crops in the Low-carbon Scenario--------------------------------- 173 Map 29: Dynamics of corn crop and soybean in the Low-carbon Scenario--------------------------------- 173

Map 30: Dynamics of planted forests and pastures in the Low-carbon Scenario (2010 – 2030). Yellow = reimaned constant; blue = crop decrement; red = crop increment ----------------------------------------- 174

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 7: Total GHG emissions in CO2 equivalent (millions of tons) by state resulting from agricultural land use---------------------------------------------------------------------------------------------------------------- 93

Map 31: Forest regrowth in the low-carbon scenario---------------------------------------------------------- 175

Map 32: Area used for agriculture, pasture, and reforestationby region----------------------------------- 175

Map 33: Comparison of cumulative deforestation, 2007-30-------------------------------------------------- 176 16

Map 34: Total area deforested, 2010-2030---------------------------------------------------------------------- 177 Map 35: Total cumulative emissions from deforestation, 2010-2030-------------------------------------- 179

Map 36: Total cumulative emissions from land use (Agriculture, Livestock, Deforestation, and Reforestation) 2010-30------------------------------------------------------------------------------------------------------- 188 Map 37: Land-use map for the year 2007 in Refernce Scenario (1 x1 Km resolution)-------------------- 197

Map 38: Schematic map of Brazil showing the different regions in the country and their boundaries for the analysis of results (above). Below, normal performance of the number of emissions sources in the different regions obtained with data from the AVHRR sensor (Advanced Very High Resolution Radiometer) from 1998 to 2008, present in the satellites of the NOAA series (National Oceanic and Atmospheric Administration)--------------------------------------------------------------------------------------------- 198

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 39: Figures (A), (B), (C) and (D) show the locations of deforestation from 2007 to 2030 in the reference (REF) and low carbon (BC) scenarios. Regions with forest remnants are also shown during that period (in green). Figures (E) and (F) show the optic depth of average aerosol from 2007 to 2030 in the reference (E) and low carbon (F) scenarios, where the current lines represent the average wind field over Brazil------------------------------------------------------------------------------------------------------------- 201

Boxes

Box 1: EGO DYNAMIC (Environment for Geoprocessing Objects)---------------------------------------------66

Box 2: Moving towards a “Legal Scenario”: Main Areas to Protect------------------------------------------- 142

Box 3: Uncertainties for Economic land-use Scenarios-------------------------------------------------------- 192

Box 4: Calculating Marginal Abatement Costs------------------------------------------------------------------- 213

Acronyms and Abbreviations ABRAF

Brazilian Association of Plantation Forest Producers (Associação Brasileira de Produtores de Florestas Plantadas)

ANEEL

National Agency for Electric Energy (Agencia Nacional de Energia Elétrica)

ANP

ARPA

BDMG BEN

BLUM

BNDES CAN

CBERS CCC CCS

CDE

CDM

CEAF CEIF

CEPEL CER

CETESB CFL

CGEE CH4

CIDE CMN

National Association of Motor Vehicle Manufacturers (Associação Nacional dos Fabricantes de Veículos Automotores)

National Agency of Petroleum, Natural Gas, and Biofuels (Agência Nacional do Petróleo, Gás Natural, e Biocombustíveis) Amazon Region Protected Areas Program Minas Gerais Development Bank National Energy Balance Brazil Land Use Model

National Bank of Economic and Social Development (Banco Nacional de Desenvolvimento Econômico e Social) National Confederation of Agriculture and Livestock China-Brazil Earth Resources Satellites

Fuel Consumption Account (Conta de Consumo de Combustíveis)

Socio-environmental Commitment Register

Energy Development Account (Conta de Desenvolvimento Energético) Clean Development Mechanism

Center for Alternative Energy Strengthening (Centro de Energias Alternativas de Fortaleza) Clean Energy Investment Framework

Research Center for Electrical Energy (Centro de Pesquisas de Energia Elétrica) Certified Emissions Reduction

São Paulo State Waste Management Agency (Companhia de Tecnologia de Saneamento Ambiental) Compact Fluorescent Lamp

Center for Strategic Management and Studies Methane

Contribution on Intervention in the Economic Domain (Contribuição de Intervenção no Domínio Econômico) National Monetary Council (Conselho Monetário Nacional)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

ANFAVEA

17

CNG

Compressed Natural Gas

CO2

Carbon Dioxide

CONAB

18

CONPET CPTEC CSR

CTEnerg

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

CT-Petro CU

DEGRAD DETER EGO EIA

EMBRAPA EPE

National Crop Supply Agency

National Program for the Rationalization of the Use of Oil and Natural Gas Derivatives (Programa Nacional de Racionalização do Uso dos Derivados de Petróleo e Gás Natural)

Center for Weather Forecasts and Climate Studies Remote Sensing Center

Sector Energy Fund of the Ministry of Science and Technology (Fundo Sectorial de Ciência e Tecnologia para Energia) Oil and Natural Gas Sector Fund of the Ministry of Science and Technology (Fundo Sectorial de Ciência e Tecnologia para Petróleo e Gás) Conservation Unit

Mapping of Forest Degradation in the Brazilian Amazon Detection System for Deforestation in Real Time Environment for Geoprocessing Objects Energy Information Administration

Brazilian Agricultural Research Corporation (Empresa Brasileira de Pesquisa Agrícola) Energy Planning Company (Empresa de Planejamento Energético)

ESCO

Energy Efficiency Service Company

FGEE

Guarantee Fund for Electric Energy Projects

FAPRI FDI

FGTS

FINAME Agricola FINEM FINEP FNP

FUNAI GDP GEF

GHG

Food and Agricultural Policy Research Institute Foreign Direct Investment

Social Security (Fundo de Garantia do Tempo de Serviço)

Agency for the Acquisition of Machines and Equipment (Agência de Financiamentos para Aquisição de Máquinas e Equipamentos) Equipment and Machinery Financing (Financiadora de Equipamentos e Máquinas)

Agency for the Funding of Projects and Studies (Financiadora de Estudos e Projetos) FINEP Consulting & Trade (FINEP Consultoria & Comércio) National Foundation for Indigenous People Gross Domestic Product

Global Environment Facility Greenhouse Gas

GTL

HFC

IBAMA IBGE IBP

ICMBio ICONE IEA

IGP-DI INPE INT I-O

IPAM IPCC IPI

IRR

KfW LNG

LULUCF MAC

MACC MCT

MELP MEPS MIPE MMA MME

Gross National Product Gas-To-Liquid

Hydrofluorocarbon

Brazilian Institute of Environment and Renewable Natural Resources (Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis)

19

Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística) Potential Biomass Index

Chico Mendes Institute of Biodiversity Conservation (Instituto Chico Mendes de Conservação da Biodiversidade) Institute for International Trade Negotiations International Energy Agency

General Price Index-Domestic Availability (Índice Geral de Preços-Disponibilidade Interna)

National Institute for Space Research (Instituto Nacional de Pesquisas Espaciais) National Technological Institute (Instituto Nacional de Tecnologia) Input-Output

Amazon Institute for Environmental Research (Instituto de Pesquisa Ambiental da Amazonia) Intergovernmental Panel on Climate Change Industrial Products Tax Internal Rate of Return

German Development Bank Liquefied Natural Gas

Land Use, Land-Use Change, and Forestry

Marginal Abatement Cost

Marginal Abatement Cost Curve

Ministry of Science and Technology (Ministério de Ciência e Tecnologia) Long-term Expansion Model

Minimum Energy Performance Standard Integrated Energy Planning Model

Ministry of the Environment (Ministério do Meio Ambiente)

Ministry of Mines and Energy (Ministério de Minas e Energia)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

GNP

M-Ref MSR MT N

20

NAPCC NIPE NPV NRC

Residential Energy Demand Projection Model

Ministry of Transport (Ministério dos Transportes) Nitrogen

National Action Plan on Climate Change

Interdisciplinary Center for Strategic Planning Net Present Value (Valor Presente Líquido) National Research Council

N2O

Nitrous Oxide

PAS

Sustainable Amazon Program

OECD

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Refining Study Model

PAC PFC

PLANSAB PME PNE

PNLT

PNMC PPA

PPCDAM PPP

PROALCOOL PROBIO

PROCEL

PRODES

PRODUSA PROESCO PROINFA

Organisation for Economic Co-operation and Development Government Accelerated Growth Plan Perfluorocarbon

National Sanitation Plan (Plano Nacional de Saneamento Básico) Monthly Employment Survey

National Energy Plan (Plano Nacional de Energia)

National Logistics and Transport Plan (Plano Nacional de Logistica e Transporte)

National Plan on Climate Change (Plano Nacional sobre Mudança do Clima) Permanent Preservation Area

Plan of Action for the Prevention and Control of Deforestation in the Legal Amazon (Plano de Ação para Prevenção e Controle do Desmatamento na Amazônia Legal) Public-Private Partnership National Alcohol Program

Project for the Conservation and Sustainable Use of Brazilian Biological Diversity National Electrical Energy Conservation Program (Programa de Combate ao Desperdício de Energia Elétrica)

Amazon Deforestation Monitoring Program (Programa de Cálculo do Desflorestamento da Amazônia) Programa de Estímulo a Produção Agropecuária Sustentável)

Support Program for Energy Efficiency Projects (Programa de Apoio a Projetos de Eficiência Energética)

Incentive Program for Alternative Electric Energy Sources (Programa de Incentivo às Fontes Alternativas)

PROLAPEC

Agriculture-Livestock Integration Program (Programa de Integração Lavoura-Pecuária)

PROPASTO

National Program for Recuperation of Degraded Pastures

PROPFLORA R&D

REDD RGR

National Program for the Strengthening of Family Agriculture (Programa Nacional de Fortalecimento da Agricultura Familiar)

Program for Commercial Planting and Recovery of Forests (Programa de Plantio Comercial e Recuperação de Florestas) Research and Development

Reducing Emissions from Deforestation and Degradation Global Reversion Reserve (Reserva Global de Reversão)

RSU

Urban Solid Residues (Residuos Sólidos Urbanos)

SF6

Sulfur Hexafluoride

SAE SFB

UFMG UFRJ

UNFCCC

UNICAMP USP

WTI

21

Secretariat of Strategic Affairs (Secretaria de Assuntos Estratégicos) Brazilian Forest Service

Federal University of Minas Gerais (Universidade Federal de Minas Gerais) Federal University of Rio de Janeiro

United Nations Framework Convention on Climate Change State University of Campinas University of São Paulo

West Text Intermediate

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

PRONAF

Units of Measure

22





Ce

CO2e

ETE

gCO2e

Gt

Gt CO2e GW

GWh ha kg

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

km

Carbon Equivalent

Carbon Dioxide Equivalent Sewage Treatment Plant

Grams of Carbon Dioxide Equivalent Billions of Tons

Billion Tons of Carbon Dioxide Equivalent

Gigawatt

Gigawatt Hour Hectare

Kilogram

Kilometer

km

Square Kilometer

m

Cubic Meters



MW

Megawatt



tCO2e

2

kW m

3

Tg

Tg CO2e

MWh

ppm

TWh

Kilowatt Meter

Teragram

Teragram Carbon Dioxide Equivalent Megawatt Hour

Particles per Million

Tons of Carbon Dioxide Equivalent Terawatt Hour

Currency Exchange

1 US Dollar (US$) = 2.20 Brazilian Reais (R$)

Acknowlegments

The team was assisted by Letícia Hissa (UFMG), Leila Harfuch, Marcelo Melo Ramalho Moreira, Luciane Chiodi Bachion and Laura Barcellos Antoniazzi (ICONE), Geraldo Martha Junior, Roberto D. Sainz, and Magda A. de Lima (EMBRAPA), Renato Toledo (Iniciativa Verde), Rodrigo Ferreira, Luiz Goulart, and Thiago Mendes (PLANTAR), Karla Maria Longo and Ricardo Almeida de Siqueira (INPE), and Mark Lundell, Adriana Moreira, Barbara Farinelli, Jennifer Meihuy Chang, Govinda Timilsina, Garo Batmanian, Fowzia Hassan, Benoit Bosquet, Alexandre Kossoy, Flávio Chaves, Mauro Lopes de Azevedo, Fernanda Pacheco, Megan Hansen, Augusto Jucá, and Rogerio Pinto (The World Bank). The team benefited greatly from a wide range of consultations with the Ministries of Foreign Affairs, Environment, and Science and Technology. Several seminars were organized, enabling consultation with representatives of the Ministries of Finance, Planning, Agriculture, Transport, Mines and Energy, Industry and Commerce.

The team acknowledges the generous support provided by the Sustainable Development Network for activities related to climate change and regional support through the Energy Sector Management Assistance Program (ESMAP).

The team is especially gratefull to Mark Lundell and Garo Batmanian for their valuable contribution throughtout the development of the study.

Special thanks to Adriana Moreira for coordinating the initial compilation of the report and to Barbara Farinelli for revising the translation, editing and coordinating the printing process.

The team would also like to acknowledge Judy Wolf and Helena Jansen for their support in translating and editing the report.

23

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

This report has been prepared by a core team led by Christophe de Gouvello (The World Bank), Britaldo S. Soares Filho (Federal University of Minas Gerias, UFMG), André Nassar (Institute for International Trade Negotiations, ICONE), Luis G. Barioni and Bruno J. R. Alves (Brazilian Agricultural Research Corporation, EMBRAPA), Osvaldo Martins and Magno Castelo Branco (Iniciativa Verde), Manoel Regis Lima Verde Leal (Centro de Energias Alternativas e Meio Ambiente, CENEA), Fábio Marques (PLANTAR), Júlio Hato and Sérgio Pacca (University of São Paulo, USP), and Saulo Ribeiro Freitas (National Institute for Space Research, INPE).

Executive Summary

Brazil’s commitment to combat climate change had already begun when the country hosted the United Nations Conference on Environment and Development, also known as the Rio Earth Summit, in June 1992. The resulting United Nations Framework Convention on Climate Change (UNFCCC) led to the creation of the Kyoto Protocol. Today, Brazil remains strongly committed to voluntarily reducing its carbon emissions. On December 1, 2008, President Luiz Inácio Lula da Silva launched the National Plan on Climate Change (PNMC), based on work of the Interministerial Committee on Climate Change, in collaboration with the Brazilian Forum on Climate Change and civil society organizations. The PNMC calls for a 70-percent reduction in deforestation by 2017, a particularly noteworthy goal given that Brazil has the world’s second largest block of remaining native forest. On December 29, 2009 the Brazilian Government adopted Law 12.187, which institutes the National Climate Change Policy of Brazil and set a voluntary national greenhouse gas reduction target of 36.1 to 38.9 percent of projected emissions by 2020.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

24

This report presents the partial results related to land use, land-use change and the forestry sector from a larger multisectoral low-carbon study for Brazil1.

As the world’s largest tropical country, Brazil is unique in its greenhouse gas (GHG) emissions profile. In prior decades, the availability of large volumes of land suitable for crop cultivation and pasture helped to transform agriculture and livestock into key sectors for sustaining the country’s economic growth. In the past decade alone, these two sectors accounted for an average of 25 percent of the national GDP. The steady expansion of crop land and pasture has also required the conversion of more native land, making land-use change the country’s main source of GHG emissions today. At the same time, Brazil has used the abundant natural resources of its vast territory to explore and develop low-carbon renewable energy.

Yet Brazil used to be one of the largest GHG emitters from deforestation and would probably continue to be so if not for the government’s recent adoption of a series of measures to protect the forest. Although drastically reduced in recent years, deforestation could continue to be a potentially large emission source in the future.

At the same time, the country is likely to suffer significantly from the adverse effects of climate change. Some advanced models suggest that much of the eastern part of the Brazilian Amazon region could be converted into a savannah-like ecosystem before the end of this century. A phenomenon known as Amazon dieback, combined with the shorter-term effects of deforestation by fires, could reduce rainfall in the central-west and northeast regions, resulting in smaller crop yields and less available water for hydropower-based electricity2. Urgent solutions are thus needed to reduce Brazil’s vulnerability to climate change and to enable the implementation of adaptation actions in the country. Like many other developing countries, Brazil faces the dual challenge of encouraging development and reducing GHG emissions. President Lula echoed this concern in his introduction to the National Plan, stating that actions to avoid future GHG emissions 1 2

Brazil Low-carbon Case Study, World Bank, June 2010. “Assessment of the Risk of Amazon Dieback,” World Bank, 2010.

should not adversely affect the development rights of the poor, who have done nothing to generate the problem. Efforts to mitigate GHG emissions should not add to the cost of development, but there are strong reasons to shift toward a low-carbon economy. Lowcarbon alternatives would offer important development co-benefits, ranging from reduced congestion and air pollution in urban transport areas to better waste management, job creation and costs savings for industry, and biodiversity conservation. Countries that pursue low-carbon development are more likely to benefit from strategic and competitive advantages, such as the transfer of financial resources through the carbon market, new international financing instruments, and access to emerging global markets for low-carbon products. In the future this may create a competitive advantage for the production of goods and services, due to the lower emission indexes associated with the life cycle of products.

25

The overall aim of this study was to support Brazil’s efforts to identify opportunities to reduce its emissions in ways that foster economic development. The primary objective was to provide the Brazilian government with the technical inputs needed to assess the potential and conditions for low-carbon development in key emitting sectors.

To this end, the World Bank study adopted a programmatic approach in line with the Brazilian government’s long-term development objectives, as follows: (i) anticipate the future evolution of Brazil’s GHG emissions to establish a Reference Scenario; (ii) identify and quantify lower carbon-intensive options to mitigate emissions, as well as potential options for carbon uptake; (iii) assess the costs of these low-carbon options, identify barriers to their adoption, and explore measures to overcome them; and (iv) build a low-carbon emissions scenario that meets the same development expectations. The team also analyzed the macroeconomic effects of shifting from the Reference Scenario to the low-carbon one and the financing required. To build on the best available knowledge and avoid duplicating efforts, the study team undertook a broad consultative process, meeting with more than 70 recognized Brazilian experts, technicians, and government representatives covering most emitting sectors and surveying the copious literature available. This preparatory work informed the selection of four key areas with great potential for low-carbon options: (i) land use, land-use change, and forestry (LULUCF), including deforestation; (ii) transport systems; (iii) energy production and use, particularly electricity and oil and gas; and (iv) solid and liquid urban waste.3

In order to estimate the emissions Brazil would generate in these four key areas over the study period, the study team defined a “Reference Scenario” that is later compared with the projected “Low-carbon Scenario”. It is worth noting that the Reference Scenario is based on a different methodology than the one used by the Brazilian government 3

Certain industrial sources of nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), and other non-Kyoto GHG gases are not covered by this study. Without a recent complete inventory, it is not possible to determine precisely the share of other sources in the national GHG balance. However, based on the first Brazil National Communication (1994), it is expected that they would not exceed 5 percent of total Kyoto GHG emissions. Not all agricultural activities were taken into account when estimating emissions from that sector; and crops taken into account in the calculation of emissions from agriculture represent around 80% of the total crop area.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Study Overview

Reference-scenario results for these main areas show that deforestation remains the key driver of Brazil’s future GHG emissions through 2030. Modeling results indicate that, after a slight decrease in 2009–11, deforestation emissions are expected to stabilize at an annual rate of about 400–500 Mt CO2.

Despite its significant decline over the past four years, deforestation remains Brazil’s largest source of carbon emissions, representing about two-fifths of national gross emissions (2008). Over the past 15 years, deforestation has contributed to reducing Brazil’s carbon stock by about 6 billion metric tons, the equivalent of two-thirds of annual global emissions.4 Without the Brazilian government’s recent forest protection efforts, the current emissions pattern from deforestation would be significantly higher.5 The drivers of deforestation occur at multiple levels. In the Amazon and Cerrado regions, for example, the spatial dynamics of agricultural and livestock expansion, new roads, and immigration determine the pattern of deforestation. On a national or international scale, broader market forces affecting the meat and crop sectors drive deforestation.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

26

in its national GHG inventory. In particular, having focused on these four areas, the Reference Scenario built by this study does not cover 100 percent of all emission sources of the country and therefore should not be considered a simulation of future national emissions inventories.

Agricultural production and livestock activities also produce direct emissions, together accounting for one-fourth of national gross emissions. Agricultural emissions mainly result from the use of fertilizer and mineralization of nitrogen (N) in the soil, cultivation of wetland irrigated rice, the burning of sugar cane, and use of fossil fuel– powered agricultural equipment. Livestock emissions result mainly from the digestive process of beef cattle, which releases methane (CH4) into the atmosphere. Models and Reference-scenario Results

To estimate future demand for land and LULUCF emissions, the study developed two complementary models: i) Brazilian Land Use Model (BLUM) and (ii) Simulate Brazil (SIM Brazil). BLUM is an econometric model that estimates the allocation of land area and measures changes in land use resulting from supply-and-demand dynamics for major competing activities.6 SIM Brazil, a geo-referenced spatialization model, estimates future land use over time under various scenarios. SIM Brazil does not alter BLUM data; it finds a place for land-use activities, taking into account such criteria as agricultural aptitude, distance to roads, urban attraction, cost of transport to ports, declivity, and distance to converted areas. SIM Brazil works at a definition level of 1 km2, making it possible to generate detailed maps and tables. Under the Reference Scenario, about 17 million ha of additional land are required to accommodate the expansion of all activities over the 2006–30 period. In Brazil as a 4

5

6

From 1970 to 2007, the Amazon lost about 18 percent of its original forest cover; over the past 15 years, the Cerrado lost 20 percent of its original area, while the Atlantic Forest, which had been largely deforested earlier, lost 8 percent. After peaking at 27,000 km² in 2004, deforestation rates have declined significantly, dropping to 11,200 km² in 2007, the second lowest historical rate recorded by the PRODES deforestation observation program (INPE 2008). These include six key crops (soybean, corn, cotton, rice, bean, and sugar cane), pasture, and production forests; the model also projects the demand for various kinds of meat and corresponding needs for hay and corn.

To estimate the corresponding balance of annual emissions and carbon uptake over the next 20-year period, these and related models calculated land use and land-use change for each 1-km2 plot at several levels.7 Results showed that land-use change via deforestation accounts for the largest share of annual LULUCF emissions—up to 533 Mt CO2e by 2030. Direct annual emissions from land use only (agriculture and livestock) increase over the period at an average annual rate of 346 Mt CO2e. Carbon uptake offsets less than 1 percent of gross LULUCF emissions, sequestering 29 Mt CO2e in 2010, down to 20 Mt CO2e in 2030. Over the 20-year period, LULUCF gross emissions increase one-fourth, reaching 916 Mt CO2e by 2030. The net balance between land use, land-use change, and carbon uptake results in increased emissions, which reachs about 895 Mt CO2e annually by 20308.

Low-carbon Options for Emissions Mitigation and Carbon Uptake

Avoiding deforestation offers by far the greatest opportunity for GHG mitigation in Brazil. Under the resulting Low-carbon Scenario, avoided emissions from deforestation would amount to about 6.2 Gt CO2e over the 2010–30 period, or more than 295 Mt CO2e per year.

Brazil has developed forest-protection policies and projects to counter the progression of pressure at the frontier and has experience in economic activities compatible with forest sustainability. Shifting to a Low-carbon Scenario that ensures the growth of agriculture and the meat industry—both important to the Brazilian economy—would also require acting on the primary cause of deforestation: the demand for more land for agriculture and livestock. This study proposed a dual strategy to drastically reduce deforestation: (i) eliminate the structural causes of deforestation and (ii) protect the forest from illegal attempts to cut it. Eliminating the structural causes of deforestation would require a dramatic increase in productivity per hectare. Increasing livestock productivity could free up large quantities of pasture. This option is technically feasible since Brazil’s livestock productivity is generally low, and existing feedlots and crop-livestock systems could be scaled up. Moreover, the use of more intensive production systems could trigger higher economic returns and a net gain for the sector economy (Chapter 7). The potential to release and recover degraded pasture is enough to accommodate the most ambitious growth scenario. 7 8

Micro-region, state, and country. When calculating national carbon inventories, some countries consider the contribution of natural regrowth towards carbon uptake; therefore, although this study does not compute this contribution in the carbon balance of LULUCF activities, it would be fair to add that information for comparison purposes. If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by 109Mt CO2 per year, thus reducing net emissions.

27

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

whole, the total area allocated for productive uses, estimated at 257 million ha in 2008, is expected to grow 7 percent—to about 276 million ha—in 2030; 24 percent of that growth is expected to occur in the Amazon region. In 2030, as in 2008, pastures are expected to occupy most of this area (205 million ha in 2008 and 207 million in 2030). Growth of this total amount over time makes it necessary to convert native vegetation for productive use, which mainly occurs in frontier regions, the Amazon region, and in Maranhão, Piauí, Tocantins, and Bahia on a smaller scale.

Model-based projections indicate that, under the new land-use dynamic, deforestation would be reduced by more than two-thirds (68 percent) in 2030, compared to projected levels in the Reference Scenario. In the Atlantic Forest, the reduction would be about 90 percent, while the Amazon region and Cerrado would see reductions of 68 percent and 64 percent, respectively. Accordingly, in 2030, annual emissions from deforestation would be reduced nearly 63 percent (from about 530 Mt CO2 to 190 Mt CO2) compared to the projected Reference Scenario. In the Amazon, the level of deforestation would fall quickly to about 17 percent of the historic annual average of 19,500 km2 observed in the recent past, thus complying with the NPCC goal of reducing deforestation in the Amazon region by 72 percent by the year 201711.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

28

The combination of reducing pasture area and protecting forests can lead to a sharp decline in deforestation emissions. This was demonstrated in 2004–07, when new forest-protection efforts, combined with a slight contraction in the livestock sector and resultant pasture area,9 led to a 60-percent reduction in deforestation (from 27,000 km² to 11,200 km²). Such a rapid reduction resulted from deforestation and its associated emissions being related to the marginal expansion of land for agriculture and livestock activities,10 without which there would be no need to convert additional native vegetation and incidentally generate GHG emissions. If the effort to reduce pasture area and protect forests were neglected, emissions from deforestation would resume immediately. To protect against illegal cutting, the forest should be further protected against fraudulent interests. The Brazilian government has made considerable efforts in this area, particularly under the 2004 Plan of Action for the Prevention and Control of Deforestation in the Legal Amazon (PPCDAM).

The study also proposed ways to reduce direct emissions from agricultural production and livestock activities. For agriculture, the study proposed an accelerated dissemination of zero-tillage cultivation. Compared to conventional farming systems, zero-tillage involves far fewer operations and can thus reduce emissions caused by altering soil carbon stock and using equipment powered by fossil fuels. Done effectively, zero-tillage cultivation can help control soil temperature, improve soil structure, increase soil water-storage capacity, reduce soil loss, and enhance the nutrient retention of plants. For these reasons, expansion of zero-tillage cultivation is accelerated in the Low-carbon Scenario, reaching 100 percent by 2015 and delivering 356 Mt CO2e of avoided emissions over the 2010–30 period.

To lower direct emissions from beef-cattle farming, the study proposed shifting to more intensive meat-production systems, as mentioned above. It also proposed genetic improvement options to reduce CH4, including improved forage for herbivores and genetically superior bulls, which have a shorter life cycle. The study projects that the combination of improved forage and bulls, along with increased productivity, would reduce direct livestock emissions from 272 to 240 Mt CO2 per year by 2030; that is, maintaining them close to 2008 levels. 9

The 2005–07 period witnessed the first decline in herd size (from 207 million to 201 million head), following a decade-long increase, together with a slight contraction in pasture area (from 210 million to 207 million ha). 10 Unlike other sectors, whose energy-based emissions are usually proportional to the full size of the sector activity, emissions from deforestation are related only to the marginal expansion of agriculture and livestock activities. 11 Over the 1996–2005 period, the historical rate of deforestation in the Amazon region was 1.95 million ha per year, according to the PNMC.

A New Land-use Dynamic

Building a Low-carbon Scenario for land use involves more than adding emissions reductions associated with mitigation opportunities; it must also avoid the potential for carbon leakage. For example, increasing forest recovery leads to carbon uptake, but it also reduces the land area otherwise available for expanding agriculture and livestock activities. This, in turn, could cause an excess in demand for land use, which could generate deforestation, inducing a lower net balance of carbon uptake. To avoid carbon leakage, ways must be found to reduce the global demand for land for other activities, while maintaining the same level of product supply found in the Reference Scenario.

In the Low-carbon Scenario, the amount of additional land required for emissions mitigation and carbon uptake totals more than 53 million ha. Of that amount, more than 44 million ha—over twice the land expansion projected under the Reference Scenario—is for forest recovery. Together with the additional land required under the Reference Scenario, the total volume of additional land required is more than 70 million ha, more than twice the total amount of land planted with soybean (21.3 million ha) and sugar cane (8.2 million ha) in 2008 or more than twice the area of soybean projected for 2030 in the Reference Scenario (30.6 million ha) (Table 1). 12

In areas with optimal conditions, forest recovery can remove 100 tC per ha on average in the Amazon Region. (Saatchi, 2007). In the reference scenario, its contribution is limited in terms of quantity. 13 The study model used meteorological and climatic variables (e.g., rainfall, dry season, and temperature) and edaphic (soil and topography) variables to estimate potential biomass. 14 If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by 112Mt CO2 per year on average.

29

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The study also explored two major carbon uptake options: (i) recovery of native forests and (ii) production forests for the iron and steel industry. For forest recovery, the Low-carbon Scenario considered compliance with legal actions for mandatory reconstitution, in accordance with the laws of riparian forests and legal reserves.12 In this sense, the Low-carbon Scenario engendered a “Legal Scenario”. Using these defined areas for reforestation, the study modeled their potential for CO2 removal.13 Results showed that the legal scenario has high carbon-uptake potential: a cumulative total of 2.9 Gt CO2e over the 20-year period or about 140 Mt CO2e per year on average14. For production forests, the Reference Scenario assumed that the thermo-reduction process would be based on coke (66 percent), non-renewable charcoal (24 percent), and renewable charcoal (10 percent), based on estimates that reflect the current situation. Two Low-carbon Scenarios were developed. The first only reflects the maintenance of the current participation of charcoal in iron and steel production (approximately 34 percent), but with a completely renewable origin. The second – considered in the overall calculations of this report – was more daring, so that the hypothesis of competition for the use of land for planted forests would be taken into consideration in a rather conservative way, assuming a total substitution of non-renewable charcoal by 2017 and the use of renewable charcoal for up to 46 percent of total iron and steel ballast production by 2030. With this, the volume of greenhouse gas uptake or “sequestration” would be between 500 and 700 MtCO2 in 2030, or from 321 to 517 MtCO2 more than in the Reference Scenario.

Table 1: Summary of additional land needs in the reference and Low-carbon Scenarios Scenario 30

Additional land needed (2006–30)

Reference Scenario: additional Expansion of agriculture and livestock production to meet volume of land required for the the needs anticipated in 2030: expansion of agriculture and • 16.8 million ha livestock activities Elimination of non-renewable charcoal in 2017 and the participation of 46 percent of renewable charcoal for iron and steel production in 2030: • 2.7 million ha

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Expansion of sugar cane to increase gasoline substitution Low-carbon Scenario: addition- with ethanol to 80 percent in the domestic market and al volume of land required for supply 10 percent of estimated global demand to achieve an average worldwide gasoline mixture of 20 percent ethamitigation measures nol by 2030: • 6.4 million ha

Total

Restoration of the environmental liability of “legal reserves” of forests, calculated at 44.3 million ha in 2030: • 44.3 million ha 70.4 million additional hectares

To increase livestock productivity to the level needed to release the required volume of pasture, the Low-carbon Scenario considered three options: (i) promote recovery of degraded pasture, (ii) stimulate the adoption of productive systems with feedlots for finishing, and (iii) encourage the adoption of crop-livestock systems. The increased carrying capacity that results from the recovery of degraded areas, combined with more intensive integrated crop-livestock systems and feedlots for finishing are reflected in an accentuated reduction in the demand for land, projected at about 138 million ha in the Low-carbon Scenario, versus 207 million ha in the Reference Scenario, for the year 2030. The difference would be enough to absorb the demand for additional land associated with both expanded agriculture and livestock activities in the Reference Scenario and expanded mitigation and carbon uptake in the Low-carbon Scenario.

A National Low-carbon Scenario

The Low-carbon Scenario constructed for Brazil in the global, multi-sectoral study is an aggregate of the Low-carbon Scenarios developed for each of the four sectors considered in the study. In each sector, the most significant opportunities to mitigate and sequester GHGs were analyzed, while less promising or fully exploited options in the Reference Scenario were not considered further. In short, this national Low-carbon Scenario is derived from a bottom-up, technology-driven simulation for single subsectors (e.g., zero tillage with straw or reduction of deforestation), based on in-depth technical and economic assessments of feasible options in the Brazilian context, and sectorlevel optimization for two of the four main sectors (land use and transport). This national Low-carbon Scenario has been built in a coordinated way to ensure

This Low-carbon Scenario represents a 37-percent reduction in gross GHG emissions compared to the Reference Scenario over the 2010–30 period. The total cumulative emissions reduction over the period amounts to more than 11.1 Gt CO2e, equal to approximately 37 percent of the cumulative emissions observed under the Reference Scenario. Projected gross emissions in 2030 are 40 percent lower under the Low-carbon Scenario (1,023 Mt CO2e per year) compared to the Reference Scenario (1,718Mt CO2e per year) and 20 percent lower than in 2008 (1,288 Mt CO2e per year) (Table 2, Figure 1). In addition, forest plantations and recovery of legal reserves will sequester the equivalent of 16 percent of reference-scenario emissions in 2030 (213 Mt CO2e per year) 16. Table 2: Comparison of emissions distribution among sectors in the reference and Lowcarbon Scenarios - 2008-30 Reference 2008

Reference 2030

Low-carbon 2030

Mt CO2e

%

Mt CO2e

%

Mt CO2e

%

Energy

232

18

458

27

297

29

Deforestation

536

42

533

31

196

19

Sector

Transport Waste

Livestock

Agriculture

Total Gross Emissions Carbon uptake

Total Net Emissions

149 62

237 72

1,288 -29

1,259

12 5

18 6

100 -2

98

245 99

272

111

1,718 -21

1,697

14 6

16 6

100 -1

99

174 18

249 89

1,023 -213 810

17 2

24 9

100 -21 79

The two areas where the proposed Low-carbon Scenario succeeds most in reducing net emissions are reducing deforestation and increasing carbon uptake. The main drivers are (i) reduction of total land area needed, via significant gains in livestock productivity, to accommodate expanded agriculture and meat production and (ii) restora15

Three seminars were held over the past several years (September 14–16, 2007; April 30, 2008; and March 19, 2009) to present and discuss the study methodology, intermediate results, and near-final results with representatives of 10 ministries. Sectoral teams also interacted on various occasions with technical-experts and public-agency representatives. 16 If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by 112MtCO2 per year on average, thus reducing the net emissions.

31

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

full consistency among the four main sectors considered. To ensure transparency, the methods and results were presented and discussed on various occasions with a range of government representatives.15 But this Low-carbon Scenario is not presumed to have explored all possible mitigation options or represent a preferred recommended mix. This scenario, which simulates the combined result of all the options retained under this study, should be considered modular—as a menu of options—and not prescriptive, especially since the political economy between sectors or regions may differ significantly, making certain mitigation options that at first appear more expensive easier to select than others that initially appear more attractive economically.

Figure 1: GHG mitigation wedges in the Low-carbon Scenario, 2008-30

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

32

tion of legal forest reserves and production forests for producing charcoal for the steel industry. By 2017, the proposed Low-carbon Scenario would reduce deforestation by more than 80 percent compared to the 1996–2005 average, thereby ensuring compliance with the Brazilian government’s December 2008 commitment.

Low-carbon Scenario, 2008-30

It is more difficult to reduce emissions in the energy and transport sectors, since they are already low by international standards, due mainly to hydroelectricity for power generation and bioethanol as a fuel substitute for gasoline in the current energy matrix. As a result, these sectors’ relative share of national emissions increases more in the Low-carbon Scenario than in the Reference Scenario.

Meeting the Challenge of the Low-carbon Scenario

Implementing the proposed Low-carbon Scenario requires tackling a variety of challenges in each of the four areas considered. The combined strategy of releasing pasture and protecting forests to reduce deforestation to 83 percent of historically observed levels involves five major challenges. First, productive livestock systems are far more capital-intensive, both at the investment stage and in terms of working capital. Having farmers shift to these systems would require offering them a large volume of attractive financing far beyond current lending levels. Thus, a large volume of financial incentives, along with more flexible lending criteria, would be needed to make such financing viable for both farmers and the banking system. A first attempt to estimate the volume of incentives required indicates an order of magnitude of US$1.6 billion per year, or US$34 billion during the period. Second, these systems require higher qualifications than traditional extensive farming, which is used to move on to new areas as soon as pasture productivity has degraded, eventually converting more native vegetation into pasture. Therefore, the financing effort should be followed by the intensive development of extension services. A third challenge is preventing a rebound effect: The higher profitability of needing less land to produce the same volume of meat might trigger an incentive to produce

Fourth, several attractive options in the Low-carbon Scenario to mitigate emissions or increase carbon uptake amplify the requirement of freeing up pasture to prevent carbon leakage. For example, while replanting the forest to comply with the Legal Reserve Law would remove a large amount of carbon dioxide (CO2) from the atmosphere, this area would no longer be available for other activities. The equivalent additional amount of pasture would need to be freed up; otherwise, a portion of production would have to be reduced or more native forest would eventually be destroyed elsewhere. A more flexible legal obligation regarding forest reserves would make the goal of accommodating all agriculture, livestock and forestry activities without deforestation less difficult, but it might also mean less carbon uptake.

Final Remarks

Brazil harbors considerable opportunities for GHG emissions mitigation and carbon uptake, positioning the country as one of the key players to tackle the challenge posed by global climate change. This study has demonstrated that a series of mitigation and carbon uptake measures are technically feasible and that promising efforts are already under way. Yet, implementing these proposed measures would require large volumes of investment and incentives, which may exceed a strictly national response and require international financial support. Moreover, market mechanisms would not be sufficient for Brazil to harvest the full range of opportunities to mitigate GHG emissions. Public policies and planning would play a pivotal role, with land competition and forest protection management at the center.

33

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

more meat and eventually convert more native forest into pasture. Such a risk is especially high in areas where new roads have been opened or paved. Therefore, the incentive provided should be selective, especially in the Amazon region, and should be given only when it is clearly established, based on valid and geo-referenced land ownership title, that the project will include neither conversion of native vegetation nor areas converted in recent years (e.g., less than 5 years).

1 Introduction

34

The urgent need to combat global climate change has been firmly established. An overwhelming body of scientific evidence, including the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2007) and a recent review on the economics of climate change led by Nicholas Stern (Stern 2007), underscore the severe risks to the natural world and global economy. According to Stern, how we decide to live over the next 20–30 years—how we treat forests, generate and use energy, and organize transport—will determine whether the risks of global climate change can remain manageable (Stern 2009).

Managing Risk: Target Levels

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Failure to hold greenhouse gas (GHG) concentrations below certain levels would imply a great risk to our planet. Recent studies have put forward various target levels, all of which would need emissions to peak soon. The IPCC (2007) concluded that stabilizing GHG concentrations at 550 particles per million (ppm)—the level at which it may be possible to hold the rise in global mean temperature under 3°C above pre-industrial levels—would require concentrations to peak no later than 2030 and then fall drastically by 2050; in this scenario, the IPCC estimates global emissions would need to be reduced to about 29 Gt CO2e by 2030.

Another recent study, conducted by the United Nations Framework Convention on Climate Change (UNFCCC), projects that emissions will reach 61.5 Gt CO 2 e by 2030. In this scenario, annual emissions from Annex I (industrialized) countries would go from 21 Gt CO 2 e to just 22.1 Gt CO 2 e by 2030, while the bulk of global emissions—50–70 percent of the emissions-mitigation potential—would come from non-Annex I (developing) countries. Despite the range of uncertainty, developing countries clearly have a vital role to play in shaping international policies and actions to cut emissions back to the required scale.

The Brazilian Context: Key Role of Forests and Other Sectors

It is difficult to imagine an effective solution to stabilizing GHG concentrations at the required scale without Brazil playing a prominent role. The Amazon rainforest, which covers more than half the country, is a reservoir of about 100 billion tons of carbon, sequestering more than 10 times the amount of carbon emitted globally each year. Given Brazil’s large forested areas—second only to those of Indonesia—it is perhaps not surprising that most of the world’s emissions from deforestation emanate from these two countries.

At the same time, Brazil is likely to suffer from the adverse effects of climate change. Some advanced models suggest that much of the eastern part of the Brazilian Amazon region could be converted into a savannah-like ecosystem before the end of the century. This phenomenon, known as Amazon dieback, combined with the shorter-term effects of deforestation by fires, could reduce rainfall in the Central-West and Northeast regions, resulting in smaller crop yields and less water available for hydropower-based electricity.

Apart from land use, land-use change, and forestry (LULUCF), Brazil accounts for only 2.3 percent of global GHG emissions; but until a few years ago, that percentage used to rise another 3 percent when considering LULUCF. Indeed, the LULUCF sector is pivotal, accounting for about two-thirds of Brazil’s gross CO2e emissions (2008), with two-thirds of that amount represented by deforestation alone. By contrast, Brazil’s energy sector has a per-capita carbon intensity of only 1.9 tCO2 per year—about half the global average and less than one-fifth the average of OECD countries. Were it not for Brazil’s previous large investments in renewable energy, the country’s current energy matrix would be far more carbon intensive, with presumably twice the amount of energy-sector emissions and national emissions 17 percent higher.

Four sectors are key contributors to Brazil’s GHG emissions. First and foremost is LULUCF, which covers the forestry dimensions described above. In addition, there are three other major emitting sectors: (i) energy, (ii) transport, and (iii) waste management. In 2008, the respective emissions contributions of these three sectors were 18, 14, and 5 percent. While waste management’s contribution was low in 2008, it has increased more than 60 percent over the past two decades.

A National Commitment to Combat Climate Change

Climate change has long been a vital part of Brazil’s national agenda. In June 1992, Brazil hosted the United Nations Conference on Environment and Development, known as the Rio Earth Summit, which resulted in an agreement on the UNFCCC and, in turn, the Kyoto Protocol. Since then, Brazil has played an active role in the international dialogue on climate change. In 2007, the Brazilian government created the Secretariat for Climate Change within its Ministry of Environment. The following year, President Luiz Inácio Lula da Silva launched the National Plan on Climate Change (PNMC), which put the issue at the forefront of the national agenda. On December 29, 2009 the Brazilian Parliament adopted Law 12.187, which institutes the National Climate Change Policy of Brazil and set a voluntary national greenhouse gas reduction target of between 36.1 percent and 38.9 percent of projected emissions by 2020.

Like other developing countries, Brazil faces the dual challenge of encouraging development while reducing GHG emissions. President Lula echoed this concern in his introduction to the PNMC, stating that actions to avoid future GHG emissions should not adversely affect the development rights of the

35

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

As the world’s largest tropical country, Brazil is unique in its GHG emissions profile. In prior decades, the availability of large volumes of land suitable for cultivating crops and pasture helped to transform agriculture and livestock into key sectors for sustaining the country’s economic growth. In the past decade alone, these two sectors accounted for an average of 25 percent of national GDP. The steady expansion of crop lands and pasture has also required the conversion of more native land, making landuse change the country’s main source of GHG emissions today. At the same time, Brazil has used the abundant natural resources of its large territory to explore and develop renewable energy, having built numerous large hydropower plants and scaled up bioethanol production as a gasoline substitute, which accounts for the low carbon intensity of its energy matrix.

36

poor, who have done nothing to cause the problem. Recognizing the need for a low-carbon pathway to growth, Brazil has chosen to benefit from the Clean Development Mechanism (CDM), an innovative financial mechanism originally proposed by Brazil, which is defined in Article 12 of the Kyoto Protocol. To date, Brazil has initiated more than 300 projects under the CDM.

1.1

Context of the Low-carbon Study

Objective and Approach of the Low-carbon Study

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

To support Brazil’s integrated effort to reduce its GHG emissions and promote longterm economic development, this study seeks to construct a transparent and internally consistent Low-carbon Scenario that could be used by the Brazilian government as a tool for evaluating the elements necessary for building a low-carbon path towards growth. This study on Brazilian emissions is one of the five case studies focusing on specific countries that contribute to the preparation of the Clean Energy Investment Framework (CEIF).

The study emphasized two main aspects: first, it used information from the literature and from existing studies as much as possible to effectively leverage the wealth of information. Second, the process emphasized a consultative and iterative approach that involved extensive discussions and exchanges of information with specialists in the field and representatives from the Brazilian government. The team researched the literature exhaustively, and, in a thorough consultative process, met with over 70 acclaimed Brazilian specialists, technicians and government representatives. The consultative process, combined with the Bank’s comprehensive knowledge of Brazilian institutions, enabled the team to create partnerships with centres of excellence that are recognized for their national and international expertise in the sectors concerned.

General Approach of the Methodology used in the Study

The study team analyzed the existing opportunities in each of the four sectors identified as the main GHG emitters: land use, land-use change, and forestry (LULUCF); energy; transport; and waste. This summary report only presents the part on land use, land-use change, and forestry. In the complete study, the team created a Reference Scenario for all four sectors until 2030 based on current projections and available models for each sector. For the energy and transport sectors, the team used existing long-term national and urban plans. However, due to a lack of similar plans for LULUCF and waste management, new models and equations were developed consistent with macroeconomic and demographic projections for the energy and transport sectors until the year 2030 . For the LULUCF sector, the team used two complementary models: (i) Land Use Model for Brazil (BLUM); and (ii) SIM Brazil, a geo-referenced spatialization model for allocating land use for specific locations and years, developed by the Remote Sensing Centre (CSR) of the Federal University of Minas Gerais (UFMG). For the waste management sector, the team worked with the Environmental Agency of São Paulo State (CETESB) to develop sets of equations for modeling waste disposal.

The study then evaluated the mitigation and carbon uptake options, assessing all the relevant sub-sectors for each sector; determined the viability of the options investigated; and finally, constructed Low-carbon Scenarios for each sector.

1.2

Approach of the LULUCF Summary Report

37

The next step was to analyze the mitigation and uptake options based on an analysis of options for reducing pressure from deforestation and protecting the forests, for mitigating emissions from agriculture and livestock, and for carbon sequestration. An economic analysis was also conducted to reduce the costs of the options proposed. For these analyses, the team adapted the “cunha” concept developed by Pacala and Socolow (2004), which increases the scale of a particular area or technology to ensure significant reductions in the GHG emissions that can be deduced from the Reference Scenario. Due to the systemic nature of the LULUCF sector, the team concluded that only using the “cunha” approach was not sufficient. For this sector, they analyzed the country’s potential for carbon uptake on a large scale and for avoiding GHG emissions in other countries through greater ethanol export. For some sub-sectors, including deforestation and land use, the team needed to make new projections whose results would be significantly different from the Reference Scenario, although the same premises would be used (for example: demand, inflation, and fuel price forecasts). An evaluation of the viability of the options identified was then conducted, including barriers that limit or prevent the implementation of the options analyzed and measures for overcoming them, and environmental and economic benefits.

Lastly, a Low-carbon Scenario was developed based on the projection of new land use and land-use change (including the additional extension of land necessary for mitigation and uptake options), on the estimate of the reduction of deforestation, and on projections of emissions reductions. Due to the limited resources, this study did not have a fifth phase, which is still necessary to evaluate the sustainability of the Lowcarbon Scenario, including its macroeconomic impact.

This report is divided into five parts: an executive summary and an introduction on the low carbon study and the main questions pertaining to land-use related GHG emissions; a chapter on the LULUCF Reference Scenario; a chapter on the LULUCF Lowcarbon Scenario; and an analysis of costs for transitioning from the Reference Scenario to the Low-carbon Scenario proposed.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The report summarizes specific studies on emissions resulting from deforestation, agriculture, livestock, and ethanol and charcoal production for the iron and steel manufacturing sector. To develop the LULUCF low-carbon scenario, land use and land-use change were projected in a way that was consistent with the projected liquid and solid biofuels; developed geo-spatial models for soil use; made projections of deforestation, adapting existing modeling exercises; and emissions projections.

2. Reference Scenario

Not surprisingly, the land use, land-use change, and forestry (LULUCF) sector accounts for more than two-thirds of Brazil’s gross CO2e emissions. Of this amount, approximately two-thirds are the result of deforestation, with the remainder being from agricultural production and livestock activities. Conversion of forest land for other land uses results in GHG emissions from the soil, while the digestive process of ruminants results in methane (CH4) emissions. A key challenge for the sector is to identify opportunities to curb the net balance of GHG emissions from deforestation while fostering economic growth.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

38

Brazil’s forests represent an enormous carbon stock. The Amazon, a reservoir of about 47 billion tons of carbon, permanently sequesters more than five times the amount emitted globally each year. At the same time, in 2010, Brazil was the world’s second largest emitter of carbon dioxide (CO2) resulting from deforestation—often driven by the need to convert land for agricultural production and livestock pasture.

This chapter describes the background and development of the LULUCF Reference Scenario.

2.1 Emissions from Land Use, Land-use Change, Deforestation, Agriculture and Livestock 2.1.1 Effects of Land Use and Land-use Change on Emissions

There are three major ways that land use and land-use change contribute to carbon emissions: (i) conversion of forest land for other land uses (agriculture, grassland, settlements, etc.), (ii) agricultural production, and (iii) livestock activities. In addition, carbon uptake via reforestation activities affects net GHG levels.

2.1.1.1 Deforestation

According to the results of this study, deforestation was responsible for 40 percent of Brazil’s gross emissions in 2008. When forest biomass is destroyed, mainly by fire and decomposition, carbon is emitted into the atmosphere. Brazil has been converting forested areas at a rapid pace (approximately 420,000 km² over the past 20 years). The Amazon lost approximately 18 percent of its original forest cover between 1970 and 2007, the Cerrado lost about 20 percent of its original area between 1990 and 2005, and the Atlantic Forest lost approximately 8 percent over the same period (INPE 2009). Between 1990 and 2005, Brazil’s carbon stock was reduced by 6 billion metric tons, largely as a result of deforestation , an amount that is the equivalent of one year of global emissions, with all sources combined. Since peaking at 27,772 km² during the 2004-2005 period, Brazil’s de-

While the spatial dynamics of livestock and agricultural expansion in the Amazon determine the pattern of deforestation at the regional level, deforestation is also affected by more wide-ranging dynamics. National and international market forces drive the development of Brazil’s meat and crop sectors. Depending on price trends, an array of agricultural and livestock activities compete for land. Many geographical studies have shown that the resulting spatial dynamics are on a national scale. Over the past three decades, soybean cultivation has expanded over 1,500 km from south to north (de Gouvello, 1999). A recent geo-statistical analysis shows that livestock-related activities are the primary reason for the conversion of forest areas, followed by the expansion of agricultural production and other phenomena, including migration, opening of paved roads, and land speculation as the main drivers of deforestation (Soares-Filho et al., 2009).

2.1.1.2 Agricultural Production

GHG emissions from agricultural production are caused mainly by changes in soil carbon stocks, and to a lesser extent by fertilizers and residues, cultivation of wetland irrigated rice, burning of agricultural residues, and use of fossil fuels for agricultural operations. According to the results of this study, direct emissions from agriculture accounted for about 6 percent of gross national emissions in 2008. Variations in soil carbon stock correspond to the loss of organic matter in the soil as a result of a particular land use.

2.1.1.3 Livestock Activities

The main source of livestock emissions in Brazil is methane (CH4) from the digestive process of ruminants. According to the results of this study, direct emissions from livestock activities accounted for about 18 percent of gross national emissions in 2008. Livestock emissions are related predominantly to beef-cattle farming. According to the Initial National Communication to the United Nations Framework Convention to Climate Change, methane emissions from the beef-cattle subsector were responsible for over four-fifths of the total amount of enteric emissions caused by Brazilian livestock in 1994. Thus, this study emphasized emissions from, and mitigation alternatives for this subsector.

39

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

forestation rates have declined sharply to 11,200 km² in 2007, the second lowest historical annual rate estimated by the deforestation assessment program (PRODES) since the year 1988, according to INPE (2008) . This trend continued the following years, a partial reflection of the higher valued Brazilian currency, the Real (R$), compared to the U.S. Dollar (US$), which has made export-based production less profitable. Implementation of the Plan of Action for the Prevention and Control of Deforestation in the Legal Amazon (PPCDAM) improved the enforcement of environmental laws through an increase in monitoring capacity, and more rigorous conservation policies for the Amazon rainforest, have contributed to this reduction.

2.1.1.4 Forestry-based Carbon Uptake

40

Apart from GHG emissions sources associated with land use and land-use change, trees remove CO2 from the atmosphere and store it in the trunk, branches, leaves, flowers and fruits, thus generating negative emissions. In Brazil, carbon uptake occurs mainly in the natural re-growth of degraded and production forests. According to the results of this study, it was estimated that forestry-based carbon removal offsets about 4 percent of national gross emissions in 2008.

2.1.2 Land Use and Land-use Change Simulation Methodology

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Exploring options for mitigating deforestation emissions first requires projecting future deforestation, which, in turn, requires simulating future land use and land-use change. To establish the Reference Scenario, the study developed two models: i) Brazilian Land Use Model (BLUM) (Box 1) and (ii) Simulate Brazil (SIM Brazil) (Box 2). These complementary models were used sequentially. BLUM projected land use and land-use change through 2030. SIM Brazil then allocated this land use and land-use change to specific locations and years, but it was necessary to first determine the area available for the expansion of agriculture and livestock activities.

2.1.2.1 Area Available for the Expansion of Productive Activities

Land use and occupation in Brazil were characterized using a combination of remote sensing equipment and secondary data, measuring area allocated for pasture and area available for the expansion of farmland, and estimating the area of each municipality with environmental liability. The characterization of the area allocated for pasture was essential as it represents the stock of land already converted for productive purposes that could be used for farms and forests if these sectors expand. As the productivity of pastures in Brazil is very low, the intensification of pasture area is one of the most important ways to make the expansion of farmland and production forests viable without affecting the agricultural frontier. The area available for productive use (farm, livestock and production forests) was defined, assuming that there would not be any additional deforestation, in other words, considering only the area available for pastures that could be converted for other uses (or more intensive use), considering those pastures in areas with impediments not suitable for farmland (Figure 2).

Figure 2: Calculation of available area for the expansion of productive activities

Source: UFMG

The total amount of area suitable for the expansion of agriculture and production forests that does not need to be cleared is 126 million hectares. If one excludes the Amazon biome and other forests, this total is 89 million hectares. These amounts (Figure 3) apply to pasture in areas without impediment that are suitable for agriculture and forests. Three pieces of information from this analysis are essential for the economic land-use model: total pasture area; available area for the expansion of agriculture and production forests; and calculation of the area that needs to be reforested for the legal scenario.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

41

Figure 3: Land use by class, excluding the Pampa, Caatinga and Pantanal biomes

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

42

Source: UFMG

For the land-use projection model, the pasture area to be converted for agriculture corresponds to a restricted maximum area to be occupied by agriculture and forests in the projections. This ensures that projections for each micro-region do not result in an expansion of farmland and forests beyond the amount of pasture available. Thus, information on the area available is relevant for projections for the Low-carbon Scenario, which is based on the assumption that any agricultural and forest expansion cannot cause deforestation, and would need to be accommodated in pastures that are suitable for these activities. The convertible pasture area was also important for obtaining data projected for micro and macro-regions. Each micro-region was restricted by the maximum area allowed for agriculture and forests, which was defined by the area of pasture to be converted.

However, information on convertible pastures was not used in Reference Scenario projections, as it did not reflect land allocation for the expansion of farms and forests only in the pasture area. In other words, the Reference Scenario considered that any additional projected demand for land would lead to a conversion of residual vegetation by micro-region in areas without impediment (Table 3). In this scenario, projections of total expansion area (the sum of the farm areas, pastures and production forests) cannot exceed the area of residual vegetation as presented below. Additional demand for land in the Reference Scenario was much less than the available land with residual vegetation.

Table 3: Comparison between total pasture area and area of residual vegetation convertible into farmland/forests in the regions of the BLUM model (1000 ha)

South

Total Pasture

Convertible Pasture for Farms/Forests

Residual VegetationConvertible into Farms/Forests

18,146

5,681

6,721

Southeast

44,053

30,335

Northeast Coast

10,801

0

Central-West Cerrado Northern Amazon

MAPITO and Bahia

Brazil

51,200

52,551

32,138

208,889

42,553

16,415

30,114

39,079

167,017

126,014

260,586

8,365

43

0

40,319

2.1.2.2 Economic Land-use, Agriculture and Livestock Modeling: BLUM Model As part of an institutional partnership for the Low Carbon Study, the Institute for International Trade Negotiations (ICONE) developed a land-use projection model for Brazil – the BLUM (Brazilian Land Use Model) together with the Food and Agriculture Policy Research Institute (FAPRI) of the Center for Agricultural and Rural Development (CARD) at the University of Iowa. For the analyses, the model divides the country into six main regions based on their homogeneity in the production and marketing of agriculture and livestock, as well as the division of biomes (Map 1): (1) South – Paraná, Santa Catarina, Rio Grande do Sul; (2) Southeast – São Paulo, Rio de Janeiro, Espírito Santo, Minas Gerais; (3) Central-Western Cerrado – Southern Mato Grosso, Goiás and part of Mato Grosso within the Cerrado and Pantanal biome; (4) Northern Amazon – Part of Mato Grosso in the Amazon biome, Amazonas, Pará, Acre, Amapá, Rondônia, Roraima; (5) MAPITO and Bahia – Maranhão, Piauí, Tocantins, Bahia and (6) Northeast Coast – Alagoas, Ceará, Paraíba, Pernambuco, Rio Grande do Norte, Sergipe. Earlier projections obtained in the six regions were divided into micro-regions of the IBGE. This division is necessary for calculating the balance of GHG emissions of the livestock sector and for the spatialization of the results.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

BLUM Region

Map 1: Map of the Main Regions of the Land-use Model

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

44

Source: Based on data from IBGE. Artwork: ICONE

The BLUM is divided into two modules that are interlinked to make land-use projections: (1) supply and demand and (2) land allocation. The first module is based on partial balances of the supply and demand of selected products for each year. Demand consists of three components: domestic demand, net exports (exports minus imports) and final stock (only the demand for milk, eggs and meats does not include the final stock variable). Supply is made up of two components: production and initial stock (this is also only for grains and sugar cane and its derivatives17).

The quantities supplied and demanded are calculated simultaneously based on the microeconomic principle of market balance, according to which the supply and demand of each product are equal. This balance occurs when there is a price that leads to the convergence between the supply and demand during the same period of time. The model uses Microsoft Excel 5.0 as a platform for its operations and the price is adjusted 17

In the case of sugar cane, only the stocks of its derivatives were considered: sugar and ethanol.

annually depending upon the excess demand for each product. The process continues until a balance is achieved and the excess is zero.

The demand for each product is estimated nationally based on econometric equations. The explicative variables of the domestic demand were generally: income per capita, population, price of the product in Brazil and trends, among others, with these variables considered differently for every product. For the demand for beef, for example, the domestic prices of competing meat, such as chicken and pork were also considered in the consumer’s decision. For net exports, global economic growth, domestic prices in US dollars and, in some cases, domestic production and the international fuel price were considered explicative variables in the equations.

45

• The equilibrium price is obtained when the supply and demand are equal for a specific year and product. In this way, prices, demand and supply are endogenous to the model. The shocks given to the model in the Low-carbon Scenario are introduced exogenically via supply or demand. In the case of ethanol, for example, as will be discussed later, a shock is given to exports, and new balances in the market are observed for all products.

• The land allocated for each activity and year is the result of the market equilibrium. For operations, the area of a farm in a given region and for a specific period is a function of the expected profitability, which, in turn is calculated based on productivity, the projected cost for that year and the price of the previous year. • The model works with prices for producers and consumers following the same tendency over time. This means that the change in demand in relation to price variations is based on prices estimated by producers.

• The model assumes perfect availability of resources for investments and working capital, which means that it is not impacted by a credit crisis for the supply and demand. For the results used in this project, given that 2009 was a credit crisis year, some specific adjustments were made for the 2009 production with the aim of reproducing the expectations for this year more precisely. • The farms’ regional productivity and Total Recuperable Sugar factor (Açúcar Total Recuperável - ATR) are projected as tendencies over time. The model is still not ready to capture climatic impacts or different levels of fertilizer use in productivity.

• Prices are established nationally and transmitted to regions using price transmission coefficients estimated by regressions. Although it is not the object of this study, the impact of transport infrastructure improvements on regional production may also be evaluated. • Production costs were divided into three categories: fixed, variable and transport. However, the costs of animal products only include variable costs. • The Reference Scenario projects the evolution of the cattle herd, assuming that there will be no significant productive gains over time through improvements in zootechnical indices such as age at the time of slaughter and rate of animal repositioning. However, since the model has no endogenous zootechnological

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In sum, the model is based on the following central hypotheses:

indices, the equations of profits and the repositioning of bullocks are altered to adjust their responses to price variations.

One of the decisions related to methodology was the choice and selection of products covered by the BLUM Model, bearing in mind that it would be impossible to determine supply and demand frameworks for all of Brazil’s agricultural products, and that there is a concentration of land use for some products. Calculations for designing the land-use model take into consideration the total demand for products; demand for grains; domestic demand for cotton, rice and beans; domestic demand for corn; domestic demand for soybean, soybean meal and oil; net cotton exports, soybean meal and oil; net exports of corn, rice and beans; demand for ethanol and sugar; demand for beef, pork, chicken and eggs; demand for milk and dairy products; national supply and production of each product; production of corn, soybean, soybean meal and oil, cotton, rice, beans, sugar and ethanol; allocation of planted area; allocation of area with grains, corn, soybean, cotton, rice and beans; beef supply; projections for the herd, slaughter and average weight of beef cattle at time of slaughter; pasture area; supply of pork; and supply of chicken and eggs. Table 4 presents a series of land-use data for products covered by the BLUM model and Table 5 summarizes the sources of data and information used by ICONE in the land-use model.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

46

Land-use projections for agriculture and livestock were made using the BLUM Land-Use Projection Model, which is an econometric model that operates at two levels: national supply and demand for every final product, and regional supply and area allocated for agricultural products, meaning that a set of parameters was estimated based on a temporal data base. The parameters are generally: price elasticity of demand and income elasticity, price elasticity of supply, and cross elasticity.

Table 4: Brazil - Area allocated and production of products covered by the BLUM model Area Allocated (ha) Cotton Rice Beans – 1st crop Beans – 2nd crop Corn – 1st crop

Corn 2nd crop winter Soybean Sugar cane Production forest Pasture Total

Production (1,000 ton)

2006

2007

2008

2006

2007

2008

2,694

3,052

2,857

1,893

2,106

1,991

844

3,018

1,529

9,632

3,332

22,749 6,179

5,269

1,080

2,967

1,035

9,421

4,634

20,687 6,964

5,455

1,066

2,881

1,143

9,656

5,052

21,334 8,235

5,874

2,724

11,722 1,578

31,332

11,183

55,026

3,899

11,316 1,234

36,311

15,059

58,392

457,246 549,905 n.a.

208,889 206,323 205,381 n.a.

264,136 261,618 263,479 n.a.

n.a.

n.a.

n.a.

Source: IBGE; CONAB; UFMG/ICONE/EMBRAPA. Note: n.a. = not applicable

4,108

12,108 1,523

39,922

18,664

60,052

687,758 n.a.

n.a.

n.a.

Table 5: Data sources

Instituto Brasileiro de Geografia e Estatística – IBGE (Brazilian Institute of Geography and Statistics)

Companhia Nacional do Abastecimento - CONAB (National Commodities Supply Corporation)

Data utilized

Reference

Beef cattle herd, swine herd, slaughter of fowl, www.ibge.gov.br swine and beef cattle. Population estimate.

Planted area, area harvested, prices, costs, supwww.conab.gov.br ply (balance of supply and demand).

Empresa Brasileira de Pesquisas Agropecuárias – EMBRAPA Prices and production of www.embrapa.gov.br (Brazilian Agricultural Resear- swine and fowl. ch Corporation) Ministério do Desenvolvimento, Indústria e Comércio International commercial www.mdic.gov.br – MDIC (Ministry of Develop- data. ment, Industry and Trade)

Instituto de Pesquisa Econômi- Macroeconomic data from ca Aplicada – IPEA (Institute for Brazil and the agriculture www.ipeadata.gov.br Applied Economic Research) and livestock sector.

Centro de Estudos Avançados em Economia Aplicada – CE- Accompanies prices and www.cepea.esalq.usp.br PEA (Center for Advanced costs. Studies in Applied Economics)

Banco Nacional do DesenvolviCredit data and investmento Econômico e Social – ments in the sugar alcohol www.bndes.gov.br BNDES (National Bank for Ecosector. nomic Development)

Agroconsult

Scot Consultoria

47

Costs and productivity of www.agroconsult.com.br sugar cane farms and fac(Fábio Meneghin) tories.

Stratification of beef cattle www.scotconsultoria.com.br herd, beef cattle slaughter, (Maurício de Palma Nogueiprices and profitability of ra) livestock.

União da Indústria de Cana-deAçúcar – UNICA (Brazilian Sug- Sugar and alcohol market. ar Cane Industry Association)

www.unica.com.br

Associação Nacional dos FabriAnnual vehicle sales per cados de Veículos Automotores fuel type, scrappage cost www.anfavea.com.br – ANFAVEA (National Autocurve. makers Association)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Source

48

Agência Nacional do Petróleo, Gás Natural e Biocombustíveis Prices of gasoline, diesel, – ANP (National Petroleum, www.anp.gov.br energy market. Natural Gas and Biofuel Agency) Food and Agricultural Policy International microecowww.fapri.org Research Institute - FAPRI nomic data and modeling.

Instituto Rio Grandense do ArRice market. roz – IRGA

Associação Brasileira das Indústrias de Milho – Abimilho Corn market. (Brazilian Corn Industries Association)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Associação Brasileira das Indústrias de Óleos Vegetais – Soybean market. ABIOVE (Brazilian Plant Oil Industry Association)

Sindicato Nacional das Indústrias de Alimentação Animal Animal feed market. – Sindirações (National Union for Animal Feed Industries) Leite Brasil (Brazil Dairy)

Data on dairy cattle herd.

União Brasileira de Avicultura – UBA (Brazilian Aviculture Data on aviculture. Union)

Associação Brasileira da Indústria Produtora e Exportadora de Carne Suína – ABIPECS Data on swine herd. (Brazilian Association for the Port Production and Export)

www.irga.rs.gov.br www.abimilho.org.br www.abiove.gov.br www.sindiracoes.org.br www.leitebrasil.org.br www.uba.org.br

www.abipecs.org.br

PNE data – export and Empresa de Pesquisa Energé- consumption of ethanol, tica – EPE (Energy Research demand for biodiesel from www.epe.gov.br Company) soybean, GNP Brazil and Global GNP and fuel price. Source: ICONE

Macroeconomic projections: Land-use projections are based on a macroeconomic scenario, which shows the trends of the Global GNP, Brazilian GNP, Brazilian population, inflation, exchange rate and fuel price for a 22-year period: 2009 to 2030. For the Reference Scenario, GNP projections, fuel price, and exchange rate refer to the B1 scenario, “Surfing the Marola” of the National Energy Plan 2030 (PNE-2030). According to PNE projections, Brazil should grow 3.7 percent between 2009 and 2020, and 4.5 percent from 2021 to 2030. The overall GNP growth rate will be 3 percent per year for the

entire period projected. For projecting the country’s population, data from the Brazilian Institute of Statistics (IBGE) were used. The macroeconomic scenario considered is important, as it entails components of demand and cost equations. Table 6 summarizes the macroeconomic scenario used for 2006, 2008, 2018, and 2030. Table 6: Macroeconomic projections Unit

GNP Brazil

% per year

Global GNP

% per year

Population Brazil

Million

Fuel Price

US$/barrel

Nominal Exchange Rate

R$/US$

Inflation Rate

% per year

2006

2008

5.39%

3.53%

3.70%

4.50%

67.00

63.50

53.07

42.67

4.07%

186.77 2.17 4.72

2.48%

191.87 1.66 6.02

2018

3.00%

214.94 3.35 3.36

Source: National Energy Plan (PNE 2030) and ICONE

2030

3.00%

236.74 4.77 2.46

2.1.2.3 Allocation of Area for Agriculture and Livestock Activities The land allocation module for the different products and regions is a component used to estimate the production of every product (grain and sugar cane) in each region, thus determining one of the components of the Brazilian supply of each product. Equations for the allocation of area are estimated for grain and sugar cane for each region and, considering that productivity per hectare was estimated as a tendency, the production of each of these products is the result of the increase in the productivity and area of each region. Brazilian production of each product is the result of the sum of the regional production of that product. Beef production is calculated based on estimates of the number of animals slaughtered and the estimated average weight of the carcasses. The allocation of area in each region in the case of grains and sugar cane was estimated considering the regional profitability of each crop and the competing crops as explanatory variables (which are negatively related). This means that the regions that present the greatest returns expected for each product will have a greater allocation of land for that product. In addition, area allocation equations for most grains and sugar cane consider the actual area estimated during the previous period as the explanatory variable, thus avoiding major variations in the estimated areas.

Equations for the allocation of pasture area were structured differently for grains and sugar cane. The size of the area allocated for pasture in each region is obtained based on the areas used for other crops (and not expected profitability) and the estimated development of the herd. This option was chosen after different attempts to estimate pasture land were made. As there are different technological levels and production systems, allocating the area based on the profitability of the livestock did not produce satisfactory results. It was thus decided to estimate the pasture areas as described above. In addition, since there is no historical model for pasture areas in Brazil,

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Variable

49

ICONE put together a historical series based on the annual series of beef cattle herds by region (IBGE, 2008b) and thus the area was projected based on the variables mentioned above, given that this is also an innovative model in that respect.

To specify the explanatory variables of the land-use equations for each product and region, which determine land allocation by product in each of the six regions, a land competition matrix was developed, as described in Table 7. This matrix was defined based on agricultural aptitude criteria (EMBRAPA, 2008a; EMBRAPA, 2008b; NIPE/ CGEE, 2005) and the trends of the planted areas observed between 1997 and 2008 (CONAB, 2008; IBGE 2008a). The area of each crop corresponds to its own expected profitability and the expected profitability of competing crops (through cross-price elasticity). It should be emphasized that the historical returns also show activities that “take” and “give” land in land competition models. Livestock, for example, is historically an activity that “gives” land in all regions, which means that it can be assumed that farm areas compete with pasture areas, but not vice versa.

Region and Competing Product

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

50

The area allocated for production forests constitutes exogenous projections for the land-use model, based on projections from the iron and steel and paper and cellulose sectors of the Grupo Plantar and the National Energy Plan – NEP/2030 (Brazil, 2007).

Table 7: Land competition matrix in Brazil

Product (dependent variable) Cotton

South

Southeast

Sugar cane

Soybean

Soybean Corn

Sugar Cane Corn

Soybean Corn

CentralSoybean Soybean West CerCorn Corn rado Northern Soybean Soybean Amazon Corn Corn Northeast Coast

MAPITO and Bahia

Soybean Corn Rice Bean

Corn Sugar Cane Soybean

Sugar Cane Corn

Sugar Cane Soybean

Corn

Soybean

Cotton Sugar Cane Corn

Corn Cotton Corn

Rice

Cotton Sugar Cane Soybean

Cotton Soybean Bean

Source: ICONE

Beans Soybean Corn Rice

SoybeanCorn Soybean Corn

Corn Soybean

Soybean Corn Rice

Soybean Corn Rice

Soybean Corn Rice

Soybean Corn Rice

Soybean Corn Rice

Pasture Corn Soybean Rice Bean Sugar Cane Corn Soybean Bean Sugar Cane Corn Soybean Cotton Bean Sugar Cane Corn Soybean Rice Bean Bean Sugar Cane Corn Soybean Cotton Bean

The co-linearity observed between the historic series of profitability of some crops, such as soybean and corn for example, which should also be highlighted, made it necessary to exclude one of these variables or to create a new variable in the land-use equations for certain regions. Specifically, for regional projection equations for areas planted with sugar cane, the profitability of soybean as an explanatory variable had to be excluded. On the other hand, for estimating pasture area, it was decided to add up areas of soybean and corn, creating a new explanatory variable for the land-use module. Since equations for the allocation of area for grain and sugar cane were based on the crops’ expected profitability, the greatest profitability compared to other crops and pastures will lead to the expansion of the area allocated. For example, consider a hypothetical situation where the most profitable crop in a specific region is sugar cane, followed by soybean, maize and lastly beef cattle, with an increase in demand for all of these products. Even with competition between the crops due to their profitability, there will be an increase in the areas allocated for crops, with the exception of pasture area. This is due to the greater potential to increase the productivity of beef cattle compared to other products, which is done by decreasing pasture size and maintaining – or increasing – herd size. There is thus no completely proportional compensation for area between grain and sugar cane if there is pasture that can be converted for agricultural purposes. In other words, the area estimated for a given crop can be reduced in one region and increased in another, but this has to do with profitability, and is not the result of a compensation process between the different regions.

In the case of areas allocated for pasture, the expansion of the area for grains and sugar cane inevitably leads to a reduction in pasture area if an increase in the size of the cattle herd doesn’t preclude it. However, history has shown that pasture area does not normally increase in regions with strong competition for land from grains and sugar cane. In fact, the opposite holds true in productive regions that are not traditionally used for the abovementioned products. Thus, if the demand for beef increases and if there are regions with stable or decreasing herds, implying a reduction in pasture area, the size of the herd will inevitably increase in regions along the agricultural frontier, leading to an increase in pasture area. It should be noted that a reduction in pasture area in certain regions will only lead to an increase in pasture area on the agricultural frontier if there is an increase in the size of the beef cattle herd.

The BLUM treats increases in on-farm productivity as a tendency that reflects past gains. For a more substantial gain in productivity compared to the Reference Scenario, it would be necessary to consider technological changes in the model (such as exogenous technological shocks), which implies changes to the structure of the entire farm in terms of production costs. This is a very strong hypothesis, as productivity levels of Brazilian and international farms are comparable, and profits will have already been

51

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

It is important to note that the regions are independent and equations for the allocation of area for grains, sugar cane and pasture are different in each region. However, as total production of each product should equal total demand, if a crop’s area is reduced in one region, the price of the product derived from this crop tends to increase, and the other region then makes up for this effect by expanding the area allocated for the crop. The model’s rationale was based on the principle that equilibrium prices of supply and demand would determine crop profitability in each region and consequently the area allocated for each crop (in the case of grain and sugar cane).

incorporated in the future as a linear tendency in time.

Considering that the BLUM estimates the allocation of area for the six large regions described above, individual crop substitution levels for short periods of time are not captured. The model’s objective is to estimate the allocation of area as a function of the competition between crops and pasture area. Projections of area allocation thus measure land-use change resulting from the supply and demand dynamic for all the products that compete for land.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

52

Moreover, the large volume of pasture area in Brazil, some of which is relatively unproductive, shows potential for agricultural expansion. Thus, livestock production may be increased by using better zootechnical indices and improving pasture quality, while reducing the area used. Since this is the most realistic technological change for farms, expansion should occur in pasture areas. In other words, according to microeconomic theory, considering, in simple terms, a technological tree with capital and land factors, and that livestock uses a relatively large amount of land and little capital, a slight variation (increase) of the capital factor would cause a more than proportionate change (reduction) to the land-use factor18. Furthermore, by improving livestock productivity, we will be working on the existing Production Possibilities Frontier (PPF/FPP), which is different with grains, where it would be necessary to use technological innovations that are not yet available, thereby displacing the PPF. Thus the preference for focusing on beef cattle in Brazil for most of the gains in productivity would be justified.

The relationship of cause and effect due to the expansion of one crop over another over time and implications from substituting between grains, sugar cane, and pastures can be measured a posteriori using the results of the model if certain assumptions are made. Given the fact that the equations for the allocation of area in one region are independent from other regions, a set of assumptions that relate land-use changes in traditional regions to those on the frontier must be established, which is the only way to measure the indirect effect of land-use change. However, it is important to emphasize that to analyze the expansion of the agricultural frontier and measure the indirect effect of a specific crop, two different aspects must be considered. First of all, there is an increase in the area on the frontier due to the loss of crop area in other regions, which could be considered the actual indirect effect of the crops. Secondly, the frontier expands to some extent as a result of pasture development due to the insufficient increase in beef cattle production despite increases in the demand for meat.

In addition, the model’s results could be used to measure the indirect effect of allocating pasture area. This will only occur if the beef cattle herd is redistributed between regions as a result of the expansion of other agricultural activities, and after discounting increases in productivity due to the intensification of beef-cattle grazing. If this redistribution does not occur, there will be no indirect effect.

An important result from the model has to do with the total area used for agrosilvipastoral activities (considering the products selected in this analysis). If this area increases over time, areas with native vegetation will be converted into productive areas. This surplus in area allocation is the result of a combination of two factors: (a) an 18

For this affirmation, the existence of the basic concept of decreasing marginal yield should be recognized.

increase in the size of the cattle herd in regions on the agricultural frontier (Northern Amazon, MAPITO, and Bahia), with a simultaneous reduction in traditional agrosilvicultural areas, which could be interpreted as an indirect effect, and (b) the expansion of crops on the agricultural frontier, which is a direct effect. Besides land competition, there are interactions between the sectors analyzed, as well as between products and sub-products. For example, between the meat and grain sectors, the demand for feed from the meat, milk, and egg supply (corn and soybean meal, basically) is one of the components of the domestic demand for corn and soybean. In the case of the soybean complex, soybean meal and oil are components of the domestic demand for soybeans, which is determined by the crushing margin. Similarly, ethanol and sugar are components of the demand for sugar cane. The methodological diagram below (Figure 4) summarizes the dynamic of the land-use model developed for this study.

53

Source: ICONE

As mentioned earlier, results obtained in the model (in the six main regions) were divided into IBGE micro-regions. The criteria used were based on the history of the planted area for each product selected, considering data on the limits of land available for the expansion of productive activities.

2.1.3 Land-use Reference Scenario

The Reference Scenario was developed based on the BLUM land-use projection model, using the Brazilian agriculture and livestock expansion pattern observed in the past. Thus, with this scenario, there are no exogenous shocks for any variable considered in the model. The Reference Scenario serves as a basis for comparison with alternative scenarios that consider the expansion pattern of livestock, agriculture, energy and transport sectors with lower levels of GHG emissions (Low-carbon Scenario).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 4: Methodological land-use diagram

According to the PNE 2030, internal ethanol consumption increases significantly during this period, going from 12.8 billion liters in 2006 to 59.2 billion in 2030. Exports should already reach a maximum of 15.8 billion liters in 2020, and drop to 13 billion liters in 2030. In the case of biodiesel, according to the PNE 2030 scenario, and based on 2007 observations, diesel consumption will increase approximately 228 percent, going from 42,784 thousand tons in 2007 to 97,876 thousand tons in 2030. In addition, the minimum percentage of the biodiesel mixture in diesel oil will go from 2 percent in 2008 to 12 percent in 2035. Soybean participation in biodiesel production should drop from 88 percent in 2008 to 35 percent in 2035. The result of this scenario is the production of 802.9 thousand tons of soybean biodiesel in 2008, which will increase to 4,133 thousand tons in 2030.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

54

It should be emphasized that the demand for ethanol and biodiesel, as well as liquid ethanol exports are exogenous to the model. The consumption scenario for these types of energy was extracted from the National Energy Plan 2030 (PNE 2030), produced by the Energy Planning Company (EPE) in 2006 and concluded in April 2007 (Brazil, 2007). Given the availability of the most recent data during the development of the current project, the amounts projected in the PNE were updated until the 2008 harvest, after which the variation projected in the PNE 2030 was adopted. Ethanol stocks and production are endogenous to the model (see Section 3.2).

Land allocation for planted forests is determined exogenically from the model and represents a restriction on the growth of the other crops. Total area occupied by forests is based on PNE projections, which state the total area occupied by eucalyptus, pine and other tropical woods for all of Brazil in 2010, 2015, 2020, and 2030. To calculate the size of the area from year to year, the constant increase between the periods of time was monitored. Based on the historical series of area occupied by planted forests in each region (1997-2007), the Brazilian projection was divided by region. The participation of each region in Brazil counted as much as the history of the growth of such participation. Total area occupied by planted forests in Brazil would go from 5.2 million hectares in 2006 to 8.45 million in 2030, a 60 percent increase. In terms of regional dynamics, the area that stands out is the southern region, which would more than double during that time, reaching 3.7 million hectares in 2030, surpassing the Southeast and becoming the largest region. Another noteworthy locality is MAPITO and Bahia, whose area would grow 124 percent, reaching 1.5 million hectares in 2030 (Table 8).

Table 8: Projection of areas occupied by production forests (million ha) Regions of the BLUM Model

2006

2008

2030

South

1,670

1,914

3,712

Northern Amazon

140

149

167

Central-West Cerrado Northeast Coast

MAPITO and Bahia Brazil

2,452

2,669

319

2,493

374

-

533

-

688

-

768

5,269

1,545

5,874

8,450

Source: PNE, ICONE

Areas considered available for agricultural expansion in the Reference Scenario were those that could be converted into pasture and areas with residual vegetation. Only pastures and residual vegetation without impediments according to the UFMG classification were considered, meaning no legal impediments (CUs and TIs), accentuated slopes or unsuitable soils. However, the legal impediments of the Permanent Preservation Areas (PPA) and Legal Reserves (LR) were not taken into consideration.

Results of general land-use projections were produced for each large region of the BLUM model for agriculture, pastures and production forests in the Reference Scenario. Although supply and demand projections are not presented here, it is important to emphasize that they are part of the model’s exit data and are determining factors in overall land allocation in Brazil for each activity. As shown in Table 9, the projected demand for land in Brazil for the year 2018 for the products analyzed will be 263.2 million hectares, meaning that there will be an increase of 1.7 percent in relation to the 259.3 million hectares used for the same products in 2006. This increment is even greater for 2030, with an increase of 6.5 percent in total agricultural area compared to 2006, moving as high as 276.1 million hectares. Thus, between 2006 and 2030, there will be a 16.9 million hectare expansion in the area occupied by livestock and agriculture as a result of the conversion of native vegetation. The Northern Amazon presents the greatest growth for that period, at 24 percent. Table 9: Productive land use (crops, pasture and forests) in the different regions of Brazil (1000 ha) Region

2006

2008

2018

2030

Brazil

259,275

257,297

263,222

276,126

Central-West Cerrado

61,756

61,087

61,843

62,994

South

Southeast

Northern Amazon Northeast Coast

MAPITO and Bahia

34,173

54,845

56,639

14,567

37,295

33,561

53,517

57,695

14,622

36,815

Source: ICONE

55

33,614

53,747

61,826

14,913

37,678

34,238

53,960

70,405

15,233

39,296

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Southeast

56

Although agriculture and livestock have expanded considerably in absolute terms, this increase could be considered negligible in annual terms. In other words, 16.9 million hectares of deforestation in 24 years means an average annual amount of 700 thousand hectares, well below the average deforestation observed in the Legal Amazon alone over the past 10 years, which was about 2 million hectares.

The decrease in pasture area in 2030 is accompanied by an increment in the beef cattle herd by 13.9 percent during the same period, indicating that there will be a 14.9 percent productivity increase in the sector, moving from 0.99 to 1.13 heads per hectare. Much of this increase in the herd is in the Northern Amazon, where the total increase will be 20.7 head accompanied by an increment in pasture area of 12.1 million hectares (Table 10). Table 10: Land use (1000 ha) in the six regions of the model for the Reference Scenario

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

South

Cotton Rice

Beans 1a Beans 2a Corn 1a

Corn 2a

Soybean Sugar Cane

Production

Forest

Pasture

2006 13

1,241 536

282

3,706

Southeast

2030 5

1,447 341

288

2006 90

122

332

349

2030

2006

2030

66

402

672

211

59

20

108

136

234

1,363

7,056

501

2,134

2,303

8,377

11,474

1,716

1,944

1,670

2,831

2,452

2,707

483

18,146

1,598

1,292

13,264

290

3,944

44,053

273

355

3,584

967

Central-West

39,565

680

Northeast

zonia

Coast

2006 60

227

521

134

174

1,010

2,708

North-Ama-

621

341

2030 71

2006 30

Bahia

2030 38

526

62

82

73

1,289

1,166

597

669

1,406

1,611

8,322

10,167

2,461

4,076 110

979

1,214

319

910

140

327

0

310

51,200

1,594

48,395

113

52,551

Source: ICONE

64,624

10,801

MAPITO and

10,812

2006 249

800

305

763

2030 546

842

584

551

1,084

1,187

1,872

2,939

688

1,365

371

160

32,138

398

1,435

30,399

The increase in the productive area along the agricultural frontier may be the result of two different but related phenomena. Firstly, there is a significant increase in the size of the herd on the frontier due to its stabilization in traditional regions and the increase in the demand for meat. In the Northern Amazon, MAPITO and Bahia, an increase in the size of the herd is expected between 2006 and 2030, of 44 and 13 percent, respectively. This may be considered an indirect effect of crop expansion, which occupies pasture areas in central-southern Brazil. In addition, an impact on the frontiers due to the increase observed in crop cultivation is more accentuated in proportional terms in MAPITO and Bahia than in the Northern Amazon, where the expansion of pasture area is much greater. In MAPITO and Bahia, an increase of 1 and 1.2 million hectares of soybean and sugar cane between 2006 and 2030 is projected, respectively. In the Northern Amazon, soybean increased 1.6 million hectares during the same period, while pasture areas increased 12 million hectares (Table 10).

The same may be observed in the Southeast, although in different proportions. In this region, there will be a greater increase in the area used for sugar cane, from 3.9 million hectares in 2006 to 7.1 million hectares in 2030. On the other hand, variations in areas used for other crops are not as noticeable, supporting the hypothesis that the region is reaching the limit of expansion in the area used for agriculture. Pasture areas in the Southeast will be reduced 4.5 million hectares, with a drop in the beef cattle herd to 2.9 million head, but still with a growth in livestock productivity in the region. The Southeast is the second most important region for dairy production. The dairy cattle herd is practically stable (Tables 10 and 11), with a 6.7 million ton increase in milk production. Meat production in the region also increased 221 thousand tons during that period. Thus, the reduction in the herd does not mean a loss in production capacity thanks to higher levels of technology. Table 11: Dairy cattle herd (1000 head) – Reference Scenario

Region

2006

2008

2018

2030

Brazil

20,942.81

22,813.01

24,471.55

27,732.54

3,078.42

3,347.27

3,805.55

4,530.30

South Southeast Central-West Cerrado Northern Amazon Northeast Coast MAPITO and Bahia

3,406.60 7,186.67 2,636.85 1,749.15 2,885.12

4,102.30 7,091.95 3,428.19 1,876.15 2,967.16

Source: ICONE

5,458.26 6,865.94 3,828.04 1,882.13 2,631.64

6,466.67 6,997.04 4,524.86 2,255.77 2,957.90

Pastures in the southern region will decrease 4.9 million hectares between 2006 and 2030, while the size of the herd will remain practically constant. A substantial growth in the soybean crop and production forests is expected: 3.1 and 1.1 million hectares during the same period respectively (Table 10). Thus, the South will maintain its considerable participation in soybean production and production forests in the country, with a rather constant total area for agrosilvipastoral activities.

The Northeast Coast showed slight land-use variations during the period considered. Production forests showed the greatest increase in area – 310,000 hectares between 2006 and 2030; followed by sugar cane – 235,000 hectares; and corn – 204,000 hectares (Table 10). This implies that the region is also at the limit of its land occupa-

57

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

While beef-cattle raising is increasing, reducing pasture area in the Central West Cerrado by 2.8 million hectares between 2006 and 2030, soybean and sugar cane will be taking over more area in this region, about 1.8 and 1.1 million hectares, respectively. This indicates that much of the expansion of crop areas will occur on pasture land. The 1.3 million hectare increase in the area for second harvest corn, which will take place between 2006 and 2030, is also noteworthy. Although it doesn’t affect land competition, this increase is very important, as it implies an increment in total corn production and thus less need for land for first harvest corn (Table 10).

tion, principally due to the edapho-climatic restrictions that impede the productive use of much of its area.

Thus, pasture expansion observed in some regions in the Reference Scenario is mainly due to the effects of herd expansion, while direct competition with agriculture on pastures is a less critical factor. Although results clearly indicate that agriculture is shifting to pasture areas, this does not mean that the latter have to move towards the frontier to compensate. Pastures expand on the frontier because of the commensurate increase in the demand for meat, and opportunity costs for herd expansion are lower in this region, thus resulting in an increase in pasture area. This becomes more clear when results per region are observed: the increase in pasture land in the Northern Amazon is greater than the loss of pasture land in other regions, which is in turn the result of competition between agriculture and production forests.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

58

It is important to understand how to interpret the phenomenon of herd stabilization in the frontier regions. Projections for the Reference Scenario basically replicate the general trends observed from 1996-2008, the period for which data on the herd, agriculture and production forests were obtained. What has been observed during this time is that the main determining factor for the expansion of pasture areas in the Northern Amazon region is the increase in the size of the beef cattle herd. This expansion has been rather constant from 1996 to 2006. However from 2006 until 2008 the size of the herd decreased in all regions as a result of an upsurge in the slaughter rate due to the increase in meat exports without a commensurate gain in efficiency in zootechnical indices, such as the rate of repositioning and lower slaughter age. With no significant improvements in zootechnical indexes, the low prices for the animals that were observed the first half of the 2000s motivated ranchers to sell beef cattle for slaughter, reducing the potential to reposition the herd, and driving prices further down.

Variations in the demand for land for other crops (corn, second-crop winter corn, first and second harvest beans, rice, and cotton) will not be as acute as for soybean, sugar cane and livestock. According to the results presented in Table 12 and Figure 5, the area for first harvest corn should increase 660,000 hectares between 2006 and 2030. However, the area for second harvest corn will increase 2.3 million hectares in the same period, mostly in the Central-West Cerrado and South.

Table 12: Land use (1000 ha) for Brazil - Reference Scenario Products

2006

2008

2018

2030

Cotton

844

1,066

1,320

1,399

Beans – 1 harvest st

Beans – 2nd harvest

Corn – 1 harvest st

Corn – 2 harvest nd

Soybean

Sugar cane

Production forests

Total agriculture summer

Pasture

Agricultural Area + Pasture

3,018

2,881

2,898

3,231

9,632

9,656

9,663

10,292

2,694

1,529

3,332

22,749

6,179

5,269

50,386

208,889

259,275

Source: ICONE

2,857

1,143

5,052

21,334

8,235

5,887

51,903

2,380

1,281

5,402

26,023

10,594

7,740

60,814

205,381 203,003

257,284 263,817

2,394

59

1,328

5,608

30,601

12,700

8,450

69,793

207,060

276,853

Figure 5: Evolution of the demand for land in Brazil by crop in theReference Scenario 2006-30 (million ha)

The increase observed in the area planted with first and second harvest corn may be explained for the most part by the increase in the demand for feed, due to a boost in pork and chicken production, which increased 79 percent and 66 percent, respectively, during the period analyzed. The area with first harvest beans decreased 0.3 million hectares, and there was a decrease of 0.2 million hectares for the second bean crop. However, bean pro-

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Rice

duction in Brazil should increase 3.5 to 4.9 million tons between 2006 and 2030, due to the increase in productivity expected during that time (Table 12).

To interpret the results, maps were used representing increments and decrements for each crop modeled during the period studied: sugar cane, cotton, rice, beans, silviculture, corn, soybean and pastures. There was a considerable expansion of sugar cane principally in northeastern Paraná, Goiás, central western São Paulo (where growth increased), the Minas Gerais Triangle, Central Tocantins, Mato Grosso do Sul, and the Northeast Coast (where growth increased, like in São Paulo). Other areas of expansion were also found in the states of Bahia, Santa Catarina, Rio Grande do Sul, Rio de Janeiro, Espírito Santo, Piauí, Maranhão, and Mato Grosso. In other words, sugar cane is expanding in all the states where agriculture is present (Map 2).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

60

It is important to emphasize that the BLUM land-use model projects productivity dynamics and trends over time. Increases in productivity based on past patterns were examined for the Reference Scenario. Productivity was found to increase an average of 0.69 and 2.10 percent per year for the crops considered. According to the PNE’s exogenous estimates, production forests will occupy an area of 8.5 million hectares in 2030, representing an increase of a little over 3 million hectares compared to 2006. Most of this growth occurs in the southern region – approximately 1.2 million hectares (Table 12), repeating the expansion tendencies observed in the past.

Cotton shows an elevated spatial dynamic (Map 2), and areas where it will be cultivated on a constant basis between 2010 and 2030 are concentrated in southeast Mato Grosso. Extensive areas of crop expansion can be seen in southwest Bahia and southeast Mato Grosso. Cotton is decreasing in western Bahia and in the state of Mato Grosso. The demand for land for cotton is actually fluctuating rather than constant. When demand decreases, areas originally used for cotton are used for pasture and probably other crops during the simulation. Map 2: Dynamics of areas where sugar cane (left) and cotton (right) are grown in the Refer-

ence Scenario (2010-2030). Yellow = crop permanence; blue = decrement; red = incr

ement

Rice-growing areas (Map 3) are widely dispersed and less concentrated. Cultivation occurs on a constant basis principally in parts of Rio Grande do Sul, Maranhão, Santa Catarina, Mato Grosso, Piauí and Pará, and is also expanding in these states (in areas close to where it is already being grown), as well as in Bahia (where it is practically non-existent). There are also areas where it is declining, such as in a small part of Rio Grande do Sul, Mato Grosso, Tocantins, Maranhão, and Piauí. The spatial dynamics are due to fluctuations in demand in the micro-regions, as national demand has generally remained rather constant.

61

ence; blue = crop decrement; red = increment

Areas where bean growing (Map 3) was more stable during the study period are concentrated in the Northeast (Ceará, Rio Grande do Norte, Paraíba, Pernambuco, Sergipe and Alagoas), where there are also small areas of crop increment. In the states of Maranhão, Piauí, Tocantins, and Bahia, there are areas where the crop has both increased and decreased, with expansion occurring mainly in the state of Tocantins. The state of Paraná in particular has areas where cultivation has both decreased and increased.

Corn (Map 4) is widely cultivated throughout the country and in most of the states where it has remained stable or increased, except for the state of Mato Grosso, where there was also an area where the original crop decreased. Soybean (Map 4) represents one of the most widespread crops in terms of area in Brazil, mostly in the South, Central South, Minas Gerais Triangle and in parts of the states of Bahia, Piauí, and Maranhão given that its tendency in the Reference Scenario is to increase in demand. Thus, there are practically no areas where the crop is decreasing. Areas of expansion predominate, some of which indicate crop intensification in regions bordering the Amazon.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 3: Dynamics of areas where rice (left) and beans (right) are grown for the reference scenario (2010-2030). Yellow = crop perman

Map 4: Dynamics of areas where corn (left) and soybean (right) are cultivated for the Reference Scenario (2010-2030). Yellow = crop permanence; blue = crop decrement;

red = crop increment

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

62

Silviculture, which is practiced originally in the southern, southeast and northeast states, as well as in Pará and Amapá, remains relatively constant between 2010 and 2030. Spots indicating where cultivation has expanded are sparsely distributed throughout the territory, and appear to recede in central-northern São Paulo State (Map 5). Pasture areas (Map 5) are an important aspect in the simulation model when they are part of the three possible transitions (native vegetation > pasture; pasture > crops; or crops > pasture). They appear to have expanded relatively little in central-southern Brazil, due to direct competition with agriculture and a constant demand for pasture. Areas where pasture areas are more stable include the states of Minas Gerais (except for the Minas Triangle region), Bahia (except for the western region), Ceará, Rio de Janeiro, part of Rio Grande do Sul, a large part of Mato Grosso do Sul and in the states of Sergipe, Alagoas, Pernambuco, Rio Grande do Norte, and Paraíba. On the other hand, pasture areas have also expanded considerably in the Amazon, mainly due to deforestation.

Map 5: Dynamic of planted forest (left) and pasture (right) areas for the Reference Scenario (2010-2030). Yellow = pasture permanence; blue=pasture decrement; red

=pasture increme

63

2.1.3.1 Division into Geographic Micro-regions Results obtained for the allocation of area for the reference and Low-carbon Scenarios in each of the six main regions were spatialized according to the IBGE level of micro-regions in order to identify the more dynamic regions that partially determine future areas of expansion of livestock activities. The criteria used were the growth history of each crop and the area available for agricultural use. In the case of sugar cane, since logistics are the greatest restriction for this sector (limiting the distance between factories and canebrakes), the locations of working factories, those under construction and those being planned determine the spatialization of sugar cane production by micro-region over time. The history of the area with planted forests was obtained based on an estimate of forest production data from IBGE (IBGE, 2008c). The area available for agriculture was estimated by the UFMG and included convertible pastures, or those in areas without any impediment (either legal and/or with steep slopes and unsuitable soils19). Areas of residual vegetation without impediment were also considered for livestock, but only in the case of the Reference Scenario.

A criterion was developed for prioritizing the different uses, which varied depending on the region, but always with sugar cane coming first – due to its more precise location between the micro-regions – followed by the other crops in different order, then production forests, and lastly, pastures. Prioritization between crops was defined, taking into consideration the ranking of the region’s most important crops in terms of area planted over the past 10 years.

19

Conversion Units (CU) and Indigenous Lands (IL) are among the legal impediments. Land-use restrictions focus on the rugosity and types of soil.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

nt

2.1.3.2 Spatialization of Land-use Change and Deforestation: SIMBRASIL Model

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

64

To evaluate emissions resulting from deforestation, a national land use and land-use change spatialization model was constructed (SIMBRASIL), with its accountability for CO2 emissions. For the reference scenario, with regard to deforestation, it was considered that a cycle that enabled the continuity of expansion tendencies and dynamics for agricultural crops and other land uses in Brazil, independent of their consequences on deforestation, with no restrictions, would surpass the legal limits (generating environmental liability). This scenario was used as a basis to construct the low-carbon scenario, which included this concern, with the aim of decreasing deforestation by reducing the demand for land for agriculture and livestock, which by definition could not generate deforestation beyond the legal limit. Reducing deforestation and recuperating environmental liability through forest restoration will lead to a reduction in GHG emissions in the low-carbon scenario compared to the reference scenario. The development of the spatially explicit model for land-use change, deforestation and carbon emissions resulting from such conversions evolved in three stages as described in Figure 6 and took into consideration its compatibility with the BLUM model. Figure 6: Architecture of the LULUCF study, with an emphasis on the components that include the deforestation factor

Source: UFMG

A spatially explicit model for land-use change and soil cover was developed (SIMBRASIL – available at www.crs.ufmg.br/simbrasil) after calculating the areas available for livestock expansion and planted forests. The objective of this model is to spatialize projects for agricultural expansion and demand for land at the micro-regional level using a 1 km2 cell, with a model by ICONE for Brazil as a whole (BLUM Model) for the two scenarios in question: the reference and Low-carbon Scenarios. The first part

of this model was developed based on material from the data base (Figure 7), which includes variables presented in Table 13. Figure 7: Example of the data base prepared for simulations of land-use change and cover

65

Variable

Description

Hydrography

Principal Permanent Rivers (MMA)

Suitable areas

Obtained from the previous stage of the study

Declivity

Elevation

Shuttle Radar Topography Mission – NASA calibrated by EMBRAPA

Shuttle Radar Topography Mission – NASA calibrated by EMBRAPA

Plant Cover and Obtained from the previous stage of the study Land Use Infrastructure

Ports, Waterways, Railways (Ministry of Transport)

Highways

Divided into two classes: Paved and Unpaved (Ministry of Transport; CSR)

Population

Data on urban population per municipal seat (IBGE, Demographic Census - 2000)

Protected Areas Include the two uses of Federal and State CUs: total protection and sustainable use of Indigenous Lands and Military Areas (MMA, IBAMA, CSR)

All of the aforementioned data were rasterized at a spatial resolution of 1 km2, the equivalent of matrixes of 4500x4500 cells. In addition to spatial data, the model incorporates land-use projection tables provided by BLUM for a basket of crops (sugar cane, soybean, maize, cotton, rice and beans), production forests and pastures. However, for each micro-region, there is one land distribution vector passed per year for each of the specified uses and the model seeks to allocate this distribution, the basis being the agricultural aptitude of the land for each crop modeled and production cost factors estimated by infrastructure and consumer market proxies.

The model was implemented based on the Ego Dynamic platform (Box 1), which was designed to operate on an annual basis on two spatial levels: IBGE micro-regions and 1 km² raster cells.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 13: Description of data developed for the implementation of SIMBRASIL

Box 1: EGO DYNAMIC (Environment for Geoprocessing Objects)

Written in C++ language and Java, the software has a library of operators called functors, which may be understood as a process that acts on a set of entry data, to which a finite number of operations is applied, producing a new set of data as a final product (Rodrigues et al., 2007). They are currently implementing the most common spatial analysis operations in the Geographic Information System (GIS) and a series of others that are created specifically by spatial simulations, including transition functions and calibration and validation methods.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

66

The spatially explicit simulation model was implemented on the EGO Dynamic platform (Dynamic – Environment for Geoprocessing Objects), an integrated software program that consists of an environmental modelling platform. Through this platform, which was developed by researchers from the Federal University of Minas Gerais in 1998 (Soares-Filho et al., 2009) it was possible to develop a diverse range of spatio-temporal models that require analytical operations and/or complex dynamic operations, such as network iterations, feedback, multi-scale approaches, map algebra and the application of a series of algorithmic complexes for the analysis and simulation of phenomena in time and space.

The advantage of the software (its current version 1.2.3. is available on www. csr.ufmg/dinamica) is its flexibility, as it enables the user to construct models by

linking up functors, which, once in order, establish a data flow chart. Through the graphic interface of the EGO Dynamic, it is possible to create models by simply connecting the operators through their entrances and exits (ports), which represent connections with specific types of data such as maps, tables, matrixes, mathematical expressions and constants. The models created appear as a diagram, whose execution follows a data flow chain. Three main models were developed through the software that was relevant to the study in question: calculation of available land for expansion, simulation of land-use change and carbon emissions from land use and land-use change.

The execution of the spatially explicit simulation model for land use and land-use change involved the development of two sub-models. The purpose of the first (Figure 8) is to produce a map based on land use, allocating lands according to the BLUM crop classes (sugar cane, soybean, maize, cotton, rice, beans), plus production forests and pasture. Inputs for the model include the land-use map, the micro-regions map and tables on the demand for land for each crop produced by ICONE for BLUM. First, the model identifies the anthropized usable area according to the original land-use map, or potential agricultural expansion areas. Then it calculates spatial favorability maps for this expansion, integrating data on agricultural aptitude (Assad and Pinto, 2008) and other criteria such as: distance, roads, urban appeal, transfer cost to ports, declivity, and distance and area converted.

Figure 8: First part of the spatially explicit model for land-use change and soil cover - land allocation

The model then calculates the rates of transition for crops and reforestation, dividing the areas projected per year by the appropriate pasture area. In case the pasture area is not sufficient, the model converts the area with native vegetation and wild grasses into pasture areas. If there is a decrease in the quantity of an agricultural crop in a micro-region, the model returns this area to the pasture stock. Thus, the model always uses the class of pasture as a temporary stock for transitions between forests, pastures and crops. The spatial allocation uses an automatic cellular mechanism to create marks on the landscape (Soares-Filho et al., 2002). In the event that the model is not successful over time in allocating areas projected by BLUM in a specific micro-region, the residual is passed on to the neighboring micro-regions for an iterative process by area in a later phase, thus generating the best estimate possible in terms of areas with BLUM projections.

In the second part, the updated land-use map from 2006 is the main input. Other inputs include the map of micro-regions, maps of transition probabilities calculated earlier by complementary models20 and the tables prepared by the BLUM. The simulation model is shown in Figure 9.

20

There are four transition probability maps: Probability of conversion to crops (set of 6 maps + 1 for planted forest), Probability of conversion to pasture, Probability of return, and Probability of regrowth.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

67

Figure 9: Spatially explicit land-use change and soil cover model -simulation of land-use change

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

68

The simulation model for the land-use change in question interacts online with the model that projects Amazon deforestation (Soares-Filho et al., 2008), which uses the demand for land for livestock and agriculture modeled by the BLUM as an input. The model aims to incorporate the indirect causes of deforestation as well as the direct conversion of the demand for land for agriculture and pasture. For the study, the deforestation projection model uses three variables that are established as constant values for the evaluation of the two scenarios: regional migration rates, and protected areas and infrastructure (i.e. paved and unpaved roads); besides two other variables that model the pressure of agricultural expansion: the rate of expansion of areas occupied by crops and the growth rate of the beef cattle herd, according to BLUM projections for the two scenarios in question. Inputs include the map of micro-regions and tables on the extension of protected areas, on original forested areas, on crops and on the number of head of cattle in the herd, provided by BLUM, as well as tables on average road density per micro-region.

2.1.4 Calculation of Emissions Associated with Land use, Land-use Change and Deforestation in the Reference Scenario

Having defined the economic and territorial magnitude of agricultural and livestock activities and their locations, it is possible to calculate the greenhouse gas emissions associated with these activities. This section presents the sequence of emissions from livestock and agriculture. Emissions due to deforestation and uptake associated with the forest area are presented in the next sections.

2.1.4.1 Emissions from Livestock Brazil is one of the world’s main beef suppliers. According to the USDA (2008), the country was responsible for 15.5 percent of global beef production (59.3 million tons of carcass equivalent) and 24.9 percent of exports (7.7 million tons of carcass equivalent) in 2008.

69

In the case of Brazil, emissions are predominantly related to the beef cattle subsector. According to Brazil’s initial national statement at the United Nations Framework Convention on Climate Change (MCT, 2004), and the first Brazilian inventory of anthropic greenhouse gas emissions (MCT, 2006), total methane emissions resulting from enteric fermentation nationwide were estimated to be 8.8 Mt for the year 1990, and those resulting from animal waste management systems, estimated to be 0.3 Mt, amounted to 9.1 Mt. In 1994, methane emissions from livestock were estimated to be 9.8 Mt, with 9.4 Mt attributed to enteric fermentation and 0.4 Mt to animal waste management systems. Only annual emissions from enteric fermentation represent 92 percent of the total amount of methane emissions from the agriculture and livestock sector. In 1994, the beef cattle category was responsible for 81 percent of methane emissions from livestock in Brazil, the dairy cattle category contributed 13 percent and other animal categories contributed 6 percent of emissions. The approximate 7 percent increase in emissions in the sector during the period 1990-1994 was predominantly due to the increase in the size of the beef cattle herd.

Brazilian beef cattle are mainly raised in extensive pastoral systems with a carrying capacity rate and performance much lower than its potential (IBGE, 1998). These facts imply the possibility of reducing emissions by using more technological production systems (e.g. feedlot systems, crop-livestock integration and feedlots) which generate an increase in the animals’ performance, and consequently improved CH4 and N2O emissions per product unit. In addition, the adoption of more intensive production systems will result in a reduction in the demand for land for cattle, making the occupation of the land by other livestock activities possible without the necessity of opening up new areas. It is also important to note that the restoration of low productivity areas (degraded) may represent a considerable drain of CO2 through the increase in carbon stocks in the organic material in the soil, as productive pastures tend to present higher levels of C in the soil (Guo; Gifford, 2002; Cerri et al., 2003). One projection of the direct emissions from Brazilian cattle was made by Barioni et al. (2007). The estimate considers a moderate increase in productivity between 2007 and 2025 and points to an increase in efficiency sufficient to counterbalance the higher production necessary to meet the demand.

Greenhouse gases most commonly associated with livestock activities are carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O) (IPCC, 2006). However, CO2

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Almost all of Brazilian production comes from herds in pasture systems. The size of the national beef cattle herd – about 200 million head – and the equally sizable pasture area occupied by livestock (over 170 million ha) has caused some concern about the potential impact of Brazilian beef cattle on the environment. In addition, the lower the animals’ performance, and the longer the time before slaughter, the higher the amount of methane emissions produced per ton of meat.

Anaerobic fermentation is known to produce CH4, and is a necessary process for eliminating excess hydrogen (Van Soest, 1994). The reduction of carbon to produce CH4 is done by a sub-population of microorganisms called methanogenic bacteria. The growth and abundance of these bacteria in the ruminal environment are aided by the presence of slowly degrading fiber, and are inhibited by the presence of starch and its final product, lactic acid, with the concomitant drop in pH. Generally speaking, the more fibrous the food (for a specific level of ingestion), the greater the quantity of CH4 emitted, and the higher the protein content (and the greater the nitrogen excretion), the higher the nitrous oxide emissions. The more food ingested, the higher the daily CH4 and N2O emissions for a specific diet. Oils and fats also considerably reduce methane production. Thus, diet can also influence CH4 production, with the proportion of energy in the diet lost being inversely proportionate to its quality, particularly its energy density.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

70

emissions from animal respiration are generally unknown due to the assumption that this carbon is derived from photosynthesis and therefore only represents a return to the atmosphere. However, CH4 produced by enteric and manure fermentation, and N2O released by the nitrification/denitrification of the nitrogen excreted are sources of net emissions. These molecules assume greater importance due to their higher potential for global warming, which is 21 and 310 times more than CO2 for CH4 and N2O, respectively.

Ruminants excrete nitrogen in their feces and urine. Both types of excretion include an endogenous component (from the animal itself), and a food component (as the N from the diet is not used by the animal). N excretion may generally be estimated as the difference between N consumption and its retention in body tissue. For example, a 300 kg animal consuming 7 kg of dry matter/day containing 10 percent of crude protein (N x 6.25) consumes 0.112 kg N/day. If this animal gains 0.9kg/day, with the gain containing 20 percent protein, retention would be 0.029 kg N/day, in other words, the animal would be excreting 0.083 kg N/day. The proportion of N in the feces that is converted into N2O varies depending on the manure management system and environmental conditions. In liquid manure management systems (in lagoons, for example), fermentation is anaerobic and CH4 and N2O production, which is considerable, increases with the ambient temperature. In the solid management system, and even more with the animal’s direct deposit in pasture areas, aerobic degradation occurs, with much lower production of these gases. In this case, which better reflects the Brazilian situation, between 0.1 and 0.3 percent of N from the dung is converted into N2O (Loyon et al., 2008).

2.1.4.1.1 Methodology

The methodology used for modeling greenhouse gas emissions followed the premise that production, and consequently the herd, are limited by the national demand for meat. In this approach, mitigation alternatives that bring about an increase in herd productivity result in a reduction in the number of animals for a specific demand that has been determined exogenously. In the present study, the estimate was provided by the ICONE consulting firm. Methane and nitrous oxide emissions from ruminants are basically a function of the quantity of food ingested and the quality of the diet. However, an increase in ingestion

generally results in an increase in the animal’s performance, leading to a consistent reduction in methane emissions per production unit, shortening the animals’ life cycle or reducing the number of matrices necessary for producing animals for slaughter.

To determine the composition of the national herd and the proportion of productive systems, an approach was adopted using exogenous estimates provided by ICONE, related to the level of balance between meat production and demand. These data were used to project the size of the herd necessary to meet the demand based on the level of productivity projected (as a function of the composition of productive systems). Estimates of available pasture areas produced by an economic competition land-use model were used as entry data to project the productivity level per area (balanced production / pasture area). In this approach, mitigation alternatives that lead to an increase in herd productivity result in the reduction in the number of animals and not in an increase in total production. Figure 10 presents a diagram of the approach used. Figure 10: Information flow in the analytical model

71

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In tropical pastures, which predominate in Brazil, most of the animals’ productivity is associated with greater food consumption and better food quality. These conditions are related to improved production systems. Given current productivity levels, Brazil has an unequaled opportunity to increase herd productivity through the adoption of different productive systems. Thus, analyses for the projection and identification of emissions mitigation alternatives study prototype farms with a variety of productive systems that reflect different levels of land-use intensification and animal productivity. Characteristic indexes of productivity, herd composition, investment structure and cash flows are attributed to each of these productive systems.

72

To estimate the quantity of food ingested and methane emissions, the animal’s weight, physiological stage, breed and performance (weight gain, birth rates and milk production) must be determined. Since the animals’ characteristics are heterogeneous in the herd, it is advisable to divide the herd into categories and to calculate emissions, as well as ingestion and emissions (IPCC, 2006) for each category. Categorizing the herd enables distinct diets to be attributed to each group, a common on-farm practice, and facilitates herd productivity calculations, which are necessary for estimating the number of animals and composition of the herd. Considering the predominance of beef cattle and the use of dairy cattle for slaughter later on, average milk production was adjusted to include the national dairy cattle herd. In the analysis model developed for this study, the herd was divided into nine categories of animals for calculating methane and nitrous oxide emissions and to determine production costs for each productive system. Table 14: Categories of animals considered in the analysis of livestock emissions

Maximum age máxima (Ix)

F1

0

12

Bullocks (up to 1 year old) M1

0

12

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Minimum age mínima (In)

Category

Variable

Cows

F0

Bulls Heifers up to 1 year old

Heifers 1 to 2 years old Heifers 2 to 3 years old Calves 1 to 2 years old Calves 2 to 3 years old

Calves over 3 years old

M0

F2 F3

M2 M3 M4

12 24 12 24 36

24 36 24 36 -

Herd composition, in other words, the percentage of the number of animals in each category, is calculated based on zootechnical indices from the simulated productive system. The percentage of animals in the herd in a dynamic equilibrium is calculated by equations 1-7, based on the zootechnical performance indices attributed to the productive system. The number of births is calculated using equations 1 and 2, based on the number of cows, and the birth and mortality rates of growing heifers.

Where N is the number of births, a is the birth rate (adimensional), ωp is the mortality rate until the first delivery (adimensional), ωi is the mortality of each category for heifers and bullocks (adimensional), and; t(Fi) is the amount of time the animals remain in that category until the first delivery (months).

The amount of time they remain in the category is calculated differently for males and females as shown in Equations 3 and 4. 73

Where Ix is the maximum age (months) of the animals in each category and βp and βa are the ages from the first delivery until the time of slaughter (months) defined by the system, respectively.

The quantities of animals in the growth categories (Fi, Mi) are therefore calculated based on the quantity of animals in the immediately lower age category, of the same gender and mortality rate as that category (Equations 6 and 7).

Lastly, the number of bulls is calculated based on the bull/cow rate (q) defined by the system in question, according to Equation 8. The number of females and males slaughtered (AF and AM, respectively) is calculated according to Equations 9 and 10.

Carcass production (PC) is then calculated for equation 11. Where CEF and CEM are carcass weight at time of slaughter for males and females, respectively, obtained based on trimestrial research on the slaughter from IBGE35and projected by ICONE.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

To calculate the number of heifers and bullocks, the proportion of 50 percent is assumed for males and females and the constant mortality rate ω0 during the year (Equation 5).

Calculation of CH4 and N2O Emissions

74

Three levels of models were applied for each scenario according to IPCC recommendations (2006). The Level 1 model follows the recommendations exactly, applying the factor of 56 kg CH4/year for each animal after weaning, plus 1 kg CH4/year for manure; as well as the factor of 0.36 kg N excreted/day, together with its rate of conversion into N2O (0.2%; Loyon et al., 2008). For Level 2, the distribution of the animals in the different herd categories is calculated as follows: cows, heifers, 1-2 year old bullocks, 2-3 year old bullocks, 1-2 year old calves, young bulls, 2-3 year old young bulls, young bulls over 3 years old, and mature bulls, as described earlier. Thus, the performance of each category in each production system is defined based on published data (FNP, 2008).

Based on weight and weight gain, liquid energy requirements for maintenance (ELm) are:

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Where PV = live weight (kg), Cfi = a coefficient that varies according to the animal category, MJ day-1 kg-1 (0.386 for lactating cows, 0.370 for bulls, and 0.322 for other categories). Moreover, there are equations for adjusting the ELm for activity and movement, with no increase for animals in cow barns; 17 percent for animals in rich pastures, small corrals, with flat topography; and 36 percent for animals on deficient pasture area, large corrals, with sloping topography. Net energy requirements for the gain (ELg) are calculated based on weight and weight gain:



(13)

Where PV = adult live weight (kg), GPD = daily weight gain (kg/day), and C = coefficient with values of 0.8 for females, 1.0 for neutred and 1.2 for non-neutred males (NRC, 2000). For lactating cows, net energy requirements for lactation (ELI) are calculated based on the production of milk and its fat content (NRC, 1989): Where L is milk production (kg/day) and G is the proportion of fat in the milk (%).

For gestating cows, the net energy requirements for gestation (ELp) are calculated based on the demand for/of maintenance:

The compositions (digestibility and gross protein content) of the diets for each category are defined according to commonly observed amounts. The digestible energy values (ED, %) are converted into ELm and ELg: 75

For adult animals, the IMS is calculated according to: Based on this set of equations, it is possible to estimate gross energy intake (GEI/ IEB) in each animal category: To calculate CH4 emissions from enteric fermentation, the factor (Ym) is applied as recommended by the IPCC (2006) for animals that consume diets that are low in ED (6.5%) or high in grain and ED (3.0%): Besides enteric fermentation, there are CH4 and N2O emissions from animal manure. For CH4, the factor of 1 kg CH4•animal-1•year-1 remains the same. For N2O emission, N excretion (EN/NE) is calculated based on IMS, gross protein content of the diet and its digestibility, as well as endogenous nitrogen excretion: N2O production is calculated in the same way as in Level 1, using the rate of conversion of 0.2% (Loyon et al., 2008): Level 3 calculations are similar to those of Level 2, except that the equations and coefficients are more specific to Brazilian conditions. For example, the value of Cfi is reduced 10 percent, given the lower maintenance requirements of zebu cattle. However, the ingestion equation includes maintenance, lactation, gestation and weight gain requirements:

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Ingestion of dry matter (IMS, kg/day) for growing animals and for finishing is calculated based on live weight and net energy for diet maintenance and activity:

76

Based on the ED, the fiber content in neutral detergent (FDN) as well as lignin (Lig) in the diet is calculated based on data from the NRC (2000):

These amounts are used to estimate CH4 emissions from enteric fermentation, using equation (27) proposed by Ellis et al. (2006): CH4 and N2O emissions from manure are equal at Level 2.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Once the amount of each productive system in the national herd was determined over time, CH4 and N2O emissions were calculated as the sum of emissions generated by each productive system.

Emissions Estimates by Prototypical Systems

“Prototypical” farms were defined to study the heterogeneity of productive systems and the possibility of mitigating emissions through changes in the adoption of different types of productive systems, with the following four systems making up a complete system (cow-calf, stocking and finishing): Complete cycle in degraded pastures Complete cycle in extensive pastures

Extensive cow-calf in pastures + supplemented stocking and finishing in crop-livestock integration Extensive cow-calf in pastures +supplemented stocking and finishing in feedlots.

Typical zootechnical indices for the simulated productive systems were attributed for each prototypical farm (Table 15). The productivity of the prototypical farms (representing each productive system) was thus calculated based on these indices using Equations 1-11.

Table 15: Zootechnical coefficients considered for each productive system

Digestibility of diet during breeding

Complete cycle in extensive pastures

56.0

62.0

Extensive cow-calf + supplemented stocking and finishing, integrating farming and livestock

Extensive cowcalf + supplemented finishing in feedlots

60.0

60.0

62.0

62.0

Digestibility of diet during stocking

58.0

Milk production

1100

1400

1400

1400

Mortality until 1 year old

7%

5%

5%

5%

Digestibility of diet during finishing Lactation period Birth rate

Mortality between the ages of 1 and 2 Mortality between the ages of 2 and 3 Mortality over 3 years old

58.0 7

55% 2% 2%

1%

60.0 60.0 7

60% 2% 1%

1%

62.0 7

75% 2% 1%

1%

72.0 7

75% 2% 1%

1%

Rate of cull cows

15%

15%

15%

15%

Male carcass weight

230

250

250

265

Yield male carcass

52%

50%

52%

50%

52%

52%

160

185

185

185

Weight of adult cows

Relationship bull/female Female carcass weight Age at first delivery

Yield female carcass Weight at birth

Weight at weaning (males)

Weight at beginning of finishing Weight gain during stocking

Weight gain during finishing

420 30

200 36 30

379

0.25

0.40

420 30

200 30 32

379

0.30

0.60

420 30

200 30

50% 32

379

0.40

0.60

77

420 30

200 30

50% 32

379

0.40

1.20

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

 

Complete cycle in degraded pastures

78

Since CH4 and N2O emissions in this study are calculated based on the productive system, and it is assumed that the quantity of CH4 and N2O emissions is dependent on the system and not on the region where the system is located, the proportions of the productive systems were attributed at the national level.

As shown in Table 16, independent of the level of calculation, the quantity of CO2-e emitted per kilo of carcass equivalent decreases as the system intensifies, being higher in a degraded pasture system, and lower in recuperated pasture systems with supplementation, but with finishing in feedlots. In view of this, the accelerated intensification of beef cattle-raising is implicit in the Low-carbon Scenario. Table 16: Greenhouse gas emissions per animal and per carcass equivalent in kg in different production systems

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Productive system Degraded pasture Extensive pasture ILP

1

Feedlot

2

Emissions per animal in the herd (kg/year) CH4

N2O

CO2-e

56.38

0.20

1.25

51.71

51.73 51.53

0.22

0.21 0.21

1.15

1.15 1.15

Emissions/product (kg CO2-e/ kg carcass) 29.65 21.89

18.76 17.64

Extensive cow-calf and finishing, in crop-livestock integration. Extensive cow-calf and finishing in feedlots.

For the construction of the reference and Low-carbon Scenarios, the size of the national herd was estimated using an approach similar to that described by Barioni et al. (2007). In this approach, a projection of the amount is adjusted numerically to meet the projected demand for meat (in kg carcass-equivalent). Data adopted for the size of the national herd, meat production, and pasture area were the result of simulations from a land-use competition model developed by the ICONE consulting firm.

Due to the lack of published national statistics on the proportion and geographic distribution of productive systems in the construction of low carbon and Reference Scenarios, the size of the national herd was adjusted numerically to meet the projected need for meat (in kg carcass-equivalent). The area, amount and production of each type of system considered were generated based on the following: (a) the sum of meat production from productive systems equals the production projected by ICONE in 2008 and 2030; (b) the sum of the pasture area occupied by productive systems equals that projected by ICONE in 2008 and 2030, and; (c) the sum of the amount in each productive system equals that projected by ICONE in 2008 and 2030. In 2008 and 2030, the proportion of the systems was interpolated in a linear fashion.

A substantial linear gain in productivity due to genetic improvement, estimated at 0.3 percent per year (Lobo et al., 2009), was considered.

With the aim of providing a basis for calculating the carbon balance in the pasture area for the group responsible for agricultural emissions, carrying capacities provided by ICONE for each municipality were used as a proxy for the level of intensification of livestock in the municipality. For mapping low productivity areas, the municipalities were arranged according to carrying capacity and pasture area, adding those with less and more carrying capacity until the pasture area equals 30 percent of the pasture area of the region. All of the municipalities with carrying capacity lower than that which surrounds 30 percent of the area were considered municipalities with low productivity. A portion of the areas with low productivity was calculated based on the area of the municipalities for each micro-region. Estimates were provided as inputs for the analysis of other topics related to LULUCF. The quantification of degraded areas also enabled an estimate of emissions over time in spatial terms for other groups. Like production systems in extensive pastures, farm-livestock integration and feedlots have very similar emissions rates per head, enabling emissions estimates for degraded pasture systems for that and other areas for the remaining amount using the following model: The amount of degraded and non-degraded areas (Bd and Bnd) may be calculated as:

Where Ld is the rate of carrying capacity for degraded pastures.

Emissions in systems with degraded (Ed) and non-degraded (EnD) pastures may be estimated as: Where Ed is the emissions coefficient per animal (Mg CO2-e head-1 year-1) in a beef cattle production system with degraded pastures, and End is the emissions coefficient per animal (Mg CO2-e head-1 year-1) in a beef cattle production system with non-degraded pastures. Total emissions for the micro-region (Et) are thus Et = Ed + End.

2.1.4.1.2 Reference Scenario Results

The expected evolution of livestock production and productivity in the Reference Scenario is considered, so that the projection of the Reference Scenario until 2030 includes changes in the composition of productive systems due to variations in the demand for beef, area, and herd.

79

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Where E represents the national herd (head of cattle), A is pasture area in the country (ha) and Pd is national meat production (t e-carcass/year). The subscript k defines the productive system. PPk is the proportion, in number of animals, of the k-nth productive system (a-dimensional) and TLk is the carrying capacity of the k-nth productive system (head/ha).

80

The projection of meat production in Brazil for both reference and Low-carbon Scenarios anticipates a 35.6 percent increase by 2030, going from 9.7 million tons of carcass equivalent in 2008 to 13.15 tons in 2030. For the Reference Scenario, a change in pasture area from 205.38 million hectares to 207.06 is projected and from 201.41 million head to 243.2 million head for the herd (Table 17). The development of the area occupied by productive systems in the Reference Scenario is presented in Figure 11, which suggests an increase in the number of cattle, increases in crop-livestock and feedlot integration, and a reduction of the herd in extensive pasture systems and in degraded pastures. Table 17: Estimates of area, herd, proportion of the herd in productive systems, and emissions for the Reference Scenario

Productive System

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Degraded pastures Extensive pastures ILP1

Feedlot Total

2

Area (million ha) 2008

2030

59,524

22,610

5,500

7,141

132,173 156,858 8,182

20,500

205,380 207,060

Herd (million head) 2008

22,379

2030

8,500

Proportion (% of the herd) 2008

2030

11,04

3,49

Emissions Mg CO2-e 2008

2030

27.974

10.625

11.500

14.932

155,510

184,539

76,69

75,88

178.837 212.219

14,879

37,182

7,34

15,29

17.111

10,000

202,768

12,985

243,205

4,93

100,00

5,34

42.759

100,00 235.421 280.536

Figure 11: Variation in the pasture area occupied by type of productive system in the Reference Scenario (million ha)

2.1.4.2 Agricultural Emissions

The quantification of GHG emissions from the soil is an inventory exercise and therefore the IPCC methodology (1966; 2006) was used as a basis. In the case of GHG emissions from the use of fossil energy, technical coefficients were used for crops, for which the hours of labor for each agricultural operation were converted into energy units (Megajoules – MJ) and the energy was converted into GHG using as a reference the quantity of CO2 equivalents (CO2, N2O and CH4) released from burning diesel oil for generating the respective quantity of energy for the agricultural operation (Boddey et al., 2008). GHG were estimated for the more common production systems for cotton, rice, bean, corn, soybean, and sugar cane crops.

2.1.4.2.1 Evaluation of CO2 Emissions from Changes in Soil C Stocks

CO2 emissions or uptake resulting from changes in the soil C stocks were estimated using the IPCC methodology (2006), according to which CO2 flows are estimated indirectly through the balance of the net variation in soil carbon stocks due to land-use changes. To estimate net changes in the carbon stocks, the IPCC methodology calls for an estimate of the stocks up to 30 cm deep, distributed by type of use and soil category for a 20-year period. Twenty years is assumed to be enough time for the soil C stock to arrive at a level of equilibrium for a specific land use. To calculate C stocks from 2010 to 2030, land-use data since 1990 had to be used.

For the methodology to function correctly, the total area of each region to be considered must be the same throughout the entire period. This study estimates GHG emissions in areas used for agriculture as well as for pasture and planted forests, called occupied areas, which are all part of the Reference Scenario. Thus, differences between the years of contraction or expansion of the occupied areas were considered complementary areas, which in reality correspond to the area “under other uses” (e.g. permanent agriculture) and under “native vegetation”. The largest area occupied by agriculture and pastures, between 1990 and 2030 for any given region, corresponded to a zero complementary area. For 1990, it is estimated that, since the complementary area always exceeded the area estimated for “other uses” for that year (data from IBGE on temporary summer agricultural areas that were not considered, and permanent agriculture areas), the difference corresponded to the area with native vegetation (Table

81

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Greenhouse gas (GHG) emissions in agricultural areas were evaluated, taking into account emissions from soils and fossil energy used in agricultural operations. The following were considered GHG emissions from land use: loss of C from the soil; methane emissions from wetland rice fields and biomass burning; and nitrous oxide emissions from burned biomass, fertilizer use, plant residue decomposition, and N mineralization due to the reduction of carbon stock in the soil. In the case of fossil energy used in agricultural areas, GHG production resulting from the energy consumed in iron and steel production for agricultural machines and tools, and from burning diesel oil for agricultural operations such as plowing, harrowing, pulverization, manure application, etc., were considered. The energy used to manufacture and transport such inputs as herbicides, pesticides, and fertilizers, is dealt with in the energy chapter of the Low Carbon Study, and published in another Summary Report. It is also considered in the calculations and will be reflected in the twofold accountability of GHG.

18). Starting in 1991, the area’s expansion, making it larger than the sum of the area occupied with “other uses” in 1990, coincided with the reduction of native vegetation. The reduction of the area signified the increase of “other uses”. Table 18: Areas under different uses and total area in 1990 by state

82

Land use State RO AC AM RR PA

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

AP

TO

MA PI

CE

RN PB PE

AL SE

BA

MG ES RJ

SP

PR SC

RS

MS

MT GO DF

Farm

Pasture

336.89

3,907.44

41.04

4,411.01

1,946.61

10,642.99

10.29

2,579.81

1.41

118.26

0.00

2,709.77

335.14

19,260.40

1,582.17

2,993.38

75.12 9.41

362.01 1.10

1,372.60 863.83

227.85

776.01

991.66

782.98

174.76

660.90

11.30

7,307.78

114.37

8,161.45

12,441.67

7,530.28

27.84

447.59

0.00

1,585.01

5.32

597.03

436.23

510.92

3,404.29

2,729.40

2,312.39 848.15

1,193.56

1,025.70

16,118.15

269.08

1,243.02

3,110.03 231.41

4,234.49

5,558.22

1,776.44

0.08

3.01

24.63

15.11

167.08 0.00

412.49

832.69 99.44

13.54

1,060.51

297.43

5,072.59

2.24

2.91

161.96

403.53

6,932.01

7,466.69

597.00

4,668.65

1,389.58

4,909.67

172.74 25.88

28,387.26 764.04

0.00

19,595.62

0.00

5,432.87

0.00

0.00

0.00

0.00 0.00

0.00

9,378.31

4,683.61

2,415.22

3,619.95

4,378.10 1,795.34

1,774.76

0.00

22,513.87

0.00

2,501.01

0.00

0.00

0.00

42,335.18 1,884.95

16,966.83

14,406.63

181.08

3,467.00

0.00

28,947.21

19.98

52.41

67.83

89.47

0.00

3,705.60

0.00

33,185.03

21,287.09

963.03

2,753.65

2,561.86

3,225.61

561.55

23,477.38

91.23

1,054.33

713.13

2,586.13

2,437.54 76.30

84.94

505.21

1,707.78

10,246.60

2,377.35

1.11

760.20

30,585.35

6,073.59

1,821.76

Other uses

Native vegetation

Total area by state

Planted forests

630.14 72.65

752.11 0.00

426.91 0.00

0.00

0.00

0.00 0.00

0.00

5,676.23

16,950.32

36,057.12 23,797.28 238.16

Thus, Figure 12 shows the total area occupied (provided by ICONE), as well as the complementary area calculated for Brazil through the years, which is also used to calculate CO2 emissions/uptake from land-use change.

Figure 12: Area of the country occupied by agriculture, pasture and planted forests, and complementary area in the form of native vegetation and other uses, from 1990-2030

For Equation 35, the soil C stock is calculated for the existing land use for a specific year and 20 years earlier (for example: if the area under a specific land use did not change for the current year compared to 20 earlier, the difference between the C stocks is zero, so there are no CO2 emissions or uptake). With ∆C being the variation of C stocks in the soil (SOC) for a specific land use, calculated for the desired time (SOC0) and 20 years earlier (SOC0-20); D is the time during which the balance of C in the soil is affected, in other words, 20 years. The soil C stock from a specific land use is calculated in relation to the carbon stock in a reference condition (in other words, native vegetation that was not degraded or improved), as shown in Equation 36: Where SOCref is the soil C stock under native vegetation for the area of interest; F is the change factor of soil C; A is the area under a specific use. The change factor for soil C stock is a product of three factors: (1) land use; (2) land

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

83

84

preparation; and (3) residue input. The change factor for soil C stock indicates the amount of C from the native vegetation that remains in the soil after 20 years of land use. The lower the value of F, the greater is the loss of soil C due to the respective land use. According to the IPCC (2006), there is no value for F estimated for different annual crops. However, for the present study, estimates for F are made for each of the six crops analyzed empirically, based on consultations with specialists and published work on C uptake per production system.

By analyzing the variations of soil C stock for the different land uses in a given region, one arrives at an estimate of C emissions or uptake. For example, if there is an increase in land use for a specific area with crops with a low F value compared to what was observed 20 years before, one arrives at a CO2 emissions estimate through the reduction of the soil C stock.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Using this method, the following hypotheses can be made: (i) over time, the C in the soil reaches a level of equilibrium that is specific to soil, climate, land-use, and management practices; and (ii) linear changes in soil C stocks occur during the transition towards a new equilibrium. Changes in soil C stock by crop

The methodology proposed by the IPCC (2006) enables CO2 emissions to be estimated based on the different land uses and recognized areas (country, state, region, micro-region, etc.). C emissions or uptake are calculated based on the balance of C stocks for every land use depending on the size of the area for each one. Thus, the final result is expressed in Mg CO2 for the total area. However, the model used to calculate total emissions from agriculture in this study requires an emissions factor per crop for each year of study, expressed as Mg CO2eq ha-1. A new methodology was developed to achieve this.

Figure 13 serves as an example: during year N, three crops (A, B and C) occupied a specific part of a known area. During year N+20, crop A lost land to crops B+C. Production systems for these crops may also have changed over time (ex.: conventional planting vs. zero tillage), in such a way that the respective change factors related to soil C stock, FA, FB and FC, have also changed from year N to year N+20. Carbon emissions or uptake from CO2 through the change of soil C stocks by crop, based on land-use change, are thus calculated as shown in Equation 37: ES (emission or uptake of C) = CF + ∆CV





(37)

Where, CF is the C stock in the area that has not changed over the years and ∆CV is the expansion of the crop area. It is assumed that the crop that expands its area is the one that causes the changes in the soil C stock. Thus, for a crop that only loses area, the ∆CV is zero.

Figure 13: Fictitious land-use change scheme for three crops (A, B, and C)

To calculate CF, the carbon stock calculation methodology was used, like in the IPCC (2006), with an alteration in the use of the change factor for soil C stock (F) (Equation 38). CF = (Cref x AN+20x ∆F)/(Dt)

(38)

Where, Cref is the C stock under native vegetation (C in the reference area); AN+20 is the area occupied by crops in the year N+20; and ∆F is the difference in F between years N and N+20. Thus, the management of the area does not change over time and the Fs remain the same, with ∆F being zero, hence CF is also zero; Dt is time (according to IPCC, 20 years).

To calculate ∆Cv, one must first calculate the sum of the total area that underwent a land-use change for crops in the area or region being studied, which could be the sum of the areas that were reduced, or those that increased, as they are equal in the module. Afterwards, for each crop whose area increased, the fraction of this increase (fr) should be calculated in relation to the total area reduced, or increased (always in a module). For crops whose area increased, the calculation of the emissions or uptake of carbon for this crop from the change in soil use (∆Cv) should consider its integrated effect on each crop whose area was reduced, using Equation 39: ∆Cv = {∑ [fra x (-Ari) x (FN+20(g)-FN(p))) x Cref}/∆t

(39)

Where fra is the fraction of the increased area; Ari is the reduced area for crop i; FN+20(g) is the change factor of soil C stock in time N+20 for the crop that gained area; and FN(p) is the change factor in soil C stock in time N for the crop that lost area. Estimate of soil C stock under native vegetation for each region

Carbon stocks under native vegetation were estimated for regions limited by the crossing of soil classes and vegetation. As the soil map for Brazil (MCT – www.mct.gov. br) shows many different categories, a simplification was done based on IPCC criteria

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

85

86

(2006) including texture, clay activity and drainage. For this study, the soil map was simplified in the following categories: Latossols (all latossols); other soils with low activity clay (Argissols, Cambissols, Planossols, Plintossols, etc.); soils with high activity clay (Chernossols, Vertissols, Alissols); Arenous soils (Entissols, Espodossols, etc.), and Hydromorphic soils (Map 6A). The Organossol category was initially considered separately, but since it occupies a very small area in the country, it was included in the Hydromorphic soils category. For the map of vegetation, the Pampas, Cerrados, Shrubland, Amazon Rainforest, deciduous forest and Atlantic Forest (Map 6B) were considered.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 6: Simplification of the soils map for Brazil with six soil categories (A); map of vegetation, with the six categories considered; (B) visualization of soil C stocks under native vegetation in Brazil (C)

Thirty soil x vegetation combinations were created using simplified soil and vegetation classifications; each was attributed a value of soil C stock based on available published data and soil data bases in the EMBRAPA Agrogas Network, whose map appears in Map 6C. The stocks calculated for the southern and southeastern regions, which were 51.12 Mg C ha-1 and 47.12 Mg C ha-1, are different from those reported for the same regions in the country’s first greenhouse gas inventory (www.mct.gov.br), which were 60.5 and 40.3 Mg ha-1, respectively. Data estimated in this study for other regions seem reasonable. Since the set of data available in this study was inferior to that used in the country inventory, stocks from the inventory from the southern and southeastern regions were considered. Thus, soil C stocks under native vegetation in the country were established for each region of the BLUM Model, as per Table 19. Table 19: Soil C stock under native vegetation for each region of the BLUM model

South

Southeast

BLUM Region

Central-West

Northern Amazon Northeast Coast MAPITO-Bahia

60.5 40.3 40.1 46.5 26.2 35.0

Original C Stock (Mg/ha)

It is not easy to attribute values to the change factors for soil C stocks (F) per crop. For example, it is impossible to separate soybean and corn production systems. Soybean production areas are generally rotated with corn in the summer. In the winter, wheat and oats are the most prevalent crops in Southern Brazil, while second-crop winter corn predominates in other regions. Thus, soybean and corn make up one production system, with few options used effectively for winter crops. For this study, the effect of soybean was considered the same as corn and included the use of winter crops. In the case of cotton, monoculture production systems are still used, and production areas are concentrated in the Cerrado, specifically in Mato Grosso, Bahia and Goias (Lamas, 2008). It is a crop that produces a little less residue than soybean and corn, which would justify a smaller F. However, they are also rotated with soybean. Similarly, the first harvest bean crop produced little residue and was then followed by the second harvest bean crop, principally in Minas Gerais, the largest producer in Brazil (EMBRAPA Arroz e Feijão [Rice and Beans]) – www.cnpaf.embrapa.br). It was found to have a similar effect as the cotton plant with regard to the soil C stock. Wetland rice is also a monoculture, and has great potential for soil C accumulation (IPCC, 2006) when irrigated or flooded. Under dry conditions, rice is used in areas that have been recently deforested, or that have been under pasture for long periods of time. For production from residues with a high C/N ratio, and depending upon the conditions of use, a slightly higher value of F is attributed to the rice crop than that of soybean-corn and cotton. The sugar cane crop is continuous and has great potential for maintaining soil C reserves (Cerri et al., 2007; Soares et al., 2009). Estimate of change in soil C stock for each region

Soil C stocks for a specific land use require the change factor for soil C stock, in addition to the size of the area under use and the C stock under native vegetation. These factors were not defined for the crops used in this study, so factors for each type of use were estimated based on criteria suggested by the IPCC (2006) and the knowledge of researchers who are part of the study team.

Statistics on the use of zero tillage in Brazil are not official and the area has not been monitored over the past five years. In the current scenario, about 77 percent of the area under zero tillage is associated with corn and soybean crops, and does not exceed 25 million hectares (Dr. Maury Sodda, Federação Brasileira de Plantio Direto na Palha [Brazilian Federation for Zero Tillage into Crop Residues) - FEBRAPDP, Ponta Grossa, PR, personal communication). Considering the crops used in this study, with the exception of sugar cane and pastures, this area is equal to 66 percent of the area planted in 2008. According to FEBRAPDP, this proportion should be maintained if there is no need to change, i.e. if zero tillage is done in a way that is not recommended (Saturnino & Landers, 1997), and problems of soil compacting, pests and diseases prompt some farmers to revert to the conventional planting system. Thus, in the Reference Scenario, or baseline, it is considered that 66 percent of the area planted with agricultural crops, except for sugar cane, will be kept under zero tillage until 2030. Zero tillage will be used for 77 percent of the area planted with corn and soybean and 8 percent of the area

87

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The stock change factor depends on the type of land use implemented (continuous agriculture, pasture, irrigated rice, etc.), soil disturbance (conventional preparation, zero tillage, etc.) and quantity of residue deposited (basically the quantity of straw and roots returned to the soil after each cycle).

88

planted with other crops used in this study. The area under zero tillage considered for the period from 1989 to 2005, was the one available in FEBRAPDP (www.febrapdp. org.br). Conventional soil preparation systems used for planting beans generally reduce the soil C stock (Zinn et al., 2005; Fernside & Barbosa, 1998) more than in the absence of disturbance in the zero tillage system (Zinn et al., 2005; Cerri et al., 2007).

In the present study, the change factor for soil C stock was between 10 and 20 percent, depending on the region, with greater variations for zero tillage (Table 20). In this case, it should be considered that winter crops are diversified and the premises of the system are complied with, thus allowing soil C accumulation (Boddey et al., 2006). Factors considered for conventional planting mean that land use reduces the C stock to levels that are 40-70 percent of those under native vegetation, as observed in samplings done in the country (Zinn et al., 2005). Table 20: Change factors for soil C stock for each type of land use

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Land use

South

Southeast

CentralWest

Northern Amazon

Northeast Coast

MAPITOBahia

Cotton

0.50-0.60 0.40-0.50

0.40-0.50

0.40-0.50

0.45-0.55

0.40-0.50

Corn

0.50-0.60 0.40-0.50

0.40-0.50

0.40-0.50

0.45-0.55

0.40-0.50

Rice

Beans

Soybean

Sugar Cane Degraded Pasture

0.70-0.90 0.45-0.55 0.50-0.60 0.40-0.50 0.50-0.60 0.40-0.50 0.85-0.90 0.85-0.90 0.80

0.80

Planted For0.80 est

0.80

Productive Pasture Other uses

Native vegetation

1.00 0.85 1.00

1.00 0.85 1.00

0.45-0.55 0.40-0.50 0.40-0.50 0.85-0.90 0.80 1.00 0.80 0.85 1.00

0.45-0.55 0.40-0.50 0.40-0.50 0.85-0.90 0.80 0.90 0.80 0.85 1.00

0.50-0.55 0.45-0.55 0.45-0.55 0.85-0.90 0.80 0.90 0.80 0.85 1.00

0.45-0.55 0.40-0.50 0.40-0.50 0.85-0.90 0.80 1.00 0.80 0.85 1.00

Note: The impact factor for crops varied depending on the proportion of adoption of zero tillage. The impact factor for the sugar cane crop varied depending on the proportion harvested without burning.

Although there are some statistics on pasture area in the country, it is not clear what proportion is considered low productivity or in a process of degradation. In the early 90s, some studies suggested that about 50 million hectares of pasture in the Cerrado areas were showing some degree of degradation (Macedo & Zimmer, 1993; Macedo, 1995). There are no recent estimates available, but it is believed that over 50 percent of the cultivated pasture area is degraded. Under these conditions, the pasture produces less organic matter and the soil tends to present lower C stocks than soil under native vegetation or productive pasture (Fernside et al., 1998; Fisher et al., 2007; Zinn et al., 2005), although reports of soils from pasture areas with more C than those under na-

Lima et al. (2008) concluded that 34 years after removing native vegetation for establishing eucalyptus plantations, the soil C stocks were reduced to 77 percent of the original. No other related studies have been conducted on Brazilian soils. For the representativeness of the study of Lima et al. (2008), the amount of 0.8 was considered for the change factor for soil C stock for planted forests. The area maintained under “native vegetation” has a stock change factor equal to 1. Areas under “other uses”, meaning abandoned areas, areas with other crops, or even those not used for agriculture, had change factors for soil C stocks equal to 0.85. Estimate of N2O emissions from the change in soil C stocks

When the soil loses C due to land use, a certain amount of mineral N is released into the soil. It is considered that 1 percent of mineralized N is released into the atmosphere as N2O (IPCC, 2006). According to the IPCC method (2006), N2O emissions are computed only when the balance of soil C stock for a given region is negative, or when there are CO2 emissions due to a particular land use. To calculate N2O emissions this way, changes in soil C stocks are computed as described above. Since the organic material in the soil has a C/N ratio of about 12 (Sisti et al., 2004), every 12 units of C that mineralize until CO2 release a unit of N in mineral form. Soon thereafter, 1/12 of the quantity of C in the soil lost from a given region represents the quantity of N mineralized. N2O emissions correspond to 1 percent of mineralized N. Estimate of N2O emissions from the soil from the increase in available N with the application of nitrogenated fertilizers and decomposition of residues

N2O emissions increase with the rise of the quantity of mineral N in the soil (Jantalia et al., 2006). According to the IPCC (2006), 1 percent of the N applied as fertilizer or mineralized from residues left after each crop cycle is emitted as N2O.

Indirect N-N2O losses are estimated based on the amount of N that enters the soil through fertilizer and the mineralization of residues. Since there are no data that enable the processes of loss per region to be differentiated, a study of the indirect losses would imply the use of a constant loss factor of N (nitrate lixiviation and ammonia volatilization) for all areas, and an indirect emissions factor would be applied to this lost quantity, the same for all regions.

In the case of fertilizers, it was considered that 1 percent of the quantity of N applied is emitted as N-N2O. If the losses are considered, the factors result in a 10 percent loss of N by ammonia volatilization, with 1 percent lost as N2O, 30 percent of N lost through nitrate lixiviation, and 0.75 percent lost as N2O. This means that, assuming that these processes occur in equal intensity for dryland crops fertilized with N in the country, the

89

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

tive vegetation are not uncommon (Fisher et al., 2007; Zinn et al., 2005). In this study, pasture area is considered underutilized based on data on the herd and the pasture area, which would correspond to a carrying capacity of less than 0.55 head ha-1. Based on the work of Fearnside & Barbosa (1998), the change factor of the soil C stock under productive pasture for Amazônia is 0.9. For the South, Southeast, Central-West and MAPITO, the factor is equal to 1 (Fisher et al., 2007). For the Northeast it is 0.9 due to the more limiting climatic conditions. According to Fisher et al. (2007), degraded pasture area would have about 20 percent less C stocked in the soil than productive pastures. Thus, the change factor for soil C stock considered for underutilized pasture was 0.8.

90

emissions factor considered (direct and indirect losses) is 1.325 percent. In the case of residues and mineralized N in the soil, the 1 percent factor was used for direct emissions. As these sources do not involve emissions from ammonia volatilization, indirect emissions are attributed to nitrate lixiviation. Thus, the emissions factor considered is 1.225 percent. In the case of wetland rice cultivation, 0.3 percent was the emissions factor used for direct emissions (IPCC, 2006). The factor considered for fertilizers for this condition is therefore 0.625, and 5.25 for other sources. Estimate of CH4 emissions in wetland rice production areas

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In this study, the wetland rice production area corresponds to the entire area planted with rice in Southern Brazil. Other regions were considered highland rice producers. In this case, methane production was only calculated for the rice-growing areas of the southern region. Results of research conducted in that region by Dr. Magda Aparecida de Lima, coordinator of one of the sections of this study, suggest that the methane emissions factor of 210 kg ha-1 would be reasonable for use in this study (Lima, 2009). The use of a single factor follows the Tier 1 of the IPCC methodology of 1996, which is annual and is not disaggregated according to the irrigation regime. In fact, ICONE’s landuse survey does not define how the land will be used in rice cultivation, so the use of the more complex methodology would not reduce the uncertainty of the study. In the wetland rice system in Southern Brazil, it was shown that the use of zero tillage reduced methane emissions by about 15 percent (Lima, 2009), an amount considered to be the effect of zero tillage on this crop in the present study. Estimate of N2O and CH4 emissions from burning sugar cane straw for the harvest

The IPCC methodology (2006) was adopted for this estimate, which calculates N2O and CH4 production as a function of the quantity of dry biomass produced and burned. An 80 percent efficiency rate for burning was taken into account (amount available for sugar cane in IPCC, 2006). With this method, each ton of cane straw burned produces 2.7 kg of methane and 0.07 kg of N2O. The total amount of dry straw (MS) produced per hectare of cane burned was 13 tons, based on a study done with different plant varieties (Xavier, 2006). These calculations are represented in Equations 40 and 41:

2.1.4.2.2 Greenhouse Gas Production from the Use of Fossil Energy

This type of GHG production occurs in agriculture from the burning of fossil fuels to generate energy for the synthesis, processing and transport of inputs, as well as from the execution of agricultural operations (soil preparation, planting, crop treatments, and harvesting).

Nitrogenated fertilizer is especially “costly” in terms of fossil energy consumption, as it is produced using the Häber-Bosch process, at high temperatures and pressure,

Agricultural operations include diesel oil for operating machinery and equipment, as well as fossil energy used in the production, maintenance and eventual disassembly and preparation of equipment (according to international standard ISO 14040), for which the Pimentel (1980) methodology was used as a basis. Agricultural operations for each crop were studied for the year 2008 through contacts with specialists and cooperatives, such as COCAMAR (PR), Fundação MT, COMIGO, etc. The total amount of energy expended for each agricultural operation for each crop was converted into CO2 equivalent, considering that the total amount of GHG emitted by burning diesel fuel, a standard fuel for generating all the energy for these operations. According to IPCC (2006), each GJ of energy generated by burning diesel oil releases 73.5 kg of CO2eq.

2.1.4.2.3 Synthesis of Emissions from Agricultural Activities

Total CO2eq emissions from agriculture amount to 2,047.9 million tons for the 2010 to 2030 period (Table 21) in the Reference Scenario. CO2 emissions occur from the emission of fossil energy during crop management, approximately 343.5 Mt CO2, and from the reduction in soil C stocks (585.2 Mt CO2), due to different types of land use, which vary over time. Table 21: CO2, CH4 and N2O accumulated emissions from agriculture from 2010-2030, expressed in CO2eq for the Reference Scenario Source

CO2

N2O

Variation in soil C stocks

585.2

94.8

Harvest residues

-

402.9

Fossil energy Fertilizers

Burned cane

Irrigated rice Total

343.5 -

-

-

928.7

-

175.9 12.0 -

685.6

Mt CO2eq -

CH4

-

-

86.5

347.1

433.6

Total

680.0 343.5

175.9

402.9 98.5

347.1

2,047.9

N2O emissions amount to 685.6 Mt CO2eq, of which 175.9 Mt CO2eq correspond to N added to the soil through fertilizers; 94.8 Mt CO2eq from the N mineralization in the

91

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

and fed by natural gas. Fossil energy used in urea production was calculated at 5.9 MJ kg-1 N, compared to 3.2 and 5.9 MJ kg-1 of P and K, respectively (Laegreid et al., 1999). Other fossil energy inputs come from herbicides and insecticides. Crops such as cotton, soybean and sugar cane consume a large quantity of pesticides and herbicides and the use of zero tillage substantially increases herbicide consumption even more. Due to the complicated synthesis of herbicides, manufacturing these products industrially requires large quantities of fossil energy, estimated at 452 MJ per kg of active ingredients. However, since the manufacturing and distribution of these products is discussed in other sections of this study, and published in another Summary Report, this section only covers emissions from agricultural operations.

soil; 12.0 Mt CO2eq from sugar cane burning; and 402.9 Mt CO2eq from N from harvest residues, with soybean residues contributing 51 percent of this total amount. Figure 14 shows how CO2, N2O, and CH4 emissions occur over the years, expressed in CO2 equivalents. The greatest variations are found with land use-related CO2 emissions. These also include fossil fuel emissions over time, which tend to be relatively constant, compared to the variations of land use-related CO2 emissions. Emissions reductions from burning during the cane harvest had little impact on total CO2eq. It has been noted that, although there are variations, emissions actually increase over time due to the expansion of the area occupied by agriculture, especially soybean, whose increase in the planted area, mainly pastures, has a strong impact on the reduction of soil C stocks, emitting CO2 and N2O. Moreover, as mentioned earlier, residues from this crop contribute a great deal of N2O emissions.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

92

Methane emissions were estimated to be 433.6 Mt CO2eq, only 86.5 Mt CO2eq of which are from sugar cane burning. Eighty percent of methane emissions are from rice grown under flooded conditions.

Figure 14: CO2, N2O, and CH4 emissions from agriculture during the 2010-2030 period, expressed in CO2 equivalents in the Reference Scenario

The states of Mato Grosso and Rio Grande do Sul had the greatest accumulations of GHG emitted from agriculture from 2010-2030 (Map 7). In Rio Grande do Sul, methane emissions from wetland rice represent about 50 percent of total emissions. In Mato Grosso, emissions are the result of different types of agricultural land use. The states of Paraná, São Paulo, Goiás and Minas Gerais, where agricultural activity is intense, also

show total emissions of about 100-300 Mt CO2eq. In conclusion, it is estimated that total emissions from land use under agriculture will increase about 42 percent, from 74 Mt CO2eq to 111 Mt CO2eq from 2010 to 2030.

2.1.5 Emissions from Deforestation Brazil has the largest area of tropical forests in the world (representing 56 percent of the national territory – Table 22). There is tremendous diversity in the forest formations throughout the country, which are distributed in its six different biomes, the main ones being tropical forests (dense and open), which occur mainly in the north; araucaria forests in the south; seasonal forests (deciduous and semi-deciduous), found principally in the southeast; tropical Atlantic Forest with a coastal distribution; shrubland in the northeast; campinaramas in the northwestern part of the state of Amazonas and in Roraima; a variety of savannas and forest formations in the Cerrado in the

93

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 7: Total GHG emissions in CO2 equivalent (millions of tons) by state, resulting from agricultural land use

94

country’s central region and production forests that represent 1 percent of the forest cover of Brazil, principally in the Atlantic Forest biome. With the greatest diversity on the planet, Brazilian forests are essential for maintaining ecological balance on both regional and global scales, including the rainfall regime, freshwater supply, biodiversity conservation, preservation of traditional crops, and mitigation of climate change. Table 22: Land use in Brazil between 1990 and 2005

Forests

Type

Other uses (agriculture, livestock, urban, infrastructure, etc)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Water depth Total

1990

(thous. km²)

2000

(thous. km²)

2005

(thous. km²)

5,200.27

4,932.13

4,776.98

195.32

159.32

159.32

3,155.29 8,514.88

3,423.43 8,514.88

Source: IBGE (2006)

3,578.58 8,514.88

Nevertheless, Brazilian forests are being converted for a multitude of purposes at a rapid pace (about 420 thousand km over the past 20 years21). The Amazon Rainforest lost about 18 percent, or a total of 720 thousand km of its original forest area between 1970 and 2007 (INPE, 2009); the Cerrado lost about 20 percent of its original area between 1990 and 2005, and the Atlantic Forest about 8 percent during the same period (SOSMA, 2005). Given the magnitude and complexity of the Brazilian territory, there is a variety of causes and processes involved in the conversion of the native plant cover. The deforestation of the Cerrado biome during the abovementioned period is generally attributed to the expansion of grain and livestock cultivation (Machado et al., 2004; Eva et al., 2004; Ferreira et al., 2007), while deforestation in the Atlantic Forest biome during the same period occurred due to real estate speculation and the uncontrollable growth of large urban centres (Texeira et al., 2009). The causes of the deforestation of the Amazon Rainforest are complex (Soares-Filho et al., 200822) and involve inter-related regional, national and global socioeconomic and political factors (Soares-Filho et al., 2005; Nesptad et al., 2006). Among the principal causes are the region’s original colonization policies and fiscal incentives for the development of activities that triggered an intense migration to the area. Later on, other processes also played a role, such as the expansion of the international market for agricultural and livestock products, supported by the strengthening of the value of the Brazilian real in relation to the dollar; the expansion of wood and livestock exploitation; the increase of agrobusiness; infrastructure development with the opening up and paving of roads; and the absence and inefficiency of state policies that are unable to halt illegal deforestation and regularize land tenure in the region (Soares-Filho et al., 2008). 21 22

About 28.4 thousand km²/year. Soares-Filho, B.S. et al. Nexos entre as dimensões socioeconômicas e o desmatamento: a caminho de um modelo integrado.In: Mateus Batistella; Diogenes Alves; Emilio Moran, (Org.). Amazônia. Natureza e Sociedade em Transformação. 1 ed. São Paulo: Edusp, 2008, v. 1.

As a result of these enormous changes, and despite the recent drop in deforestation rates in the Amazon since 2005, as confirmed by INPE (2009), the country’s great heritage of forest resources is threatened (Table 23). Table 23: Risk of extinction of arboreal forest species in Brazil in 2000

Not threatened with extinction

7,559

Vulnerable

187

Critically threatened Threatened Total

34

100

7,880

Source: FAO (2005)

Quantity

95.9% 0.4%

%

95

1.3% 2.4%

100.0%

The role of forests in CO2 emissions, the main gas that contributes to the greenhouse effect, is pivotal, as forests harbor substantial carbon reserves. It is estimated that Brazilian forests store a total of approximately 54 billion tons of carbon. In this context, deforestation in Brazil is the process that contributes most to CO2 emissions, with a total of 70 percent of national emissions. The accountability of emissions from deforestation brings Brazil up to fifth place in the list of emitting countries, representing 5 percent of the global total (CAIT_WRI, 2007).

Estimated deforestation in the Reference Scenario (Figure 15 and Map 8) is higher than estimates from the BLUM model projections. As explained in the methodology, SIMBRASIL studies Amazon deforestation rates that are calculated based on the spatial lag regression model, as well as projections of the demand for land for cultivation. The objective is to incorporate the indirect effect of agricultural expansion into the land-use dynamic in the Amazon.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Types

Figure 15: Deforestation dynamic in the three main biomes in Brazil in the Reference Scenario (km2/year)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

96

It is important to note that the modeling for the reference scenario does not incorporate the possible effects of the new objectives of reduced deforestation as announced in the National Plan on Climate Change, such as the strict observance of the Forest Code. These objectives, as well as the observance of new laws on areas of permanent preservation and legal reserves, will be considered within the framework of a legal scenario that constitutes one of the elements of the low-carbon scenario proposed by this study, whose results will be presented later on. Map 8: Deforestation in the Reference Scenario (2010-2030)

Deforestation-related emissions from the conversion of forests into pastures were calculated using the IPCC methodology (2003). In this case, the difference between carbon stock from from plant cover from period 1 (t1) and period 2 (t2), in the case of pasture, was used. Since biomass from the different plant physiognomies varies spatially, a carbon stock mosaic was assembled (Map 9) and adopted as a basis for calculations. The amounts for this mosaic are between 0 and 276.5 tons of C/hectare. The amount of 4 tons of C/hectare was attributed for pastures. Map 9: Carbon Stock Mosaic

This data was compiled from various sources. A biomass map was used for the Amazon (Saatchi et al., 2007). For the rest of the country, data was compiled associating an average biomass value according to the recommendations of the Brazilian Emissions Inventory (MCT, 2004) for each plant physiognomy from the PROBIO map (MMA, 2007). CO2 uptake from regenerating secondary forests and production forests was estimated using maps showing the biomass sequestration potential of natural vegetation and forestry, respectively, provided by the Initiativa Verde.

Emissions from land-use change are responsible for the positive balance obtained (Figure 16). These emissions are principally the result of the deforestation of forest remnants. About 66.4 to 81.3 percent of total emissions are due to land-use change from the deforestation of the Amazon, and 59 to 67 percent of total projected emis-

97

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In addition to the spatially explicit land-use change model, there is an online emissions accountability model for land use and land-use change. This model incorporates a table of carbon emissions/uptake for each use, and accordingly, the land-use transition. The data was compiled from various sources. A biomass map was used in Amazônia (Saatchi et al., 2007), while for the rest of the country, data was compiled associating an average biomass amount to each plant physiognomy from the PROBIO map (MMA, 2007), based on recommendations from the Brazilian Emissions Inventory (MCT, 2004). The uptake of CO2 via secondary regenerating forests and production forests was estimated using maps showing biomass removal potential through the use of natural vegetation and silviculture, respectively, provided by the consulting firm Initiativa Verde.

Figure 16: Emissions from land use in the Reference Scenario

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

98

sions. Deforestation in the rest of the country is responsible for 30.8 to 13.4 percent of total emissions from land-use change. Emissions from changes in other uses are attributed to other types of land conversion, whose participation is between 2.75 and 5 percent of the total amount between the initial and final years of the study, with a maximum participation of 6.38 percent in 2010.

2.2 Carbon Uptake Through Reforestation

This section presents the databases and the model used to analyze carbon uptake through reforestation, the purpose of which was to evaluate carbon uptake potential estimates in the Cerrado and Atlantic Forest biomes. For the systematization of the basic information, the databases were organized in the SHP format, which were later exported for the spreadsheet in XLS format.

A potential plant biomass model was developed for the Cerrado and Atlantic Forest biomes, estimating the atmospheric carbon removal potential through forest restoration in permanent preservation and legal reserve areas, and establishing carbon uptake potential through the abovementioned reforestation for each micro-region

located in these biomes. Parallel to the application of this model to the restoration of native vegetation, the carbon uptake potential through the planting of energy forests, managing the average annual increment also by micro-region (tCO2e/ha/year), was estimated as well.

2.2.1 Methodology

99

The modeling procedures follow the pattern described by Iverson et al. (1994), who developed a spatial model for potential carbon uptake in forests. This means estimating the potential quantity of plant biomass above and below the soil, excluding anthropic interventions and natural disturbances such as fire, storms, and excessively long dry periods. The model developed for carbon uptake through reforestation assumes that the density of plant biomass that a specific region can support is dependent on climatic, topographic and edaphic conditions, but without taking into consideration the cumulative impact of anthropic activities such as pollution, wood extraction, landuse change, etc.

The application of this biomass potential model is summed up in Equation 42, featuring the use of four main layers: IBP=I(ICMW)+I(rainfall)+I(topography)+I(soils)

(42)

Where ICMW = Índice Climático Modificado de Weck (1970)(Weck’s Modified Climatic Index), which includes figures such as temperature and length of the growing season; Precipitation = annual rainfall averages for each locality; Topography = altitude and gradient of land; Soils = classified according to texture and fertility.

An index was attributed to each layer (I) with the maximum value of 25 points, so that the maximum value possible for the model would be 100 points. The climatic index and average annual precipitation represent half of the IBP. The altitude and gradient variables form the topography layer, given that altitude received a maximum number of 13 points and the gradient a maximum of 12. Soil type (texture and fertility) represents the remaining 25 percent of the model (Figure 17).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

2.2.1.1 Details of the Potential Biomass Model

Figure 17: Diagram of potential carbon removals by reforestation for the Cerrado and Atlantic Forest biomes

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

100

Levels

Weck’s Modified Climatic Index

Weck (1961, apud Weck, 1970) developed an empirical model based on climatic data to determine the potential productivity of forests in Germany. Later on, the researcher expanded his work to include the tropics and developed the following empirical formula related to his index (Weck, 1970) (Equation 43), known as Weck’s Climatic Index (ICW):

(43)



Considering that dt(Celsius) is the diurnal difference between the maximum and minimum average temperatures of the hottest month of the growing season; S (hours) is the average day length during the growing season; P1(dm) is the number of months during which the average annual precipitation is less than 200 mm; P 2(dm) is the number of months during which the average annual precipitation exceeds 200 mm; G(months) is the length of the growing season, which corresponds to the number of months without hydrous deficit; H(%) is average humidity relative to the air; and Tm(Celsius) is the average temperature of the hottest month of the growing season. This index is based on the following premises: In the tropics, there is less respiration if the nocturnal temperature is low (dt).The net productivity of biomass is directly proportionate to day length. The relationship between net productivity and amounts of precipitation is not linear. A continuous increase in precipitation above 2000 mm/year will correspond to a successive reduction in the increase of net productivity. Net productivity is directly proportionate to the duration of the growing season.

Net productivity is directly proportionate to the relative humidity of the air (H)

which, in turn, is highly dependent on amounts of precipitation and existing plant cover.

Pluviometric precipitation has less of an effect on net productivity if the temperature during the growing season rises. Applying it to the estimate of current biomass in the tropical forests of Asia, Iverson et al. (1994) modified the Weck Climatic Index as follows in Equation 44:

101

These authors used the modified index as a basis for evidence that the proportion of biomass production per unit of area for biomass density is constant in all biomes or climatic regions (Brown & Lugo, 1982) in mature tropical forests. Moreover, total biomass is the result of the integration of net production as a function of the time it takes to reach maturity.

In this study, since the objective is to determine biomass potential, relative air humidity was excluded from the index (H), as it is a variable that is highly correlated with existing vegetation. Thus, the WMCI was used in the following simplified formula (Equation 45):

Growing Season (G)

(45)

In the tropics, the growing season corresponds to periods during which there is no hydrous deficit. This variable is strongly associated with the dry season, but presents variations depending on plant cover, type of land, and hydrographic basin. As the purpose of this study is to provide an estimate for plant biomass production in the Cerrado and Atlantic Forest biomes, it was considered that the periods of hydrous deficit would be the months without rain, as this would be an easy-to-obtain and highly reliable variable. Thus, the growing season (G) is defined as (Equation 46): G = 12 - S

(46)

where S = dry months; Without dry phase – considered an absence of hydrous deficit; Sub-dry – period of deficit equivalent to one month; 1-2 dry months – deficit equivalent to two months; 3 dry months = 3 months of deficit; 4 to 5 dry months = considered 5 months of hydrous deficit. Solarimetry

Data on daily sunshine (hours) and average day length (hours) during the growing season were used at this level. This section, together with rainfall and temperature data, was used for Weck’s Modified Climatic Index (1970). WMCI values were divided into 25 classes in a non-linear fashion. More values were

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

(44)

Figure 18: Points attributed to the WCMI values in the model. Modified by Iverson et al. (1994)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

102

grouped together in lower classes because the vegetation is more sensitive to WMCI in the dry extreme; a pattern similar to the one used by Holdridge (1967). The first 16 classes showed an increment of 25 units; the next 6 classes an increment of 50 units; and the last 3, an increment of 100 units (Figure 18).

Rainfall

The correlation between precipitation classes and biomass density was assumed to be positive up to the amount of 3,200 mm/year, at which point this correlation begins to have a negative effect on biomass (Brown & Lugo, 1982). According to observations made by Brown et al. (1993), 400 mm annually is the minimum amount needed to support arboreal formations (Figure 19). Figure 19: Points attributed to the amounts of rainfall in the model, according to Iverson et al. (1994)

Altitude

Different authors report that altitudinal zoning changes the vegetation patterns, mainly through the climatic variations associated with each altitude class. Therefore, an altitude layer was included in our model. The altitude classes were based on the 103 suggestions of Iverson et al. (1994) and were divided into five classes according to the general variations in vegetation depending on the altitude of the Cerrado and Atlantic Forest biomes: 0–15 m – Coastal forest – mangrove

51–500 m – Sub-mountainous formation 501–1.500 m – Mountain formation + 1.501 – High mountain formation

The points attributed to each altitude class are shown in Figure 20. The 0-15 meter altitude class received fewer points due to the fact that these plant formations occur throughout the coast and normally have lower biomass values than lowland forests. Figure 20: Points attributed to altitude classes in the model, modified by Iverson et al. (1994)

Slope

According to Iverson et al. (1994), slope is one of the variables whose correlation with forest biomass is extremely variable. Considerable amounts of biomass were already found on relatively sloped lands compared to the amounts found in adjacent flat areas. Thus, the slope of the land gets relatively low points in this model, going from

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

16–50 m – Lowland formation

12 points (up to a 10 percent gradient) to 8 points (gradients greater than 20 percent) (Figure 21). Figure 21: Points attributed to the degree of incline of the land, modified by Iverson et al. (1994)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

104

Soils

Different edaphic factors affect the biomass distribution patterns in tropical forests (Whitmore, 1984). Forest productivity is generally related to soil fertility, but this potential is much more affected by climatic factors and texture, as large amounts of biomass have been reported in the Amazon region in forests growing on nutrient-poor soils, but with suitable texture (Laurance et al., 1999; Saatchi et al., 2007). However, to standardize the study, the model adopted the IBGE soil fertility map in Brazil, where different degrees of fertility assume the amounts presented in Table 24.



Table 24: Points of the different IBGE fertility classes Points 25 22

Fertility

Average to high

Low to medium

18

Low to very low

7

Very low

15 12

Low

Low to very low

Thus, for the development and application of the model, the following data need to be examined: Rainfall;

Solarimetry; Soils;

Temperature;

Growing season;

Weck’s Modified Climatic Index; Altitude;

Declivity.

105

Data bases used in the model

Sources used in the data bases vary. The environmental data base stores information extracted or processed from official cartographic data (IBGE), thematic maps (EMBRAPA and JRC) and spatial models (SRTM, WorldClim). Environmental databases

Data availability for the different environmental layers in Brazil in 2008 is much greater than that found by Brown & Iverson (1994) for Southwest Asia, or by Brown & Gaston for tropical Africa; for the Cerrado and Atlantic Forest, data is currently available in greater spatial resolution and thematic diversity. Two important sources, which were not available in the 90s and are available today for almost the entire planet, are based on SRTM and WorldClim topography.

The SRTM Base (Shuttle Radar Topography Mission) consists of a topographic investigation done through orbital radars. The international topographic mission was lead by NASA, and its measures were realized in the year 2000. The resulting digital land model has approximate spatial resolution of 60 m and a margin of error of ± 6.2 in South America.

The WorldClim database was developed by the University of California and published in 2005. Data from tens of thousands of meteorological observation posts around the planet were geo-referenced and interpolated, resulting in climatological maps derived from temperature and precipitation data. The data base has 55 grids that describe each month: minimum temperature, maximum temperature and precipitation, and bioclimatic variables that are relevant for ecological modeling. The grids cover almost the entire globe, except for the polar regions, with a spatial resolution of 30’.

Although this is a consistent data base established for the international scientific community, we feel that an analysis of the specific area of study might be more accurate. It was thus decided to review the WorldClim data in order to evaluate its accuracy for the study area. In parallel, the re-sampling of the digital model of the SRTM land was audited for a 30’ slope. The schedule for the WorldClim audit included the following steps:

Tabulation of available climatic data from the EMPRAPA database “Brazilian climatic database” for 35 small cities that are distributed regularly along WorldClim Zone

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The environmental data bases consist of a set of grids (geo-referenced matrixes) standardized by the overlapping of layers, with auxiliary data in vector format, such as hydrography, official boundaries of the Cerrado and Atlantic Forest biomes, political division of the territory and official mapping of the Brazilian ecosystems.

106

34, which includes the study area. Seventeen meteorological posts were selected from this initial group, and these were described with regard to coordinates, altitude, average temperature, annual precipitation and number of months with precipitation lower than 50 mm.

Consultation with WorldClim for the cell amounts (pixels) which include coordinates from the 17 meteorological posts selected. The amounts consulted were: altitude (re-sampled from SRTM), average temperature, annual precipitation, number of months with less than 50 mm of precipitation (this last layer was produced by the reclassification of precipitation for each month [amount50mm=0] and for the summary of all the months of the year). The amounts were collated and tests comparing the results were conducted (ANOVA). All the consultations at WorldClim, analyses of reclassification and overlap were conducted using the software ArcView 9.3.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

For all the comparisons realized, the WorldClim database is not statistically different from the EMBRAPA environmental database (Levene test, p>0.92). Thus, the WorldClim database can be used for modeling, as can be observed in Figure 22. Figure 22: Graph of the box-plot where the distribution of the amounts for altitude, rainfall, and months of hydrous deficit may be observed for the EMBRAPA and WorldClim databases

The climatic variations considered as described in the potential biomass variation were organized in grids measuring 30° latitude by 30° longitude. The spatial resolution is 30 seconds (0.93 km at the equator), resulting in a grid with 3,600 X 3,600 cells. WGS 84 datum was adopted once Brazil officially abandoned the use of SAD 69 and adopted SIRGAS. WGS 84 datum is standard for WorldClim and SRTM, and presents only about a 107 1cm difference for SIRGAS. Table 25: Entries in the environmental database

Variables

2

Slope

Digital land model

1

4

SRTM-NASA

Annual precipitation

(mm)

Monthly precipitation > 2m – 2m

6

(%)

(m)

Duration of dry weather

5

Base

Altimetry

II Precipitation

3

Unit

(month) (month)

Monthly precipitation < 2m, >2m=2 (month) III Temperature

Average temperature of the hottest (°C) month

7

IV Solarimetry

8

Daily insolation V Soil

9

10

11

Fertility

VI Soil cover

Map of plant cover in Brazil VII Complementary Data

Limits of biomes of interest

Biomes

SRTM-NASA WorldClim WorldClim WorldClim WorldClim WorldClim

(hs)

Solarimetric atlas

(FAO)

JRC / EMBRAPA

Classes

Databases - Maps

IBGE/EMBRAPA

IBGE

The study is limited to the Cerrado and Atlantic Forest biomes. According to the IBGE, the Cerrado biome has an area of 2,063,001 km2 and the Atlantic Forest biome 1,112,170 km2. The two have very different floristic compositions: arboreal formations predominate in the Atlantic Forest, and savannas and cerrado fields are more common in the Cerrado, presenting, nevertheless, very distinct potential for carbon uptake. Map 10 shows the limits of Cerrado and Atlantic Forest biomes.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

ID

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

108

Map 10: Boundaries of the Cerrado and the Atlantic Forest, extracted from the Map of Brazilian Biomes, produced in 2004 by IBGE in cooperation with the Environment Ministry. This map indicates the boundaries adopted for the area of the study.

Altitude

The relationship between altitude and biomass density is not linear: the extreme altitudes have low amounts of biomass density potential, such as in coastal formations and mountain tops. The high altitudes tend to have little impact on the study area, as Brazil’s altitudes are low on average and the highest are well below the limit for forest formations (3,750m).

The altitude with the greatest biomass productivity would be between 16 and 750 meters, which covers almost the entire country. Only areas in bright blue or brown have restrictions for biomass development (Map 11).

Map 11: Altimetry based on the digital SRTM land model. Original data offer an average altitude with 3’’ resolution. The model presented was re-modeled for 30’’.

Declivity

According to Iverson et al. (1994) large amounts of biomass density can also be found in areas of accentuated declivity, but there is a tendency for flatter areas to present higher average densities than sloping areas. The flatter areas are shown in yellow 109 in Map 12. The low declivity levels found are justified by the smoothing effect of the remodeling for 30’’; however this is not an average declivity model.

Rainfall

This is one of the main variables for modeling biomass potential. The methodology proposed by Iverson et al. (1994) anticipates two entries for this variable: average annual precipitation participates in Weck’s Modified Climatic Index and is alone in the final overlap (Map 13). Map 13: Average annual precipitation in millimetres, from the WorldClim climatic database. Annual precipitation is directly related to potential productivity; biomass density tends to be greater in blue areas than in red.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 12: Declivity in percentages based on the digital land model re-modeled for 30’’

According to Iverson et al. (1994), an increment in precipitation above 3,000 mm would have a negative effect on biomass productivity. However, average precipitation in the study area does not exceed that amount.

110

Growing season

The growing season corresponds to the number of months where monthly rainfall is over 50 mm. During periods of low hydrous availability, plants close their stomata to avoid loss of humidity, and often lose their leaves. CO2 absorption from the atmosphere is either non-existent or low at this time, when hardly any plant growth occurs. In the Cerrado biome, the growing season lasts a maximum of nine months, such as in Campo Grande and Cuiabá. The most common duration is seven months, as is the case in Brasília, Goiânia and Belo Horizonte. The growing season in the Caatinga lasts less than five months, but it is outside of the study area (Map 14).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 14: Length of the growing season indicated by the sum of months with higher precipitation than 50 mm, obtained from the climatic WorldClim model

Average values obtained for the growing season in the Atlantic Forest are much higher than those of the Cerrado, although they vary greatly. Some parts of the Atlantic Forest present low values, like the area between the cities of Rio de Janeiro and Belo Horizonte, and between Aracaju and Maceió.

This grid was calculated based on the previous map so that the variation in the distribution of rainfall in the WMCI may be incorporated. Weck’s Modified Climatic Index states that the duration of the growth period is directly proportionate to net productivity. In Equation 47 (WMCI), the letter “G” represents the number of months of the growing season.

Average temperature of the hottest month of the year

This item takes into consideration the average temperature of the hottest month of the growing season. It negatively influences carbon uptake potential because the higher the temperature, the more respiration and less net absorption of CO2 (Map 15).

Map 15: Average temperature of the hottest month of the year (Celsius)

Database - Soils

The soil map used was the IBGE fertility map. A hierarchical fertility organization may be applied for most classes, which would be directly proportionate to the contribution of the edaphic component to biomass potential.

The predominance of “very low fertility” soils in the Cerrado is very high, above 70 percent. This class is very rare in the Atlantic Forest, as it is associated more with coastal regions, suggesting a high fertility x potential biomass correlation. On the other hand, more fertile soils from the Rio Paraná basin are equally divided between the Cerrado and Atlantic Forest (Map 16). Map 16: Soil fertility map for Brazil

Database – plant cover

This map shows the soil cover classification for Brazil. Both anthropic and natural areas adhere to the FAO classification system. The mapping was done by EMBRAPA

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

111

112

as an integral part of the GLS2000 (Global Landcover 2000), a Joint Research Center project, responsible for managing an accurate soil cover data base for International Conventions (Climate Change, Combatting Desertification, Ramsar, and Kyoto Protocol), and serving as an initial register (Map 17). There is a predominance of agricultural areas in the focus biomes. According to the FAO legend, the Cerrado’s features fall under the savanna classification, and are not characterized as forest formations in most of the remnant areas. For the Atlantic Forest biome, natural areas are classified principally as humid and swamp forests, but also with savannas. This map does not participate directly in the IBP calculation, but serves as a reference for calibration as it indicates potential biomass distribution, which is still greatly obscured by anthropic pressure.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 17: Map of plant cover in Brazil for 2000



Map 18: Map of Brazilian ecosystems, IBGE, representing an estimate of the distribution of “original” plant formations with a simplified legend indicating areas of transition

2.2.1.2 Carbon Uptake Potential through the Restoration of the Legal Reserves The potential for carbon uptake from the Atlantic Forest biome was between 183 and 661 tCO2e/ha. The micro-regions that show the greatest potential are generally located in the southern region and in the state of São Paulo, but specifically in the Serra do Mar. As expected, there was less potential in micro-regions with greater hydrous deficit and consequently a shorter growing season, such as in Sergipe. For the Cerrado biome, the carbon uptake potential was between 195 and 467 tCO2e/ha in Minas Gerais and Mato Grosso do Sul, respectively (Map 19).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

113

Map 19: Map of carbon uptake potential through the forest restoration of the Legal Reserve in the Cerrado and Atlantic Forest in tCO2/ha

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

114

2.2.1.3 Carbon Uptake Potential through the Restoration of Riverside Forests

With regard to carbon uptake potential in riverside forests in the Atlantic Forest biome, the lowest amount observed was in the state of Sergipe (231 tCO2/ha) and the highest amount was in Paraná (720 tCO2/ha). For the Cerrado biome, carbon uptake potential was between 220 and 552 tCO2/ha in the micro-regions of Barra (BA) and Ituverava (SP), respectively (Map 20). Map 20: Map of carbon uptake potential through the restoration of riverside forests in the Cerrado and Atlantic Forest biomes in tCO2/ha

Uptake through forest vegetation was only studied outside of the Amazon Rainforest, due to the uncertainty about its natural carbon balance (Nobre et al., 2001). However, secondary forests play a considerable role in the Cerrado and Atlantic Forest biomes, as a good part of these remnants constitute secondary forests that are about 50 to 60 years old. A logistical function for calculating potential biomass accumulation was assumed that takes into consideration plot location, and current and peak age. The latter 115 variable is assumed to be 200 years (Figure 23) for natural forests, while for reforested areas it is 20 years. Annual uptake, however, takes into account the regeneration of these forests at an average initial age of 60 years.

Note: The local potential in this example is equal to 100 t/ha.

2.2.1.4 Carbon Uptake Potential through Energy Forest Plantations in the Cerrado and Atlantic Forest Biomes CO2 absorption values by energy forests were defined in the productivity data obtained in the CCAP (2006), which presents minimum and maximum productivity (tCO2/ha/year) for this type of reforestation in Brazil. Although data from ABRAF (2007) shows significantly higher average productivity rates, they were initially not considered for the modeling, as they incorporate land management and correction practices that were not integrated into this model.

Lower amounts were generally observed in the northeast of Brazil, with a minimum of 28.85 tCO2/ha/year in the micro-region of Brumado in Bahia. The northeast average was about 39.83 tCO2/ha/year. For the other extreme, the greatest potential for annual

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 23: Logistical function of biomass uptake using local biomass potential and age of vegetation as parameters

carbon renewal was observed in the state of Paraná, in the micro-region of Capanema. The average for the southern part of the country was 48.45 tCO2/ha/year, while the average for the Southeast was 41.28 tCO2/ha/year (Map 21).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

116

Map 21: Forest productivity (tCO2/ha/year) for the Cerrado and Atlantic Forest biomes

2.2.2 Reference Scenario for Forest Restoration

Even with credit lines at reduced interest rates for native forest restoration, this activity has not been adopted voluntarily. For most rural property owners, reforestation or even simply fencing off areas of riparian forest or legal reserves implies a loss of productive areas. Even with non-recoverable financing, farmers are reluctant to give up PPAs for forest restoration and the subsequent elimination of this environmental liability. This is particularly relevant for small rural properties, of which permanent preservation areas may occupy a large part. Forest restoration generally occurs due to legal obligations in most cases, with terms of behavioral adjustments and forest compensations.

Thus, on the scale of this study, the evolution of native forest restoration in the Reference Scenario may be considered negligible. The largest private initiative of this type was in the State of São Paulo, with the restoration of 12.7 thousand hectares of riparian forest, implemented by the AES Tietê along the boundaries of reserves where it is the electricity generation concessionaire. Restoration occurs at a rate of 250 hectares/ year23. At the public level, the Government of São Paulo has made major efforts to restore native vegetation through the Projeto de Recuperação de Matas Ciliares (Riparian Forest Restoration Project - PRMC), whose resources come from the Global Environment Fund (GEF) and is expected to restore 1,500 hectares of riparian forest. However, the two initiatives combined do not cover 1.5 percent of the state’s current deficit, 23

http://www.aestiete.com.br/content/doc/ANEXO_III_Reflorestamento.pdf

which is over 1 million hectares.

Based on these observations, it was determined that although the contribution of forest restoration in the Reference Scenario is quantitatively rather limited, the experiences themselves are essential from the qualitative point of view in order to understand how to overcome the obstacles identified.

117

The Reference Scenario for the additional use of renewable charcoal was created based on the analysis of two groups of causal factors and their respective impacts on the participation of the three thermo-reduction agents in the Brazilian iron and steel industry. The first group, defined as the main group, constitutes the fundamental axis of the analysis, which is the generalized maintenance of the different obstacles identified, including investments; management; institutional and technological obstacles; and market gaps. The second group, the auxiliary group, is subordinate to the first and involves the degree of legislative deterrence or control over the use of non-renewable charcoal. In this group, two types of sub-scenarios were analyzed: (i) the enforcement of a legislative structure that tolerates the large-scale use of non-renewable charcoal and (ii) the enforcement of a legislative structure with a low tolerance level for the large-scale use of non-renewable charcoal.

The first group deals with the degree of scarcity of renewable charcoal and exposes the insufficient supply of inputs and absence of additional policies and incentives. The objective of the second group is to determine the thermo-reduction agent to be used in a scenario with scarce renewable charcoal, in other words, either coal or non-renewable charcoal. The second group thus assumes a purely auxiliary role, as the lack of additional planted forest, as shown in the first group is a strong indicator of the occurrence of net GHG emissions and the non-occurrence of net uptake. These two legal subscenarios essentially serve as key references regarding specific sources of emissions to be avoided in a low-carbon scenario.

Two possible Reference Scenarios were identified based on these two groups and are presented below. The resulting emissions appear in the summary report on energy, which only deals with the impacts of land-use change.

Reference Scenario with a low level of legal restrictions: In this Reference Scenario, the principal group of premises was adopted in combination with the sub-scenario (i) of the auxiliary group, which was the scarcity of planted forests combined with a low level of legal restrictions. As a result, it was assumed that a 3.7 percent increase per year for iron and steel production would be in keeping with the actual participation of thermo-reduction agents on the market24 in which 66 percent of the process of thermo-reduction necessary for producing iron and steel would continue to be based on the use of coke, 24 percent on the use of non-renewable charcoal, and 10 percent on the use of renewable charcoal, as illustrated in Figure 24 below:

24

According to estimates presented in the report on other topics of this study. The assumption of 3.7% growth per year was adopted based on the National Energy Plan.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

2.2.3 Non-renewable Charcoal and Planted Forests for Renewable Charcoal

Figure 24: Reference Scenario for charcoal with a low level of legal restrictions; participation of thermo-reduction agents in the Brazilian iron and steel-producing market

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

118

Source: Research by AMS, IBGE, ABRAF, SINDIFER

Therefore, the renewable charcoal deficit will increase in absolute terms, despite a modest growth in the planted area due to the proportional distribution over time.

Reference Scenario with a high level of legal restrictions: In this Reference Scenario, the principal group of premises was adopted in combination with the subscenario (ii) of the auxiliary group of premises: scarcity of planted forests for the production of renewable charcoal combined with a high level of legal restrictions. As a result of the increased legal restrictions regarding the use of non-renewable charcoal, the gradual decrease in the use of this thermo-reduction agent was assumed until its complete elimination starting in 2017. This scenario is based on the growing tendency to enforce legal deterrence against the use of non-renewable charcoal observed throughout the country, especially in the state of Minas Gerais, which is responsible for over 60 percent (SINDIFER) of iron and steel production using charcoal in Brazil25.

Although this scenario takes into consideration the increased rigor in the applicable legislation regarding the use of charcoal, it is important to note that the prevalence of the main group of barriers means that planted forests will continue to be scarce. Thus, the market segment that is currently based on non-renewable charcoal would then be based on coal. This scenario is based on the economic premise that the simple deterrence of the use of non-renewable charcoal does not automatically generate a relative increase in the supply of renewable charcoal. Figure 25 below illustrates the Reference Scenario with a high level of legal restrictions. 25

There are already stringent regulations and legal restrictions in Minas Gerais regarding the use of non-renewable charcoal, considering a period of transition that could last 7 to 10 years. These restrictions are the result of a sustainability pact signed between the productive sector, the state government and different non-governmental organizations.

Figure 25: Reference Scenario for charcoal with a high level of legal restrictions: participation of thermo-reduction agents in the Brazilian iron and steel-producing market

Source: Research in AMS, IBGE, ABRAF, SINDIFER

Given the growing tendency to impose legal restrictions on the use of non-renewable charcoal, this scenario is more likely to occur than the previous one. There is a stronger argument that one cannot assume that iron and steel producers, regardless of scale, would make the necessary investments in the expansion of the sector based on a non-sustainable option from the environmental and legal point of view (non-renewable charcoal). It is therefore possible and more conservative to assume that the opportunity cost for new investments in the iron and steel sector must be based on the use of mineral coke or renewable charcoal from new investments in planted forests.

Estimates of net and adjusted emissions generated in the aforementioned Reference Scenarios were calculated on this basis. The conclusion was reached that uptake in the scenarios with a high or low level of legal restrictions was the same, as the volume of planted forests is the same in both (Table 26).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

119

Table 26: Projection of CO2 Emissions and Uptake in the Reference Scenario (use of coal and/or non-renewable/renewable charcoal) - 2010 to 2030 (in thousand tCO2) 2010 Emissions 120 (Net) Emissions (Gross) Uptake*

2012

2014

2016

2018

2020

2022

57,917 62,283 66,977 72,025 77,453 83,291 89,568

2024 96,319

2026

2028

103,579 111,385

64,096 68,927 74,122 79,708 85,716 92,176 99,124 106,595 114,629 123,268 6.2

6.6

7.1

7.7

8.3

8.9

9.6

10.3

11.1

11.9

2030 119,780 132,559

Source: Adapted from data presented in “topic O” report (industry-related emissions)

*Note: Uptake by planted forests for the production of renewable charcoal in the Reference Scenarios

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 26: CO2 emissions projection for the Reference Scenario (charcoal)

Source: Adaptation of data presented in the “topic O” report (industry-related emissions)

2.3 Reference-Scenario Emissions Results

The study team generated an integrated Reference Scenario for LULUCF based on subsectoral analyses, using the emissions calculation methods indicated above, which were then integrated into the SIM Brazil model. Using these models made it possible to generate maps and tables that registered annual emissions and carbon uptake over the study period, calculated for each 1-km2 plot and integrated by micro-region, state, and country (Figure 27).

12.8

Figure 27: Reference Scenario results, emissions from land use and land-use change, 2009-30

Emissions from land-use change via deforestation account for the largest single share of total emissions from LULUCF—up to 533 Mt CO2e per year by 2030. Direct annual emissions from land use (agricultural production and livestock activities) increase over the period up to an annual rate of 383 Mt CO2e. The model shows a decrease in the annual carbon uptake rate, from 28 Mt CO2e in 2010 to 20 Mt CO2e in 2030. For the entire period considered, the net balance between land use, land-use change, and carbon uptake results in increased emissions, reaching about 895 Mt CO2e annually by 203026.

26

When calculating national carbon inventories, some countries consider the contribution of natural regrowth to carbon uptake; therefore, although this study does not compute the contribution in the carbon balance of LULUCF activities, it would be fair to add that information for purposes of comparison. If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by 109MtCO2 per year, thus reducing net emissions.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

121

3 Mitigation and Carbon Uptake Options 122

Based on the projected evolution of LULUCF-sector emissions in the Reference Scenario (Chapter 2), the study explored opportunities for reducing emissions and scaling up carbon uptake.

The study proposes a Low-carbon Scenario for land use and land-use change in Brazil focused mainly on (i) containing national demand for land for crop and pasture expansion in order to reduce emissions from deforestation, (ii) scaling up the identified mitigation options for agriculture and livestock, and (iii) maximizing carbon uptake potential associated with legal forest reserves and production forests.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Sections 3.1 to 3.6 identify mitigation options through agricultural production (zero tillage), charcoal (forestry-based carbon uptake), carbon uptake potential through reforestation (Cerrado and Atlantic Forest) and livestock and deforestation, respectively. Each of these five sections examines obstacles to the adoption of the respective mitigation measures, exploring ways to overcome them and accompanying measures.

3.1 Mitigation Options in Agriculture: Zero tillage

Zero tillage was found to be the most promising strategy in the agricultural sector for reducing agriculture-related GHG emissions. The adoption of the zero tillage system, which entails the elimination of soil disturbance, crop rotation and soil cover maintenance (Saturnino & Landers 1998), could increase soil C stocks to a higher level than the conventional land preparation system (Sisti et al., 2004; Diekow et al., 2005), close to native vegetation levels (Jantalia et al., 2007).

Monitoring of C in the soil has been done using the IPCC methodology (1996, 2006), which establishes a depth of 0-30 cm as a reference. Variations of C stocks depend on the area’s history. Published reports resulting from studies that evaluated the effect of zero tillage for cereals and soybean on the accumulation of C in the soil found that rates were 0.5 MgC ha-1 year-1 for the most superficial soil layer. However, this figure is difficult to extrapolate for the entire country, as areas under the same production system for over 20 years (IPCC, 2006) are not experiencing any more changes in soil C stock. Although other studies have shown greater variations in the length of time needed for establishing stocks (Coleman et al., 1997), 20 years may be enough for the tropical region, where the C cycle is faster.

Another important point has to do with the way zero tillage is conducted. In Southern Brazil, soybean and wheat are the main crops used in summer and winter, respectively, while in the Cerrados the system is based on the soy-corn mini-harvest. Problems related to pests and diseases in zero tillage may appear in monocultures in summer as well as winter (Derpsch, 1997), which is why crop rotation is essential for ensuring its success. Crop rotations under this system are also essential for reducing problems related to erosion, pests, and diseases, and to make the best use of accumulated organic matter in the soil. While zero tillage may increase soil C stocks, it can also provide conditions for greater denitrification activity and increase N2O emissions (Smith & Conen, 2004), although

this does not seem to occur in all soils. Jantalia et al. (2008) did not find any differences in the N2O in soils under zero tillage compared to those under conventional planting. With well-draining soils, such as latossols, which are common in most agricultural areas in the country, the use of zero tillage does not favor N2O emissions (Rochette, 2008).

In addition to favoring soil C accumulation, the use of zero tillage can reduce meth123 ane emissions from wetland rice, a recommended mitigation strategy for the production system (Wassmann et al., 2000). Studies done in wetland rice areas show different levels of reduction in CH4, amounting to an average of 48 percent (Table 27). The effect of zero tillage may be explained by the increase in electron receptors (Hanaki et al., 2002) and phototrophics (Harada et al., 2005), which reduce methanogenous activity. Research on wetland rice systems in Brazil shows that the absence of soil rotation in the zero tillage system can reduce methane emissions by about 15 percent (Lima, 2009). Local (time measured)

Philippines (cycle) Japan (2 years) Japan (2 years)

Conventional planting g CH4 m-2

27.2

48.4 45.8

Zero tillage

% Reduction in methane emissions

References

19.6

28

Wassmann et al., 2000 Shao et al., 2005

15.6 18.8

68 59

China (year)

117.9

19.7

83

China (cycle)

17.9

15.7

13

China (year) Japan (year)

117.9 17.9

Average

68.4 10.2

42

Hanaki et al., 2002 Hanaki et al., 2002 Shao et al., 2005

43

Harada et al., 2005

48

--

Benefits of zero tillage

Xiang et al., 2006

The zero tillage system is characterized by the elimination of soil disturbance, maintenance of soil continuously covered with crop residues, and the use of crop rotations. One of the great benefits of zero tillage is the decrease in soil erosion. In Southern Brazil, which is more mountainous, loss of soil from erosion can be substantial. Studies show that soil loss is reduced an average of 70 percent, and water loss an average of 20 percent (Figure 28) with zero tillage. Maintaining residue on the soil helps reduce the effects of wind erosion, although the extent of this type of erosion in the country is not clear.

One of the changes caused by zero tillage is related to soil structure. The use of machinery on the soil can cause compaction, principally in the more superficial layers, which can be serious depending on the crops used in the rotation. Species such as forage gramineae from the Brachiaria and Panicum genus can remedy this situation due to their deep and abundant root system.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 27: Methane emissions from conventional planting and zero tillage in irrigated rice areas in different locations and a comparison of emissions reductions between the two uses

Well-conducted zero tillage has positive effects on capping the soil temperature, improving the structure and capacity of water storage, and increasing nutrient retention sites in the layer occupied by plant roots (Gassen & Gassen, 1996). 124

The reduction of agricultural operations for soil preparation, which took a month or more before one could plant, was another benefit of zero tillage. It made it possible to have two or three harvests per year, thus economizing on fuel and labor for operations and maintenance (Figure 28).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 28: Percentage of reduction of soil and water losses from zero tillage (ZT) compared to conventional planting (CP)

(Adapted from De Maria 1999)

Table 28 shows costs for investments and O&M for agriculture in the Reference Scenario, as well as the accumulated income earned between 2010 and 2030. Approximately 68 percent of the revenue includes total costs for the Reference Scenario. Values for the Low-carbon Scenario appear in the same table, with a 100 percent adoption rate for zero tillage. Total costs represent 44 percent of the income for the Low-carbon Scenario. Investment costs are reduced 29 percent and O&M costs, 8 percent. Cost reductions stem from the elimination of tools and materials, such as fences and plows, and the decrease in the use of fuel (about 40 l/ha for each harvest).

Table 28: Cumulative costs and revenue in the reference and Low-carbon Scenarios with the adoption of zero tillage from 2010 to 2030 Cumulative Proposal

considered

Zero tillage

Reference Scenario Options Investment Cost

473,851,746.00

O&M

2,324,026,541.00

Mitigation or Carbon Uptake Options Income

4,114,575,626.00

Investment Cost 335,574,435.00

Income (without

O&M

2,129,349,512.00

carbon)

125

5,618,152,902.00

Thus, the adoption of zero tillage for mitigating greenhouse gas emissions will not generate additional costs, as it will automatically lead to an increase in income.

In the Low-carbon Scenario, 100 percent of the cotton, rice, beans, corn and soybean production area will be converted into zero tillage starting in 2015. The principal result of this process is the reduction of CO2 emissions from land use, which is 237 Mt CO2eq (Table 29). The reduction of N2O in the soil through the mineralization of organic N is 55 Mt CO2eq. Zero tillage in areas with irrigated rice contributes to reducing emissions by 10 percent compared to the Reference Scenario, or 43 Mt CO2eq. The amount of fossil energy saved is 21 Mt CO2eq, mostly from economizing on the use of diesel oil in agricultural operations. Total avoided emissions amount to about 356 Mt CO2eq, the equivalent of a 17 percent reduction compared to the Reference Scenario. Table 29: Greenhouse gases produced in the Low-carbon Scenario: adoption of zero tillage in 100 percent of the agricultural area from 2015 to 2030

Emission Source CO2 produced with a reduction in soil C stock

N2O from fertilizers, residue (including burning of sugar cane) and mineralization of nitrogen in the soil with a reduction in C stocks CH4 produced from wetland irrigated rice and burning of sugar cane Use of fossil energy to power agricultural operations Total

GHG Emissions CO2e)

(Mt

Difference compared to the lowcarbon scenario Mt CO2e

% reduction

348.4

236.8

40.5

631.0

54.6

8.0

390.8

42.8

9.9

322.4

21.1

6.3

1,692.5

355.5

Results per unit of the federation are presented in Maps 22 and 23 below.

17.0

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

3.1.1 Emissions Reduction Potential Associated with Zero Tillage

Map 22: Mitigation by crop, 2010 to 2030

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

126

Cotton (Mt CO2e)

Rice (Mt CO2e)

Bean (Mt CO2e)

Corn (Mt CO2e)

Soy (Mt CO2e)

Sugar cane (Mt CO2e)

Positive Effect Negative Effect

Source: EMBRAPA, Word Bank Brazil Low Carbon Case Study

Map 23: Total emissions from agriculture, 2010 to 2030

3.1.2 Obstacles Limiting the Expansion of Zero Tillage The use of zero tillage requires three basic actions to ensure the system’s sustainability: a) continuous planting without traditional soil rotation; b) the use of crops that leave a sufficient amount of litter to cover the soil the entire year; and c) crop rotation in summer and winter to break cycles of pests and diseases and improve nutrient recycling in the soil.

Surveys on the use of zero tillage in Brazilian agriculture, with the support of rural extension agencies and farmers, show that the system was widely adopted up until the beginning of this decade. The following years, adoption estimates for zero tillage in production areas were limited to the analysis of trends, resulting in an estimate of 25 million hectares in 2005. Consultations with specialists from the Brazilian Federation of Zero Tillage into Crop Residues and the Zero-Tillage Farmers Association of the Cerrado (APDC) provided information on the stagnation of zero tillage adoption. Farmers seemed to be increasingly using conventional practices, such as light terracing, subsolators, or even completely returning to conventional planting (Dr. John Landers, APDC, personal communication). However, there is some agreement among specialists that the area under a “well-executed” zero tillage system, based on the premises outlined, is actually well below the 25 million hectares that appear in the statistics.

Changing from conventional planting to zero tillage has never been easy. Myths about the use of zero tillage, such as the risk of soil compaction and low efficiency liming, and the occurrence of pests and diseases as a result of poorly planned systems or not following recommendations, discourage farmers from even trying. There are also

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

127

still a number of technical obstacles that have yet to be surmounted. 128

Access to technology. Farmers do not generally adopt technologies that they are not familiar with. They also resist acquiring the necessary knowledge, which is a major obstacle for small-scale farmers, who are responsible for much of grain production such as beans and corn, and who, for economic and cultural reasons, have little to no access to professional support to help adapt their production systems. This is less of an obstacle for large-scale farmers.

Costs of conversion/economic advantage. Depending on each situation, the beginning of the zero tillage system can be more onerous due to the need for machines and more inputs, and there is no consensus that the use of zero tillage is always economically advantageous for all parts of the country.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Available knowledge. The degree of knowledge and technologies available for areas with more amenable climates in Southern Brazil means that the use of zero tillage is much more widely disseminated there. Nevertheless, for regions including the north/northeast of the State of Paraná and other parts of the country, especially the Cerrados, more research has been requested on cover crops for the period following the summer harvest, which ensures that enough residues are produced to cover the soil during the year.

Logistics and infrastructure. The Brazilian farmer generally suffers greatly from the handling and storage of farm products. The higher value of soybean means that there is no way to store crops such as corn, one of the most important options for crop rotation in the summer. There is also no guarantee that these alternative cereals used in summer rotations will be purchased. These factors only serve to stimulate the soybean monoculture and weaken one of the links for the success of zero tillage: diversification.

3.1.3 Proposals for Overcoming Obstacles

The obstacles that hinder the expansion of zero tillage in the country need to be surmounted. The following measures are recommended to achieve this:

Encourage basic research and technology in order to generate a continuous information flow to ensure the sustainability of zero tillage in different parts of the country.

Restructure the rural extension system by training technicians who act as a link between research institutions, universities and the different actors in the productive sector. It is essential that technical universities and schools include the zero tillage system in the minimum professional training curriculum.

Facilitate differentiated priority credit for farmers who adopt the zero tillage system; e.g. increase agricultural credit with lower interest rates for farmers who practice zero tillage; offer rural insurance with the possibility of reduced premiums depending how long it takes to adopt the system, etc.

Increase storage area and guarantee the purchase of suitable products for zero tillage, such as corn and rice. Develop financial “hedge” instruments for prices for essential inputs for the zero tillage system (e.g. herbicides).

3.2 Carbon Uptake through the Increase of Planted Forests for Renewable Charcoal As discussed in Chapter 2, Brazil’s main available options for carbon uptake are planted forests and native forest recovery—particularly reforestation of riparian forests and legal reserves. The next two sections identify the carbon removal potential 129 of these options, first for production forests and second for native forest recovery, and analyze and explore ways to overcome the obstacles to their implementation.

One mitigation option considered here was the additional use of renewable charcoal in the Brazilian iron and steel production sector. This section deals with the potential of net greenhouse gas uptake or “sequestration potential” as a function of a possible increase in renewable charcoal production.

The National Plan for Climate Change studies this type of mitigation option and reference is made to the need to double the actual area of planted forests in Brazil (MMA, 2008). However, since the Plan is still at an early stage, no exact amounts have been mentioned regarding the additional use of renewable charcoal in the iron and steel production sector. This topic has also been studied by the Productive Development Policy, which is still being developed and coordinated by the Ministry of Development, Industry and Commerce (MDIC).

As mentioned earlier, the additional use of renewable charcoal as a thermo-reduction agent in the iron and steel production process may result in two types of climatic benefits: (i) emissions reductions in the industrial process and (ii) an increase of carbon stocks generated by additional stocks of sustainable forest plantations. A low-carbon scenario was developed within the framework of this study (LCCCS ). This report compares the Reference Scenario with the uptake potential of the new forest plantations in two situations. In the first (Scenario 1), the hypothesis would be that the participation of charcoal in iron and steel production would be maintained at current levels, approximately 33 percent until the year 2030, assuming that all of the charcoal used by the sector would be from plantation forests. At the present time, less than half of the charcoal used in the sector comes from planted forests (see AMS, 2009, ABRAF). In the second situation (Scenario 2), participation would increase from the current 33 percent to approximately 46 percent by 2030. In both cases, charcoal would continue to be the main thermo-reduction agent used in iron and steel production, but in the second case, the relative participation of charcoal would increase about 13 percentage points. However, since both scenarios are very ambitious and are dependent on different structural changes in current production conditions (to be discussed later on), it was decided to adopt them as hypotheses for projections to keep the demand for land in Brazil at a more conservative level, while incorporating a considerable expansion in areas of planted forest. But these are not the most likely scenarios in the absence of new measures. For the two situations, the following premises were used: A substantial decrease in investment, management, institutional and technological barriers, generating a significant increase in the supply of planted forests for renew-

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

3.2.1 Carbon Uptake Potential Associated with the Increase in Renewable Charcoal Production

The main obstacles to using renewable charcoal for iron and steel production are: (i) scarcity of planted forests, (ii) higher transaction costs for renewable charcoal compared to mineral coke and non-renewable charcoal and (iii) technical and logistical limitations of using renewable charcoal in large blast furnaces. Given the aforementioned technical and operational limitations, it will be necessary to promote the injection of charcoal powder in large blast furnaces that run on coke (especially in the integrated sector) to stimulate new productive arrangements based on the use of smaller blast furnaces (especially in the independent sector). In conclusion, maintaining the current arrangements would make potential changes impracticable for different companies in the sector, especially in the integrated sector, which makes up the better part of Brazilian iron and steel production.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

130

able charcoal through the different measures presented earlier in this report, including the use of the Clean Development Mechanism of the Kyoto Protocol or similar instruments. With the decrease in the aforementioned barriers, the stock of planted forests would represent an area of 2,352 to 2,663 million hectares in Scenario 1 and 3,276 to 3,663 million hectares in Scenario 2.

As a result of the implementation of the new measures, such as those suggested here, and the gradual increase in the availability of planted forests starting in 2010, there may be a measured change in the participation of each thermo-reduction agent in the growth projected for the sector starting in 2018 (3.7 percent per year ). In Scenario 1, the annual growth of iron and steel production using coal will be about 2.8 percent, and the annual growth of production using renewable charcoal will be approximately 0.9 percent. In Scenario 2, the reverse will occur, in which case the annual growth of iron and steel production using coal will be about 1 percent, and the annual growth of production using renewable charcoal will be about 2.7 percent. In other words, both Low-carbon Scenarios were completely based on the sector’s expansion, including the consolidation of the different investment decisions. Possible co-benefits and negative effects:

In addition to the climatic benefits, a variety of co-benefits may be attributed to the expansion of renewable charcoal for iron and steel production. One indirect but extremely important co-benefit is its contribution to reducing pressure on native forests in Brazil. Historically, native forests have met most of the demand for wood in the country, which has contributed to the deforestation of native forests in different biomes.

Another co-benefit is the significant reduction in the country’s dependence on outside energy sources due to its external dependence on coal. Approximately 80 percent of the material is imported due to the scarcity as well as the properties of coal produced in the country (Brito, 1990). If the measures taken help avoid the use of coal in the future to some extent, there may also be a significant reduction in emissions from other polluting gases, such as SO2/Sox, as well as in the net consumption of atmospheric O2. In the process of manufacturing iron and steel using coal, “1,376 kg of O2 are consumed for each ton of pig iron produced. On the other hand, when charcoal is integrated (…) there is practically no O2 removal from the atmophere” (Bonezzi, Cadeira-Pirez and Brasil Junior, 2004; apud Sampaio, 1999). According to Castro (2000), for iron and steel production using charcoal from planted forests “the productive cycle shows a negligible consumption of 8 kg O2/ ton of pig iron, with the permanent restoration of 14,120 kg

O2/ ton of pig iron in the atmosphere”.

The effects of activities related to the production of planted forests can also be presented as co-benefits if they occur on degraded or less productive lands, and if they comply with Brazilian environmental legislation. These possible gains include great potential to contribute to sustainable development in rural areas by generating new jobs, or the establishment, monitoring and preservation of areas of native biomes adjacent to the plantations, contributing to biodiversity conservation, compared to what might occur in degraded areas. Potential negative effects of the hypotheses of Low-carbon Scenarios are directly linked to some of the basic premises adopted in this report, such as respect for environmental legislation, especially with regard to deforestation-related provisions. Negative environmental impacts may occur in cases where the law is not complied with, either in the socio-environmental management of forest plantations and carbonization practices on different scales, or in the unsustainable conversion of native forests in production areas. Potential negative effects would need to be studied within the framework of a more thorough analysis, including the implementation of the Brazilian environmental legislation as a whole, which is beyond the scope of this report.

There is a definite need to examine these risks, although there are fewer now than in the past. It is also important to emphasize that a conclusive analysis should compare the substantive consequences of not complying with the law with the potential negative effects of using alternative products, in other words, the possible negative effects of using coal and charcoal from unsustainable deforestation practices. If not, the analysis could either not reflect, or reflect in an unbalanced way, possible trade-offs regarding the use of the three thermo-reduction agents in question. Quantification of the potential net uptake of CO2 in the Low-carbon Scenarios:

The hypothesis of the implementation of the Low-carbon Scenarios would result in the generation of net CO 2 uptake from the atmosphere commensurate with additional stocks of planted forests. Estimates of the uptake potential were based on the average pluriannual stocking capacity of 190 tCO 2 e per hectare, as presented above and in Chapter 2. This is a conservative approach in that average CO 2e stock per hectare only includes the average quantity of live biomass during the seven years of the wood planting and harvesting cycle . Expected gains in productivity are also considered with possib l e o p e ra t i o n a l o r te c h n o l o g i c a l i m p rove m e n t s i n fo re s t p l a n t a t i o n s , and in the processes of carbonization and thermo-reduction over time, in a way that is coherent with the gains in productivity obtained in the past . Considering that the total area required may vary between 2,352 and 2,663 million

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

On the other hand, to produce a ton of pig iron, it was found that “10 kg of SOx produced by the trajectory of coal” is also emitted (Castro, 2000) and about 9.5 SO2 from the same process (Bonezzi, Cadeira-Pirez and Brasil Junior (2004), apud Sampaio (1999)). These emissions occur basically due to the chemical composition of coal 131 which contains sulfur and “other undesirable substances such as heavy metals, which are only partially removed from combustion emissions. The combination of these substances with water vapor in the atmosphere can form sulfuric acid precipitation” (Castro, 2000).

hectares in Scenario 1, the stocking potential in 2030 would be between 446.8 MtCO2e and 499.6 MtCO2e, as shown in Figure 29 and Table 30. Figure 29: CO2e stock from forest plantations for renewable charcoal in Scenario 1

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

132

Table 30: CO2e uptake from forest plantations for renewable charcoal in Scenario 1 Low-carbon Scenario 01 Uptake of tCO2e

 

Reference Scenario

Low-carbon Scenarios (by productivity)

Low

Medium High

Net uptake potential in the highest productivity scenario (Highest) in tCO2e

2010

2015

2020

2025

2030

144,364 146,258 175,394 179,407 178,503

159,973 170,484 264,639 376,007 446,875

170,506 178,120 276,493 396,691 473,202

181,642 189,335 293,902 419,567 499,681 37,278

43,077

118,508 240,160 321,178

Total area required in Scenario 2 would be between 3,276 and 3,663 million hectares and the stocking capacity would vary between 622.4 MtCO2e and 695.9 MtCO2e, in 2030, according to Figure 30 and Table 31 below:

Figure 30: CO2e stock in forest plantations for renewable charcoal in Scenario 2

Table 31: CO2e uptake in forest plantations for renewable charcoal in Scenario 2 Low-carbon Scenario 02 Uptake of tCO2e  

Reference Scenario

Low-carbon Scenarios (by productivity)

Low

Medium Higher

Net uptake potential of the higher productivity scenario (tCO2e)

2010

144,364

2015

6,258

159,973 179,677

170,506 187,726

181,642 199,545 37,278

53,287

2020

175,394

343,677

359,071

381,680

206,286

2025

2030

179,407

178,503

592,806

695,992

531,261

560,484

413,399

622,440 59,110

517,489

Net uptakes that were hypothetically generated by the two Low-carbon Scenarios in 2030 were estimated based on the difference between the stocks of planted forests in the respective scenarios and stocks in the Reference Scenario that are the equivalent of approximately 1 million hectares in 2030. Thus, the maximum net uptake potential in 2030 would be approximately 321.1 M tCO2e in Scenario 1 (see Table 30) and 517.4 MtCO2e in Scenario 2 (see Table 31), as illustrated in Figure 31 below.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

133

Figure 31: Comparison of CO2e stock in Scenarios 1 and 2 and the Reference Scenario.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

134

3.2.2 Obstacles to the Expansion of Production Forests for Renewable Charcoal

Brazil is one of the few countries capable of producing iron and steel using charcoal on a large scale27. However, the increase in the use of renewable charcoal from planted forests in an effort to avoid using mineral coke or non-renewable charcoal in the future is now faced with a number of obstacles and market flaws, which are listed below:

Lack of adequate funding: Although forest plantations in Brazil are highly productive compared to other countries, they require substantial long-term investments (ex: land, labor, etc.). The first profits are usually seen only after the seventh year, meaning that the necessary loans for this activity require a grace period of at least seven years, and a minimum duration of 10 years for fast-growing species such as eucalyptus. This credit structure is non-existent in commercial Brazilian banks and rather rare in public ones. Most federal funding programs (PROPFLORA, PRONAF) are geared towards small-scale production, which, although important, is not enough to reverse the deficit of planted forests in Brazil. Difficulty in obtaining credit: In addition to the scarcity of financing, access to credit is also a major obstacle. Many banks still have difficulty accepting planted forests as a loan guarantee, although it is allowed for other agricultural crops. Often the land alone 27

Brazil, 2007

can be considered a guarantee. On the other hand, the irregular situation of some firms in terms of environmental licensing often makes it more difficult to grant funding, further proof of the need for coordination between public funding policies and economic agents within the framework of environmental licensing processes.

Another important program is the BB Florestal which, besides investing in forests, also covers cost defrayal and marketing. Despite its rather considerable resources, with nominally about R$76 million in operation in 2008, contracts are concentrated in the state of São Paulo, with over 73 percent of the total amount of applied resources (ABRAF, 2009) and are being used for other industrial sectors.

Some state-level experiences have met with relative success, such as the Proflorestas Program of the Minas Gerais State Development Bank (BDMG), but suffer from the same scarcity of resources. There are still some resources available in the so-called Constitutional Funds (FNO-BASA; FNE-BNB; FCO-BB), but there is no program that specifically targets the use of charcoal from planted forests for iron and steel production.

Higher transaction costs than alternative products, and the capital market’s greater aversion to charcoal from planted forests: Transaction costs from planting and managing production forests are significantly higher for renewable charcoal than for the main alternative products. Compared to the global substitute (mineral coke), renewable charcoal has going against it: a long maturation period, with a 14-21 year production cycle; greater need for labor for planting and carbonization processes; high costs for land and difficulty obtaining environmental licencing and financing. On the other hand, coal is a global commodity with an established international market, whose production costs and logistics structures are widely known. The product also benefits from increasing returns to scale and network externalities (see Krugman & Obstfeld, 2000). Compared to non-renewable charcoal from deforestation, renewable charcoal may face dishonest and often illegal competition, giving rise to negative externalities. Nonrenewable charcoal does not require substantial investments for land and plantations, thereby drastically reducing its production costs. In this context, with its market shortfalls, international investors tend to have greater risk aversion to renewable charcoal, a long-term investment with higher transaction costs than coal.

Lack of security in the sustainable supply of renewable charcoal: Brazil has experienced a historic deficit of wood from forests that have been planted for different

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

According to ABRAF (2009, p. 93), “the forest sector currently has lines of credit 135 at its disposal for small-scale forestry projects that are implemented by public federal banks.” These funds are concentrated basically in two programs, PROPFLORA and Forest PRONAF, developed by the National Bank for Economic and Social Development (BNDES) and by the partnership between the Environment Ministry (MMA) and the Agricultural Development Ministry (MDA), respectively. Although substantial, such resources are not enough to meet the financial needs necessary for reversing the deficit of planted forests in the country. In 2007, PROPFLORA spent approximately R$52 million, while PRONAF spent about R$12 million the same year (ABRAF, 2009: 94-95). The amount needed for financing within the framework of the Low-carbon Scenario is approximately US$ 6 billion.

136

purposes, especially for the supply of renewable charcoal28. From 1967 to 1988, the federal government had a fiscal incentive program (FISET), which, despite a number of problems, contributed to the significant increase of the area with planted forests (Brazil, 2007). With incentives coming to an abrupt end, the prevalence of the abovementioned obstacles became even more pronounced, followed by an increase in the deficit of renewable charcoal on the market. On the other hand, the opening up of the Brazilian market in the early 90s facilitated access to coal (Brazil, 2007). The country’s chronic renewable charcoal deficit, commonly known as “forest blackout”, makes companies in the sector highly vulnerable. This problem is to a great extent the result of the different obstacles listed in this section, as well as market flaws due to negative information, and the respective challenges in evaluating the risks of alternative products (coal, and non-renewable and renewable charcoal), as mentioned in the last section, exposing companies to periods of shortage in renewable charcoal.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Historic variations in the price of pig iron, which – strictly in the financial sense – could make the establishment of plantations more attractive, have not resulted in a proportionate increase in plantation establishment. On the contrary, the current gap in plantation establishment has actually grown. Throughout history, positive changes in the price of pig iron had no impact, doing nothing to reverse the deficit situation between the annual establishment of plantations and the effective consumption of reduction agents. Rather, the deficit increased when the price of iron increased significantly (research by SINDIFER/ALICEWEB/AMS 2007, 2008). This rather inflexible relationship between the establishment of plantations and their final use corroborates the sizeable risks perceived in investing in plantations, which helps explain the predominance of coal, an easily accessible and abundant global commodity.

Prevalence of inefficient carbonization technologies: Most of the conversion of wood into charcoal in Brazil is still done using inefficient technologies that emit large amounts of GHGs, including CH4. These are still common processes even if the wood comes from renewable sources. In cases of illegal wood carbonization practices using wood from deforestation, the environmental damage can be even greater, with clandestine situations making working conditions even worse.

Social communication: There is a considerable lack of information and communication between the economic players involved in the productive chain and civil society with regard to positive impacts and ways to mitigate possible negative ones associated with large-scale wood cultivation and charcoal production. This could result in problems for public policy formulation and in the definition of regulations. Risks related to regulations: Despite the fact that the production logic behind planted forests (silviculture) is the same as for other crops, the regulations for the sector are 28

Different governmental and non-governmental organizations have published reports on the status of plantations and the sources of wood supply, as well as on specific deficits of plantations for the purpose of producing charcoal in Brazil, including the Instituto Brasileiro de Geografia e Estatística (Brazilian Institute for Geography and Statistics - IBGE), Banco de Desenvolvimento Social e Econômico (National Bank for Economic and Social Development), the Environment Ministry (MMA), the Brazilian Silvicultural Society (SBS), the Instituto Brasileiro de Pesquisa Florestal (The Brazilian Institute for Forestry Research - IPEF), the Associação de Silvicultura de Minas Gerais (Silvicultural Association of Minas Gerais - AMS, antiga ABRACAVE), the Universidade de Viçosa (UFV), Universidade de São Paulo (ESALQ/USP), STCP Engenharia (Engineering), the Associação de Defesa do Meio Ambiente de Minas Gerais (Association for Environmental Defense of Minas Gerais -AMDA), among others.

Environmental regulations and tree-planting laws in Brazil are generally complex and the environmental licencing process takes at least six months, despite recent attempts to simplify it. Some characteristics of the regulations are inherent to the nature of the subject: land use in Brazil and the need to combine economic and socio-environmental development. Nevertheless, although these characteristics are seen as natural and unavoidable obstacles, they seem to carry more weight than the regulatory characteristics used for alternative inputs. Investors who opt to import coal, for example, are not subjected to this type of environment-related procedure, which impacts the analysis of opportunity costs. Another example is the necessity of acquiring a large quantity of land to preserve legal reserve and permanent preservation areas, as explained earlier. Measures such as those mentioned above are necessary for guaranteeing the sustainability of the process, but they influence decisions on the use of the different thermo-reduction agents. Internalization of environmental costs for the use of charcoal is not necessarily accompanied by an equivalent marginal income to complement the environmental benefit generated. Therefore, the evaluation of the opportunity costs for the use of renewable charcoal often suffers, characterizing the trade-offs between economic and environmental aspects.

3.2.3 Measures for Overcoming Obstacles

Table 32 shows some suggestions for overcoming the aforementioned obstacles.

137

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

different. One example is the need for a licence to harvest and transport wood from planted forests, which is not the case for the harvesting of other crops, thereby generating additional limitations. Overcoming this type of obstacle is not only linked to the simple need for less bureaucracy, but also to measures that ensure the wood’s origin control.

Table 32: Measures proposed to surmount obstacles

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

138

Identification of proposed measures

Category

(i)

Rectifying

(ii)

Rectifying

(iii)

(iv)

(v)

(vi)

Incremental

Rectifying

Incremental

Incremental

Description Revise current national financing instruments, including the BNDES and the BB Florestal and Constitutional Funds, with the objective of facilitating and expanding the availability of credit for the productive chain in the use of renewable charcoal in iron and steel production, including forest plantations, carbonization technologies, use of by-products and thermo-reduction. Encourage public and private banks to recognize forest assets as guarantees during the risk evaluation process for operations in the sector.

Encourage and support the use of the Clean Development Mechanism of the Kyoto Protocol as an additional source of resources and financing, using methodologies that cover most of the productive chain.

Revise the sector’s regulations in light of the nature of wood cultivation, aiming at synergy between federal and state entities, with the objective of improving the environmental licencing process, especially: simplification of the licencing process for wood cultivation in areas that are degraded, underutilized or previously used for other crops; and simplification of harvesting and transport licencing, without harming environmental integrity.

Strengthen the monitoring structure for the illegal use of non-renewable charcoal from illegal deforestation practices, based on (i) command and control mechanisms within the governmental framework and (ii) measures that stimulate the use of products based on sustainable forest cultivation, and the depreciation of the value of products from deforestation practices, from buyers in the productive chain all the way to the final consumer. Include renewable charcoal (solid biofuel) and its deriviatives (tar and biogas from carbonization) in the Brazilian biofuel policy and in the agendas of the respective agencies for its regulation, stimulation and promotion in Brazil and abroad.

(vii)

(viii)

Incremental

Incremental

Develop an environmental communication and education program in a partnership between government and civil society, including the private sector, with the objective of informing the Brazilian population about alternatives such as the sustainable use of planted forests for iron and steel production.

139

Stimulate applied research on more efficient processes for converting wood into charcoal and for making the best use of by-products from the process (e.g. use of tar and combustion gases).

In addition to climate-related environmental co-benefits, and the possible negative effects analyzed above, an analysis of the gains and losses for the different stakeholders involved in the Low-carbon Scenario is presented as follows: Companies from the iron and steel production chain:

Charcoal producers: With the increase in demand, companies that produce renewable charcoal from planted forests would benefit from the possible creation of a structured input market. This market is not very structured at present, and most firms that use non-renewable charcoal (from native forests) internalize production. On the other hand, non-renewable charcoal producers would incur considerable losses, as incentives for forest plantations combined with the strict legislation would practically eliminate activities based on deforestation and the use of non-renewable charcoal. In an efficient market economy, this flow of gains and losses would not result in net losses of taxes and jobs. On the contrary, those involved in unsustainable activities could be absorbed by new sustainable ones. However, the implementation of public policies that support this transition would be necessary to ensure the logistics of the process and reduce resistance from the political side (see Energy Report).

Coal producers: Given that the Low-carbon Scenario contemplates a significant absolute growth in iron and steel production using coal, and that most of the coal used in iron and steel production in Brazil is imported, national producers would not suffer any significant negative impacts, as there is already room for a considerable increase in national production, subject to technical feasibility.

Iron and steel industry: Iron and steel producers who use or can use charcoal would naturally also benefit from the new policies, as the implementation of the Low-carbon Scenario would reduce uncertainties and obstacles associated with the use of renewable charcoal in iron and steel production, increasing supply security and decreasing investment risks. Those who would lose out in the long run are producers who deliberately use non-renewable charcoal on a large scale in order to avoid considerable investments, generating unfair competition. As in the case of coal producers, coal-based iron and steel production would not be negatively impacted, as the fundamental premises of the Low-carbon Scenario do not involve any change in investments that are already consolidated or being implemented.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Implications: “Stakeholders” - winners and losers

Climatic benefits associated with renewable sources of charcoal production and the potential increase in the participation of this thermo-reduction agent in the market by 2030 could help bring about a balance in emissions resulting mostly from national production, which would continue to be based on coal. If emissions from the Brazilian iron and steel industry were analyzed as a whole, national iron and steel production using coal could actually benefit from the increase in the use of renewable charcoal, as the sector’s carbon intensity could do an “about-face” and establish itself as one of the lowest, if not the lowest, in the world. If the net uptake potential were to be considered and attributed to the sector, the positive impact on the balance of emissions would be even more significant. Therefore, charcoal from renewable sources could be a complementary and non-exclusive alternative to the use of coal, within the context of a partnership for the sustainable development of the national iron and steel sector.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

140

Even in Low-carbon Scenarios, coal will continue to be the predominant thermo-reduction agent due to an increase in production and the lack of techno-economic feasibility of its use in most cases. However, the main target, as well as the principal challenge of any hypothesis on the Low-carbon Scenario would be the elimination of the different obstacles while generating additional incentives so that future expansions, which have yet to be planned, may be based on renewable charcoal. Should this not occur, any hypotheses for Low-carbon Scenarios would be impracticable and the Reference Scenario that appears in item 2.2.3 of this study would prevail. If current economic, technological, logistical and institutional conditions are maintained, the potential increase in the relative participation of renewable sources of charcoal would not be viable and any restrictions on the use of coal could negatively affect the competitiveness of the national iron and steel industry.

Nevertheless, despite the positive marginal costs, it is essential that the positive climatic externality generated by the Low-carbon Scenario be priced and converted into marginal revenue so that the scenario does not generate economic losses for the different companies in the sector, further proof of the importance of the CDM, the carbon market, and other similar mechanisms in this context. Other members of civil society:

Other likely winners in a new scenario with the policies proposed would be the rural populations in regions located in the immediate vicinity of iron and steel production hubs (within a 300-500 km radius). These populations would benefit from the increase in rural job opportunities related to the productive/ forest chain of renewable charcoal, such as solid biofuel. It would thus be of the utmost importance to ensure the implementation of best working practices in regions where silviculture and charcoal production activities are expanding. In addition, municipalities where forest plantations and charcoal production are established could benefit from the increase in tax revenue due to the expansion of economic activities in rural areas. Government:

A major benefit for public management, which can be attributed to the implementation of the Low-carbon Scenario, would be the increase in the rasterability, transpar-

3.3 Carbon Uptake through Native Forest Recovery

As illustrated in Chapter 2, there is some potential for CO2 uptake through the natural regrowth of degraded forests, which has already been mentioned in the Reference Scenario. But because of the botanical obstacles mentioned earlier, the carbon-capture potential associated with natural regrowth remains limited. Despite these challenges, various studies and projects have demonstrated that forest plantings can promote the accelerated reestablishment of native plant cover. Such plantings induce microclimatic changes favorable to germination, the establishment of plantlets and the generation of a layer of litter and humus, which increases soil fertility. In addition, shade from young trees helps to suppress invasive grasses. Because of the large areas of degraded ecosystems, such as abandoned pasture and croplands, where native forest recovery activities could be implemented, such activities can represent significant carbon uptake potential in Brazil.

To assess CO2 uptake potential through native forest restoration, the study developed a biomass potential model in the most promising biomes, the Cerrado and Atlantic Forest. These biomes, which had large forested areas in former times, have suffered greatly from deforestation over the past two centuries.

3.3.1 Carbon Uptake Potential Resulting from a“Legal Scenario” for Forest Restoration

To estimate the carbon uptake potential through forest restoration, the establishment of goals for these activities is required. As a result of consultations with government representatives, this study adopted compliance with forest law as a target with regard to forest preservation areas and reserves. The costs of implementation are analyzed in Chapter 7.

The greatest reforestation potential for carbon uptake in Brazil considered in this study revolves around a “Legal Scenario” involving compliance with and enforcement of laws governing the management and use of riparian forests and legal reserves (Box 2). A two-step calculation is required to estimate that potential: (i) determining the area needed to comply with the legislation, and (ii) estimating the CO2 uptake potential resulting from native forest restoration in this area. To estimate the amount of land needed for reforestation to comply with the Legal Reserve Law, the study used the area of the municipality as the basis for calculating the percentage of legal reserve. The study excluded conservation units, (CUs), indigenous lands, PPAs of major watercourses, areas with declivity above 15 percent, unfit soils and urban areas. Legal reserve percentages defined by the Forest Code were used. Also excluded were areas with native plant cover, including secondary vegetation, savannas, and forests. The area left equaled the intended area for forest recovery in compliance with the Legal Reserve Law.

141

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

ency and monitoring capacity in relation to the fiscal and socio-environmental aspects of the productive chain in iron and steel production, especially on a small scale. This could result in the significant reduction of government inspection costs, especially with regard to measures that seek to control biomass origin.

Box 2: Moving towards a “Legal Scenario”: Main Areas for Protection Permanent Preservation Areas 142

Permanent Preservation Areas (APP) are forested areas found on the banks of rivers, lakes and other water bodies, which preserve the water resources, prevent soil erosion, maintain the landscape and geological stability, ensuring the welfare of human beings. In the case of riparian forests in Brazil, the width of APP depends on the width of the river (Table A). Table A: Comparison of the width of the river and the APP: Width of the river (m) Up to 10

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

10-50

50-200

200-600

Larger than 600

Width of APP (m) 30 50

100 200 500

Legal Reserves

Legal Reserves are areas in Brazilian rural properties (except APP) that are vital to the sustainable use of natural resources, conservation and rehabilitation of ecological processes and biodiversity conservation. The percentage of land set aside as legal reserve varies by biome: • 80% in rural property located in the Legal Amazon;

• 35% in rural property located at the Legal Amazon and Cerrado areas;

• 20% in rural property located in forest areas or other forms of native vegetation in other regions of the country, especially the Atlantic Forest.

To estimate the uptake potential, the study team assumed that the legal reserve areas to be restored would be reforested gradually until 2030, when full legality would be achieved. Starting in 2010, 1/21 of the total area for reforestation would be deducted every year from the area available for agricultural production. The environmental liability for the country was estimated at about 44 million ha, about one-third of which would be located in the Amazon region (Table 33).

Table 33: Area needed for reforestation under Brazil’s Legal Reserve Law, by state

Mato Grosso do Sul

Mato Grosso

Goiás

Distrito Federal

Maranhão

Piauí

Rio Grande do Norte

Paraíba

Pernambuco

Alagoas

Sergipe

Bahia

Rondônia

Total for Brazil: 44,344,390 ha 

3,398,792

9,465,888 2,611,730

0

40,959

0

State Acre

Amazon

Roraima

Pará

Amapá

Tocantins

3,062

Paraná

27,167

Santa Catarina

58,239

91,861

118,800

242,079

4,794,589

Rio Grande do Sul Minas Gerais

Espirito Santo

Rio de Janeiro São Paulo

Area for reforestation (ha) 721,161 34,848

46,757

11,369,199 0

1,644,537

1,711,257 398,679

1,184,241 2,682,095

205,436

178,087

3,314,927

Sources: ICONE, UFMG.

The study estimated the carbon uptake potential for the Legal Scenario at about 2.9 Gt CO2 over the study period; that is, about 140 Mt CO2e per year (Figure 32).29 Figure 32: Carbon uptake potential of forest recovery activities and production forests

29

143

If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by an average of 112Mt CO2 per year.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Area for reforestation (ha)

State

Map 24: CO2 uptake potential through forest restoration by 2030 and total CO2 uptake potential

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

144

In the meantime, since biomass accumulates over an extended period of time in recently planted native forests, often over 100 years before peaking, the forest restoration of the legal reserve is a measure whose uptake potential is beyond the scope of this study. Map 24 shows the uptake potential obtained for 2030 and total potential, per state.

It is important to note that enforcing forest legal reserves implies releasing the corresponding land currently occupied by other activities (i.e., crops or pastures). This means that the land use and land-use change projected in the Reference Scenario (Chapter 2) would need to be revised. Such a revision would be significant since the area released for legal enforcement of the forestry law would equal more than twice the estimated deforested area under the Reference Scenario. This runs the risk that the benefits gained from carbon uptake resulting from forestry activities could be partially lost via increased conversion of native vegetation to accommodate crops and pastures displaced by restored legal reserves.

3.3.2 Obstacles to Forest Restoration and Ways to Overcome Them

Obstacles to forest restoration can be divided into two main categories, ecological and economic, and are described as follows.

Ecological Obstacle: The main ecological factor of natural recomposition is that the predominant species should regenerate naturally, without the need for long-term planting. However, depending on the degree of degradation of the ecosystem, and if regeneration does not occur, the colonization of the area by arboreal species and the secondary succession would be negatively impacted. The following obstacles are considered limiting factors for natural regeneration in areas such as pastures and abandoned agricultural fields: • Absence of propagulum: lack or little availability of an adequate seed bank on

the ground, absence of dispersers and the seeds’ difficulty in reaching the soil due to the quantity of biomass from gramineae.

• Lack of plant establishment: in this scenario, even if there is an adequate seed bank, seed predation and the herbivorous consumption of young plants, in addition to competition with gramineae, makes the natural reestablishment of the 145 plant cover difficult. Although these factors impede the natural regeneration of native forests, a number of scientific studies have shown that forest plantations can eliminate these barriers, facilitating and accelerating the reestablishment of native plant cover. The positive effect of plantations is in the micro-climatic changes they generate, favoring plant germination and establishment, and creating a layer of litter and humus that increase soil fertility. In addition, the shade of the young trees helps suppress the growth of invasive gramineae.

Economic Obstacle: Currently observed in the State of São Paulo, which has a deficit of over a million hectares of riparian forests. Despite the fact that the state government created reforestation programs, for example the Secondary Forest Restoration Project, with funding from GEF, and that the Federal Government has made lines of credit available for the native forest restoration on rural properties, less than a thousand hectares in total have been registered for the native vegetation restoration by rural property owners. As mentioned earlier, the high costs of restoration, combined with the loss of productive area, are the main economic obstacles to the implementation of large-scale forest restoration activities. The voluntary participation of property owners in reforestation programs that focus on the restoration of these areas is rare. Forest restoration generally occurs in most cases due to legal obligations that entail changes in behavior and practices, and forest compensation. However, the following measures can provide major resources to help develop forest restoration projects:

Establishment of volunteer market for emissions compensation: The volunteer market for corporative greenhouse gas emissions compensation has great potential to contribute to the restoration of forest areas that are part of the Low-carbon Scenario (legality). This market is currently not regulated, and is without clear emissions factors in many inventories. Moreover, there is no standard for carbon uptake potential through planted trees for compensation, as this type of compensation is customarily awarded through actual tree planting. Also, the organizations and companies that award the compensation have no idea where the trees are planted, so the creation of legal mechanisms that restrict tree planting for GHG-related compensation to permanent preservation areas would be essential for implementing the long-term legal scenario.

Simplification of the environmental licencing of forest restoration activities: The requirements of forest activities that aim at restoring degraded areas through the use of native species must be simplified with regard to the environmental licencing of this

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

• Other factors: in addition to the aforementioned causes, burning, over-exploitation of the areas and the absence of symbiotes and pollinators are also considered major obstacles to natural regeneration.

activity, eliminating bureaucracy and facilitating the implementation of this type of project.

Stimulate the CDM forest program modality: The CDM forest program modality is considered major progress within the CDM in the implementation of large-scale forest restoration programs through different activities. However, due to the perceived risks and long validation and approval process, it has not developed to the extent it should have, by combining new restoration areas. These areas should be added to the Program of Activities during project implementation, without going through the validation and approval process to which a traditional CDM project is subjected. Thus, government support to the CDM forest program would be a public policy that could generate additional resources for native forest restoration programs, and is also essential for overcoming resistance to this modality.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

146

Derogation of the Forest Servitude Regime: The Forest Servitude regime established by the Forest Code permits the compensation of the environmental liability of a specific property with another property located in the same micro-basin if it has a forested area that exceeds the limits established by the Forest Code. However, with the Atlantic Forest Law, which prohibits the deforestation of any area that is in an intermediate or advanced stage of regeneration, this regime takes on a different character, as the owner registers forest excess on his property to comply with the obligation of the other property owner, who in turn does not reforest within his property.

Foster the market for non-wood forest products: The harvesting and marketing of native fruits, plant resins, honey, and other non-wood forest products considerably increases the value of native forests and should stimulate the restoration of legally protected forest areas, generating an income for local communities. These efforts should be encouraged.

Fiscal incentives: Rural property owners whose properties do not present deforestation in the Permanent Preservation Area (PPA) and the Legal Reserve (LR) receive fiscal incentives that are as close as possible to the opportunity costs of the land. These incentives may be in the form of a tax exempt status or even the availability of lines of credit at reduced interest rates and with longer grace periods. This measure would motivate rural property owners with vegetation coverage liabilities to eliminate them by reestablishing natural plant cover on their properties. Environmental education: Environmental education and awareness-raising programs for rural populations and land owners, teaching them about the importance of forests for the environment and for the entire productive chain in the agriculture and livestock sectors. In addition, awareness-raising campaigns should be conducted with appropriate forest restoration models through educational programs, while building capacity among the rural population and land owners regarding the importance of forests in our society.

3.3.3 Reforestation Support Policies

Within the framework of reducing and eliminating carbon emissions in the Cerrado and Atlantic Forest biomes, the principal legal instruments that provide any kind of guarantee for the maintenance of these biomes and the preservation of the carbon

stock are the Forest Code and the Atlantic Forest Law. The Forest Code

The Forest Code establishes two distinct types of areas within the property that cannot be deforested: the Permanent Preservation Area (PPA) and the Legal Reserve (LR). The PPAs are defined based on geographic aspects, such as riparian forests, which are 147 30-3,000 meter wide areas of forested land that are adjacent to a body of water; and forests located on top of hills and slopes with declivity above 45 degrees. The Legal Reserve (LR) is a part of the property, excluding the PPAs, that cannot be deforested. The limits of the Legal Reserve are defined depending on the biome such as: • 80 percent of rural property located in the Legal Amazon;

• 20 percent of rural property in forested areas, or other forms of native vegetation in different parts of the country, principally the Atlantic Forest.

Forest Servitude Regime

The Forest Code has a flexibility mechanism that is frequently used in the regularization of the Legal Reserve: the Forest Servitude Regime. Through this mechanism, the rural land owner who does not have the minimum property required by the Legal Reserve legislation can compensate for this deficit through the Forest Servitude Regime. This means that the property owner can transfer a forest area that is equal to his deficit from another property that has an excess of vegetation, as long as it is located in the same micro-basin. The Atlantic Forest Law

The Atlantic Forest Law goes beyond the Forest Code, establishing limits for the deforestation of vegetation that exceeds the Legal Reserve. In practice, it imposes limits on deforestation for areas of vegetation at different stages of regeneration including:

• Advanced regeneration: may be deforested only in exceptional cases, when necessary for the implementation of projects that are of public interest, scientific research and preservation practices; • Mid-level regeneration: in addition to what is stated above, when necessary for the implementation of activities and crop-livestock-related uses for the smallholder farmer and traditional communities, that are essential for their subsistence;

• Initial regeneration: in states where the total remainder of Atlantic Forest is less than 5 percent, this stage of regeneration will comply with the same legal regime for vegetation at mid-level regeneration.

It should be emphasized here that, at any stage of regeneration, the cutting and removal of vegetation should be authorized beforehand by the relevant state agency. National Forest Program - NFP

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

• 35 percent of rural property located in the Cerrado area in the Legal Amazon;

148

Launched by the Federal Government in 2000, the general objective of the National Forest Program – NFP is “the promotion of sustainable development, reconciling utilization with the protection of ecosystems and making the forest policy compatible with other sectors in such a way as to promote the increase of internal and external markets and the institutional development of the sector”.

Thus, the NFP combines the environmental, social, and economic aspects of the Brazilian forest sector, including, among its specific objectives: • To stimulate the sustainable use of native and planted forests;

• To encourage reforestation activities, namely on small rural properties;

• To restore permanent preservation forests on legal reserves and in modified areas; • To support economic and social initiatives of populations that live in forest areas;

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

• To penalize illegal deforestation and the predatory extraction of forest products and sub-products, containment of accidental burning and forest fire prevention; • To promote the sustainable use of national, state, district or municipal production forests; • To support the development of forest-based industries;

• To expand internal and external markets for forest products and sub-products;

• To promote social, environmental and economic aspects of the services and benefits provided by public and private forests; • To promote the protection of forest biodiversity and ecosystems.

The NFP makes lines of credit available through the funds listed below. Principal lines of credit

PRONAF – National Program for the Strengthening of Family Agriculture

The purpose of the National Program for the Strengthening of Family Agriculture (PRONAF) is to financially support agriculture and livestock and non-livestock-related activities through the direct employment of the labor force that includes the rural farmer and his family. It has a specific modality for the forest sector, the PRONAF-FLORESTA, which deals with the funding of projects involving silviculture, agricultural systems and sustainable extractivist exploitation for rural farmers. • The following are eligible for credit in PRONAF-FLORESTA: • Family farmers and rural laborers who:

• Cultivate a plot of land as owner, homesteader, lessor, partner or concessionaire of the National Agrarian Reform Program;

• Reside on or near the property;

• Do not have title to an area greater than four fiscal modules, quantified according to current legislation;

• Earn at least 80 percent of the family income from crop-livestock and non-croplivestock operations; • Perform and run overall operations, outsourcing only when necessary based on seasonal demands for agricultural and livestock operations;

• Earn gross family income between R$1,500.00 and R$10,000.00, excluding re- 149 muneration from social welfare benefits as a result of rural activities. Family farmers and rural laborers who: • Reside on or near the property;

• Do not have title to an area larger than four fiscal modules, quantified according to the current legislation;

• Earn a minimum of 80 percent of the family income from the crop-livestock and non-crop-livestock operations of the establishment; • Perform and run overall operations; maintaining 2 permanent employees, outsourcing whenever the seasonal nature of the activity requires;

• Earn gross family income over R$10,000.00 annually and as much as R$30,000.00 excluding remuneration linked to social welfare benefits stemming from rural activities.

PRONAF-FLORESTA has an interest rate of 4 percent per year and a 12-year time frame for repayment, with an 8-year maximum grace period. Funding limits are R$6,000 for the beneficiaries of group (A) and R$4,000 for those of group (B). • PROPFLORA – Commercial Planting and Forest Restoration Program

• The credit line made available by BNDES, PROPFLORA has the following general objectives:

• Planting and maintenance of forests for industrial use;

• Recomposition and maintenance of areas of permanent preservation and legal forest reserves;

• Implementation and maintenance of forest species for the production of wood to be burned for use in the process of drying agricultural products;

• Implementation of silvopastoral (combination of livestock with forests) and agroforestry (combination of agriculture with forests) projects; and • Planting and maintenance of palm forests for biofuel production.

Beneficiaries include rural farmers (physical or legal entities), associations and cooperatives. The ceiling for beneficiaries is R$200,000 per year, with an interest rate of 6.75 percent per year. The time frame for repayment and grace periods adhere to the

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

• Cultivate a plot of land as owner, homesteader, lessor, partner or concessionaire of the National Land Reform Program;

following regulations: 150

• Up to 144 months, from the grace period until the date of the first cut, can be increased from 6 months to 96 months, for forest planting and maintenance projects to be used for industrial purposes and in the production of wood to be used for burning for the process of drying agricultural products; • Up to 144 months, including a 12-month grace period for projects for the restoration and maintenance of permanent preservation areas and legal reserves; • Up to 48 months, including an 18-month grace period for other projects for establishing forest seedling nurseries.

Constitutional Finance Funds

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The 1988 Federal Constitution earmarked 3 percent from income taxes and any type of profit, as well as industrialized products, to be applied to financing programs for the productive sectors of the northern, northeastern and central-west regions. By using part of the income tax for the more needy regions, the Union prompted the creation of the Constitutional Fund for Financing the North (Fundos Constitucionais de Financiamento do Norte - FNO), Northeast (Nordeste - FNE) and Central-West (Centro-Oeste - FCO), with the objective of promoting the economic and social development of those regions through programs that fund productive sectors.

Rural farmers, individual firms, legal entities and associations and production cooperatives that develop activities in the crop-livestock, mineral, industry, agro-industry, tourism, infrastructure, commerce and service sectors may solicit financing from the FNO through the Banco da Amazônia S.A. for the northern region; the FNE through the Banco do Nordeste do Brasil in the Northeast; and from the FCO through the Banco do Brasil S.A, in the central-west. Credit lines made available through each of these forestsector-related funds are summarized as follows: FCO – Constitutional Fund for Financing the Central-West

FCO PRONATUREZA: This credit line made available by the Bank of Brazil is part of the Constitutional Fund of the Central West, where a good part of this region is within the limits of the Cerrado biome. Its potential beneficiaries are rural farmers (physical or legal entities), associations and cooperatives. This line of credit has the following general objectives: • Sustainable forest management;

• Reforestation for energy and wood; • Agroforestry systems;

• Restoration of degraded areas;

• Acquisition of machines and equipment;

• Integrated rural and industrial projects; • Promotion of the market.

FNE – Constitutional Fund for Financing the Northeast

Made available by the Bank of the Northeast, this credit line has a line item, the FNE Verde, to be used to finance productive activities, with an emphasis on environmental conservation and protection in productive activities in general including organic ag- 151 riculture and livestock operations, such as the conversion of traditional systems into organic ones, forest management, reforestation, agro-silvicultural and agroforestry systems, alternative energy generation, collection and recycling systems for solid residues, environmental studies, implementation of environmental management systems and certification, clean technologies and restoration of degraded areas. The actions of the FNO cover the states of Acre, Amapá, Amazonas, Pará, Rondônia, Roraima and Tocantins. This fund offers credit at interest rates that vary depending on the borrower, from 8.75 to 14 percent per year. Operations related to the industrial, agro-industrial, tourism, infrastructure, commercial and service sectors include sustainable forest management activities, which correspond to line item FNO Floresta. Demonstrative Project Sub-Program - PDA

Implemented by the Environment Ministry starting in 1996 within the framework of the Pilot Program to Conserve the Brazilian Rainforest (PPG7), the PDA supports innovative initiatives of civil society organizations in the sustainable use and preservation of natural resources in the Amazon and Atlantic Forest biomes, aiming at improving the quality of life of the populations involved.

The program has supported initiatives in Amazonia and the Atlantic Forest, and in their associated ecosystems. Between 1996 and 2003, the PDA supported 194 projects, with 147 in the Amazon and 47 in the Atlantic Forest. The projects developed actions in areas including agroforestry systems and environmental restoration (including nursery construction), forest and water resource management and environmental preservation. The PDA currently supports projects that are divided into three components: 1. Alternatives to Deforestation and Burning Project (PADEQ): with 49 projects contracted in the states of Pará, Mato Grosso, Rondônia, Roraima and Tocantins;

2. Consolidation: aims at strengthening experiences previously supported by the PDA through the more integrated consolidation of environmental, economic, social and institutional sustainability, and currently supports 31 large projects, with 12 in the Atlantic Forest and 19 in the Amazon;

3. Actions to Conserve the Atlantic Forest: 99 approved large and small-scale projects throughout almost all the states where this biome is present.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

FNO – Constitutional Fund for Financing the North

3.4 Mitigation Options for Livestock Activities

152

Over the past few years, research on mitigating greenhouse gas emissions caused by ruminants has been stepped up and different technological options have been explored, particularly with regard to manipulating the animals’ diets. However, since Brazilian livestock is raised predominantly on pastures, many of the mitigation options tested in developed countries are not applicable on a large scale.

Due to the magnitude of degraded pasture areas in Brazil, their restoration has taken on a key role. The reduction in the productivity and quality of fodder and carbon stocks, combined with the low level of animal productivity, result in a higher level of emissions per product unit in this system. In addition, this type of system is extremely demanding with regard to the need for land, and its maintenance or expansion is associated with the need to open up new areas with native vegetation.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

More intensive systems, such as integrated crop-livestock and feedlots also have great potential to mitígate emissions and increase animal productivity, while reducing emissions per product unit as well as the demand for land.

This alternative will be prioritized due to the large amount of emissions caused by deforestation in Brazil. The potential to reduce emissions if these more technological systems are associated with research programs and incentives for the genetic improvement of both fodder species and animals will also be taken into consideration. A technical summary on the main alternatives that are applicable to Brazilian production systems is presented in the following section.

3.4.1 Main Options Considered for Mitigating Emissions from Livestock Pasture restoration

It is estimated that about 60 percent of pastures in the Cerrado region are currently at some stage of degradation. Pasture degradation results in a decrease in the soil carbon stock, a reduction in carrying capacity, an increase in soil loss due to erosion, and a significant increase in CO2-e emissions per kg of meat. However, the restoration of these areas may reverse these characteristics, resulting in an increase in carbon sequestration from the soil as well as carrying capacity, a decrease in soil loss from erosion, and a reduction in CO2-e emissions per kg of meat. Adoption of integrated crop-livestock systems

Crop-livestock integration consists of the implementation of different productive systems with grains, fibers, meat, dairy and fuel in the same area, in association, either sequentially or in rotation. On-farm land use is alternated in time and space between agriculture and livestock (Vilela, 2008).

The integration of meat and grain production systems is one of the viable options for the development of alternatives for re-establishing the productive capacity of cultivated pastures. According to the author, out of all the benefits of the potential synergy between pastures and annual crops, the following stand out: a) improvement of the physical, chemical and biological properties of the soil; b) a break in the cycle of disease,

pests and damaged plants; c) a reduction of economic risks through the diversification of activities, and d) a reduction in the cost of restoration/renovation of degraded pastures.

According to Kichel et al. (2003), integrated crop-livestock systems can attain carrying capacity rates of 3 animals/ha when combined with supplementing and feedlots, whereas the national average is approximately 1 head/ha. In a recent report, Martha Junior et al. (2006) concluded that the amount gained in terms of live weight on the pasture the first year in integrated farming-livestock systems was between 9 and 40 arrobas of carcass/ha/year, depending on the edaphoclimatic conditions and local management, while extensive systems typically produce between 3 and 5 arrobas of carcass/ha/year. In systems of stocking and finishing males only on pasture, Magnabosco et al. (2003) obtained carrying capacity rates of 2.68 animal units/ha and 1.48 units during the dry season, with an average weight gain of approximately 6@/year, enough to make the slaughter of pasture animals viable in less than 30 months. Expanding the adoption of finishing in feedlots

The use of this technology for terminating animals in a short period of time (60 to 120 days) before slaughter has a number of advantages, among which are their lower age at time of slaughter, an increase in carcass weight and meat quality, improvement in the herd’s rate of reproduction, and greater land-use efficiency. Typically, the animals are fed a moderate diet or canned concentrated grains, together with more voluminous feed (ex. silage corn or sorghum, chopped cane). Thus, an animal that gains 1.6 kg/day in a feedlot produces in only 90 days what an animal that gains 0.4 kg/day in a pasture would produce in a year. However, in most regions of Brazil, extensive livestock is subject to seasonal variations due to differences in temperature and precipitation. During periods of fodder scarcity, the animals’ performance is low or even negative, which needs to be recuperated during the next rainy/hot season. Thus, using feedlots in the final phase of termination can reduce the age at time of slaughter by another year, with a significant reduction in CH4 production from enteric fermentation. Improvement of fodder quality

According to the FAO (2007), the potential to mitigate greenhouse gas emissions through the genetic improvement of fodder species has been relatively unexplored. However, according to the report, there are strong indications that this approach could, a priori, meet with a relatively high degree of success.

Two aspects are under discussion with regard to the genetic improvement of tropical fodder species for the reduction of methane emissions. The first has to do with nutritional quality. An increase in digestibility and in the amounts of soluble carbohydrates, as well as its voluntary consumption by the animal, can generate a significant reduction in the amount of methane produced per product unit. For example, Lovett et al. (2004) demonstrate that increasing the soluble carbohydrate content in forage grass by 33 g/kg generated a 9 percent reduction in methane production in vitro. It is

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The increase in animal productivity by maintaining pasture areas in better condition makes it possible to improve zootechnical indices, while reducing greenhouse gas 153 emissions per product unit.

estimated that there may be a reduction of about 15-28 percent in emissions from the use of improved fodder species (FAO, 2007). 154

The second aspect has to do with the occurence of antimethanogenic compounds in fodder plants. Johnson & Johnson (2002) cite different compounds that appear to have such an effect. Woodward et al. (2001) determined a reduction in methane emissions for sheep and dairy cows consuming fodder that was rich in condensed tannins. Ulyatt et al. (2002) found that methane emission was substantially reduced for sheep and dairy cows that consumed kikuyu grass (Pennisetum clandestinum), under specific conditions, suggesting the presence of compounds that have not been identified thus far. Genetic improvement of the beef cattle herd

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Zootechnical indexes (e.g. birth and weaning rates, weight gain, age at first pregnancy and at time of slaughter, etc.) obtained in Brazil are still lower than those of developed countries for different reasons, such as the quality of tropical fodder, the climate, the presence of diseases and parasites, and the herd’s genetic potential. The latter aspect is due to the fact that the taurine cattle used in countries with temperate climates have undergone genetic selection over centuries. Their productive potential is therefore high. In comparison, zebu cattle used in Brazil and in other tropical countries have only been utilized for 130 years in the Americas. Genetic selection did not take place in India, their country of origin, until 20 years ago, nor in Brazil. Today there are different genetic evaluation programs in Brazil, including the Zebu Cattle Improvement Program (ABCZ/EMBRAPA), the Nelore Program of Brazil (National Association of Breeders and Researchers) and the PAINT (Lagoa da Serra), among others. Although these programs are all different, each one studies aspects such as weight, sexual precociousness, and maternal ability.

None of these programs includes the nutritional efficiency aspect, although it is of significant economic interest. Moreover, more efficient animals consume less feed and produce less methane with the same performance (Nkrumah et al., 2006). It is estimated that for most characteristics of economic relevance, genetic progress could reach 1 percent per year, although in Brazil progress is generally about 0.3 percent per year (Lobo et al., 2009). Genetic improvement has not achieved its potential for several reasons: low rate of technology adoption, lack of availability of improved animals, nonutilization of the most advanced selection techniques, prioritization of other characteristics (eg. fur, breed, etc). This means that the Brazilian herd is much larger than necessary for satisfying the demand for meat, which implies that the production of CO2-e/kg of meat is way above that of competing countries. For example, Brazil and the United States are the biggest meat producers in the world, with 9.47 and 11.98 million tons in 2007, respectively, although the beef cattle herd in the USA is half the size of the Brazilian herd (Nass, 2010). Part of this difference is related to production systems, which are extensive in Brazil and extensive/intensive in the USA, with most of the animals being kept in feedlots. Another reason for this difference is the Brazilian herd’s lower genetic potential. An alternative for reducing CO2-e emissions while increasing meat production would be to improve the genetic quality of the national herd.

3.4.2 Obstacles and Proposals for Overcoming Them Adoption of more productive systems

There is also the activity’s low rate of return, which would require low interest rates to make it more attractive. The rather favorable economic performance of croplivestock integration justifies the actions the Brazilian government has taken over the last five years (PROLAPEC, PRODUSA) to promote the adoption of these systems, to reduce business risks and increase rural incomes, and to restore degraded pasture areas, thereby enabling the expansion of crop-livestock systems in already anthropized areas.

Incentive policies for the early slaughter of animals can also generate productivity gains and emissions reductions. A good example of this is the Early Bullock Program in Mato Grosso do Sul. In this program, animals slaughtered in meat processing plants with carcass and teeth typification (as a way of evaluating the animal’s age), and the minimum desired carcass weight, provide the registered cattle breeder with a financial incentive through the reduction of the ICMS payment from 16.67 percent to 66.67 percent. The adoption of a similar policy at the national level could be an incentive for the adoption of more intensive systems, not only in the stocking and finishing phases, but also in the cow-calf phase, as the better remuneration for animals for slaughter will be reflected in improved remuneration for the calves, enabling better intensification of the cow-calf phase. In addition, positive externalities such as a reduction in clandestine slaughter, better carcass conformity and more tender meat are expected for this type of policy. Another aspect to be considered in the adoption of more intensive systems is a greater need for effective management. Public policies that promote rural extension and training for cattle breeders are important for surmounting this obstacle. Genetically improved fodder species

Since all the scenarios evaluated in the present report examine the extensive use of pasture, and cattle-farming in Brazil currently occurs predominantly in this system, the use of fodder species with less potential for methane production for ruminants would have a significant impact on methane gas emissions in the atmosphere. However, the priority of fodder species improvement programs in Brazil currently is to develop material with favorable agronomic characteristics and resistance to pests and diseases; they do not include the evaluation of emissions levels. On the other hand, ongoing research is testing evaluation techniques for methane production in vitro for fodder plants.

It is currently estimated that a research program on genetic improvement for the launching of a cultivar would cost approximately R$4 million over a 12-year period.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The first obstacle for pasture restoration and the adoption of more intensive systems is the need for initial investment capital for the transition to the productive sys- 155 tem. Since the activity is not very economically attractive (Table 34), the availability of credit is essential, particularly for financing the purchase of animals to increase carrying capacity. In case financing does not include the purchase of animals, the breeder will most likely underutilize available fodder resources due to the lack of capital for purchasing animals.

156

Within the context of genetic improvement with a focus on management and quality, public policies that promote the financing of projects in this area, which has not been a priority in crop-livestock research thus far, would steer efforts towards research universities and institutions in order to select fodder species that are of higher nutritive value, as well as improved management strategies for its use, resulting in the launching of cultivars with better methane emissions potential for ruminants. Use of genetically superior bulls

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Genetic improvement, which has a longer period of return, is often not considered a priority by breeders in extensive systems. Thus, programs that provide incentives for evaluating bulls, and subsidies for acquiring tested animals of good lineage, may contribute to greater sector efficiency in the medium term and also help reduce GHG emissions. Assuming that 2.3 million bulls are needed to maintain the national herd (a bull-to-cow ratio of 30:1), a 50-percent premium for improved animals above their slaughter value, and four years of useful life for the bull, the total value of subsidies for the national herd would amount to R$350 million per year. Positive externalities for adopting such a measure include increased productivity, better quality carcasses and an increased calving rate (assuming andrological testing of improved bulls). The mitigation potential for direct emissions from livestock, combined with the mitigation options proposed depends on the scenario adopted to substitute low-productivity production systems with higher productivity production systems: the greater this substitution, the greater the reduction of direct emissions from livestock. Since more productive systems enable the same level of production to be achieved on smaller pasture areas, this scenario was constructed based on the need to free up pasture areas as part of the strategy for reducing deforestation, which will be presented in the next section. As a result, calculating the potential for mitigating emissions resulting directly from livestock can only be done once this deforestation reduction strategy has been quantified at the end of the next section, after determining the amount of meat production allocated for each productive system in the Low-carbon Scenario.

3.5 Reduction of Emissions from Deforestation

Deforestation appears to be the main source of emissions in the Reference Scenario. While significant, the mitigation and carbon uptake potential described in the foregoing section remains limited compared to the large volume of GHG emissions resulting from deforestation. As mentioned above, a main trigger of deforestation is the need to convert native vegetation into land to accommodate crops and pasture expansion. The land-use modeling developed by this study makes it possible to estimate the volume of additional land needed and associated deforestation in the Reference Scenario. To avoid emissions from deforestation, ways would need to be found to reduce global demand for land, while maintaining the same level of products supply as the Reference Scenario. In systemic terms, mitigation of emissions through land-use change could be achieved by absorbing the expansion of these activities via the increased productivity of other ones.

Brazil’s major agricultural activities already show high levels of productivity and consequently do not offer opportunities to increase productivity on the scale required

to absorb these additional levels of demand for land. For example, the productivity of a soybean plantation in Brazil was 2.86 tons per ha in 2008, compared with 2.81 tons per ha in the United States (Table 34). Table 34: Average productivity of selected crops in different countries (tons per ha), 2008

Argentina

Bangladesh

China, People’s Republic of EU-27 India

Indonesia Mexico

Pakistan

Paraguay Thailand

United States

Uzbekistan, Republic of Brazil

Soybean 2.78 1.61

1.06

2.62 2.81

2.86

Corn

5.17

5.67

2.3

3.22

9.46

3.99

Cotton

1.30 0.57 0.65 0.99

0.83

1.49

Rice

3.93 6.43

3.31 4.66

2.76 4.20

Beef-cattle farming shows much greater potential for increasing productivity per hectare, which can be applied to a much larger pasture area, since pastures occupy 207 million ha compared to 70 million ha for agricultural activities in 2030 in the Reference Scenario. Consequently, increasing the technological level and the intensification of livestock-raising can play an essential role in reducing the need for land for this activity, while releasing the land required for the expansion of other activities.

In Chapter 2, in the section related to the calculation of emissions from livestock in the Reference Scenario (see 2.1.4.1.2 Emissions estimates by prototypical systems), we saw that there are low-productivity productive systems, particularly: • Complete cycle in degraded pastures • Complete cycle in extensive pastures

• There are also high-productivity production systems in Brazil, but only on a limited scale, particularly:

• Extensive cow-calf in pastures + supplemented stocking and finishing in croplivestock integration

• Extensive cow-calf in pastures + supplemented stocking and finishing in feedlots.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Country 

157

Crop (tons per ha)

The BLUM/SIMBRASIL tool that was developed in this study to model land use and its future changes helps qualify the substitution of low productivity systems with more productive systems year after year, and simulates the location of pastures that could be vacated to accommodate the economic growth of projected crop-livestock systems in the Reference Scenario, as well as new land uses considered in the Low-carbon Scenario. Details of this quantification appear in the next chapter, which has a land-use scenario that is compatible with GHG emissions mitigation and uptake proposals considered in the Low-carbon Scenario of this study.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

158

The modernization of the Brazilian livestock sector through the accelerated expansion of systems 3 and 4, to substitute for the low productivity systems 1 and 2, would enable the same amount of meat to be produced in a much smaller pasture area. Due to the large area currently occupied by low productivity systems – about 200 million ha – this substitution opens up the possibility of vacating very large volumes of pasture area compared to the expansion of other agricultural activities, which are already highly productive and only occupied about 52 million hectares in 2008. A 5 percent increase in the agricultural area, or 2.5 million hectares, corresponds to only 1.25 percent of pasture area used by low productivity livestock systems. Thus, pasture areas may be vacated to accommodate the expansion of agricultural activities, virtually eliminating the need to clear new areas.

4 Low-Carbon Land-Use Scenario in Brazil

Earlier sections presented opportunities for avoiding GHG emissions and carbon uptake associated with land use and land-use change, particularly emissions from agricultural production and livestock activities, and carbon uptake via production forests and native forest recovery. But putting together a low-carbon land-use scenario is not simply a matter of adding (in the case of avoiding emissions) or subtracting (in the case of uptake) the volumes of greenhouse gases associated with these opportunities. For example, while increasing the land area allocated for forest recovery and production forests leads to carbon uptake and a reduction in emissions from iron and steel production, it also decreases the amount of land that is otherwise available for the expansion of agriculture and livestock activities. The potential conversion of more native vegetation areas for the expansion of these agriculture and livestock activities would generate carbon leakage. To avoid this situation, ways must be found not only to reduce the additional amount of land needed under the Reference Scenario, but also to release land for the mitigation and removal activities envisioned while maintaining the same level of products.

4.1 Additional Needs for Land for Carbon Uptake Activities and Biofuel Export

In the Low-carbon Scenario, more than 53 million ha is the amount of additional land needed for total emissions reductions and carbon uptake. Of that amount, more than 44 million ha—twice the land expansion projected under the Reference Scenario—is for forest recovery under Brazil’s legal reserve law. The total volume of additional land required is over 70 million ha, more than twice the total amount of land planted with soybean (21.3 million ha) and sugar cane (8.2 million ha) in 2008 or more than twice the area of soybean projected for 2030 in the Reference Scenario (30.6 million ha) (Table 35).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

A key conclusion from the study’s investigations on emissions mitigation is that in order to reduce deforestation, the main source of emissions, enough land from existing pastures must be vacated to accommodate new activities and thus avoid the conversion 159 of native vegetation.

Table 35: Mitigation and carbon uptake options for a Low-carbon Scenario and associated needs for additional land Scenario 160

Additional land needed (2006–30)

Reference Scenario: additional Expansion of agriculture and livestock production to volume of land required for the meet the needs anticipated in 2030: expansion of agriculture and >16.8 million ha livestock activities

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Low-carbon Scenario: additional Elimination of non-renewable charcoal in 2017 and the volume of land required for miti- participation of 46 percent of renewable planted chargation measures coal for iron and steel production in 2030: > 2.7 million ha

Total

Expansion of sugar cane to increase gasoline substitution with ethanol to 80 percent in the domestic market and supply 10 percent of estimated global demand to achieve an average worldwide gasoline mixture of 20 percent ethanol by 2030: > 6.4 million ha Restoration of the environmental liability of “legal forest reserves”, calculated at 36.2 million ha in 2030: > 44.3 million ha 70.4 million additional hectares

One possible consequence is that land-use expansion for activities that promote lower levels of emissions, fossil-fuel substitution (as detailed in Chapter 4), or even carbon capture may provoke an excess in demand for land use, which could in turn generate deforestation, causing a lower net carbon uptake balance.

4.2 Toward a New Pattern of Productivity for the Livestock Industry

The study simulated the new distribution of productive systems for livestock that should be promoted in order to liberate enough pasture land to accommodate all the needs for additional land due originating from crop expansion in the Reference Scenario, and for the implementation of new emissions reduction and carbon uptake options proposed under the Low-carbon Scenario.

Figure 33: Mitigating measures for the construction of the Low-carbon Scenario

Source: ICONE (2009)

To increase livestock productivity per hectare—thereby absorbing agricultural expansion and other low-carbon activities without causing deforestation, while reducing emissions per unit of meat—five options were considered: (i) promoting the recovery of degraded pasture; (ii) stimulating the adoption of productive systems with feedlots for finishing; (iii) encouraging the adoption of crop-livestock systems; (iv) developing genetic improvement programs for higher quality, lower emissions forage adapted to Brazil, and (v) developing incentive programs for the use of genetically superior bulls. The projected effects of the productive systems considered for the reference and Low-carbon Scenarios are compared below (Figure 35).

A change in the pasture area is projected for the Low-carbon Scenario, from 205.38 million hectares to 137.82 million hectares. A variation in the herd is also estimated to go from 201.41 (IBGE, 2009) to 214.27 million head. These variations helped estimate the proportion of productive systems for 2008 and 2030 as presented in Figure 34.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

161

Figure 34: Change in pasture area occupied according to type of productive system (million hectares)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

162

Reference Scenario

Low-carbon Scenario

The new distribution of livestock production per production system can also be seen in Figure 35 below by the number of head of cattle in the different systems in the reference and Low-carbon Scenarios. Figure 35: Variation in number of head of cattle in productive systems, 2009-30

Reference Scenario

Low-carbon Scenario

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

163

4.3 Mitigation Potential of Direct Emissions from Livestock in the Low-carbon Scenario 164

The variation in the systems’ composition generated substantial gains in land productivity. Projections indicate an increase in productivity from 47.22 kg of carcass equivalent /ha in 2008 to 63.51 and 95.42 kg carcass equivalent /ha in 2030, respectively, for the reference and Low-carbon Scenarios, enabling a reduction of 68,239 million hectares in the pasture area.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 36: Projection of Brazilian herd productivity between 2009 and 2030 for the reference and Low-carbon Scenarios

Figure 37: Projection of pasture area in Brazil from 2009 to 2030 (Low-carbon Scenario)

As a result of the substantial rise in meat production projected for the period, there was an increase in direct emissions for the sector. On the other hand, there is a considerable reduction in direct emissions in the Low-carbon Scenario (34.1 thou Mg of CO2-e per year in 2030) due to the reduced emissions per production unit in the Low-carbon Scenario (Figure 36).

The transition from a lower to a higher productivity system alone has little effect on GHG emissions per animal (1.25 tCO2e in the degraded-pastures scenario versus 1.15 tCO2e in other scenarios). But higher productivity in more intensive systems generates a significant reduction in the size of the herd projected for 2030 (208 million head in the Low-carbon Scenario versus 234.4 million in the Reference Scenario), which would in turn generate significant emissions reductions per unit of meat (Figure 38) and in the total value (Figure 39).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

165

Figure 38: Comparison of methane emissions from beef-cattle raising (Mt CO2e per year), 2008–30

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

166

Figure 39: Comparison of methane emissions per unit of meat (kg CO2e per kg), 2008–30

The combination of improved forage and genetically superior bulls, together with the proposed increase in livestock productivity, would reduce direct livestock emissions from 273 to 240Mt CO2 per year by 2030, thereby maintaining emissions at about the 2008 level.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

167

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 25: Number of heads of cattle

168

Map 26: Total cumulative emissions from livestock, 2010-2030

4.4 A New Land-use Scenario for the Low-carbon Scenario With the data provided by BLUM on land requirements for the Reference Scenario, the land-use change simulation model was constructed again. In addition to the entry data, the difference between the reference and Low-carbon Scenarios is that for the latter, when there is environmental liability in the micro-region, the deforestation rate is reduced to zero and environmental restoration is implemented through forest restoration. It is important to note that in both scenarios, the projection model for deforestation in the Legal Amazon is activated, simulating additional deforestation from indirect causes. With new data provided by the economic modeling team on the need for land in the Low-carbon Scenario—the development of which is based on a wide array of improvements in zootechnical livestock indices and the subsequent decrease in pasture area, more area for sugar-cane production, restoration of environmental liability with regard to legal reserves and PPAs, and greater use of charcoal for ironworks—the landuse change simulation model adopted in the Reference Scenario was run again.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

169

170

Increased carrying capacity rates associated with greater herd productivity, as a combined effect of the recovery of degraded areas and the adoption of more intensive livestock stocking and finishing systems (integration of crop-livestock systems and feedlots), are reflected in an accentuated reduction in the demand for land, projected at about 137.82 million hectares in the Low-carbon Scenario, compared to 207.06 million ha in the Reference Scenario for the year 2030 (Table 36). The difference would be sufficient to absorb the demand for additional land associated with the expansion of agriculture and livestock in the Reference Scenario, as well as the expansion of mitigation and uptake activities in the Low-carbon Scenario (Figure 40). Table 36: Comparison of land-use results for the reference and Low-carbon Scenarios (millions of ha)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Reference scenario

Low-carbon scenario

2030

Var. 2030– 2006

Difference (2030) between lowcarbon and Reference Scenarios

8.92

(57)

5.90

2.72

Land use

2006

2008

2030

Var. 2030– 2006

Grains (harvest)

38.94

37.79

47.92

8.98

47.86

5.27

5.87

8.45

3.18

11.17

Sugar cane

Production forest Pasture

Total area for agriculture and livestock1 Restoration Balance

Herd (per 1,000 head) 1

2

6.18

8.24

12.70

6.52

19.19

13.01

6.49

208.89

205.38

207.06

(1.83)

137.82

(71.07)

(69.24)

259.27

257.28

276.13

16.85

216.04

(43.23)

(60.08)

205.890

201.410

234.460

28.570

208.000

2.120

(26.46)

-

-

-

-

44.34

44.34 1.11

2

44.34

(15.74)

Total area allocated to cotton, bean (1st harvest), corn (1st harvest), soybean, sugar cane, production forest, and pasture.

Represents expansion of agricultural area between 2006 and 2008 in the northern and northeastern regions. Source: ICONE

Figure 40: Evolution of Brazil’s demand for land by crop, 2006-30 (millions of ha)

Reference Scenario

Low-carbon Scenario Source: Adapted from ICONE (2009)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

171

Map 27 also shows the dynamics of the cotton crop in the Low-carbon Scenario. Like in the Reference Scenario, yellow represents areas where cotton is cultivated in 2007 and 2030, remaining constant. Areas in blue indicate where the cultivation of the product declined, and areas in red show where it expanded during the period modeled. The map shows that there were no significant changes vis a vis the Reference Scenario for cotton, as the demand for the product did not change between the two scenarios. Map 27: Dynamics of sugar cane cultivation (left) and cotton (right) in the Low-carbon Scenario (2010 – 2030). Yellow = crop permanence; blue = crop decrement; red = crop increment

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

172

The first map (Map 27) shows that sugar cane areas undergo major land-use changes in relation to the Reference Scenario due to the great increase in demand for land for sugar cane for the expansion of ethanol production. However, geographic distribution patterns are maintained, with some intensification and, in traditional areas, spread more towards the central-west and interior of Bahia.

Results for rice are shown in Map 28. There have not been any significant changes in relation to the Reference Scenario. Geographic distribution patterns, showing the crop’s expansion and regression, remain practically unchanged.

The dynamics of the bean crop in the Low-carbon Scenario are also shown in Map 28. There are no noticeable changes of any significance in relation to the Reference Scenario, as there have been practically no variations in the need for land for this product.

Map 28: Dynamics of the rice (left) and bean crops (right) in the Low-carbon Scenario (2010-2030). Yellow = crop permanence; blue = decrement; red = increment

Map 29 shows the results for corn. There is a significant change in the occurrence of this product compared to the Reference Scenario. Areas where the crop has decreased are in the western part of the states of Rio Grande do Sul, Santa Catarina, São Paulo and northern Paraná. On the other hand, there it has also increased in other parts of the same states, as well as in Minas Gerais.

With regard to soybean, a few changes can be seen (Map 29). Just like in the Reference Scenario, there are some areas where the crop area has expanded and is growing near its original areas. Its geographical distribution pattern has not changed either, as it still found in the states of the south, central-west, Minas triangle and western Minas, western Bahia, Piauí, and Maranhão. A decrease in the soybean crop may be observed in São Paulo, which occurs to a lesser degree in the Reference Scenario, and can be justified by the competition with sugar cane. Map 29: Dynamics of corn crop (left) and soybean (right) in the Low-carbon Scenario (2010-2030). Yellow = crop permanence; blue = decrement; red = expansion

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

173

Map 30: Dynamics of planted forests (left) and pastures (right) in the Low-carbon Scenario (2010 – 2030). Yellow = remained constant; blue = crop decrement; red = crop increment

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

174

Map 30 shows the forestry dynamics. Due to the increase in the demand for production forests to neutralize deforestation for charcoal, there are many differences between the Reference Scenario and the Low-carbon Scenario. While in the Reference Scenario there are practically no areas of expansion, in the Low-carbon Scenario, these are rather clear, especially near regions with previously established plantations.

Map 30 also shows pasture dynamics, again with considerable differences compared to the Reference Scenario. Since the low-carbon scenario includes land-use intensification, pasture area becomes the main donor of cropland, especially in the central-south and northeast of the country. With the exception of a few areas of expansion scattered throughout the northeast of Minas, Rio Grande do Sul, Paraná and Santa Catarina, a decrease in pasture areas predominates in this vast part of the country,

A map of forest regrowth (Map 31) was created for the Low-carbon Scenario. Depending on the assumptions made for this scenario regarding the restoration of environmental liability according to the current Forest Code, restoration of the native vegetation is being stimulated in the micro-regions where there is environmental liability at the beginning of the simulation. This regrowth occurs up to the limit required by law for the legal reserve (PPAs were not considered in the regrowth due to the limits of spatial resolution adopted by the study). The permanence of areas of wild grass between 2010 and 2030 occurred mainly in the state of Maranhão, as well as in Minas Gerais and Bahia, although to a lesser degree. Only Maranhão presented areas where wild grass is declining. In the case of this study, the information used for developing the land-use map indicated a strong occurrence of areas with wild grass, and the same rule applies for this type of plant cover: deforestation only ceases when the legal barrier is overcome.

Map 31: Forest regrowth in the Low-carbon Scenario

Map 32 below synthesizes the evolution of the area occupied by agricultural and livestock activities in the reference and Low-carbon Scenarios. Map 32: Area used for agriculture, pasture, and reforestation by region

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

175

4.5 Reduction of Deforestation in the Low-carbon Scenario

176

A decrease in the need for land, which was calculated based on assumptions generated by the Low-carbon Scenario, will lead to a reduction in deforestation rates compared to the Reference Scenario. New soil-use and deforestation maps were produced with the same spatial emissions model for land use developed with the EGO Dynamic platform (Map 33). The model for the Low-carbon Scenario works like a legal scenario; that is, when there is environmental liability, deforestation rates are set to zero and a simulation of a regeneration process for the micro-region in question is started.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 33: Comparison of Cumulative Deforestation, 2007–30

Reference Scenario Low-carbon Scenario

Map 34: Total area deforested, 2010-2030

Model-based projections indicate that, under the new land-use dynamic, deforestation would be reduced by more than two-thirds (68 percent) compared to the Reference Scenario; in the Atlantic Forest, deforestation would be reduced about 90 percent, while the Amazon region and Cerrado would see reductions of 70 percent and 65 percent, respectively. In the Amazon region, the level of deforestation would drop quickly to about 17 percent of the historic annual average of 19,500 km2.30 (Map 34).

It was expected that, with demand for pasture land reduced to zero as projected by the ICONE module, deforestation rates would also be reduced to zero; however, that was not the case. Deforestation still continues in certain parts of the Amazon states of Acre and Pará, with the model’s incorporation of indirect causes, through spatial lag regression (as in the Reference Scenario). Thus, in micro-regions where the legal deforestation limit was not reached in 2009—where there is still room for legal deforestation and where the indirect dynamics modeled are the determining factors—deforestation will continue unabated.

Moreover, although residual deforestation is not quite at zero, the remaining amount is compatible with the 70-percent Amazon deforestation-reduction target that the PNMC set for 2017, having as its baseline the historic average of 19,500 km2 per year. Therefore, average annual amounts of 4,000 km2 produced by the model are below the 5,000 km2 per year threshold established as a final target for Brazil (Figure 41). 30

Between 1996 and 2005, the historical rate of deforestation in the Amazon region was 1.95 million ha per year, according to the PNMC.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

177

Figure 41: Evolution of deforestation in the Low-carbon Scenario (curve) (km2 per year)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

178

Source: UFMG (2009)

Figure 42: Evolution of deforestation in the low-carbon (LCS) and Reference Scenarios (RS) (thousands of ha per year)

Map 35: Total cumulative emissions from deforestation, 2010-2030

4.6. Additional Measures for Protecting the Forest from Deforestation Although the low carbon land-use scenario offers solutions for bringing the need for additional land virtually to zero, it is expected that complementary forest protection measures would also be required for two major reasons. First, the legal limit for deforestation (up to 20 percent of properties located in the Amazon region) has not yet been reached. Thus, where the complex dynamic of deforestation is powered by the financial value of the wood or cleared land (along with with the need for cropland, pasture and production plantations), deforestation would continue. Second, there may be a significant delay between the time demand for cropland, pasture or production forests is reduced and the time one could effectively observe a behavioral change among deforestation agents at the frontier (i.e., since they may continue to speculate on demand that has already dried up far upstream in the land market chain). As already mentioned in Chapter 2, this reflects the fact that other indirect factors, in addition to the concrete need for additional land for the crop-livestock sector expansion, also play a role in the deforestation process. The model therefore includes indirect causes that are not captured by the land availability variables. These results support

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

179

the urgent need to adopt additional measures to contain deforestation.

180

The Low-carbon Scenario thus proposes to implement additional forest-protection measures in forested areas where deforestation is illegal. Given the many ongoing programs and abundance of literature available on this topic, including the Plan of Action for the Prevention and Control of Deforestation in the Legal Amazon (PPCDAM), this study was limited to analyzing existing proposals. Below is a list of the principal measures, policies, programs and actions that aim directly or indirectly at reducing deforestation and its associated emissions. Protected Areas Expansion and Consolidation

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

ARPA Program – Continuation and Expansion. In 2003, the Brazilian Government initiated the implementation 31 of the Amazon Region Protected Areas Program (ARPA). Over 30 million hectares32 of conservation units (CUs) have been created as Integral Protected Areas and Protected Areas with Sustainable Use under this program by means of an initiative supported by national (MMA and ICMBio) and international (World Wildlife Fund, World Bank, and KfW) partners through the Protected Areas Fund. The program is being implemented in three stages (2003-2008; 2009-2013; and 2014-2016) and will create about 50,000 ha of protected areas (Table 37). Table 37: Snapshot of protected areas in the Amazon biome and ARPA participation Protected or military area Military area Indigenous land Total protection

Sustainable use

31 32

Total

State Federal State Federal

Area (km²)

No. 6

282 44

37

72

80

521











26,235

987,219

137,385

Portion Protected area suof biome pported by ARPA (%) (%) 0.6

23.4

201,918

233,523

1,817,355

22.5

4.8

13.2

5.5

43.0

Source: Soares-Filho et al. (2008)

-

3.3

5.5

231,072

-

80.6

26.2 16.8

DF nº 4.326. A v a i l a b l e a t < h t t p : / / w w w. m m a . g o v. b r / s i t i o / i n d e x . p h p ? i d o = c o n t e u d o . monta&idEstrutura=154>. Last access on 10/05/2009.

The PRODES program – Monitoring the Brazilian Amazon by Satellite, implemented by the National Institute for Space Research (INPE) since 1988 is funded by the Ministry of Science and Technology, with the collaboration of IBAMA and MMA. The analyses, carried out mainly based on the use of images from the TM sensor onboard the North American satellite Landsat and provides annual deforestation rates in the region, increments and decrements of deforested areas and specialized data in vector and raster formats. The results are widely used by the national and international scientific community and were important for raising awareness about the deforestation process in the region. DETER – (Detection System for Deforestation in Real Time), another program developed by INPE, is based on data from the MODIS sensor from the Land/Water satellite and WFI Sensor from the CBERS satellite (the data is less refined than PRODES data). The DETER system aims at the rapid monitoring of the deforestation dynamic in the Legal Amazon, in an effort to provide support to supervisory activities. There are monthly reports during the dry periods and trimestrial reports during rainier periods, due to cloud presence. However, the program only manages to identify areas larger than 25 ha. A third program, DEGRAD – Mapping of Forest Degradation in the Brazilian Amazon maps degraded (i.e. partially deforested) forest areas in the Amazon using CBERS and Landsat satellite imaging. Studies dating back to 2007 enable areas of up to 6.5 hectares to be identified at different stages of degradation (see Figure 43).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Soares-Filho et al. (2008b) confirm the importance of protected areas and of ARPA in particular in helping to avoid deforestation. A decrease in the historic rates of deforestation in the Amazon as of 2004-05 can be attributed, in part, to a series of measures that are part of the Plan of Action for the Prevention and Control of Deforestation in the Amazon, including the creation and consolidation of CUs. According to these authors, the probability that deforestation will occur around protected 181 areas is 10 times greater than in the interior. Based on an analysis of historic rates of deforestation around protected areas, the study demonstrated that there is no significant redistribution of deforestation in other areas due to the creation of protected areas. Nevertheless, the consolidation of protected areas is a strong mitigating measure against the deforestation process observed in the Amazon at a relatively low cost. The creation and consolidation of protected areas thus becomes part of a national deforestation reduction strategy once its maintenance can be ensured at relatively low cost. The same authors (personal data) estimate a cost of 1.3 to 10 billion dollars (NPV) for the consolidation and management of the network of protected areas in the Amazon over a thirty-year period. According to estimates made by Amend et al. (2008), the maintenance cost for these areas will be US$3.72 per ha.

Figure 43: Identification of forest degradation patterns in the Amazon within the framework of the DEGRAD program. Source: INPE, 2009

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

182

According to the National Research Institute management report, the following resources have been available for monitoring projects via satellite for the Amazon (including the three aforementioned programs) for the last three years (Table 38): Table 38: Resources of the INPE for monitoring the Amazon by satellite Year

Budget (R$)

Total spent (R$)

2006

1,415,506.00

456,708.55

2007 2008

2,750,000.00 2,850,000.00

Source: INPE, 200933

2,072.634.00 2,077.178.20

B. Development of Integrated Projects

PPCDAM – The Plan of Action for the Prevention and Control of Deforestation in the Amazon, coordinated by the President’s Office, is implemented through the 33

Available at

The Sustainable Amazon Program (PAS): strives for a new development landscape by focusing on environmentally sustainable, economic solutions. Its targets and directives are based on a current diagnosis of the Amazon. The program is implemented according to an agreement between federal and state governments and promotes the integration of promotion and production. One of the driving forces behind the application of resources from the Amazon Fund together with PPCDAM, it is based on the principal that more efforts are needed to ensure the sustainable development of the forest’s socioeconomic potential if impact mitigation, through the creation of CUs, does not prevent the deforestation of the Amazon. Actions and strategies must be implemented with greater local government participation. The Program also helps regulate space appropriation dynamics while providing suitable conditions for populations and communities by guaranteeing their social rights. The participation of private capital is essential for providing conditions for the implementation of these projects (PPG7, FAM, etc.) (MMA, 2008). C. Creation of Forest Protection Funds

Amazon Fund (FAM): established by decree nº 6.52734, the fund aims to secure donations through non-reimbursable investments for actions to prevent, monitor and combat deforestation, and to promote the conservation and sustainable use of forests in the Amazon biome. It entails different activities: management of public forests and protected areas; environmental monitoring, evaluation and supervision; management; economic activities based on the sustainable use of the forest; ZEE territorial and land regularization system; conservation and sustainable use of biodiversity; and restoration of deforested areas. Actions should comply with PAS and PPCDAM directives, as the fund was developed within the broader context of Brazilian public preservation policies. The fund’s resources (managed by BNDES) come from donations from certified donors, and equivalent value in tons of avoided carbon, calculated according to a methodology to be established by a technical team. The fund has a Technical Committee (responsible for testing MMA emissions calculations) and an Orientation Committee (responsible for the implementation and preservation of the fund’s initiatives and goals). Norway signed a contract to donate US$110 million to the fund and the country plans to donate the first installment of US$1 billion by 2015. D. Sustainable use of forest resources and payment for environmental services:

Public Forest Management: To promote forest conservation, the concession for the sustainable use of public forests aims to increase forest appreciation. In support of this goal, Law 11,284 was created in 2006 to regulate forest management in public 34

Available at . Last access on 09/05/2009.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

coordinated action of 13 ministries. The general aim of PPCDAM is to reduce deforestation rates in the Brazilian Amazon through a set of integrated actions including territorial and land ordinances, monitoring and evaluation to foster sustainable production activities involving partnerships between federal agencies, state governments, mayoral offices, civil society and the private sector. PPCDAM has three main axes around which activities are conducted: (i) land and territorial ordinances, (ii) environmental 183 monitoring and evaluation, and (iii) productive and sustainable activities. During 2008-11, the government plans to invest approximately US$500 million in PPCDAMrelated initiatives.

184

areas; the law also established the Brazilian Forest Service (SFB) and National Fund for Forest Development (FNDF). This legal document establishes three types of management for sustainable forest production: (i) the creation of conservation units for sustainable forest production, such as FLONAS; (ii) the non-onerous use of forests for sustainable and social development; and (iii) paid forest concessions35, based on a public bid, to guarantee access to products traditionally used by local populations. The SFB will be responsible for the public forest management system, for stimulating sustainable forest development and for the management of the fund (FNDF). In the case of concessions, decisions are made with the aim of ensuring that this process benefits society as much as possible. Thus, the choice of concessions (which cannot exceed 20 percent of the area to be conceded the first 10 years) use criteria such as best price, less environmental impact, greater socioeconomic benefits, improved efficiency and aggregation of local value, and restriction to national companies. Table 39: Implementation of the public forest management systems: benefits and losses Beneficiaries

Losses

Public Forest Management: Implementation of the Brazilian Forest System Law 11.284/06

Beneficiaries of the new public forest management system will be the local communities who live off of forest products, and who want to participate in the regional economic dynamics, formalizing their entrance into the market, expanding the multiple uses and enjoying non-onerous conditions; and private business people who prefer not to buy land, who want to use the forests legally, so that they may have access to credit, export, tourism, reforestation of degraded areas, certification, jobs and income.

Areas opened up through deforestation; it is hoped but not guaranteed that this process will be slowed down.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Measure

Source: (Azevedo, T., Tocantins, M.A., 2006)

The Forest Grant Plan annually defines the public forests that may be subject to conversion, as identified in the National Register of Public Forests. It also defines the necessary management resources, especially with regard to monitoring. The table below presents estimates of the resources necessary for implementing activities planned for 2009:

35

Do not imply rights of ownership, only use of resources.

Table 40: Summary of expenditures anticipated for Public Forest Management services in 2009 Activities Anticipated (summary)

Resources (million R$)

National Register of Public Forests

8

Support forest management activities

7.8

Creation of the National System of Forest Information (Sistema Nacional de Infomações Florestais)

5.4

Monitoring of public forests

National Forest Development Fund (Fundo Nacional de Desenvolvimento Florestal) Implementation of the SFB administrative structure Total

10 15

2.5 8

56.7

Source: Plano Anual de Outorga Florestal (Annual Forest Grant Plan), 200936

Forest Allowance (Bolsa Floresta): one of the first programs to apply the concept of paying for environmental services in Brazil. Implemented by the State Government of Amazonas, it plans monthly payments of R$50.00 to families registered by the project, and residents of state CUs. The families’ permanence in the program is linked to the development of sustainable activities in these areas, which principally revolve around the reduction of deforestation activities. The state target covers about 60,000 families in the program and extends access to indigenous communities. Program resourcescome from the State Fund for Climate Change, Environmental Conservation, and Sustainable Development, which was created by the State Law for Climate Change nº 3,13537.

Program for the Socio-Environmental Development of Rural Family Production in the Amazon (PROAMBIENTE): initiated by social movements representing smallholder farmers in association with IPAM in 2001, having been adopted earlier by the Environment Ministry and included in the Pluriannual Plan. It seeks public policy innovations for the development of smallholder farmers in the Amazon region, but is also compatible with the current environmental paradigms. It thus seeks to overcome the impression of rural credit as the only economic instrument for development, and suggests compensation for environmental services38 performed by farmers linked to the program as a tool, once they have transitioned to sustainable production systems. According to the Environment Ministry’s report, the program should already be benefitting almost 4 million families from traditional communities in 148 groups, which, in turn form 11 poles that are distributed throughout Amazonia. In this program, the Production Unit becomes the basic unit, each one represented by a community group. 36

Available at http://www.mma.gov.br/estruturas/sfb/_arquivos/paof_2009_vf_95.pdf Last access on 11/05/2009. 37 Available at < http://www.florestavivaamazonas.org.br/download/Lei_est_n_3135_de_050607. pdf > 38 For example, the reduction of deforestation, the recuperation of environmental liabilities, soil, water, and biodiversity conservation, reduction of the use of agrochemicals, reduction of the risk of fire, more sustainable energy matrix, transition to agroecology.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Forest concessions

185

E. Environmental certification

Socio-Environmental Register: The Socio-Environmental Commitment Register (CCS) is a voluntary register of properties whose owners are committed to improving the “socio-environmental performance” of their properties (http://www.yikatuxingu. org.br/projetos/ver/48). The CCS already has over 1.5 million hectares of property, a large part of which is located at the headwaters of the Xingu River. With regard to the CCS, registered properties receive preferential treatment from meat-processing plants in the region (e.g., the Independência and Bertim meat-processing plants already pay a better price for an arroba (15 kg) of beef cattle from properties listed in the CCS.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

186

Agreements made between groups from each pole establish environmental service targets to be achieved in each unit. The condition for payment to the farmers from the 11 poles is the achievement of the goals established in the agreements. The program also seeks the qualification of participating farmers through ATER (Assistência Técnica e Extensão Rural - Technical Assistance and Rural Extension), which is composed of agents selected from the community itself, and aims at the formation of a single cooperation network. In addition, the project receives support from the Brazilian National Environmental Fund – FNMA and the Embassy of the Netherlands. With regard to the amount of resources involved, in one of the pioneer poles of the program, the Transamazon Pole (which includes families from three municipalities in the state of Para), received a compensation payment for 6 months for 340 families in the amount of R$100.00 /month per family, an additional compensation of R$126.00 per family for buying material and tools, and salaries of R$380.00 for new community agents for eight months (Nepstad et al., 2007).

4.7 Balance of emissions from land use and land-use change in the Low-carbon Scenario

Based on mitigation options for direct emissions, and the reduction of emissions linked to deforestation and carbon uptake through plantations, the SIMBRASIL model calculated annual emissions for the 2007-2030 period resulting from land use and land-use change for each micro-region.

Compared to projections in the Reference Scenario (Figure 44), emissions from deforestation are considerably lower under the new land-use dynamic considered in the Low-carbon Scenario (Figure 45), at about 170-190 Mt CO2e per year over much of the period. This decrease is due to less demand for pasture area and the subsequent drop in the need to convert land via deforestation. Annual land-use emissions (i.e., agriculture and livestock) increase 310 to 340 Mt CO2e over the period, with agricultural emissions accounting for most of this increase. Still there is a 6 percent overall reduction in emissions compared to the Reference Scenario. CH4 emissions from beef cattle remain relatively stable at 236 to 249 Mt CO2e per year, since the gains from reduced CH4 production per unit of meat are offset by increased production. Of these emissions caused by different types of land use, livestock is the greatest emitter, surpassing emissions caused by the deforestation of the Amazon.

Finally, carbon uptake shows a growing trajectory, presenting an initial rate of approximately 133 Mt CO2 per year for 2010 and a final rate of 213 Mt CO2 per year for 2030, as a function of the growth in forest plantation cover and recuperation of environmental liabilities of legal reserves and PPAs. The resulting balance between use, change, and uptake shows a decrease in the amount of net emissions between 2007 and 2030, reaching a rate of approximately 321 Mt CO2e per year in 2030, a reduction of 187 nearly 65 percent compared to the Reference Scenario39.



39

If the carbon uptake from the natural regrowth of degraded forests were to be included, then the potential uptake would increase by 112Mt CO2 per year on average, thus reducing net emissions.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 44: Reference Scenario results: emissions from land use and land-use change, 2009–30

Figure 45: Emissions from land use and land-use change under the new land-use dynamic in the Low-carbon Scenario

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

188

Map 36 below compares net emissions generated by land use and land-use change in the Reference Scenario and in the Low-carbon Scenario for each unit of the federation. Map 36: Total cumulative emissions from land use (agriculture, livestock, deforestation, and reforestation) 2010-30

Emissions reduction from deforestation and carbon uptake through plantations and forest recovery are much greater than the reduction of emissions from all the other sectors considered in the global low-carbon study for Brazil (energy, transport and waste). The decrease in deforestation and increase of forest plantations are two areas where the Low-carbon Scenario proposed had the greatest success in reducing emissions. Together, these two areas represent 67 percent of the net reductions registered 189 during the 2010-30 period (Table 41). It is more difficult to reduce emissions from the transport and energy sectors, as they are already low compared to the international standards, mainly due to the considerable amount of hydroelectricity and bioethanol in the current energy matrix. Table 41: Comparison of cumulative emissions distribution among sectors in the reference and Low-carbon Scenarios, 2010-30

Sector

Mt CO2e

% of total

4,101

14

Land use

16,709

Energy

7,587

Waste

Transport Total

1,633

30,030

Low-carbon Scenario (2010–30)

Mt CO2e

% of total

3,614

19

55

9,228

25

5,763

6

100

375

18,980

Reduction

Mt CO2e

% of total

487

5

48

7,481

30

1,824

3

100

1,258

11,050

67

12

16

100

% of Reference Scenario (2010–30) 44 78

13

24

37

Consequently, the distribution of GHG emissions among the different sectors in the Low-carbon Scenario differs significantly from the distribution observed in the Reference Scenario, mainly because the amount of emissions from deforestation is reduced to approximately 70 percent compared to the Reference Scenario (Figure 46).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Reference Scenario (2010–30)

Figure 46: Comparisons of gross emissions distribution among sectors in the reference and Low-carbon Scenarios, 2008–30

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

190

4.8 Key Uncertainties for Emissions Estimates Since the reference and proposed Low-carbon Scenarios are subject to uncertainties, the results are indicative and should be used to inform stakeholders of future emissions if the study’s assumptions, which were based on a broad and ongoing consulta- 191 tive process, are verified. Some of the uncertainties result from calculations related to either the reference or Low-carbon Scenario, while others concern both. This section first outlines overall uncertainties for the four main areas and then addresses more sector-specific ones. For emissions-generating activities, both the reference and Low-carbon Scenarios depend heavily on the macroeconomic projections of the 2030 National Energy Plan (PNE 2030) published by the EPE in 2007. The plan’s B1 scenario, adopted as the reference case, estimates that the Brazilian economy’s average growth rate at 4.1 percent annually. As a consequence of the recent financial crisis, the Brazilian government expects lower GDP growth, particularly in the near term. If so, decreased supply and demand for a variety of services and products would slow the pace of deforestation and energy consumption, including the demand for transport services. However, given the longer–term timeframe of the study, medium-term projections for emissions growth under the Reference Scenario are less affected by the crisis and would remain about the same. The same short- and medium-term trends would also apply to the Low-carbon Scenario. Land-use Questions

With respect to uncertainties for projected land-use emissions, one must distinguish between the gross volume of GHG emissions and net emissions obtained after taking into account carbon uptake activities involving mainly production forests and native forest recovery. Uncertainties for gross emissions differ between the first and second stages of calculations: (i) projecting land use and land-use changes and (ii) converting the results into emissions.

The economic modeling developed for the first stage of calculations benefited greatly from the wealth of historical local data, which allowed for robust calibrations of the key parameters and equations (Box 3). Based on the results, it was assumed that the main uncertainties are linked to the abovementioned macroeconomic projections, which directly affect projections for expanded cropland and meat production and thus deforestation. If cropland and meat production expand more than expected under the Reference Scenario, then more effort will be required under the Low-carbon Scenario to release enough pasture; otherwise, the additional deforestation that would result would lead to increased emissions. For the second stage of calculations, the main uncertainties are based on available data for soil carbon content and the vegetation converted, which drive the conversion of deforestation into GHG emissions. Estimates of the above- and below-ground carbon content of biomass depend on the accuracy of the data, which can only be improved by intensive field research. The uncertainty of the data used for this national study is estimated at about 20 percent, which mainly affects the Reference Scenario, since conversion of native vegetation is at very low levels in the Low-carbon Scenario.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Macroeconomic Projections

Another uncertainty involves the expected effect of productivity gains on the growth of livestock. In the study, the Brazilian share of the international market is taken as an exogenous projection from FAPRI (Box 3). Increased productivity could improve competition and thus spur increased production. Since productivity gains mean less need for pasture area, such a rebound effect should not cause more deforestation, provided such gains are limited to the areas of the former low-productivity systems.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

192

Under the Low-carbon Scenario, an added uncertainty is the pace of releasing pasture for expanding agricultural crops to avoid deforestation and comply with the legal scenario adopted as a target for forest recovery–based carbon uptake. The rapid fall of deforestation-based emissions entails considerable efforts to improve livestock productivity to free up pastures for other activities. To the extent that the release of pasture keeps pace with the annual need for additional land for crop expansion, the conversion of native vegetation would no longer be needed; in theory, deforestation and related emissions would then be brought to zero. Key questions are whether the pace of pasture release and agricultural expansion will match and whether the necessary conditions will be created to ensure that the pace of agricultural expansion is not too rapid. Achieving the right pace on the livestock side and providing the right incentives—positive or negative—for forest protection are critical. If the required financial disbursements are not made on time, deforestation and its related emissions will continue unabated.

Box 3: Uncertainties for Economic land-use Scenarios

Uncertainties inherent to the economic modeling of future land-use scenarios are related to the modeling of (i) domestic demand (a function of income, ultimately linked to macroeconomic projections and equilibrium prices determined by the modeling), (ii) exports (a function of macroeconomic parameters and prices), and (iii) production (a function of costs and productivity per hectare). Price elasticities were calibrated from a historical series (1996–2008), while production costs and per-hectare productivity for various crops were based on data from the National Supply Company (CONAB); the Brazilian Institute of Geography and Statistics (IBGE); and Agroconsulta and Scott Consultoria, two private firms that update estimates for the sector on an annual basis. Brazilian export projections are exogeneous and were based on global projections of the Food and Agricultural Policy Research Institute (FAPRI), the same source used by the U.S. Department of Agriculture; FAPRI projections were used to calibrate export projections for 2009–18 and 2019–30. It was thus assumed that the key uncertainties are linked to macroeconomic projections. Under the Reference Scenario, projections for meat exports and pasture are relatively conservative. With the exception of the Amazon region, where significant growth in pasture is expected, volume nationwide remains fairly stable, which can be attributed to the continued stability in the global meat demand. Stabilization—or even a slight decrease in meat exports, observed over the past several years—is difficult for Brazilian industry to reverse, following the impressive development of the previous decade (1997–2006). Source: ICONE

Therefore, the carbon uptake volume indicated in this study may be at the upper limits of the range. Building flexibility into target setting would reduce the volume of carbon sequestered. At the same time, it would facilitate the effort of releasing the corresponding amount of pasture and thus mitigate the risk of inducing carbon leakage. That is, conversion of native vegetation would occur somewhere else as a result of the domino effect triggered by the induced net reduction of land available at the national level for crop and livestock expansion. In terms of carbon balance, avoiding the release of the full carbon stock of one hectare of burned forest in the atmosphere is preferable to the progressive removal of GHGs from the atmosphere through the restoration of one hectare of forest. It is thus essential to ensure consistency between efforts to release pasture and enforce the restoration of legal reserves.

4.9 Benefits Related to Reducing Aerosol Emissions Resulting from Deforestation by Burning

A study was conducted with the pupose of generating estimates for aerosol emissions from burning in the projected scenarios, the effects of land-use change and soil cover on surface flows, and lastly, how these changes affect the hydrologic cycle of South America, especially Amazonia.

Burning, which occurs mainly in the tropics, is a major source of atmospheric pollutants (Artaxo et al., 2002, Andreae, 1991). In South America, during the winter months, hundreds of thousands of fires are set principally in the cerrado and forest ecosystems. This burning occurs mainly in the central and Amazon regions, although the spatial distribution of the smoke covers an extensive area amounting to about 4-5 million km², far greater than the area where the fires are concentrated (Freitas et al., 2005, 2006, 2007). Gases are released into the atmosphere during biomass combustion, including some greenhouse gases, tropospheric ozone precursors and aerosol particles that interact efficiently with solar radiation, affecting microphysical processes, cloud formation dynamics and air quality. The effects of these emissions are widespread, affecting the composition and physical and chemical properties of the atmosphere in South America and nearby oceanic areas on a regional scale and potentially on a global scale. Emissions from burning change the atmospheric radiative balance both regionally and globally through the direct effect of aerosol particles when they reflect and

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Under the Low-carbon Scenario, the main carbon uptake potential resides in the recovery of legal forest reserves. Indeed, the proposed Low-carbon Scenario considered full compliance with the Forest Reserve Law—including an enormous effort to recover riparian and native forests—as a target for carbon uptake. This “Legal Scenario” would imply a break with the past. A fully Legal Scenario may be difficult to implement and flexibility mechanisms are already being discussed, especially regarding legal re- 193 serves, which may reduce the net area reforested. For example, in such Amazon states as Rondônia and Pará, which have already developed economic and ecological zoning, the legal reserve can be reduced from 80 percent to 50 percent, particularly for rural properties located along the main roads. In exchange, landowners would commit to fully restoring the 50 percent legal reserve, with the abated 30 percent converted into “agriculture consolidation areas.”

The balance between radiation and the hydrologic cycle can also be affected indirectly by emissions from burning through micro-physical alterations and the dynamics of cloud formation (Kaufman, 1995), due to the greater availability of cloud condensation nucleii (CCN) and ice in the atmosphere, which cause changes in cloud drop spectra (Andreae et al., 2004; Koren et al., 2004; Rosenfeld, 1999; Cotton and Pielke, 1996) and in thermo-dynamic stabilization (Longo et al., 2006). The increase in the concentration of aerosol particles results in the production of a greater number of smaller cloud drops, with two outcomes: first, the greater quantity of drops reflects more solar radiation back to space (although it cools the atmosphere), and second, the smaller size will be less favorable for rain production, as the tiny droplets tend not to stick together to form large drops that become rain. On the other hand, thermo-dynamic stabilization caused by the direct interaction of aerossol particles with solar radiation (reducing the heat in the low atmosphere by reducing solar radiation), restricts the rise of convective cells generated close to the surface, thus inhibiting cloud formation. This set of factors suggests that the effects of burning can have an impact on a local scale, with a major impact on the regional hydrologic cycle, as well as the planetary energy redistribution pattern in the tropics for medium and high latitudes.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

194

scatter solar radiation back in space, reducing the quantity absorbed by the land area, and when they absorb solar radiation, thereby heating up the atmosphere. Jacobson (2001) suggests that atmospheric warming due to black carbon aerosols could balance the cooling effect associated with other types (sulfates), and that their direct radiative force could exceed that of CH4. Thus, aerosol particles produced from incomplete combustion processes would only come second after CO2 in terms of contributing to atmospheric radiative heating.

On the other hand, land-use change can lead to variations in the balance of energy, water and momentum on the surface, due to the corresponding changes in its albedo, the evapotranspiration capacity associated with the plant cover and its spatial structure. In particular, the substitution of forested areas with deep root systems for pasture areas results in an increase in the albedo and less accessibility to deep soils with substantial water storage potential. This change generally leads to an inversion in the Bowen ratio-energy balance, producing drier, hotter and deeper planetary layer limits, mainly during the dry season. Thus, land-use changes cause alterations in the pattern of the hydrologic cycle, which can be evaluated using numerical values and land-use scenarios.

Another relevant aspect that has not been studied extensively is the effect of landuse change on dust aerosols. With more exposed soils and intense winds (which can be expected with the decrease in land rugosity when forests are replaced by pastures) there may be a significant increase in the production and removal of dust from the soil, impacting the radiative balance as well as the cloud microphysics and hydrologic cycle.

4.9.1 Methodology: Numerical Modeling with CCATT-BRAMS

The methodology used was based on numerical atmospheric modeling using the emissions model, chemical reactivity, transport and deposit of gases and CCATT aerosols (Coupled Chemistry-Aerosol-Tracer Transport) combined with the atmospheric BRAMS model (Brazilian developments on the Regional Atmospheric Modeling System).

Figure 47: Transport processes simulated by the CCATT-BRAMS, including plume rise, deep and shallow convective transport by cumulus, diffusion in the PBL, dry and wet deposition

Different types of aerosols are parametrized by the CCATT, including particulate matter generated by burning, resuspension of dust from the soil, agricultural activities and emissions of urban/industrial origin. Its CARMA diagram (Community Aerosol & Radiation Model for Atmospheres) enables the effects of long and short waves in aerosol and hydrometeor particles to be evaluated. This means that the model is able to conduct studies on the direct and indirect effects of aerosols on the radiative balance, as well as calculations of rates of heat, providing an important tool for studying the interaction between aerosols and the atmosphere (Longo et al., 2006) The effects of land-use change on atmospheric circulation are studied through the use of a series of numerical simulations whose pattern of occupation of the surface area was described by reference and Low-carbon Scenarios generated by other groups. In

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The BRAMS is a numerical meteorological model that simulates atmospheric circulation from hemispheric scales to scales of major turbulences in the planetary boundary layer. The model has a multi-layer arrangement that enables the different spatial and temporal scales (Walko et al., 2000) to be resolved simultaneously, with state-of-the-art physical parametrizations and a modern parametrization of cumulus clouds developed in the formalism of the ensemble (Grell and Devenyi, 2002). 195 The CCATT is a numerical system that is designed to simulate and study emissions, transport, deposits and physical and chemical processes associated with trace gases and atmospheric aerosols. It is a Euleriano transport model that merges completely with the BRAMS, enabling the simultaneous numeric provision of time, air quality and impact of aerosols and land-use changes on atmospheric development (Freitas et al. 2005, 2006, 2007; Longo et al., 2006, 2007). It has a transport model that resolves phenomena on the grade and sub-grade scale (the main processes are shown in Figure 47), besides a complete chemical mechanism for the prognosis of chemical reagent species.

196

this study, land-use and carbon equivalent emissions maps are used for the 2007-2030 period. The characteristics of each type of occupation shown by land-use maps were obtained from the literature. These include relevant biophysical properties such as albedo, rugosity, leaf area index and root depth. These properties enable the parametrization of the area through the simulation of sensible and latent heat flows and momentum. Different effects were considered either separately or together in each simulation, enabling an understanding of the individual impact and the potential existing feedback. The following section shows how equivalent carbon data and land-use maps were used to estimate aerosol emissions.

4.9.1.1. Calculation of Aerosol Emissions

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Two types of data were used to calculate aerosol emissions: land-use maps and carbon equivalent emissions maps for Brazil. These data include the period from 2007 to 2030 and were generated for both the reference and the Low-carbon Scenarios. Both sets of data had 1x1 km resolution and were produced by the Topic A team. Their use will enable aerosol emissions to be estimated based on the quantity of carbon equivalent available in the atmosphere for both scenarios being studied.

Emissions were assumed to be coming from biomass combustion generated by the deforestation of forest areas that were later converted for other uses (such as agriculture and pasture). These deforested areas were located using land-use maps, which illustrate the type of biome occupation present in a given area (Map 37). Using this type of annual map, it was possible to determine the annual development of forest cover, and thus to locate regions where it is disappearing (deforestation).

While land-use maps were used to locate points of aerosol emissions, carbon equivalent maps were used to determine the quantity of aerosol emitted at these points. Only positive carbon equivalent values were used, as the objective was to estimate emissions based on the amount of carbon released into the atmosphere. The necessary calculations for transforming carbon equivalent emissions into aerosol emissions follow these steps: a) Estimate quantity of carbon equivalent emitted through combustion. b) Transform this quantity into carbon dioxide emissions.

c) Obtain emissions from aerosols using the values of emissions factors available in the literature. For calculating the item, it was assumed that 85 percent of the carbon equivalent emitted was the result of combustion processes (Soares Filho, personal communication), thus we have (Equation 48): [carbon equivalent emissions from combustion] = 0.85[carbon equivalent emissions] (48)

A factor of 3.66 was adopted (Soares Filho, personal comunication) for the transformation described in item b for the conversion of carbon equivalent emitted in terms of carbon dioxide (Equation 48b): [carbon dioxide emissions] = 3.66[carbon dioxide equivalent]

(48b)

Map 37: Land-use map for the year 2007 in the Reference Scenario (1x1km resolution)

Item c was realized using values from emissions factors for carbon dioxide and aerosols with which it is possible to estimate the fraction of CO2 attributed to particulate matter and then estimate aerosol emissions. Table 42 shows some values from emissions factors associated with forest, savanna and pasture according to work done by Andreae and Merlet, 2001. Table 42: Emissions factors (g/kg) for different biomes for CO2 and aerosols (particulate matter with a less than 2.5 micrometer diameter - PM2.5) Tropical Forest Savanna Pasture

CO2

PM2.5

1664

4.9

1580

1664

Source: Andreae and Merlet, 2001

9.1 4.9

As deforestation was assumed based on biomass combustion from forest remnants, amounts associated with tropical forests, or 1580 g/kg for the carbon dioxide emissions factor and 9.1 g/kg for aerosols were used. Considering what was said in previous sections, the expression for the calculation of aerosol emissions may be written as follows (Equation 49):

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

197

198

which directly illustrates the conversion of carbon emission equivalent into aerosols emitted by combustion. The emission units are the same as for the carbon equivalent, in other words tons per hectare and per year.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

To finalize, in order to enter emissions data into the CCATT-BRAMS, it was necessary to parametrize the annual performance of these emissions. As the model generates results every 6 hours, emissions data were converted into a larger temporal resolution through performance curves for the number of emissions sources generated by the 6 different regions of Brazil, as shown in Map 38. Thus, annual emissions data were converted into daily data and inserted into the model. It should be noted that emissions generally peak between August and November, depending upon the region. Map 38: Schematic map of Brazil showing the different regions in the country and their boundaries for the analysis of results (above). Below, normal performance of the number of emissions sources in the different regions obtained with data from the AVHRR sensor (Advanced Very High Resolution Radiometer) from 1998 to 2008, present in the satellites of the NOAA series (National Oceanic and Atmospheric Administration)

4.9.1.2 Aerosol Emissions in the Reference and Low-carbon Scenarios

Both scenarios show a drop in emissions until the year 2010. In the Low-carbon Scenario, the drop is more sudden, reaching approximately 6000 tons per hectare in 2010. In the Reference Scenario, this amount is about 16,000 tons (62 percent higher). From 2010 on, emissions pratically stabilized in the Low-carbon Scenario, with close to 7600 tons in 2030. On the other hand, there was an increase in the Reference Scenario, with maximum emissions of approximately 22,400 tons occurring in 2030, although this is below the 2007 emissions level (22,800 tons). According to Table 43, emissions in the Low-carbon Scenario are 62 to 66 percent less than those of the Reference Scenario between 2010 and 2030. Figure 48: Estimate of total annual aerosol emissions in Brazil for the reference and Lowcarbon Scenarios (Table 43)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

This section discusses results obtained for emissions estimates from aerosols in the reference and Low-carbon Scenarios. Only emissions from deforestation according to the methodology were taken into account in the calculations. The annual emissions performance may be seen in Figure 48 and Table 44. 199

Table 43: Total annual aerosol emissions (tons per hectare and per year) throughout the country for the reference (REF) and low carbon (LC) scenarios. Also shown are the figures of absolute differences (LC-REF) and differences in percentages (LC-REF (%)) between emissions for the two scenarios. Year

200

LC

LC-REF

LC-REF (%)

6397

-12286

-65.8

5940

-10295

-63.4

2007

22789

22854

2010

15879

6035

2008 2009

2011 2012

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

REF

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Total

20505 18684

16723 16234 17244 17594 17996 18534 18428 17960 18532 18929 18468 19063 19282 20224 20021 20729 20604 21002 20507 22373

458305

20253 5975 6242 6655 6411 6267 6546 6669 6814 6859 6977 6972 6958 6947 7084 7238 7147 7160 7301 7584

191286

65

-252

-9844

-10748 -11002 -10939 -11585 -12267 -11882 -11291 -11718 -12070 -11491 -12091 -12323 -13276 -12937 -13491 -13457 -13842 -13207 -14789

-267020

0.3

-1.2

-62.0 -64.3 -63.8 -62.2 -64.4 -66.2 -64.5 -62.9 -63.2 -63.8 -62.2 -63.4 -63.9 -65.6 -64.6 -65.1 -65.3 -65.9 -64.4 -66.1

-58.3

Map 39 shows the spatial distribution of aerosol emissions in the country in the two scenarios, concentrating principally on the Amazon Rainforest and in its transition region with savannas. The difference in the spatial distribution between the low-carbon and Reference Scenarios is principally due to the difference in intensity of the pace of deforestation in the regions. In the Reference Scenario, mainly in the transition areas of the Amazon Rainforest, conversion of forest areas into pasture is intense, resulting in an increase in areas of emissions. The area of pasture expansion in the country went from 2.6x106 to 2.8x106 km² in the Reference Scenario.

As a visual comparison, Map 39 also shows the aerosol load simulated by the CCATTBRAMS model, represented by the optic depth of the aerosol, illustrating the result obtained by entering amounts of emissions from burning into the model. 201

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Map 39: Figures (A), (B), (C) and (D) show the locations of deforestation from 2007 to 2030 in the reference (REF) and low carbon (BC) scenarios. Regions with forest remnants are also shown during that period (in green). Figures (E) and (F) show the optic depth of average aerosol from 2007 to 2030 in the reference (E) and low carbon (F) scenarios, where the current lines represent the average wind field over Brazil

4.9.2 Results

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

202

One important aspect that emerges is that, although emissions from burning come principally from the Amazon region, the smoke emitted may be transported to regions far from the emissions sources due to atmospheric circulation. Results shown by the CCATT-BRAMS model in Map 38 demonstrate the transport of plumes from burning by low level streams. These streams flow predominantly from the Northeast to the Southeast and are found at an elevation of about 2000 meters. They originate from the change in the trajectory of the tradewinds when they meet the Andes mountain range and are responsible not only for the transport of plumes due to burning, but also for the humidity generated from the Amazon region to the southern and southeastern parts of Brazil. This aspect of transport to distant regions prompted the analysis of results from different parts of the country. The boundaries used in the respective regions were the same as those exhibited in Map 38. Region 2 covers a good part of the deforestation area and is responsible for the greater production of fire outbreak, followed by region 1 (remembering that in Map 38 the fire outbreak are normalized). Although there are also fire outbreaks in other parts of Brazil, the impact of meteorological variabilities appears much greater due to the transport of plumes from the Amazon region to regions 4, 5 and 6. In the following section, the different impacts on precipitation and temperature of the reference and Low-carbon Scenarios of the study will be presented.

4.9.2.1 Precipitation

This section discusses the differences in the amounts of precipitation observed between the reference and Low-carbon Scenarios. But before discussing these differences, and as a justification for evaluating the model’s performance, Figure 49 shows the average monthly precipitation simulated in the CCATT-BRAMS model from 2007 to 2008 in the reference and Low-carbon Scenarios compared to data on precipitation obtained by the National Water Agency (Agência Nacional de Águas - ANA), which corresponds to the climatology realized during the period between 1982 and 2005. The reason that only 2007 and 2008 are included in this analysis is that they are the initial years for the two scenarios, and it is expected that their performance is not very different from the climatology. In addition, the difference between the two scenarios with regard to emissions caused by burning and land-use change is not very perceptible.

The results show that the model was able to coherently accompany the performance of the precipitation observed in the different regions of Brazil, mainly with regard to their seasonal performance. The only exception was region 6, where the model tended to underestimate the precipitation in the middle of the year compared to ANA climatology, but that didn’t impede the analysis of the impact on the difference in precipitation between the scenarios, nor its tendency to vary through the years.

With regard to the difference between the two scenarios, Figure 50 shows the average monthly precipitation observed in the reference and Low-carbon Scenarios during the period between 2007 and 2030 simulated in the CCAT-BRAMS.

203

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 49: Average monthly precipitation in the 6 regions analyzed in the reference and Low-carbon Scenarios from 2007 to 2008 compared to data obtained from the Agência Nacional de Águas (National Water Agency - ANA), which corresponds to monthly precipitation during the period from 1982 to 2005. The margins of error represent the standard deviation for each month

Figure 50: Average monthly precipitation in the 6 regions analyzed in the reference and Low-carbon Scenarios from 2007 to 2030 (bar graph left axis). Also shown is the difference between the reference and Low-carbon Scenarios (line graph right axis)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

204

Results suggest that the impact on precipitation between the two scenarios is principally due to the effect of the aerosols. In the six regions analyzed, the greatest difference between the two scenarios occurs during the months of September and October, during which there are more emissions of particulate matter due to burning. In regions 1 and 2, which are the most affected by these emissions, the impact on the Reference Scenario was from -55 to -70 mm on average between 2007 and 2030. Particularly in region 2, the increase in the length of the dry season in the Reference Scenario is very clear. Similar to the discussion on the topic in section 4.2 on liquid radiation, the impact on precipitation also occurred in other regions, mainly due to the transport of emissions from burning by low level streams, with little impact in region 3, which is located in the Northeast (-11 mm), and moderate impact in regions 4, 5 and 6, more to the South, where the model suggested a difference of -21, -23 and -23 mm, respectively, during the month of October. The difference between the two scenarios can be seen in Figure 51, which presents the average spatial and trimestrial distribution during the period from 2007 to 2030. It also shows that the impact is not very significant between February and June, when aerosol emissions from burning are less intense. The peak occurs during the August,

September and October trimester, when the impact on the precipitation in the Reference Scenario can be -200 mm at some points. It should be noted that the Low-carbon Scenario has more precipitation because of the influence of aerosols on cloud microphysics. Environments that are more heavily loaded with particulate matter (in the case of the Reference Scenario) make it difficult for the cloud droplets to grow, and precipitation tends to be lower in these cases. This effect was parameterized in the CCATT- 205 BRAMS model.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 51: Difference in precipitation (mm) between the reference and Low-carbon Scenarios for the years 2007 to 2030 during the February, March and April (A), May, April and June (B), August, September and October (C), and November, December and January (D) trimesters.

4.9.9.2 Temperature

Figure 52: Average monthly temperature in the 6 regions analyzed in the reference and Low-carbon Scenarios in 2007 and 2008 compared to data obtained from the National Meteorology Institute (Instituto Nacional de Meteorologia - INMET), which corresponds to the monthly climatology of temperature during the period from 1977 to 2000. The margins of error represent the standard deviation for each month

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

206

This section discusses the differences observed in the temperatures between the reference and Low-carbon Scenarios. Similar to the section on precipitation, results from the model and climatology amounts were compared. Figure 52 shows the average monthly temperature simulated in the CCATT-BRAMS model during the years 2007 and 2008 in the reference and Low-carbon Scenarios, compared to the temperature data obtained by the National Institute of Meteorology (INMET) which correspond to the climatology realized from 1977 to 2000. Results show that the model was able to coherently accompany the performance of the temperature observed in the different regions of Brazil, underestimating the temperature only in region 5 and overestimating it in region 6, but with the seasonal performance preserved. Once again, these facts do not prevent the impact on the difference in temperature between the scenarios, nor its tendency to vary through the years, from being analyzed.

Figure 53 shows the average spatial and trimestrial distribution of the temperature difference between the two scenarios from 2007 to 2030. W ith precipitation, the impact is less significant from February to June. The peak occurs during the trimester corresponding to the months of August, September and October, when the temperature in the Reference Scenario can be almost 3 degrees higher at some points. Regions 1, 2 and 3 show a tendency for the temperature to increase regardless of the scenario in question, while regions 4, 5 and 6 exhibit little tendency in this direction. For example, in region 2, in the Reference Scenario in 2007, the temperature was 27 degrees. Over time this amount increased to 30 degrees in 2030. The increase in temperature through the years in regions 1, 2, and 3 is related to the increase in the sensible heat simulated in these regions. In annual terms, the Reference Scenario is about 1 degree higher in regions 1 and 2, and tens of degrees higher in other regions.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

When analyzing the results and comparing them to temperatures obtained from the reference and Low-carbon Scenarios, one particular detail stands out. Although in the Reference Scenario there is a lower penetration of radiation from solar radiation due to the greater quantity of particulate matter present in this scenario, resulting in less net radiation, results showed that the temperatures were higher than in the Lowcarbon Scenario. This occurred due to the modification in the amount emitted via latent 207 and sensible heat by area. In this context, more energy is emitted via sensible heat in the Reference Scenario, compensating for the incident radiative loss, and resulting in higher temperatures. In compensation, the Low-carbon Scenario, even with greater net radiation available, has more energy emitted via latent heat, with no temperature changes in the process. Its temperatures were thus lower. It should be noted that the difference between the amounts of latent and sensible heat between the scenarios was mainly framed by the difference in simulated precipitation between them.

Figure 53: Difference in temperature (Celsius) between the reference and Low-carbon Scenarios for the years 2007-2030 for the February, March and April (A), April, May and June (B), August, September and October (C), and November, December and January (D) trimesters

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

208

4.9.3 Summary of the Reduction of Impacts on Rainfall and Temperature Regimes in the Low-carbon Scenario

To summarize, the regions most affected by burning tend to have reduced average annual precipitation during the period from 2007 to 2030 due to the gradual

increase in emissions from particulated material during that period. It should be mentioned that this performance occurred independent of the scenario in question although the Reference Scenario presents less precipitation than the Low-carbon Scenario. An example of this performance in the reduction of precipitation occurred in the region of the deforestation arc. While it rained about 1800 mm in 2007 in the region, in 2029, this amount was almost 1200 mm in the Reference Scenario, a reduction of approximately 35 percent. In annual terms,

the average percentage difference in precipitation between the reference and Low-carbon Scenarios may exceed 30 percent, since this percentage was approximately 15 to 20 percent most of the years between 2007 and 2030.

Figure 54 shows a spatialization of the difference in the amount of accumulated rainfall during the August-September-October trimester between the reference and 209 Low-carbon Scenarios using the 2007-2030 average.

The average temperature tends to increase in regions most affected by burning during the 2007-2030 period due to the gradual rise in the amounts of sensible heat related to the decrease in precipitation during that period. Once again, this type of performance was independent of the scenario in question. Analyzing the region of the deforestation arc, while its temperature was between 26 and 27 degrees in 2007, in 2030 its average exceeded 30 degrees in the Reference Scenario. In annual terms, the average difference between the reference and Low-carbon Scenarios was around 1 degree in the regions most affected by burning and by tens of degrees in other regions. Figure 55 shows a spatialization of the difference in temperature in the AugustSeptember-October trimester between the reference and Low-carbon Scenario, using the 2007-2030 average.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 54: Difference in accumulated rainfall between the reference and Low-carbon Scenarios using the 2007-2030 average. The color scale refers to amounts in millimetres of rainfall per year

Figure 55: Difference between the reference and Low-carbon Scenarios in average air temperature, taken between 2007 and 2030. The color scale reflects the amounts in celsius

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

210

The use of a low-emissions scenario from particulate material and with less conversion of green areas into pastures resulted in a 15 to 20 percent increase in average annual precipitation between 2007 and 2030 compared to the Reference Scenario. Along similar lines, a one-degree reduction in the temperature per year, due to less sensible heat flow, was observed.

5 Analysis of Transition Costs from the Reference Scenario to the Low-carbon Scenario

This study conducted a cost–benefit analysis that enabled comparisons to be made between individual options in the Low-carbon Scenario and between the low-carbon and reference-scenario options in general.

It should be emphasized that it is not possible to conduct an exhaustive cross-sectoral economic analysis of externalities. Although the key co-benefits of certain mitigation and carbon uptake options considered under the Low-carbon Scenario could be measured in physical terms to study their sustainability, the shear number and diversity of the sectors involved virtually precludes a comprehensive analysis of externalities. Ensuring the homogeneity of the analysis inevitably means limiting it to direct and measurable costs and revenues, thus omitting important co-benefits that may be essential for shaping the decision-making process.

Making a joint assessment of the different measures considered is especially challenging, since they are implemented in diverse contexts. Some occur within the public economy framework and are implemented by local or federal government, while others are conducted by the private sector. Some generate revenue, others savings, and still others generate co-benefits and externalities. Some are capital-intensive with a timeframe that goes beyond 2030, while others involve short-term changes in operational conditions. The assessment could vary significantly, depending on whether it is from the public or private sector perspective. In order to better inform decision-makers, the study team conducted the cost-benefit analysis using both social and private sector approaches. The social approach provided a basis for making a cross-sectoral comparison of the cost-effectiveness of the 40 mitigation and carbon uptake options considered in the study. A social discount rate was used to calculate the Marginal Abatement Costs (MACs). The MACs of all proposed mitigation and carbon uptake measures were sorted by increasing value, and plotted along a single graph to facilitate a quick cross-sectoral comparison of their costs and the volume of emissions they could reduce or sequester. This graph, which combines the 40 options from the four sectors mentioned above, is presented in the main part of the low-carbon multisectoral report for Brazil. The private approach assessed the conditions under which the proposed measures could become attractive to economic agents who are deciding whether to invest in lowcarbon alternatives in lieu of the more carbon-intensive options found in the Reference Scenario. The private approach adopted in the study estimated the economic incentive that would be needed in order for the proposed mitigation measure to become attrac-

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

An economic analysis of the Low-carbon Scenario helps inform both the govern- 211 ment and society about the economic costs and benefits of choosing the development path with lower carbon emissions. It also helps to understand the conditions under which the proposed mitigation and carbon uptake options could be effectively implemented. At the same time, there is no single method for analyzing these options. A variety of perspectives can be used to inform a wide range of audiences and agents about the economic conditions under which a Low-carbon Scenario could be implemented.

212

tive. If the incentives were provided through the carbon finance market, the private approach would indicate the minimum carbon price, expressed in US$ per tCO2e, needed to make the low-carbon option attractive enough to be implemented. This does not necessarily mean that the corresponding economic incentive must be in the form of carbon revenue through the sale of carbon credits; capital subsidies for low-carbon technologies or a combination of incentives could be used. Financing conditions and tax credits can sometimes be far more efficient in channeling the corresponding incentive to make the low-carbon option the preferred choice of project developers. The “Social Approach”: Calculating the Marginal Abatement Cost Curve

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Using the social approach, the costs and benefits of the option implemented in the Reference Scenario during the 2010–30 period were subtracted year by year from the costs and benefits of the proposed low-carbon option implemented during the same period. The 2009 net present value (NPV) of the annual incremental costs and benefits were then calculated to determine the average avoided tCO2e, or MAC, during that period. The NPV was calculated using a social discount rate of 8 percent. That is the value used in the PNE 2030 for Brazil’s long-term National Energy Plan and is generally used for projects financed by the Brazilian Development Bank (BNDES). In this study, activity-level mitigation measures were analyzed individually. Portfolios of these measures were then developed at the sectoral level to construct a Lowcarbon Scenario. The associated potential for each mitigation option was adjusted to ensure internal consistency at the sectoral level to avoid the duplicate calculation of emissions reductions.

Since decision-makers may have to choose between alternatives that differ markedly in terms of cost-benefit distribution over time, particularly with regard to investment costs, 2009 values were used for calculations and comparisons (Box 4).

For purposes of comparison, the study also conducted sensitivity analyses for discount rates of 4 and 12 percent.

The results of calculations of marginal abatement costs using the social approach for mitigation and carbon uptake options for the four sectors covered by the general multisectoral study are presented in Table 44 below. The results for options related to land use and land-use change are in boldface.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

213

Table 44: Mitigation potential and marginal abatement cost of various alternatives, based on three discount rates

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

214

The “Private Approach”: Determining the Break-even Carbon Price

To assess the feasibility of implementing the mitigation and carbon uptake options from a private sector perspective, the study team calculated the incentives that would be required to make the proposed measures attractive to Brazil’s economic agents. The team applied a two-part method. First, it estimated the minimum internal rate of return (IRR) that Brazil’s economic agents could expect in the subsector where the proposed mitigation measure is implemented. Second, it estimated the required minimum incentive as the perceived revenue per avoided tCO2 that would make shifting from the reference option to the low-carbon option attractive; that is, the resulting IRR, including the incentive, would at least equal the benchmark IRR.

This data was compiled to arrive at a consensus on the rates used and observed in practice, yet these benchmark IRRs remain indicative. At the same time, they differ markedly from the social discount rate used to calculate the MAC and can change from one sector or subsector to another, confirming that the MAC presented in the above section should not be used as a proxy for the market incentive to be provided at the project level.

GHG mitigation projects with IRRs above benchmark IRRs are expected to attract market investors; conversely, those with IRRs below benchmark IRRs will likely require added incentives, such as carbon credits or other mechanisms in order to attract private financing. The level of such incentives is seen as the break-even carbon price because it represents the amount of incentive that will equate benefits and costs to achieve the required benchmark IRR. If the break-even carbon price for a GHG mitigation option is negative, the implementation of such a measure is, for the most part, already attractive, and its IRR is, in most cases, even higher than the sector’s IRR benchmark and no incentive is needed. However, if the break-even carbon price is positive, the option is not attractive and cannot generate the required benchmark IRR without incentives in the amount of the break-even cost.

Interestingly, for certain mitigation options, the value of the Marginal Abatement Cost (MAC), which uses the social discount rate of 8 percent, was less than zero; but the break-even carbon price, which uses private-sector discount rates, such as the indicative benchmark IRR, was positive (e.g., reduction of deforestation). Corresponding options, which appeared economically attractive under a social approach, are no longer attractive when using a private-sector approach. Other mitigation options, already considered expensive when viewed with the social discount rates, would have much higher costs when assessed from the private sector perspective.

40

It is important to note that, in practice, certain proposed mitigation options are components of projects and cannot be separately financed; thus, for these options, the IRRs for overall projects were used.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Because the risk levels perceived by investors differ according to type of technology, investor strategies may also vary based on the market conditions observed in particular subsectors – and necessary rates of return may differ depending on the technology.40 To establish such a benchmark IRR, the study team consulted both the institutions in Brazil that finance projects in the subsectors considered, as well as important players and entrepreneurs in the field. While issues of confidentiality prevented these institu- 215 tions from disclosing detailed information through this report, the consistency of the data provided gave the project team a sense of the robustness of the estimates thus established.

Table 45: Comparison of sector benchmark IRRs and break-even carbon prices for various mitigation options (8 percent social discount rate) Mitigation  option Residential  lighting Steam  recovery  systems Heat  recovery  systems Industrial  lighting Solar  thermal  industrial  energy Combustion  optimization Recycling Furnace  heat  recovery  system Other  energy  efficiency  measures Scalling  up  no  tillage  cropping Optimizing  traffic Reducing  deforestation  +  livestock Landfill  methane  destruction Sugarcane  cogeneration Natural  gas  displacing  other  fuels Reforestation Ethanol  displacing  domestic  gasoline Investing  in  bike  lanes Wastewater  treat.  +  methane  destruction  (res.  &  com.) Gas  to  liquid  (GTL) Ethanol  exports  displacing  gasoline  abroad Eletric  motors Existing  refineries  (energy  integration) Wind Renewable  charcoal  displacing  non-­‐renew.  charcoal Investing  in  railroads  and  waterways  vs.  roads New  refineries Commercial  lighting New  industrial  processes Existing  refineries  (incrustation  control) Transmission  line  Brazil-­‐Venezuela Refrigerators  (MEPS) Wastewater  treat.  +  methane  destruction  (ind.) Investing  in  metro Existing  refineries  (advanced  controls) Solar  heater  -­‐  residential* Bullet  train:  São  Paulo  -­‐  Rio  De  Janeiro

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

216

Abatement  cost   Carbon  incentive-­‐ incremental   (US$/tCO 2 )  (8%   social  discount   approach   rate) (US$/tCO 2 ) (120) (243) (97) (228) (92) (220) (65) (173) (55) (123) (44) (104) (35) (91) (26) (41) (14) (22) 0 0 (2) 4 0 6 3 7 (105) 8 (20) 10 39 12 (8) 24 1 25 10 33 (2) 34 2 48 (50) 72 7 75 (8) 93 21 95 29 97 19 106 (52) 122 2 174 73 209 (31) 216 (41) 223 103 251 106 371 95 431 4 698 400 7787

Benchmark   IRR  (%) 15 15 15 15 15 15 15 15 15 8 15 10 12 18 15 10 15 15 12 25 15 15 15 10 15 17 15 15 15 15 15 15 12 17 15 15 19

*Note:  Positive  MACs  residential  solar  heater  versus  negative  costs  for  industrial  solar-­‐thermal  substitution  reflect  the  lower   carbon  content  of  residential  electricity  generation  (mainly  hydropower)  versus  the  higher  carbon  content  of  industrial  thermal   energy  generation  (gas,  diesel,  coal).  

Many of the mitigation options with negative MACs would also not require incentives from the private sector perspective (e.g., most energy-conservation options in the industry). These would generate such great economies of energy that implementation, even from a private-sector perspective, would be considered a win-win situation. In such cases, mandatory standards may be an option to harvest such “low-hanging fruits.” Obviously not all mitigation options would be tackled solely from a private sector perspective; otherwise, government incentives may be provided for reasons other than GHG emissions reductions. Nonetheless, this perspective is valid to demonstrate

where incentives might be better placed or most required and where other tools, such as regulation and standards, may be more appropriate than carbon finance.

Table 46: Volume of incentive required (undiscounted) in order to achieve the emissions reductions considered in the Low-carbon Scenario from 2010 to 2030

Avoided Emissions (Mt CO2e)

Total Incentive Required (US$ MMs)

Annual Incentive Required (US$ MMs)

Energy

1,721

142,892

6,804

LULUCF

7,481

46,769

2,227

Transport Waste Total

487

1,317

$ 11,006

185,018

$

70,256

444,935

8,810

$

3,346

21,187

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In theory, every measure whose break-even carbon price falls below the market carbon price would be implemented as a result of the action of market forces. However, as mentioned earlier, the corresponding economic incentive would not necessarily be in 217 the form of carbon revenue through the sale of carbon credits; other incentives, such as financing conditions or tax credits, could be used. An estimate of the total volume of incentives needed over the study period would amount to US$445 billion or US$21billion per year on average. Transport mitigation options would require the greatest amount of average annual incentives at approximately $9 billion, followed by energy at $7 billion, waste at $3 billion and LULUCF at $2.2 billion (Table 46). Almost all of the mitigation options would require financial incentives, with the exception of energy efficiency measures.

5.1 Costs of Reducing Emissions from Deforestation

218

The two main emissions mitigation and carbon uptake options identified in this study are: (i) avoiding deforestation, estimated at 9.8 Gt CO2e over the 2010–30 period and (ii) carbon uptake through the restoration of legal forest reserves, estimated at about 1.0 Gt CO2e during the same period. The sub-sections that follow analyze the costs of transitioning from the LULUCF Reference Scenario to a proposed Low-carbon Scenario, to harvest the potential of these two major mitigation and uptake measures.41

Two key measures were analyzed in terms of investment and financing needs to quantify the costs involved in avoiding deforestation, which are (Chapters 2&3): Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

• Improving livestock productivity to free up the land necessary for other activities. It is estimated that this measure will lead to a 70 percent reduction in deforestation, declining from an annual average of 19,500 km² to roughly 4,780 km² per year (a figure slightly below the government target of 5,000 km²).

• Preserving forests. This complementary set of measures aims at protecting the forest where deforestation is illegal.42

5.1.1 Improving Livestock Productivity

There are four categories of livestock production systems: two of lower productivity (degraded and extensive pasture) and two of higher productivity (feedlot and mixed crop-livestock). In the Reference Scenario, degraded and extensive pasture account for over 90 percent of the land used for livestock activities. In the Low-carbon Scenario, these lower productivity systems are gradually replaced by the feedlot and mixed crop-livestock systems, until these higher productivity systems cover approximately 60 percent of the total land required by livestock production in 2030. The increase in beef production in the higher productivity systems would reduce the need for pasture, so the land could be used for other purposes. This would in turn reduce the pressure on the forests, resulting in lower GHG emissions.

As discussed in Chapter 3 (Table 35), 70.4 million hectares of additional land would be made available: 16.8 million hectares for crops, production forests and pasture expansion in the Reference Scenario and 53.4 million for new activities for new mitigation and carbon uptake activities in the Low-carbon Scenario (44.3 million hectares for the restoration of environmental liability in legal forest reserves, 6.4 million hectares for additional ethanol production, and 2.7 million hectares for production forests).

Compared to the lower productivity systems, high productivity systems require significantly more financial resources for investment and expenses, and offer higher returns. In terms of production costs over the 2010-2030 period, recovery of de41 42

More details are found in the LULUCF technical report and in consultants’ reports on related topics. Other measures on avoiding deforestation where it is still legal will not be calculated in this analysis. Some of the measures currently being discusssed in Brazil as well as internationally include financial incentives, sometimes called payments for environmental services, and are offered to economic agents to compensate for opportunity costs for cancelling deforestation rights.

graded pasture through the adoption of the crop-livestock system would require an additional investment of R$2,925 (US$1,330) per hectare, as well as another R$21,300 (US$9,682) per hectare to cover expenses. Adoption of the feedlot system for cattle during the same period would require R$1,144 (US$ 520) per hectare of additional investments and R$4,869 (US$2,213) per hectare for additional expenditures (Table 47). Table 47: Investments and expenditures for prototypical livestock systems (2009-30) Investment

Expenditures

Extensive pasture

2,775

4,644

Degraded pasture

2,124

Feedlot



3,267

Crop-Livestock

5,049

2,594

7,463

23,894

*

R$ additional per hectare *

Total

Investment

Expenditures

Total

7,419

651

2,051

2,702

2,925

21,300

24,225

4,717

10,730

28,943

-

1,144

Exchange rate R$2.20 = 1US$.

-

4,869

-

6,013

Based on the relative prices considered, the higher productivity systems (feedlot and crop/livestock) generate dramatically higher IRRs (7.50 percent and 15.47 percent, respectively) than those of low productivity systems (degraded and extensive pasture) (Table 48). Table 48: Economic and financial performance of prototypical livestock systems (2009-2030) System Degraded pasture Extensive pasture Feedlot

Crop-livestock

NPV* (R$ per hectare)

IRR (%)

(1,857.84)

NC**

(1,128.76) (95.19)

1,953.46

* Based on an 8 percent social discount rate. ** NC = non-calculable sufficiently negative value.

0.56

7.50

15.47

As a result, the economics of the reference and Low-carbon Scenarios differ markedly. The per-hectare cost under the Low-carbon Scenario is R$10,600, far higher than that of the Reference Scenario. Over the 2010-30 period, the per-hectare cost difference would amount to R$3,139 on average (Table 49).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

R$ gross per hectare*

Production System

219

Table 49: Investment and expenses in the reference and Low-carbon Scenarios Scenario Reference

220

Low Carbon

Total investment expenditure (Gross R$ per hectare) 2,688

2,996

5,020

7,849

Total investment expenditure (Additional R$ per hectare)

7,708

10,845

Source: EMBRAPA

2,688 308

5,020

7,708

2,829

3,137

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The economic performance of the livestock sector is far better in the Low-carbon Scenario than in the Reference Scenario. Using an 8 percent social discount rate, the overall NPV of the investment and corresponding cash flows of the Reference Scenario over the 2010-30 period, were R$18 billion (US$8 billion). By contrast, the NPV of the Low-carbon Scenario results in R$14 billion (US$6.5 billion). Compared to the Reference Scenario, the average IRR for the Low-carbon Scenario for the livestock sector increased from a negative value43 to 11.24 percent (Table 50). It is important to note that the NPV and IRR calculated here refer only to new investments made from 2010 onward in both scenarios. Neither investments made before that date nor related expenses were taken into account. Table 50: Comparable economic and financial performance of the livestock sector



Scenario

NPV (2010–30) (R$ billion)

IRR (%)

Reference

(17,782)

NC*

Low Carbon *

14,335

NC = sufficiently negative non-calculable amount.

11.24

These differences in economics are accompanied by differences in environmental performance: The Low-carbon Scenario for LULUCF does not require additional land for land use, and therefore does not contribute to deforestation and, in turn, its associated GHG emissions.

5.1.2 Forest Protection

Although the low carbon land-use scenario offers solutions for bringing the need for additional land virtually to zero, it is expected that complementary forest protection measures would also be required for two major reasons. First, the legal limit for deforestation (up to 20 percent of properties located in the Amazon region) has not yet been reached. Thus, where the complex dynamic of deforestation is powered by the financial value of the wood or cleared land (along with with the need for cropland, pasture and production plantations), deforestation would continue. Second, there may be a significant delay between the time demand for cropland, pasture or production forests is 43

The illegal appropriation of land for speculative purposes may explain why an apparently unattractive activity from the economic point of view continues unabated. However, the land title question that the “Terra Legal” (Legal Land) program seeks to address was not within the scope of this study.

reduced and the time one could effectively observe a behavioral change among deforestation agents at the frontier.

To assess investment costs and expenditures for managing and enforcing the protection of conservation units where deforestation is illegal, the study used the “Investimento Mínimo de Conservação” (Minimum Conservation Investment – MCI [IMC]), developed by the Working Group on the Financial Sustainability of the National System of Conservation Units (Sustentabilidade Financeira do Sistema Nacional de Unidades de Conservação - SNUC), created by the Ministry of Environment.44 Using the IMC tool, the study assessed the costs associated with four protection activities over the 2010-30 period: (i) protection of indigenous reserves, (ii) protection of conservation units, (iii) control along the road network and (iv) remote sensor monitoring. These activities aim to prevent intrusion into and deforestation of these areas, as well as prohibiting the transport of products resulting from from illegal deforestation. During this period, protection costs would total US$24 billion, or $1.14 billion per year on average (Table 51).

44

The IMC (Minimum Conservation Investment - Investimento Mínimo de Conservação) is a tool that is based on the financial module of the Sistema Mínimo de Conservação (Minimum Conservation System – MICOSYS), developed by D. Vreugdenhill; see D. Vreugdenhill, “MICOSYS, Aplication Honduras ‘National Parks Model’”, Evaluation Record in MS Excel, prepared by PPROBAP, COHDEFOR/PNUD/World Bank/GEF Project (2002).

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Therefore, the Low-carbon Scenario proposes to implement additional forest protection measures in forested areas where deforestation is illegal. Given the many ongoing programs and abundant literature available on this topic, including the Plan of 221 Action for the Prevention and Control of Deforestation in the Legal Amazon (PPCDAM), this study was limited to reviewing existing proposals (Chapter 3). We present here, in order of magnitude, the results of a preliminary analysis of additional costs that could arise from the need for additional forest-protection activities. These aim to ensure that the full potential to reduce deforestation will be achieved via the release of pasture land and livestock productivity gains, as proposed in the Low-carbon Scenario.

Table 51: Projection of costs for forest protection in areas where deforestation is illegal (in millions of US$) Conservation Units

222

Year

2010

2011

516 0

2018

2019

2020

2021

2022

2023

2024

2025

2026

2027

2028

2029

2030 Total

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

516

Cost

430

43

381

0

93

43

400

ture 430

430

2017

Sensing Expendi-

0

2016

Annual

Invest-

2014

0

Total

by Remote

Expendi-

430

0

Monitoring

way Network Invest-

0

2015

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

ment

Control of the High-

Expendi-

2012

2013



Invest-

Indigenous Reserves

430

430

430

430

430

430

430

430

430

430

430

430

430

430

430

430

430

9,035

ment

1,680 43

43

43

43

43

43

43

43

43

43

43

43

43

43

43

43

43

43

2,539

ture 372

391

410

419

429

438

448

457

467

476

486

495

505

514

523

533

542

552

561

9,797

ment 112 0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

112

ture

Cost

93

1

3,205

93

1

958

93

93

93

93

93

93

93

93

93

93

93

93

93

93

93

93

93

93

1,963

1

1

1

1

1

949

968

977

987

996

1

1,006

1

1,025

1

1

1

1

1

1

1

1

1

1

1

1

21

1,015

1,034

1,044

1,053

1,063

1,072

1,082

1,091

1,101

1,110

1,120

1,129

23,983

It should be emphasized that the mitigation options considered under the Lowcarbon Scenario do not include additional measures to prevent deforestation in areas where it is still legally allowed. Elaboration and quantification of such proposals were beyond the scope of this study. If such additional measures, like for instance, payments for compensating landowners for forfeiting their rights to deforest, were to be added, additional costs and benefits would have to be integrated into account in analysis, leading most probably to higher marginal abatement costs. Calculating the Marginal Abatement Cost from the Social Viewpoint

Three calculations are required to determine the MAC. The first is the year-overyear incremental cost of the Low-carbon Scenario for livestock in relation to the Refer-

ence Scenario (annual differential between the net results of the two scenarios). Next, the incremental costs for each year are calculated in current 2009 values, using a social discount rate of 8 percent. Finally, the weighted average based on the annual emissions reduction volume (from deforestation) is calculated.

The result of the calculation indicates a marginal negative cost of US$2.5 per tCO2 avoided. This suggests the adoption of more productive systems, versus the existing predominant extensive and degraded pasture systems, should produce economic gains for the beef sector, in addition to mitigating GHGs. While the projected productivity gains in the Low-carbon Scenario would almost certainly have positive economic outcomes, this initial “social viewpoint” analysis could prove misleading for those keen on learning what the real costs would be to get livestock breeders to adopt more productive systems. In reality, the conclusions differ markedly when perceived from a private sector point of view, as shown by the following preliminary results regarding the breakeven carbon price (Section 7.1.3.a.iv). When the costs of forest protection over the 2010-2030 period are included – US$24 billion- the MAC goes up to US$0.48 per CO2 avoided. Calculating the Break-even Carbon Price from the Private Sector Viewpoint

Transitioning from predominantly low productivity systems, specifically feedlots and livestock systems, would require high levels of investments and operations and maintenance disbursements of over US$ 430 billion over the 2010-30 period, or US$ 22 billion per year. Although the Low-carbon Scenario results in an IRR of 11.24 percent, only production systems – especially cattle in feedlots, with an IRR of 7.5 percent – may not be remunerative enough to be implemented on a large scale initially.

Thus, in the case of livestock production, it would be especially important to complement a social economic analysis (ex: social discount rate) with a private sector analysis. The main justification is that while the social point of view doesn’t vary between the reference scenario and the Low-carbon Scenario, the private sector point of view changes dramatically because Brazil’s livestock sector has limited access to bank credit and depends heavily on its own capital resources for investing in livestock-related technologies. The productivity of more traditional livestock systems, whose returns are often only about 0.5 percent or less, is generally insufficient for defraying the costs of bank credit.

Promoting a transition from lower to higher productivity systems could contribute to increasing the rate of return for these businesses. However, adopting high productivity systems presupposes substantially higher investments, which would require access to bank credit. Thus, the rate of return for these endeavors must at least equal the credit costs plus the expected profit to provide livestock breeders an adequate incentive.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

As previously mentioned, the proportion of higher productivity systems is greater 223 in the Low-carbon Scenario than in the Reference Scenario, which results in a positive NPV of the incremental results of R$14.3 billion, versus $18 billion NPV in the Reference Scenario. The overall IRR for the Low-carbon Scenario is 11.24 percent, and the calculation is based on the incremental costs of the implementation and expansion of the higher productivity (more cost intensive) systems and their related returns.

Therefore, IRRs have to be far higher in the Low-carbon Scenario than in the Reference Scenario. 224

The sum total of the expected rate of return, plus financing costs (i.e., long-term interest rate [LTIR] + percentage of spread ~ 10 percent +) is generally higher than the rates of return that certain productive options recommended for the Low-carbon Scenario can achieve (i.e., about 0.56 percent for extensive systems, 7.5 percent for feedlot systems, and 4.5 percent on average for the Low-carbon Scenario).

The social approach does not explain why high productivity systems need substantial incentives, while traditional production systems, which make less profit, tend to expand on their own. What at first glance seems to be a win-win situation – less land needed, and thus less pressure to clear forests and expand the agricultural frontier on one hand, and a better biological and economic performance for the livestock breeder on the other – may not be a very accurate portrayal.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In short, the expected IRRs or private discount rates related to livestock breeding in the Reference Scenario are low (approaching 0.5 percent), while those considered in the Low-carbon Scenario are significantly higher (at least 10-12 percent). If bank loans, which benefit from lower interest rates (ex: Banco da Amazonia [5-8.5 percent] or BNDES [5.75-6.75 percent]) are needed only to finance part of the overall sum required, it may be considered that, under the Low-carbon Scenario, a producer would need to achieve an average IRR of at least 10 percent, which is a rather conservative value. The study used this benchmark IRR to produce an initial estimate of the incentives a Low-carbon Scenario would require to generate substantial productivity gains in the livestock sector resulting in the release of needed pasture land to accommodate growing alternative activities without inducing pressure on forests. It should be emphasized that this study is a first attempt to determine the level of incentives required, but more studies are needed to examine the question in greater depth.

To calculate the break-even carbon price, the only incremental costs considered were those associated with the implementation and expansion of higher productivity systems. Given that the feedlot system has an IRR of 7.5 percent, which is less than the benchmark IRR used in this study (12 percent), the break-even carbon incentive required was calculated to ensure that this system would reach an IRR equal to the benchmark rate. The calculation indicates that this incentive should be about US$ 1.47 per tCO2e, or approximately US$ 9 billion over the 2010-30 period to avoid 6 Gt CO2e and ensure an IRR of 12 percent. When the costs of forest protection during the same period are taken into account– US $24 billion – the incentive to implement the overall strategy to reduce deforestation by about 80 percent of the historical observed rates rises to US$ 6 per tCO2e or US$ 36.5 billion to avoid 6 Gt CO2e (Figure 56). Using a higher IRR of 15 percent, the resulting break-even carbon incentives would be $1.88 and $6.64, including forest protection costs.

Figure 56: Marginal abatement cost (8 percent social discount rate) and break-even carbon price (considering an IRR of 12 percent) for deforestation avoidance measures

Financing Requirements 

To implement the higher productivity, livestock production systems in the lowcarbon scenario, the required financing of investments, operations and maintenance would total R$946 billion (US$430 billion) over the 2010-2030 period, with investments representing approximately 30 percent of total expenditures, or about US$21.5 billion per year (Table 51). A smaller amount would be necessary in the Reference Scenario, since these higher productivity systems are expected to expand in that scenario, albeit at a far more limited scale. Releasing another 70.9 million hectares in the Low-carbon Scenario would require another R$720 billion (US$327 billion) more for financing higher productivity systems. This would represent about US$16 billion in additional annual costs, equivalent to 72 percent of the gross value of beef production in 2008.45 As a point of reference, financing from the Brazilian government for the sector was US$3 billion in 2007, or approximately 10 percent of the estimated annual investment required by the Reference Scenario in 2010 (US$32.5 billion).

Financing requirements would be significantly lower if the Low-carbon Scenario were not to incorporate mitigation and carbon uptake measures that require extra land on top of expansion of agricultural land in the Reference Scenario (legal forest carbon uptake, ethanol for increased national consumption and for export, and production forests for the iron and steel industry). In the Reference Scenario, the additional land 45

The gross value of meat production in 2008 (based on figures for April 2008 by the IGP-DI) was estimated by the Confederação Brasileira de Agricultura e Pecuária (Brazilian Confederation of Agriculture and Livestock - CNA) at R$49,59 billion (see Rural Indicators Indicadores rurais XI (90 [Set.-Out.]):6.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

225

226

for agricultural and livestock production is 16.8 million hectares, less than a third of the total volume of land released under the Low-carbon Scenario (via high productivity systems of livestock production to accommodate both expansion of crops and all measures considered) (Table 52). Without added mitigation and carbon uptake activities, the financing required in the Low-carbon Scenario for improved livestock production to release land for crop expansion would total US$238 billion – US$108 billion more than in the Reference Scenario – and US$262 billion when estimated forest protection costs are added. Table 52: Livestock-sector investments and expenses to release land to absorb additional lands needed in the reference and Low-carbon Scenarios (2010-30)

Cumulative investments in feedlots for cattle and croplivestock (billion R$)

Cumulative expenses for feedlots for cattle and croplivestock (billion R$)

Total investment in feedlots for cattle and croplivestock (billion R$)

Reference

0

92,075

134,351

226,426

16.8*

107,699

356,397

464,095

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Scenario

Cleared Pasture area (millions of hectares)

Reference (Absorption of additional land needed) Low carbon **



*

70.4**

225,322

721,124

Additional lands needed for crop, pasture and forest expansion.

946,446

Absorption of additional land needed for the expansion of crops, pasture and forests in the Reference Scenario, plus land needed for proposed mitigation and carbon uptake in the Low-carbon Scenario.

5.2 Forest Recovery: Legal Forest Reserves

Cost

Financing measures and available lines of credit for the restoration of the deforested areas of native vegetation on rural properties were listed earlier (item 3.3.2). There are obviously additional needs for funding to achieve the legal scenario, but this is not a critical barrier at the moment, as available lines of credit for forest restoration are currently under-utilized due to other obstacles mentioned below, the main one being the loss of productive area on rural properties.

Implications (winners and losers)

In the legal scenario, forest restoration in legally protected areas means the displacement of agricultural crops and livestock activities that currently make up the type

of land use practiced in these areas. Thus, despite the fact that compliance with the legal scenario implies different benefits for the local, regional and global climate, for biodiversity conservation and for the restoration of the quality of environmental services, such as the hydrologic cycle, greater competition for land for agricultural crops is expected, raising the opportunity cost of land and possibly causing an increase in food 227 prices.

Forest restoration costs can be divided into the following components, all of which include a labor cost component:

b. Ground preparation. Includes costs for fertilizers, elimination of weeds and sauba ants and digging of appropriate holes for planting saplings; total cost is estimated at R$ 1,000 to R$5,000 per hectare. c. Planting. Includes costs for saplings and labor; costs are estimated at R$1,200 to R$ 2,300 per hectare. d. Maintenance of restored areas. Includes regular weeding and periodic and application of fertilizer where needed. These costs could account for as much as 50 percent of total costs. Final per-hectare costs would depend on the extent to which the environment has deteriorated and the levels of intervention necessary to re-establish vegetation. Four levels of intervention correspond to four scenarios (Figure 57), as follows: • Minimum: The area to be restored has great potential for natural regeneration, only requiring fencing-off to permit the re-establishment of the plant cover.

• Light: In addition to fencing-off, the area requires planting of tree species used in the forest restoration exercise. • Moderate: The ground is very compacted from years of livestock grazing and is completely colonized by gramineous plants. Required interventions include fencing-off, ground preparation, elimination of weeds and ants, and extensive planting of saplings, machinery could be used to contain costs. • Major: In addition to the conditions described above, the ground is extremely degraded and eroded and thus unsuitable for machinery; owing to ecological degradation, such an area would likely continue in a low-carbon state indefinitely.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

a. Fencing-off. Costs are estimated at R$1,500 to R$ 2,000 per hectare.

Figure 57: Variation in forest restoration costs by intervention scenario

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

228

However, intervention costs may vary considerably, primarily due to the costs of manual labor in rural areas and the purchase of inputs for machinery, whose prices tend to vary, even within the same state. Average amounts that appear in Figure 57 above are based on different estimates for forest restoration and date from articles in specialized literature.

As it is impossible to geo-spatialize forest restoration costs in the legal scenario, abatement and investment costs were simulated, using a moderate intervention scenario. The carbon uptake rate from forest restoration used was the average absorption for the Cerrado and Atlantic Forest biomes, consisting of an uptake of 98.3 tCO2/ha in 2030.

The incremental cost was not calculated in the legal scenario, only the cost of forest restoration, as the legal scenario assumes that no economic activity would occur in such areas. The average marginal cost would therefore be US$ 41.68/tCO2, while the level of incentive expected (break even carbon price) would be US$ 50.52/tCO2.

Considering that the total volume of forest restoration would be 44 million ha, based on the total marginal cost indicated above, the total non-discounted cost would be US$ 1.84 billion for the period considered, or an average of US$ 92 million per year.

Considering that the total volume of forest restoration would be 44 million hectares, the total non-discounted cost, based on the abovementioned marginal cost, would equal US$54 billion over the 2010-2030 period. The average annual cost for financing over the period would be US$2.7 billion.

5.3 Renewable Charcoal

Impacts on land use: necessary area and hypothesis on the correlation with deforestation

Two main aspects were taken into account in this section with regard to the impacts of the low-carbon scenario on land use: (i) availability of agricultural lands and (ii) possible impacts on deforestation practices due to a possible additional need to convert areas with native forests into areas with planted forests.

The amount of land needed for implementing the Low-carbon Scenario was

229

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 58: MAC and equilibrium price of carbon for CO2 uptake through legal forest restoration

230

estimated to be between 3,327 and 3,663 million hectares, depending on the different levels of productivity46. This amount represents approximately 0.35 percent of the national territory. Even when the higher amount is added to the additional need for land for the livestock and agriculture sectors, the entire requirement may be met by occupying areas currently used for pastures at different stages, indicating that the land required for supplying the Brazilian iron and steel industry with charcoal does not necessitate the conversion of areas occupied by native forests in production areas47. Marginal abatement cost

Considering a discount rate of 8 percent per year, the consolidated marginal abatement cost for the use of renewable charcoal instead of coal or non-renewable charcoal was estimated to be approximately US$ 9.00/ tCO248. When a discount rate of 15 percent is applied, the consolidated marginal abatement cost is an average of US$ 27.14 / tCO2 (Table 53).49

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 53: Marginal abatement cost

Mitigation Measure

Additional use of renewable charcoal (instead of non-renewable charcoal or coal).

Discount Rate (per year)

Marginal Abatement Cost US$ / tCO2

8%

8.95

15%

27.141

Source: Adaptation of data presented in the report on industrial sector emissions

As emphasized earlier, this mitigation measure requires investments in the forest sector (new planted forests) as well as in the industrial sector (carbonization and thermo-reduction processes in blast furnaces). However, the main difference between the costs and investments necessary for the implementation of the Low-carbon Scenario has to do with the establishment, maintenance, and harvesting of additional quantities of planted forests.

For the possible substitution of non-renewable charcoal, by definition the marginal abatement cost is at least in proportion with the investment in plantation forests for seven years, since this investment is not made when non-renewable charcoal from deforestation is used. In the industrial sector, additional investments are necessary for the charcoal-making process and for the increase in the number of blast furnaces used to make pig iron with renewable charcoal. This financial input is slightly higher than the investments needed for the expansion of pig iron production using coal50. The main economic results related to this mitigation alternative are presented in Table 54. 46

47 48

49

50

To see details on the calculation methods and the reasons for this variation, please see the land-use section (LULUCF) in the study. Ibid. According to INT estimates, specific calculations for the substitution of non-renewable charcoal and coal are practically the same, respectively: US$8.9 / tCO2 and US$ 9.0 / tCO2. Given the discount rate of 15% per year, INT estimates that the marginal abatement cost for avoiding the use of non-renewable biomass is US19.53 / tCO2e and US$ 34.75 US$ / tCO2 for coal. However, of the two sub-reference scenarios presented for charcoal, it was decided to adopt an average of US$27.14/ tCO2 in this report, given the integrated nature of this mitigation alternative, and that isolated figures are dependent on the level of legal restrictions. See detailed estimates in the report on industrial sector emissions.

Table 54: Summary of economic parameters for the 2010-2030 period

4,245.44

1 21-year cycle

-2,678.28

None

385.07

9.0

Source: Adapted from the report on industrial sector emissions

The total investment necessary for implementing the Low-carbon Scenario is estimated to be approximately US$ 4,245 billion (current value), according to the calculations presented in the report on emissions from the industrial sector (subject of another Summary Report). Considering the adjusted potential, this amount represents approximately 12.74 percent of total investments estimated for the implementation of mitigation measures throughout the Brazilian industrial sector, which are estimated to be US$ 33,331 billion. Table 55: Investments in the additional use of renewable charcoal compared with total mitigation measures in the Brazilian industrial sector, considering the adjusted potential Mitigation Measures

Investment (US$1,000)

% of Investments

Sum of mitigation measures from the entire Brazilian industrial sector

33,330,829

100.00

4,245,440

12.74

Additional use of renewable charcoal instead of coal or non-renewable charcoal

231

Source: Adaptation of the data presented in the report on industrial sector emissions

Investments necessary to ensure the viability of the additional use of renewable charcoal for iron and steel production represent over 60 percent of total investments in biomass considered in the different mitigation measures anticipated in this study, which represent approximately 20 percent of all the investments projected, as per Table 56.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Use of renewable charcoal instead of coal or nonrenewable charcoal

Total Investment (VP) (US$ million)

Total Abatement Internal Avoided Cost (AdjustNumber Net Income Rate of Emised Potential) of Years of (million Return sions (US$/tCO2) Investment US$) (%) (million (Rate 8% / tons CO2) year)

Figure 59: Percentages of investment distribution by group of measures

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

232

Source: Industrial sector emissions report

The table below shows the main hypotheses used for the technical-economic analysis, developed based on the industrial sector emissions report (the subject of another Summary Report from this study). Considering that steel and iron production using renewable charcoal is responsable for 90.2 percent51 of the emissions reductions that may be attributed to biomass substitution, the table below was adapted to the amounts presented by the report on emissions from the industrial sector, reflecting this proportion. Adding the amounts from investments and costs referring to biomass substitution and the elimination of non-renewable biomass, the following hypotheses were put forth: Table 56: Hypotheses of the technical-economic analysis

Baseline (Amount Present in 106 US$)

Mitgation Measures

Investment

Cost Energy / O&M

Use of renewable charcoal instead of coal or nonrenewable charcoal

4,652

341,333

51

Mitigation Options (Amount Present in 106 US$)

Income Investment

0

8,897

Cost of Energy

338,630

Cost of O&M Income

1,135

Source: Adaptation of data presented in the report on industrial sector emissions

COPPE, 2009

0

5.4 Emissions Abatement with Zero Tillage The sociological and environmental significance of the zero tillage system within the context of the national agriculture policy seems not to be fully recognized by the government. The country’s current policy does not penalize farmers who harm the 233 environment, such as in conventional planting where tons of soil are lost due to erosion, silting up rivers and lakes, or even causing soil degradation due to the loss of organic material. From a cultural point of view, the farmer is comfortable not changing his production practices, as he doesn’t have to learn anything new, take risks or make investments.

a. Create incentives for basic research and technology to continuously generate information that ensures the sustainability of zero tillage in different parts of the country.

b. Restructure the rural extension system, training technicians so that they may serve as a link between research institutions, universities and different segments of the productive sector. c. Establish priority credit, facilitated and differentiated for producers who adopt the zero tillage system; ex. raise the limit for agricultural credit, with lower interest rates, geared towards producers who practice zero tillage; provide agricultural insurance, with the possibility of reducing premiums depending how long it takes to adopt the system. d. Expand storage facilities and guarantee the purchase of relevant products for zero tillage, such as corn and rice. e. Develop financial instruments that “hedge” the prices of essential inputs for zero tillage (e.g. herbicides).

Cost calculation

Variables required for estimating the marginal cost of emissions abatement through zero tillage were developed as follows:

Discount rate. To discount cash flows for agricultural activities, a yield curve is used rather than single discount rate. The idea behind this is that the capital used is made up

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Obstacles to the expansion of zero tillage in the country must be surmounted and the following public policies are recommended to achieve this:

Exchange rate. Over the past few years, the exchange rate between the Brazilian real and the US dollar has been extremely volatile, which made it difficult to predict its development. It was thus decided to use market estimates for the dollar for the end of the period, based on the Central Bank’s Focus research, for the next few years until 2013, the last year available. On that basis, the hypothesis was that the dollar would adjust to inflation, which is estimated to be 4.5 percent per year until 2035.

Cost of land. The cost of land was estimated based on land prices according to the Agricultural Economics Institute of São Paulo (Instituto de Economia Agrícola de São Paulo [IEA]). Average land prices in the municipalities of São Paulo are available in reais per hectare and were collated in June every year from 1995 until 2008. They were updated by the IGP-DI to the prices of March 2009. The choice of the São Paulo data base distorts the calculation somewhat, given that these are the most expensive lands in the country. They have the best infrastructure and proximity both to the consumer market and the port of Santos, Brazil’s largest port, and suffer from pressure from real estate speculation. However, this is still the best option as it is the most reliable data source in Brazil. Its performance can be observed in Figure 60.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

234

of a combination of different “zero coupon bonds” with annual maturation from 2008 until 2035. A yield curve was constructed by adding risk-free rates for every year, provided by the NTNB (Notas do Tesouro Nacional - National Treasury Notes) with a corresponding due date, like the risk premium, which varies from 5 percent to 4 percent during that time period, multiplied by the β of 1.5 usually used in agribusiness projects in Brazil. The risk premium for the Brazilian economy and the β from the crop-livestock sector were obtained through interviews with specialists from the Banco Nacional do Desenvolvimento Econômico e Social (Brazilian Development Bank).

Figure 60: Cost of land in the state of São Paulo between 1995 and 2008

It should be noted that even the best data available do not enable the construction of a precise model on the evolution of land prices until 2035, as the series is not long enough. It was thus decided to use the historic method of mean values in real terms to estimate future land prices. Therefore, the amount of R$12,593 per hectare was used for the year 2008, which is the average for the last available prices registered. The amount of R$10,562 was used for the year 2009, which means the average price was R$10,017, with a 4.5 percent increase due to inflation. From that point on, the price of land was adjusted to inflation which is estimated to be 4.5 percent per year. Price of the main commodities. The calculation of the representative price in the study for commodities (soybean, corn, rice, beans, cotton) must be useful for calculating the total “marginal” cost of carbon captured with the conversion from traditional planting to zero tillage. Thus, the increase in the physical quantities of total agricultural production should result in the value of agricultural production. The most obvious way to obtain this here is to calculate average prices for the products considered for each year for the participation of each of these products in the total amount produced.

For this calculation, estimates of quantities produced were based on studies done by ICONE. The prices used and the estimates for the period from 2008 to 2035 were extracted from the data base of prices paid to producers from São Paulo State, which is maintained by the Instituto de Economia Agrícola de São Paulo (Agricultural Economics Institute of São Paulo).

The first step in estimating prices that are as yet unknown was to consider the respective historical series, starting in January 1980. The reason for dispensing with the years prior to 1980 was due to the fact that there was a structural break in the series as a result of the intensification of globalization and the international commodities trade.

Limited to the beginning of the series, the next step was the visual inspection of the five graphs below, which show the prices paid to farmers from the state of São Paulo, adjusted by the IGP-DI/FGV in reais of March 2009, per ton of product.

Figure 61 shows that the 1980s were more volatile than subsequent years. This volatility, which was also observed in international markets, was exacerbated in Brazil by economic instability. As if to illustrate this point, the country witnessed five monetary reforms and great volatility in the exchange rate from 1980 until 1994. Starting in the 1990s, commodities prices showed greater stability throughout the world and in Brazil as well, which seems to coincide with a new agricultural price pattern. However, such stability may be threatened by an imminent structural break resulting from a combination of three factors: the increase of food consumption in emerging market countries such as China, India, South Africa, Russia and Brazil; the emergence of biofuels; and the effects of the global economic crisis. These factors, which are responsible for the increased volatility observed over the last two years, have generated some skepticism as to the validity of balancing supply and demand for fore-

235

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

As can be observed in the above figure, after declining from 1995 to 2000, land prices (in real terms) rose systematically the following years, possibly reflecting the commodities boom and the increase in the liquidity of financial markets. Indications are that land values should drop steadily compared to 2008 prices for an indeterminate period of time, as a reaction to the financial crisis.

casting prices on a distant horizon. 236

With a view to establishing a reasonable forecast for the prices of products, but faced with the aforementioned obstacles, it was decided to use historic price averages, in real terms, for the five commodities, calculated from July 1994 onward. This choice is justified by its simplicity, as there is no guarantee of accuracy using more sophisticated models. The definition of the period of analysis aims to eliminate the effect of hyperinflation by limiting the beginning of the series to the moment the Brazilian real was introduced, when the country’s economy began to show more stability.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Figure 61: Variations in crop prices used in the present study

Outputs for O&M, principal inputs and investments.The data base used to calculate outputs is the estimate of costs per hectare of production, developed by CONAB/MAPA (Cia Nacional de Abastecimento/Ministério da Agricultura, Pecuária e Abastecimento – National Supply Company/Ministry of Agriculture, Livestock and Supply). The agricultural costs considered in the study are O&M costs (operations and management) for the most important inputs and investment expenditures. As the expenditures of CONAB are not classified according to this nomenclature, the first step was to develop this classification.

For calculating costs for O&M and for the main inputs, all available estimates were selected for the five crops for the 2008/2009 harvest. Estimates of the cost of zero tillage for the other crops were then separated, to be combined under the conventional planting label. The total of the estimates for the two groups, aiming at a cost estimate per hectare for zero tillage and for conventional planting, followed the logic of the calculation of average costs for the quantity produced for each commodity compared to the total quantity of commodities produced in each type of planting.

For calculating cash flow from agricultural operations, the rigidity hypothesis for the participation of O&M expenses and inputs per crop was used. Thus, based on estimates of quantities produced by each product and planting technique, the expenditures were adjusted by the change in composition of the physical quantities produced for all goods in the total production of each product and adjusted according to the inflation estimate for each year (4.5 percent per year). Outputs for investments made for land were not considered, only for improvements and equipment. Considering that CONAB estimates remuneration for improvements and equipment, or fixed costs, like 6 percent52 of half the price of new equipment, the calculation of expenditures for necessary improvements and equipment for production is simply the division of the expected remuneration resulting from the fixed capital, divided by 6 percent and multiplied by 2. For calculating expenditures for investments per hectare, it was still necessary to consider expenditures for the investment in each crop for its relative participation every year.

Investments can obviously not be made every year with the cash flow, as a substantial part of the investment is made just prior to the productive activity. Given the lack of information on the investment performance for the commodities considered, it was decided to launch the costs for investment the first year of the series. However, as the period considered in the study is extensive, and taking into account that agricultural equipment has a 10-year period of depreciation, it was decided to make new invest52

6 percent is the yield considered for the alternative use of capital.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The vector of prices of each of the commodities every year for the period considered (2008 until 2035) is composed of historical averages (1994 until 2008) adjusted by the estimated accumulated inflation. Considering that inflation is largely the effect of fiscal policy, and that there is no indication of inflation goals beyond the year 2012, the current inflation target until 2012 (4.4 percent per year) was used hypothetically to estimate future annual inflation until 2035. With these considerations, the price of the 237 physical unit (ton) that is representative for the five commodities was calculated as the average for the agricultural prices considered for the participation of each commodity in the total physical production of the five commodities.

ments every 10 years. 238

For the sake of simplification, the equipment’s residual value was not considered, given the fact that each type of equipment has a differentiated residual value and most are negligeable. Another area of simplification adopted for lack of information is that there will be no distinction made between fixed capital, improvement and equipment. Improvements depreciate over 25 years, besides leaving a residual of 25 percent of the initial value. Cost

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Costs for implementing the zero tillage system for soybean, corn, rice, beans, and cotton were obtained based on data from CONAB/MAPA (Cia. Nacional de Abastecimento/Ministério da Agricultura, Pecuária e Abastecimento) for the 2008/2009 harvest. Estimates of Brazilian agricultural production costs per hectare are given for different parts of the country. The agricultural costs considered were for O&M (operational and management costs), for the most important inputs (herbicides, fertilizers, etc.) and for investments (machinery and equipment). Expenditures for land acquisition were not considered. Costs of the conventional planting system were thus obtained for the development of the Reference Scenario. Estimates for the different cost items for each planting system were obtained by calculating the average costs for each commodity, considering the quantity produced by each one, thereby simulating what would be a “joint commodity”. Cost proportionality between the items did not vary throughout the year for each commodity.

The total amount that was not discounted from investments from 2010 to 2030, expressed in 2009 reais for the Low-carbon Scenario (100 percent of the planted area under zero tillage) is R$335.6 billion, or 70 percent of the total required for the Reference Scenario (R$473.9 billion) for the five aforementioned crops. It should be considered that this difference is valid for the aggregate commodity, but can be greater or less for each of the five crops in the different regions that contribute to the CONAB data base (Table 57). Due to the lower investments, operational costs and inputs (8 percent less), required by the Low-carbon Scenario, the marginal abatement cost for emissions is negative (-R$ 0.72/ton C), indicating that the alternative considered in this scenario, zero tillage, is economically superior (Table 58). Therefore, for the same market condition, the IRR for a scenario for 100 percent of the area under zero tillage is always greater than the IRR obtained for Reference Scenario conditions.

These results confirm earlier economic evaluations for planting systems in Brazil, according to which the use of zero tillage is always more advantageous, and its cost is 6 percent lower on average. It goes against common sense that environmentally sustainable techniques tend to be more costly than those of the market and require additional incentives in order to be adopted.

Table 57: Discrimination of costs considered in the study

DISCRIMINATION

I – ON-FARM COSTS 1 – Operation with planes 2 – Operation with machines 3 – Rental of machines/services 4 – Temporary labor 5 – Permanent labor 6 – Seeds 7 – Fertilizer 8 – Pesticides 9 – Other expenses Total On-Farm Expenses (A) II – POST-HARVEST EXPENSES 1 – Production insurance 2 – Technical assistance 3 – Outside transport 4 – Improvements 5 – Storage Total of Post-Harvest Expenses (B) III – FINANCIAL EXPENSES 1 – Interest Total Financial Expenses (C) VARIABLE COST (A+B+C = D) IV – DEPRECIATION 1 – Depreciation of improvements/installations 2 – Depreciation of tools 3 – Depreciation of machines Total Depreciation (E) V – OTHER FIXED COSTS 1 – Periodic maintenance of machines/tools 2 – Social duties 3 – Fixed-rate capital security Total of Other Fixed costs (F) Fixed Cost (E+F = G) OPERATIONAL COST (D+G = H) VI – INCOME FACTORS 1 – Remuneration expected on fixed capital 2 – Land Total Income Factors (I) TOTAL COST (H+I = J) Developed by: CONAB/DIGEM/SUINF/GECUP

Classification of Expense O&M O&M O&M O&M O&M Inputs Inputs Inputs O&M

Capital O&M O&M O&M O&M O&M Capital Capital Capital O&M O&M O&M

239

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Average Productivity:

240

However, the adoption of this planting technique in Brazil has somewhat stagnated and even decreased, possibly due to the perceived risk in changing productive systems, and the limited knowledge on the system’s correct use, among other obstacles that were already discussed. In light of all this, an economic incentive program should be considered a strategic motivating factor for farmers to overcome these obstacles. Table 58: Emissions reduction potential in tons of CO2eq, average abatement cost during the period and price to be paid per ton of C to compensate the implementation of zero tillage Net Reduction Options for the Mitigation or Up- Potential between 2010-30 (tCO2e) take of carbon

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Zero tillage

355,415,105

Average abatement cost during the period (US$/tCO2) Discount Rate (8%) - 0.72

Break-Even Carbon Price (US$/tCO2) -0.20

6 Conclusion

Brazil has gained considerable experience in forest-protection policies and projects and finding ways to generate economic activities that are compatible with the sustainability of native forests. Forest-protection projects and policies are used as barriers to counter the progression of pioneer fronts. However, a more drastic reduction in forest destruction needs more than just protection. Shifting to a Low-carbon Scenario would require acting on the primary cause of deforestation: the demand for more land for agriculture and livestock. Therefore, this study proposes a strategy that acts on two complementary fronts: (i) eliminating the structural causes of deforestation and (ii) protecting the forest from attempts to cut it down. Implementing the first part would involve working with stakeholders who use already deforested land, while the second presupposes working with those with a vested interest in new deforestation efforts. With regard to the first issue, eliminating the demand for more land would require accommodating the expansion of agriculture and the meat industry—both of which are important to the Brazilian economy—on already deforested land. This would mean a drastic increase in productivity per hectare. Technically, one option available is to increase livestock productivity, thereby giving up large quantities of pasture. This option is technically possible since current average livestock productivity is low and would entail the scaling up of already existing productive systems in Brazil (i.e., feedlots and crop-livestock systems). The potential for releasing and recovering degraded pasture is considerable and is enough to accommodate the most ambitious growth scenario. Moreover, moving from lower- to higher-productivity production systems can trigger a net gain for the sector economy since more intensive processes converge with higher economic returns (Chapter 7). But this option also presupposes four challenging points.

First, productive livestock systems are far more capital intensive, both at the investment stage and in terms of working capital. Having farmers shift to these systems would require the offer of a large volume of attractive financing far beyond current lending levels. Commercial interest rates are usually too high to make such investments attractive. Moreover, banks are often unwilling to lend to farmers, whom they 53

See “Avaliando o Risco de Colapso da Floresta Amazônica: Uma Avaliação do Banco Mundial”, by José A. Marengo, Carlos A. Nobre, Walter Vergara, Sebastien Scholtz, Alejandro Deeb, Peter Cox, Wolfgang Lucht, Hiroki Kondo, Lincoln Alves, and Jose Pesquero.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In moving toward a national Low-carbon Scenario, Brazil’s main challenge is undoubtedly to reduce deforestation. Despite the government’s recent success in implementing aggressive forest protection policies, deforestation is expected to continue 241 being the country’s largest source of GHG emissions well into the future. Moreover, several recent studies have shown than deforestation means far more than just GHG emissions. Deforestation from fires also emits aerosols that affect rainfall and temperature regimes (see Section 4.9) and a recent World Bank assessment on the collapse of the Amazon Rainforest (known as “Amazon Dieback” in English),53 show that there is a clear interaction between deforestation and the expected damage to the forest from global climate change, with its most severe progression following the same spatial pattern as deforestation. For the sake of reducing GHG emissions and halting the accelerated dieback of the Amazon forest, fires should be eliminated from the Amazon region.

242

perceive as high-risk borrowers. Thus, a large volume of financial incentives, along with more flexible lending criteria, would be needed to make such financing viable for both farmers and the banking system. Over the past five years, the Brazilian government has developed programs to stimulate the adoption of more productive systems (e.g., PROLAPEC and PRODUSA) in order to reduce business risk, increase income in the field, and renovate degraded pasture areas. A first attempt to estimate the volume of incentives required indicates an order of magnitude of US$21.5 billion per year. Second, these systems require higher qualifications than traditional extensive farming, in which farmers move on to new areas as soon as the pasture is too degraded to be productive, eventually converting more native vegetation into pasture. Therefore, financing should be accompanied by the intensive development of extension services. Public policies that promote rural extension and training for cattle ranchers would be important to overcome this barrier.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Third, a rebound effect should be avoided. In other words, higher profitability with less land needed to produce the same volume of meat might trigger an incentive to convert more native forests into pasture. Such a risk is especially high in areas where new roads have been opened or paved. Therefore, any incentive provided should be geographically selective: It should be given only when it has been clearly established, based on a valid and geo-referenced property title, that the project will not include the conversion of native vegetation nor areas converted in recent years (e.g., less than 5 years), legally or not. This study confirmed that such a stipulation would be technically possible, since enough pasture can be vacated nationally even without increasing the productivity of livestock in the Amazon region. Therefore, any subsidized financing for livestock production in the Amazon region should be made on an extremely selective and stringent basis, and the area in question should be closely monitored.

Fourth, a number of attractive options considered in the Low-carbon Scenario to mitigate emissions or increase carbon uptake emphasize the need to liberate a considerable amount of pasture. For example, full compliance with the Legal Reserve Law would result in the replanting of over 44 million ha currently allocated for other activities. While replanting the forest would remove a large amount of carbon dioxide (CO2) from the atmosphere, this area—more than twice the expected expansion of agricultural and pasture land under the Reference Scenario—would no longer be available for such activities. To avoid a “deforestation leakage” the freeing up of the equivalent additional amount of pasture would be required; otherwise, part of the production would have to be reduced to prevent the conversion of more native vegetation in another location. The same rationale applies to the expansion of any other activity that requires land (e.g., bioenergy activities involving ethanol or renewable charcoal), though on a far smaller scale. Under the Low-carbon Scenario, further expansion of all these activities taken together would require less than one-fourth of the additional land required for legal forest reserves. Thus, there is a difficult trade-off between (i) more efforts to increase livestock productivity in order to free up more land and (ii) full enforcement of the recovery of legal reserves and crop expansion. Less compliance with the current legal obligation regarding forest reserves would make it easier to achieve the goal of accommodating all activities without any need for deforestation, but it would mean less carbon uptake. The reverse is also true.

Protecting forested areas where deforestation is illegal could be achieved via an array of activities, ranging from repressive police action to sustainable use projects. In recent years, the Brazilian government has made considerable efforts in this area, particularly under the Action Plan for the Prevention and Control of Deforestation in the Legal Amazon (PPCDAM). Protection measures may include activities similar to those already put into practice under the PPCDAM, such as (i) expansion and consolidation of protected areas, (ii) development of integrated projects, and (iii) promotion of the sustainable use of forest resources. Such efforts will need to be maintained and probably amplified. If the proposed strategy is fully implemented—that is, if the demand for additional land is phased out and the forest is protected against the remaining causes of deforestation—then the contribution of Brazil’s LULUCF sector activities could be inverted from high net GHG emissions to a net GHG uptake of about 195 Mt CO2 per year by 2030.

243

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

To protect the forest against the remaining causes of deforestation, it is proposed that forested areas where deforestation is illegal be protected against fraudulent interests. It should be noted that there may be a sizeable gap between the time that demand for land decreases and the time it takes before the behavioral change of proponents of deforestation at the frontier, whether legal or illegal, is observed.

7Annex: Analysis of Low-Carbon Scenarios 244

To construct a land-use scenario that allows a Low-carbon Scenario, the study generated successive intermediary scenarios that incorporate the impacts of the different mitigation and uptake options considered. Four individual scenarios in particular were analyzed, plus one that combined the options considered in the first four scenarios: Herd optimization scenario;

Scenario to increase forest production plus cattle;

High ethanol export scenario plus cattle and production forests;

Legal scenario (or recomposition of the Legal Reserve) plus cattle; Scenario where the last four scenarios occur simultaneously.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

In order to focus on the need to convert native vegetation, all of the successive scenarios were developed so that the total area occupied by agriculture and livestock activities did not increase after the beginning of the period. The largest total area occupied by agriculture and livestock observed between 2006 and 2008 was chosen for each region as the limit for the expansion of agrosilvipastoral activities until 2030. Only the area in the Northern Amazon region was larger in 2008, while more area was occupied by agriculture and livestock in other regions in 2006. Improvement of zootechnical indexes and pasture intensification will be key variables for ensuring that the greater need for land for sugar cane and production forests, and the reduction in the productive area for the legal scenario, do not result in additional deforestation, avoiding the trickle-down effect on the agricultural frontier. Table 59 shows the four Lowcarbon Scenarios, which will form the basis for the final scenario that combines the four previous ones.

Table 59: Relationship of the Low-carbon Scenarios developed for this study Scenarios 1. Herd optimization

Action taken to avoid the domino effect

Improvements in zootechnical indicators

245

3. Herd optimization with an increase in production forests, greater ethanol exports and adoption of 2nd Pasture intensificaLow-carbon Scegeneration ethanol tion nario: mitigation Mixture of 20 percent ethanol in gasoline with Brazil measures supplying 15 percent of this market 4. Herd optimization with legal scenario (forest restoration) Pasture intensificaRestoration of environmental liability of legal forests tion calculated as 44.34 million hectares

5. Combined effect of all these measures Source: ICONE

Pasture intensification

In the first Low-carbon Scenario, called the herd optimization scenario, increases in the productivity of the beef cattle herd, with improved zootechnical indices (higher birth rate and lower age at time of slaughter), were considered. This scenario has the greatest impact on pasture area, as the model considers the herd a directly proportionate variable in determining the size of the pasture area. Thus, it is hoped that the pasture area will undergo a more rapid intensification process than what was observed in the past and in the Reference Scenario. This is essential for accommodating the greater need for land in the other scenarios.

The second scenario considered the greater need for production forests, as well as a smaller herd. The principal behind the production forest scenario is to increase the demand for charcoal for iron and steel production as a substitute for coal and charcoal from native forests. Given the great demand for energy for producing pig iron, which is the main raw material used for iron and steel production, the greater contribution of charcoal from production forests would make a tremendous impact on land use. A demand for about a million hectares of production forests was considered for pig iron production in the Reference Scenario. In the second Low-carbon Scenario, this demand will increase to 3.6 million hectares, which represents an additional capture of approxi-

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

2. Herd optimization with an increase in production forests Pasture intensificaElimination of non-renewable charcoal by 2017 and tion participation of 46 percent of renewable charcoal in iron and steel production

mately 500 million tons of CO2.

In the third scenario, besides the production forest scenario described above, it is assumed that ethanol will replace 10 percent of the gasoline consumed worldwide by 2030 and that Brazilian exports will represent 15 percent of global ethanol consumption. Such assumptions are essential for the expectations of the main consumer countries with regard to gasoline consumption, mandates for ethanol use, productive capacity and commercial regimes (Walter et al., 2008). Brazilian ethanol exports, which were at 3.5 billion liters in 2006, are expected to reach 19, 37 and 84 billion liters in 2015, 2020 and 2030, respectively. In the ethanol scenario, internal ethanol consumption does not change in relation to the Reference Scenario.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

246

The third scenario, greater ethanol exports and production forests, consists of a set of four exogenous changes in relation to the Reference Scenario. Besides the improvement of zootechnical indices, some of the aspects considered are: greater ethanol exports from Brazil, progressive adoption of the technology for second generation ethanol production, and a greater area allocated for production forests. Thus, like in the other Low-carbon Scenarios, the total area for agriculture and livestock does not change.

The adoption of second generation ethanol technology has a direct impact on the land-use model, as the use of cellulosic matter (principally sugar cane bagasse) for ethanol production reduces the demand for sugar cane for the same amount of ethanol, thus reducing the need for land. Second generation ethanol is thus gradually being adopted. In 2010, it is responsible for only 0.4 percent (0.13 billion liters) of all national production but this percentage increases progressively to 2.5 percent (1.3 billion liters) in 2015 and 6.1 percent (4.5 billion liters) in 2020, reaching 13.3 percent (17.3 billion liters) in 2030.

The fourth scenario considers the gradual reforestation of the Legal Reserve (LR) until its complete restoration in 2030. There are a number of obstacles to calculating the liability of the Legal Reserve (LR) in Brazil, especially because it should be done at the rural establishment level. In addition, the restoration of Permanent Preservation Areas (PPA) is required in order to be considered a legal scenario. Despite these obstacles, since the beginning of the study, there has been some expectation that the Lowcarbon Scenario, in the case of LULUCF, would have to be based on conditions that are very close to a legality scenario. It was then agreed to call the restoration scenario of legal reserve the“legality” one. Although the team concluded that the exact calculation would be discarded, it was decided to make an approximate calculation based on data prepared by the UFMG.

A simplified method was developed to calculate the amount of area necessary to be reforested in order to be considered a Legal Reserve, given the limited data available. The area legally defined as an LR depends on the area of each rural property and biome. Since there are no data on the size of the properties, the municipality was used as an approximation. Thus, the percentage of LR was calculated based on the size of the municipality, excluding areas referred to in the UFMG maps as Conservation Units (CU), Indigenous Lands (IL), main watercourses and urban areas. Percentages defined by the Forest Code were used: 80 percent in the Amazon biome, 35 percent in the Cerrado inside the Legal Amazon and 20 percent in the other biomes and regions.

After estimating the area to be used as an LR, the area with existing native vegetation, between secondary vegetation, savanna and forests, was eliminated. The result is the area to be reforested to fulfill the legal qualifications as an LR (Table 60). The study considered that the areas that need to be regularized will be reforested gradually, year by year. Thus, as of 2009, 1/22 of the total area to be reforested would be deducted from 247 the area available for agricultural production, until it achieves full legality in 2030. Table 60: Area necessary for the reforestation of the Legal Reserve by state in Brazil (hectares) Area for Reforestation

Mato Grosso do -3,398,792 Sul

Acre

Distrito Federal

Pará

Mato Grosso

-9,465,888

Maranhão

-40,959

Goiás Piauí

-2,611,730 0

0

-27,167

Sergipe

-118,800

Pernambuco Alagoas Bahia

Rondônia

Total Brazil

Roraima Amapá Paraná

Santa Catarina

-58,239

Rio Grande do Sul

-91,861

-242,079

Amazonas

Tocantins

Rio Grande do -3,062 Norte Paraíba

UF

Minas Gerais

-4,794,589

44,344,389

Espírito Santo Rio de Janeiro São Paulo

Source: UFMG. Developed by: ICONE

Area for Reforestation

-721,161 -34,848

-46,757

-11,369,199 0

-1,644,537

-1,711,257 -398,679

-1,184,241

-2,682,095 -205,436

-178,087

-3,314,927

The last scenario combines the four previous ones. It thus includes a herd with improved zootechnical indices, greater demand for ethanol and production forests and the recuperation of environmental liability by reforestation.

7. 1 Herd Optimization Scenario

The first Low-carbon Scenario was developed in partnership with EMBRAPA Cerrados. In this scenario, the herd increases from 206 to 208 million head between 2006 and 2030. Thus, there is a larger gain in the herd’s birth rate compared to the Reference Scenario, going from 0.77 to 0.82 bullocks for each female between 2006 and 2030.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

UF

248

This represents a 0.35 percent increase per year between 2009 and 2030. Moreover, the reproduction rate increased in relation to the Reference Scenario, with an increase of 0.80 percent per year between 2009 and 2030, and from 23 percent to 27 percent for the total cattle herd between 2006 and 2030. Despite a smaller beef cattle herd, beef production was similar to that of the Reference Scenario, which was necessary for meeting the demand for meat. Between 2006 and 2018, beef production will go from 9.9 to 11.2 million tons, and up to 13.2 in 2030 (Table 61). Table 61: Balance of supply and demand for selected products, herd optimization scenario Products

Cotton

Rice

Bean

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Corn

Soybean

Units

Thousand tons

Thousand tons

Thousand tons

Thousand tons

Thousand tons

Soybean meal Thousand tons

Soybean oil

Sugar

Ethanol

Beef

Milk

Chicken

Eggs

Pork

Thousand tons

Thousand tons

Million litres

Thousand tons

Thousand tons

Thousand tons

Million units

Thousand tons

2006

3,659

14,344

3,625

45,362

57,559

23,684

5,984

29,767

18,781

9,928

26,153

9,354

23,575

2,864

Source: ICONE

2008

5,107

12,800

3,936

61,598

63,524

25,655

6,529

34,349

28,482

9,699

28,716

10,880

23,039

3,102

2018

7,133

15,529

4,424

73,663

83,230

30,708

7,489

44,061

51,843

11,222

38,807

12,670

25,725

4,382

2030

9,120

20,611

5,432

89,351

105,444

46,097

11,425

55,852

75,533

13,163

54,071

15,737

29,312

5,606

The most important result in this scenario has to do with pasture areas. With the herd growing less in size, and the hypothesis that the total area will not increase from 2009 onward, the size of the pasture area declined significantly during the period of analysis. Between 2006 and 2018, the pasture area should decrease 10.7 million hectares and 18.8 million hectares by 2030, arriving at 190 million hectares. This implies a gain in productivity in terms of number of animals per hectare, which will increase from 0.99 to 1.09 for the entire period, representing an increment of 0.48 percent per year (Table 62).

Table 62: Land use in Brazil, herd optimization scenario (1000 ha)

844.20 3,017.83 2,694.21 1,529.39 9,632.09 3,331.81 22,748.97 6,179.26 5,269.29

208,888.89

2008

1,066.37 2,880.70 2,856.81 1,143.11 9,656.20 5,052.38 21,334.28 8,234.90 5,886.76

205,380.63

Source: ICONE

2018

1,350.66 2,910.66 2,390.40 1,280.97 9,693.51 5,373.23 25,976.84 10,579.43 7,740.00

198,217.33

2030

1,453.43 3,228.70 2,389.68 1,327.74 10,412.82 5,638.37 30,520.04 12,631.20 8,450.00

190,097.26

Although there was a significant reduction in pasture area in all regions, it was more pronounced in the Northern Amazon, comparing the reference and herd optimization scenarios. In the latter, pasture areas in the region decrease 2.7 million hectares between 2006 and 2030, contrary to the Reference Scenario, where an increase of 12 million hectares for the same period was observed (Table 63) mainly due to the reduction in the size of the cattle herd by 18.2 million head in this scenario compared to the Reference Scenario. This was the region that presented the greatest impact on the reduction in the herd compared to the Reference Scenario, from 68 million down to 49.9 million head as shown in Table 64. Table 63: Regional allocation of pasture areas, reference and herd optimization scenarios (1000 ha)

2006

2008

Reference Scenario 2030

Herd Optimization Scenario 2030

Brazil

208,889

205,381

207,060

190,097

Central-West Cerrado

51,200

50,636

48,395

47,338

South

Southeast

Northern Amazon Northeast Coast

MAPITO and Bahia

18,146

44,053

52,551

10,801

32,138

17,603

41,865

53,728

10,487

31,061

Souce: ICONE

13,264

39,565

64,624

10,812

30,399

12,606

39,678

51,879

10,196

28,401

249

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Cotton Rice Bean – 1st harvest Bean – 2nd harvest Corn – 1st harvest Winter corn Soybean Sugar Cane Production Forests Pastures

2006

Table 64: Regional distribution of cattle herd, Reference Scenario and herd optimization scenario (1000 head)

250

2006

2008

Brazil

205,886

201,410

Central-West Cerrado

56,445

55,506

South

Southeast

Northern Amazon Northeast Coast

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

MAPITO and Bahia

27,200

39,209

47,391 8,665

26,977

26,607

37,525

47,149 8,156

26,468

Source: ICONE

Reference Scenario

Herd Optimization Scenario

27,342

25,673

2030

234,460 36,266

63,238

68,064 8,958

30,592

2030

208,025 37,548

58,086

49,901 8,372

28,446

As there were no significant changes in the area allocated for crops, there will also not be any marked change in the distribution of the cattle herd between the six regions of the model. Due to the reduction in average age at time of slaughter, and the need for feed supplements for the animals, an increase in corn production of 517 thousand tons was observed in 2006, reaching 5 million tons in 2030. For the Reference Scenario, this implies an additional demand for an area of 120 thousand hectares in 2030 for first harvest corn, and an increase of 31 thousand hectares in area for second harvest corn during the same period. This exogenous increment in demand was due to an increase in the price of corn, resulting in a reduction in net exports of 152 thousand tons, as well as a reduction in demand for other uses. Chicken and pork production decreased 319 and 61 thousand tons in 2030, respectively, in relation to the Reference Scenario. Thus, with all factors combined, total corn production increased 808 thousand tons compared to the Reference Scenario.

7.2 Production Forest Scenario

As stated earlier, the central hypothesis of this scenario is the increase in the demand for charcoal from production forests as a substitute for charcoal from native forests and coal for the production of pig iron. Moreover, its starting point is its initial herd of 208 million head, resulting from the herd optimization scenario.

During the period analyzed (from 2006 to 2030) the area to be used for production forests will increase 112 percent, going from 5.3 to 11.2 million hectares. In this Lowcarbon Scenario, the area allocated for production forests in 2030 will be 2.8 million hectares larger than in the Reference Scenario, which this year is an area of 8.4 million hectares. This difference of 2.8 million hectares between the two scenarios was accommodated to a great extent in the pasture areas, which went from 209 to 188 million

hectares during the period analyzed, due to increases in productivity. The reduction of 21 million hectares of pasture accommodated, in addition to the expansion of production forests, the increase in the different crops analyzed, which maintained their occupation of the area and their production, compared to the Reference Scenario (Table 65).

251

Table 65: Regional distribution of the production forest in the reference and production forest scenarios (thousand hectares)

2006

2008

2030

Brazil

5,269

5,887

8,450

Central-West Cerrado

319

385

910

South

Southeast

Northern Amazon Northeast Coast

Mapito and Bahia

1,670 2,452 140 -

688

1,886 2,690 154 9

762

Source: ICONE

2,831 2,707 327

310

1,365

Reference Scenario Production Forests 2030

11,174 2,885 4,968 992 491

310

1,528

With regard to the distribution of areas allocated for production forests between the regions of the model in 2030, an increase in the participation of the southeast region from 33 to 44 percent of the area in Brazil was observed, comparing this Low-carbon Scenario with the Reference Scenario. This can be explained principally by the high concentration of the iron and steel industry in this region, and resulted in a decrease in the participation of the southern region, dropping from 31 percent down to 25 percent.

7.3 Ethanol Scenario and Production Forests In this scenario, with its high volumes of ethanol exports, 8 billion liters of fuel were exported, 6.5 times more than the amount observed in the Reference Scenario. Thus, the total demand consisting of exports, domestic demand and final ethanol stocks will reach 147 billion liters in 2030. Such an increase in the demand for ethanol means that the need for land for sugar cane (all uses) will exceed 19 million hectares in 2030 throughout the country, or 6.5 million hectares more than in the Reference Scenario (Table 66). An analysis of the aggregate values for Brazil show that the impact of sugar cane expansion will not significantly reduce the area occupied by other crops, as it is accom-

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Regions

Reference Scenario

252

panied by the reduction in pasture areas to a great extent. In fact, the area planted with soybean and first harvest corn, the crops that will be most affected by sugar cane expansion, will be reduced less than 0.5 percent during the projected period. Pasture area in the present scenario will be 26.5 million hectares less than in the Reference Scenario and 9.6 million hectares less than in the herd optimization scenario in 2030 (Table 62). Sugar cane expansion will thus impose a more efficient use of pasture areas for the herd optimization scenario, due to an approximately 5 percent increase in carrying capacity.

It should be emphasized that the adoption of second generation ethanol contributes significantly to reducing pressure that the expansion of ethanol exports exerts on the demand for land. In the productivity pattern projected for 2030, 182 million tons of sugar cane will be needed to produce 17 billion liters of ethanol. Assuming productivity of 100 tons of sugar cane per hectare, the production of cellulosic ethanol will reduce the demand for land for cane by approximately 1.8 million hectares54.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 66: Land use in Brazil, ethanol scenario (in 1000 hectares)

2006

Cotton

Rice

Bean1

st

Bean 2 Corn1

nd

st

Corn 2

nd

844

3,018

2008

2,881

2,857

3,332

5,052

1,529 9,632

1,143

5,608

Pasture

208,889

205,381

-

5,887

1,454

10,292

21,334

Production forest

1,453

1,328

9,656

8,235

1,399

2,394

22,749 6,179

Ethanol and production forest scenario

3,231

Soybean

Sugar cane

Herd optimization scenario

2030

1,066

2,694

Reference Scenario

30,601

12,700 8,450

207,060

Source: ICONE

2030

3,229

2,390 1,328

10,413 5,638

30,520

12,631 8,450

190,097

2030

3,242

2,414 1,322

10,333 5,609

30,417 19,188

11,174

188,049

The regional analysis indicates that, like in the Reference Scenario, a good part of sugar cane expansion occurs in the Southeast, where the area dedicated to sugar cane will reach 8.1 million hectares in 2020 and 11.1 million hectares in 2030. Although the percentage of the total area of the southeast’s participation decreased over time, this reduction will be less acute than in the Reference Scenario (Table 67).

54

A more rigorous evaluation does not allow such an interpretation, as ethanol production and export would probably be different if there is no implementation of second generation ethanol.

Table 67: Regional sugar cane distribution in the Reference Scenario, the herd optimization scenario and the ethanol scenario (in thousand hectares)

Brazil South

Southeast

Central-West Cerrado Northern Amazon

Northeast Coast

Mapito and Bahia

2006

2008

2030

6,179

8,235

12,700

501

954

1,594

483

3,944 113

979

160

694

5,120 135

1,150 182

1,292

7,056 110

1,214

1,435

Source: ICONE

Herd Optimization Scenario

Ethanol Scenario

1,297

1,605

2030

12,631 7,197

1,369 111

1,217

1,441

2030

11,149 2,594 259

1,435

2,146

The greatest variations in the sugar cane area in relation to the herd optimization scenario occur in the Southeast, Central West Cerrado, MAPITO and Bahia, respectively (Table 67). In these regions, a greater response to pasture reduction has been observed both from the pressure caused by cane expansion as well as its potential intensification.

It should be emphasized that the scenario with a large volume of ethanol exports causes practically no expansion of the area for sugar cane or any other agricultural operation in the Northern Amazon (measured by the difference in area between the present scenario and the herd optimization scenario). Thus, once technical improvements in the herd optimization scenario were implemented, the decrease in pasture areas as a result of the reduction in the cattle herd was enough to accommodate almost the entire expansion of crops generated in the scenario with the greatest ethanol exports.55

7.4 Legal Scenario (Reforestation of the Legal Reserve)

The area necessary for the reforestation of the Legal Reserve (LR) is estimated to be about 44 million hectares. Over half is located in the Northern Amazon region (Table 68) principally due to the fact that greatest percentage of LR needed, or 80 percent of the properties, is in the Amazon biome.

55

253

19,188

The scenario analized considers the progressive adoption of second generation technology in the case of sugar cane.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Regions

Reference Scenario

Table 68: Reforestation needs in order to comply with the Legal Reserve in the regions of the model (1000 ha) Region

Area to be reforested by 2030

Central-West Cerrado

7,870

South

254

Southeast

Northern Amazon Northeast Coast

MAPITO and Bahia Brazil

3,294 6,381

24,573 299

1,928

44,344

Source: ICONE

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

The area allocated for reforestation between 2009 and 2030 is completely accommodated by pasture land, which will decrease approximately 60 million hectares during this time, going from 203.6 in 2009 to 143.9 million hectares in 2030 (Table 69). The reduction in pasture area is the result of the improvement in zootechnical indices due to the herd optimization scenario, and of pasture intensification due to the expansion of crop areas and forest restoration.

Table 69 shows that since the need for reforestation will be greater in the Northern Amazon region in 2030, there will also be a larger reduction in pasture area in this region, decreasing 25.3 million hectares between 2009 and 2030. Thus, with the reduction in pasture observed during the period analyzed, it was necessary to move the herd between the different regions of the model, so that the gain in productivity in all regions was similar and compatible with the development observed in the past. Table 69: Pasture area in the regions of the model in 2009 and 2030 (in 1000 ha), in the reforestation scenario of the LR Region

2009

2030

South

17,664.65

Northern Amazon

52,574.64

27,306.56

203,600.67

143,866.39

Southeast

Central-West Cerrado

Northeast Coast

MAPITO and Bahia

Brazil

41,439.97

50,385.22

10,569.62 30,966.56

Source: ICONE

9,281.27

32,590.04 38,799.26 9,896.74

25,992.52

In conclusion, the great need for reforestation and its concentration in a few regions will require a considerable reduction in pasture areas, livestock intensification and

herd relocation, with a greater need for investments in livestock in this scenario than in the others. It will also imply a change in the geographic distribution of the country’s production-related operations, giving rise to new slaughterhouses and processing plants, logistic system for production distribution and other processes that are part of the production chain.

The production of grain and other crops in this scenario was stable compared to the Reference Scenario, made possible due to a decrease in pasture areas to accommodate all the reforestation needs, with no need to reduce other crops. It also meant that it was technically possible to comply with environmental restrictions without impacting agricultural production. However, such an arrangement requires the adoption of new techniques for livestock, the breeder’s adaptation and leads to higher production costs.

255

A separate analysis was done for the legal scenario in these two biomes, where areas of Legal Reserve without plant cover in the Cerrado and Atlantic Forest biomes have restored the vegetation at a rate of 1/22 per year of the total available until 2030.

Table 70 presents the results summarized by state. The CO2 uptake potential per hectare ranges between 68.9 and 149.6 tons, as observed in the states of Rio Grande do Norte and Paraná, respectively. For each federal unit, the total potential per state was between Rio Grande do Norte (190,532 tCO2) and Mato Grosso (261,904,694 tCO2), while the total for Brazil was 1,053,723,278 tCO2, at an average abatement cost of US$ 40.42.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Cerrado and Atlantic Forest

Table 70: Presentation of quantitative results by state for the Atlantic Forest and Cerrado

256

State

Total tCO2 (2030)

tCO2/ha

Nominal value of investment

Abatement cost (US$/ tCO2)

Break-even carbon price (US$)

Alagoas

4,074,612

108.08

928,586,767

37.91

43.95

Goiás

96,268,673

88.49

26,796,150,060

46.3

53.68

Bahia

Espírito Santo Maranhão

Mato Grosso

Mato Grosso do Sul Minas Gerais

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Paraíba Paraná

Pernambuco

Rio de Janeiro

Rio Grande do Norte Rio Grande do Sul Rondônia

Santa Catarina São Paulo Sergipe

Tocantins Total

10,896,586 7,820,631 1,484,927

261,904,694 124,773,664 94,409,049 787,757

99,308,662 2,048,635 6,534,600 190,532

72,712,093 19,874,197 22,361,522

158,714,178 4,672,513

64,885,751

1,053,723,278

108.06 91.39 87.03 92.96 88.13 84.50 88.88

149.60 84.44 88.09 68.90

147.40 102.36 134.65 114.94 94.42 98.31

2,483,712,919 2,107,755,900 420,235,831

69,393,439,992 34,871,346,416 27,518,095,595 218,299,926

16,350,124,781 597,524,371

1,827,161,315 68,113,435

12,150,222,069 4,782,298,785 4,090,412,294

34,010,904,452 1,218,874,295

16,256,623,392

256,089,882,598

37.91 44.83 47.07 44.07 46.49 48.48 46.1

27.39 48.52 46.51 59.47 27.8

40.03 30.43 35.65 43.39 41.68 40.42

43.96 51.97 54.58 51.10 53.90 56.21 53.44 31.75 56.25 53.92 68.94 32.22 46.40 35.28 41.33 50.31 48.32 46.87

7.5 Aggregate Scenario: Herd, Production Forests, Ethanol, Forest Restoration

The last Low-carbon Scenario combines all of the previous ones: herd optimization, large-scale ethanol exports, increase of production forests and legality or reforestation of the Legal Reserve. Thus, considering a zero increase for the total area, impacts on land-use change appear greater compared to earlier scenarios. This section analyzes impacts on land use, agriculture and livestock production for Brazil and the six regions considered in the model.

Table 71 shows a comparison of land-use results in the different scenarios analyzed for Brazil for 2006, 2008, and 2030. Compared to the Reference Scenario, the main

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

impact on land use in the aggregate scenario occurs on pasture lands, which went from 207 to 138 million hectares in 2030, respectively, for each scenario, with a herd of 208 million head (Table 73). Compared to a herd optimization scenario, where the herd also has 208 million head, pasture areas amounted to less than 52.3 million hectares in 2030, which can be explained by the greater demand for land for the ethanol, production forest and legal scenarios. This implies a greater increase in productivity on 257 pastures in order to accommodate such a demand; in other words, the number of head per hectare, or carrying capacity, increased significantly compared to the reference and herd optimization scenarios.

Herd (thousand head)

Agriculture and Livestock Area

Total Restoration

Pasture

Production Forest

Sugar Cane

Grains (1st harvest)

 

 

 

257,284

201,410

205,886

0

205,381

5,874

259,275

0

208,889

5,269

8,235

37,794

38,937

6,179

2008

 

2006

 

Reference

234,460

276,127

0

207,060

8,450

12,700

47,917

2030

28,573

16,852

0

-1,829

3,181

6,521

8,980

Var. 20302006

208,002

216,042

44,344

137,820

11,174

19,188

47,860

2030

Herd-Forest-Ethanol-Restoration

208,025

259,183

0

190,097

8,450

12,631

48,005

2030

Source: ICONE

2,116

-43,233

44,344

-71,069

5,905

13,009

8,923

Var. 20302006

Herd

2,139

-92

0

-18,792

3,181

6,452

9,068

Var. 20302006

208,025

259,859

0

188,049

11,174

12,631

48,005

2030

Herd with Forest

2,139

584

0

-20,840

5,905

6,452

9,068

Var. 20302006

208,099

256,019

0

180,521

8,450

19,188

47,860

2030

Herd with Ethanol

Low Carbon

Table 71: Comparison of land use results in all scenarios for Brazil

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Var. 20302006

2030

2,213

-3,256

0

-28,368

3,181

13,009

8,923

208,024

212,952

44,344

143,866

8,450

12,631

48,005

Herd with Restoration

258 2,138

-46,323

44,344

-65,023

3,181

6,452

9,068

Var. 20302006

17,603

50,636

18,146

51,200

10,801

Central-West Cerrado

Northeast Coast

32,138

52,551

44,053

31,061

10,487

53,728

41,865

30,399

10,812

64,624

48,395

39,565

-1,739

11

12,074

-2,806

-4,488

-4,881

-1,829

Var. 20302006 -8,581

Var. 20302006

Herd 2030

-5,540

Var. 20302006

2030

Herd with Forest

-5,540

Var. 20302006 2030

Herd with Ethanol

25,590

9,682

26,981

38,285

27,718

9,564

28,401

10,196

51,879

47,338

39,678

Source: ICONE

-6,548

-1,120

-25,569

-12,916

-16,335

12,606

-3,738

-605

-671

-3,863

-4,375

28,237

10,196

51,716

47,256

38,038

12,606

-3,901

-605

-834

-3,944

-6,015

27,681

9,981

51,718

43,525

36,138

11,480

-4,457

-820

-833

-7,676

-7,915

-6,666

Var. 20302006

2030

Herd with Restoration

-8,864

Var. 20302006

-25,244 25,993

9,897

-6,146

-904

-12,401 27,307

38,799

-11,463

32,590

9,281

137,820 -71,068 190,097 -18,792 188,049 -20,840 180,521 -28,368 143,866 -65,023

2030

Herd-Forest-Ethanol-Restoration

Low Carbon

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Mapito and Bahia

Northern Amazon

Southeast

13,264

2030

208,889 205,381 207,060

South

Brazil

2008

2006

Regions

 

 

 

Reference

Table 72: Comparison of results for pasture area in all scenarios for Brazil and regions

259

260

An 11 percent increase in carrying capacity in the herd optimization scenario, and an annual growth rate of 0.46 percent, was observed in Brazil between 2006 and 2030. This increment will already be significantly higher in the last Low-carbon Scenario, with an annual growth rate of 2 percent per year and a 53 percent increase in the carrying capacity during the same period. One of the main topics to be analyzed in this scenario is the impact on land use in the regions considered in the model. In the Northern Amazon, Southeast and Central-West, pasture areas will be reduced to 25, 12 and 9 million hectares, respectively, during the period analyzed, based on data from Table 73, although the reasons for this vary greatly between the regions. Table 74 shows the regional land-use results for some select products.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Table 73: Results for the cattle herd in the reference, herd optimization and aggregate scenarios (1000 head)

 

2006

Reference

2008

2030

Herd Optimization

Herd-Forest-Ethanol-Restoration

25,673

27,590

2030

Brazil

205,886

201,410

234,460

Central-West Cerrado

56,445

55,506

63,238

58,086

70,644

8,665

8,156

8,958

8,372

8,587

South

Southeast Northern Amazon

Northeast Coast

MAPITO and Bahia

27,200

39,209

47,391 26,977

26,607

37,525

47,149 26,468

27,342

36,266

68,064 30,592

Source: ICONE

208,025

2030

37,548

49,901 28,446

208,002 39,944

27,951 33,287

Table 74: Results for land use and herd for selected products in the aggregate scenario

Source: ICONE

2018

2030

1,033.71 7,524.12 1,586.56 176.25 1,217.86 921.01

1,604.67 11,146.98 2,594.07 259.31 1,435.27 2,145.76

14,26407 36,552.75 45,065.89 41,681.69 10,276.62 29,306.12

9,564.48 27,718.43 38,284.66 26,981.17 9,681.55 25,590.19

1,497.35 2,900.25 3,577.29 11,169.51 135.97 876.17

25,911 36,574 61,847 43,174 8,772 29,550

3,294.18 6,380.55 7,870.04 24,572.92 299.13 1,927.58

27,590 39,944 70,644 27,951 8,587 33,287

Almost all of the reduction in pasture area in the Northern Amazon is due to the amount of area needed to recompose the deforested areas from Legal Reserve, 24.6 million hectares. This also implies a significant reduction in the cattle herd by 22 million head in 2030 within the framework of the herd optimization scenario, although there will still be a 15 percent gain in the livestock carrying capacity in this region between 2006 and 2030, representing a 1 percent annual growth rate during that period. There will be significant land-use change in the Reference Scenario, particularly due to the zero growth hypothesis in the total area of the Low-carbon Scenarios, as well as the recuperation of the environmental liability of the Legal Reserve. Thus, it may be said that avoided deforestation in this scenario will be 37 million hectares in relation to the Reference Scenario at the end of the period in question.

261

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

2006 Sugar cane (thousand hectares) South 483.25 Southeast 3,944.35 Central-West 500.59 Northern Amazon 112.63 Northeast Coast 978.68 MAPITO and Bahia 159.77 Reforestation (thousand hectares) South Southeast Central-West Northern Amazon Northeast Coast MAPITO and Bahia Pasture (thousand hectares) South 18,145.56 Southeast 44,052.98 Central-West 51,200.45 Northern Amazon 52,550.55 Northeast Coast 10,801.06 MAPITO and Bahia 32,138.30 Herd (thousand head) South 27,200 Southeast 39,209 Central-West 56,445 Northern Amazon 47,391 Northeast Coast 8,665 MAPITO and Bahia 26,977

Livestock intensification in the Southeast is due to the expansion of sugar cane (11.2 million hectares in 2030), production forests (5 million hectares) and reforestation (6.4 million hectares). All of these products combined will result in a 16 million hectare expansion between 2006 and 2030. Despite the growth in the sugar cane area, the Southeast will reduce some of its contribution to this product in Brazil, going from 68 percent in 2006 to 62 percent in 2030. On the other hand, the Central West and MAPITO region and Bahia will increase their contributions independently by five percentage points during the same period, absorbing almost the entire loss of the Southeast.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

262

Despite the significant reduction in pasture areas, the herd remains stable in the Southeast and will increase by 12.6 million head in the Central-West in 2030 as part of the herd optimization scenario. Herd expansion in the Central-West is largely due to the significant reduction in the herd in the Northern Amazon. Productivity gains are 62 percent in the Southeast and 67 percent in the Central-West between 2006 and 2030 in the last Low-carbon Scenario. The annual growth rate for carrying capacity during this period for the two regions will be 2.22 percent and 2.20 percent, respectively, reaching 1.44 and 1.85 head per hectare in 2030.

Compared to the herd optimization scenario, the Southeast will also lose participation in sugar cane production. With regard to production forests, the region absorbed 43 percent of total Brazilian expansion due to the sizeable concentration of the iron and steel industry in this region.

In addition to sugar cane expansion, the livestock intensification observed in the Central-West will principally be the result of reforestation aimed at the recomposition of the Legal Reserve, which had a deforested area of 7.9 million hectares. In conclusion, it can be said that the aggregate Low-carbon Scenario implies, besides a zero increase in the total area utilized, a more intense reduction process for pasture areas in order to absorb the expansion of agricultural areas, production forests and reforestation in the different parts of the country, especially in the Southeast and Central West.

Figure 62 compares land-use development for grains (first harvest), sugar cane, pastures, production forests and forest restoration (the latter only for the combined scenario) for the reference and aggregate scenarios (herd-ethanol-forests-restoration) from 2006 to 2030. To summarize, it may be observed that, whereas in the Reference Scenario there is expansion in the total area used for livestock-agriculture, in the aggregate Low-carbon Scenario, the total area remains static. Thus, all crops and reforestation expand across pasture areas, necessitating a more sizeable livestock intensification process. It is worth noting that forest restoration represents 21 percent of the total area used for livestock/agriculture (areas with crops plus pastures).

Figure 62: Results of the Reference Scenario (left) and aggregate Low-carbon Scenario (right).

Source: ICONE

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

263

8 References

264

ABRAF – ANUÁRIO ESTATÍSTICO DA ABRAF: ano base 2008 / ABRAF – Brasília, 2009

Alcides, Felipe Rodrigues. Considerações Ecológicas Sobre Plantios de Eucalipto – Anais do VIII Congresso de Ecologia do Brasil, 23 A 28 DE Setembro, 2007, Caxambu, MG. Alencar, A.; et al. Desmatamento na Amazônia: indo além da emergência crônica. Manaus, IPAM, 2004.

AliceWeb - Sistema de Análise das Informações de Comércio Exterior via Internet. Available at: http://aliceweb.desenvolvimento.gov.br/.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Amend, M; C. Gascon; J. Reid; J.M.C. da Silva. Parks Produce Local Economic Benefits in Amazônia. Conservation Strategy Fund Report, 2005. Disponível em . Last access on 12/05/2009.

AMS – Associacao Mineira de Silvicultura. Available at: http://www.silviminas.com. br/. AMS – Associacao Mineira de Silvicultura. Perspectivas e tendências do abastecimento de madeira para a indústria de base florestal no Brasil – Uma contribuição à construção e companhamento dos cenários futuros. Minas Gerais. 2004. Disponível em:www.silviminas.com.br

Amthor, J.S. and M. Hustoun. 1998. Terrestrial Ecosystem Responses to Global change: A Research Strategy. ORNL/TM-1998/27. Oak Ridge National Laboratory.

Andreae, M. O. Biomass burning: Its history, use and distribution and its impact on environmental quality and global climate, Global Biomass Burning: Atmospheric, Climatic and Biospheric Implications, edited by J. S. Levine, pp. 3-21, MIT Press, Cambridge, Mass., 1991.

Andreae, M. O. Correlation between cloud condensation nuclei concentration and aerosol optical thickness in remote and polluted regions, Atmos. Chem. Phys., 9, 543556, 2009.

Andreae, M. O.; D. Rosenfeld; P. Artaxo; A.A. Costa; G.P. Frank; K.M. Longo; M.A.F. Silva-Dias. Smoking rain clouds over the Amazon, 2004. Science, 303, 1337-1342.

Andreae, M. O. and P. Merlet. Emission of trace gases and aerosols from biomass burning. Global Biogeochemical Cycles, 2001, vol.15, no 4, 955-966 p. Artaxo, P.; J.V. Martins; M.A. Yamasoe; A.S. Procópio; T.M. Pauliquevis; M.O. Andreae; P. Guyon; L.V. Gatti; A.M.Cordova. Physical and chemical properties of aerosols in the wet and dry season in Rondônia, Amazonia. Journal of Geophysical Research, 2002, 107 (D20): 49.1-49.14

Assad, E.; Pinto, H. Aquecimento global e a nova geografia da produção agrícola no Brasil. EMBRAPA/ UNICAMP: Campinas, 2008.

ATLAS Solarimétrico do Brasil. Recife: Editora Universitária da UFPE, 2000.

Azevedo, T.; Tocantins, M.A. Instrumentos econômicos da nova proposta para a gestão de florestas públicas no Brasil. In: Megadiversidade. v.2.n.1-2.2006 BEM 2006, Balanço Energético Nacional, Ministério de Minas e Energia.

Berry, E. X., Modification of the warm rain process, Proc. First. Natl. Conf. Weather 265 Modification, Ed. American Meteorological Society, State University of New York, Albany, pp. 81- 88, 1968.

Boddey, R. M.; Soares, L.H.B.; Alves, B.J.R.; Urquiaga, S.. Bio-Ethanol Production in Brazil. In: Pimentel, D. (Org.). Biofuels, Solar and Wind as Renewable Energy Systems. Berkeley: Springer, 2008, v. p. 321-356.

Boddey, R.M.; Jantalia, C.P.; Macedo, M.O.; Oliveira, O.C.; Resende, A.S.; Alves, B.J.R.; Uruguiaga, S. Potential of carbon sequestration in soils of the Atlantic Forest region of Brazil. In: Carbon sequestration in soils of Latin America. Haworth Press, Binghamilton. 2006. pp. 305-347.

Bonezzi, Cristiane Belize; Caldeira-Pires, Armando; Brasil JR., Antônio C. P. “Avaliação do Ciclo de Vida e a Competitividade Ambiental da Siderurgia no Brasil”. Available at: http://www.intempres.pco.cu/Intempres2000-2004/Intempres2004/Sitio/Ponencias/36. doc Acessado em 17/06/2009. Bonham-Carter, G. Geographic information systems for geoscientists: modelling with GIS, Pergamon, New York, 1994. 414p.

Brasil. Banco de Desenvolvimento de Minas Gerais (BDMG). Demonstrativo Oficial 2008. Disponível em: http://www.bdmg.mg.gov.br/

Brito, José Otávio. Carvão vegetal no Brasil: gestões econômicas e ambientais. Estud. av. vol.4 no.9 -São Paulo May/Aug. 1990.

Brazil. Brazil’s contribution to prevent climate change. Brasília: Ministério das Relações Exteriores; Ministério da Ciência e Tecnologia; Ministério do Meio Ambiente; Ministério das Minas e Energia. Ministério do desenvolvimento, Indústria e Comércio Exterior. 2007. Brasil, Empresa de Pesquisa Energetica – EPE. Plano nacional de energia 20230 - PNE/2030. 408 p. ISBN: 978-85-60025-02-2 Rio de Janeiro. 2007. Disponível em . Acesso em 5 dezembro 2007.

Brasil. Ministério do Meio Ambiente. Plano de metas do PNF 2004-2007. Disponível em: www.mma.gov.br.

Brasil, Ministerio do Meio Ambiente - MMA. Projeto de Conservação e Utilização Sustentável da Diversidade Biológica Brasileira – PROBIO. Disponível em . Acesso em 23 de março de 2008.

266

Brown, S. and A.E. Lugo. 1982. The storage and production of organic matter in tropical forests and their role in the global carbon cycle, Biotropica 14:161–187.

Brown, S. and A.E. Lugo. 1984. Biomass of tropical forests: A new estimate based on forest volumes, Science. 223:1290-1293.

Brown, S. and A.E. Lugo. 1992. Aboveground biomass estimates for tropical moist forests of the Brazilian Amazon, Interciencia. 17:8–18.

Brown, S.; A.J.R. Gillespie and A.E.Lugo. 1989. Biomass estimation methods for tropical forests with applications to forest inventory data. Forest Science. 35:881-902.

Brown, S.; L.R.Iverson; A. Prasad and D. Liu. 1993. Geographical distributions of carbon in biomass and soils of tropical Asian forests. Geocarto International. 4:45–59. CAIT-WRI, World Resource Institute. 2007. Available at

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Canalli, B.L. Desafios para adoção e permanência em SPD. Anais do 8º. Encontro de Plantio Direto no Cerrado, 28/06 a 01/07 2005, Tangará da Serra, MT. Editora Sanches: Tangará da Serra, MT. P. 101-107. 2005.

Castro, Luiz Fernando Andrade. Produção de Gusa em Altos-Fornos a Carvão Vegetal e a Coque. Relatório Técnico. UFMG/ Departamento de Engenharia Metalúrgica e de Materiais. Belo Horizonte, Julho de 2000.

Catharino, E.L.M.; L.C.Bernacci; G.A.D.C. Franco; G. Durigan and J.P. Metzger. 2005. Aspectos da composição e diversidade do componente arbóreo das florestas da Reserva Florestal do Morro Grande, Cotia, SP. Biota Neotropica. 6(2). CCAP. 2006. Greenhouse gas mitigation in Brazil: scenarios and opportunities through 2025. Center for Clean Air Policy – COPPE.

Cerri, C. E. P.; G. Sparovek; Martial Bernoux; W.E. Easterling; Jerry M. Mellilo; Carlos Clemente Cerri. Tropical agriculture and global warming: impacts and mitigation options. Scientia Agricola, v. 64, p. 83-99, 2007.

Coe, M.; M. Costa and B.S. Soares-Filho. The influence of historical and potential future deforestation on the stream flow of the Amazon River – Land surface processes and atmospheric feedbacks. Journal of Hidrology, 2009. Coleman, K.; D.S. Jenkinson; G.J. Crocker; P.R.Grace; J.Klir; M. Korschens; P.R.Poulton and D.D.Richter. Simulating trends in soil organic carbon in long-term experiments using RothC-26.3. Geoderma, v. 81, p. 29-44, 1997. Companhia Nacional de Abastecimento – CONAB. Central de informações agropecuárias – safras. Disponível em . Acesso em 3 de junho de 2008. CONAB. Companhia nacional de abastecimento. www.conab.org.br.

CONCAWE. 2004. Well to wheels analysis of future automotive fuels and Powertrains in the European context, January, 2004. 60 pg.

Cotton, W. and R. Pielke. Human impacts on weather and climate, Cambridge University Press, New York, 1996.

De Maria I.C. Erosão e terraços em plantio direto. Boletim Informativo da Sociedade Brasileira de Ciência do Solo. Viçosa, volume 24, 1999, p 17-21.

Derpsch, R. Agricultura sustentável. In: Saturnino, H.M.; Landers, J.N. (Ed.). O meio ambiente e o plantio direto. Goiânia : EMBRAPA, SPI, 1997. p.29-48.

Diekow, Jeferson; J. Mielniczuk; Heike Knicker; Cimelio Bayer; Deborah Pinheiro 267 Dick and Ingrid Kogel Knabner. Soil C and N stocks as affected by cropping systems and nitrogen fertilization in a southern Brazil Acrisol managed under no-tillage for 17 years. Soil & Tillage Research, Amsterdam, v. 81, p. 87-95, 2005.

Dislich, R.; L. Cersósimo and W. Mantovani. 2001. Análise da estrutura de fragmentos florestais no Planalto Paulistano – SP. Revista Brasileira de Botânica. 24(3):321332. EC, 2008, European Commission Proposal for a Directive of the European Parliament and of the Council on the Promotion of the Use of Energy from Renewable Sources, 2008/0016 (COD).

EMBRAPA. 1999. Centro Nacional de Pesquisa de Solos. Sistema Brasileiro de Classificação de Solos. Rio de Janeiro. 412 p.

Empresa Brasileira de Pesquisa Agropecuaria – EMBRAPA. Sistema de monitoramento agrometereológico - Agritempo. Disponível em . Acesso em fevereiro de 2008. EMPRESA BRASILEIRA DE PESQUISA AGROPECUÁRIA – EMBRAPA. Aquecimento Global e a nova geografia da produção agrícola no Brasil. 83p.Posigraf. Agosto 2008. EPE, 2008, Empresa de Pesquisa Energética, Plano Nacional de Energia 2030 PNE 2030.

Eva, H. D.; A.S Belward; E.E. de Miranda; C.M. di Bella; V. Gond; O. Huber; S. Jones; M. Sgrnzaroli and S.Fritz (2004). A land cover map of South America. Global Change Biology 10, 731-744.

FAO 2001. FAOCLIM 2.0 A World-Wide Agroclimatic Database. Food and Agriculture Organization of the United Nations: Roma, Itália.

Farr, T. G., et al. 2007. The Shuttle Radar Topography Mission. Rev Geophys. 45, RG2004, doi:10.1029/2005RG000183.

Fearnside, P.M.and R.I Barbosa. Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. Forest Ecology and Management, Amsterdam, v. 108, n. 1-2, p. 147-166, 1998 Federal Register, 2009, 40 CFR. Part 80, Part II, Vol. 74, No. 99, May 26, 2009.

Ferreira, N. C.; L.G. Ferreira; A.R. Huete and M.E. Ferreira (2007b). An operational deforestation mapping system using MODIS data and spatial context analysis. International Journal of Remote Sensing 28, 47-62.

Freitas, S. R.; K.M. Longo; M. Silva Dias; P. Silva Dias; R. Chatfield; E. Prins; P. Artaxo; G. Grell and F. Recuero. Monitoring the transport of biomass burning emissions in South

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

EBAMM Release 1.0, 2005, The ERG biofuels analysis meta-model.

America. Environmental Fluid Mechanics, 2005, 5 (1-2), 135 – 167 p.doi: 10.1007/ s10652-005-0243-7 268

Freitas, S. R.; K.M. Longo and M. Andreae. Impact of including the plume rise of vegetation fires in numerical simulations of associated atmospheric pollutants. Geophysical Research. Letters, 2006, 33, L17808, doi:10.1029/2006GL026608.

Freitas, S. R.; K. Longo; M. Dias; R. Chatfield; P. Dias; P. Artaxo; M. Andreae; G. Grell; L. Rodrigues; A. Fazenda and J. Panetta. The Coupled Aerosol and Tracer Transport model to the Brazilian developments on the Regional Atmospheric Modeling System (CATTBRAMS). Part 1: Model description and evaluation. Atmospheric Chemistry and Physics. Discussion Paper, 2007, 8525-8569 p. Fisher, M.J.; S.P. Braz; R.S.M. Santos; S. Urquiaga; B.J.R. Alves and R.M. Boddey. Another dimension to grazing systems: Soil carbon. Tropical Grasslands, v. 41, p. 65-83, 2007.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Food and Agricultural Policy Research Institute – FAPRI. FAPRI Models. Available at . Accessed on January 8 2007.

Fundação SOS Mata Atlântica – SOSMA: Atlas de remanescentes. 2005. Available at . Furriela, R. B. 2005 Introdução às Mudanças Climáticas Globais. São Paulo.

Gassen, D. and F. Gassen. Plantio Direto, o caminho do futuro. Passo Fundo: Aldeia Sul, 1996, 207p.

Gomes, B.Z.; F.R. Martins and J.Y. Tamashiro. 2004. Estrutura do cerradão e da transição entre cerradão e floresta paludícola num fragmento da International Paper do Brasil Ltda., em Brotas, SP. Revista Brasileira de Botânica. 27(2):249-262.

Greenland, D.J.; A. Wild and D. Adams. Organic matter dynamics in soils of the tropics-from myth to complex reality. In: Lal, R. and P.A. Sanchez. (eds.) Myths and science of soils of the tropics. Madison: SSSA Spec., 1992. p.17-34. (Pub. 29).

Greet 1.6 Greenhouse gases, regulated emissions and energy use in transportation, Argonne National Laboratory, Argonne, Illinois, USA.

Grell, G. and D. Devenyi. A generalized approach to parameterizing convection combining ensemble and data assimilation techniques. Geophysical Resource Letters, 2002, 29, 14, 1693.

Gruberruber, N.; P. Friedlingstein; C.B. Field; R. Valentini; M. Heinmann; J.E. Richey; P.R. Lankao; E.D. Schulze and C.T.A. Chen. The vulnerability of the carbon cycle in the 21st century: An assessment of carbon-climate-human interactions. In: The Global Carbon Cycle (C.B. Field and M.R.Raupach, eds). Scope, no. 62. Island Press, Washington D.C. 2004, p. 45-76. Hahn, C. M. et al. 2004. Recuperação florestal: da muda à floresta. São Paulo: SMA.

Hanaki, M.T.; T. Ito and M. Saigusa. Effect of no-tillage rice (Oryza sativa L.) cultivation on methane emission in three paddy fields of different soil types with rice straw application. Jpn. J. Soil Sci. Plant Nutr. 73: 135-143. 2002.

Harada, N.; S. Otsuka; M. Nishimura and S.Matsumoto. Influences of indigenous phototrophs on methane emissions from a straw-amended paddy soil. Biol. Fertil. Soils 41:

46-51. 2005

Hijmans R.; S. Cameron; J. Parra; P. Jones and A. Jarvisc. 2005. Avery high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25: 1965–1978. Holdridge, L. 1967. Life Zone Ecology. Tropical Science Center, San Jose, Costa Rica.

Houghton, R. A.; D.L. Skole; C.A. Nobre; J.L. Hackler; K.T. Lawrence, and W.H. Cho- 269 mentowski. 2000. “Annual fluxes of carbon from deforestation and regrowth in the Brazilian Amazon”. Nature 403: 301–304. IBAMA – Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis. Available at: http://www.ibama.gov.br/

IBGE – Instituto Brasileiro de Geografia e Estatistica. Censo Agropecuário: 2006. Rio de Janeiro, 2007. Available at: www.ibge.gov.br

IBGE – Instituto Brasileiro de Geografia e Estatistica. Produção da extração vegetal e da silvicultura 2003 volume 18. Rio de Janeiro. Available at: www.ibge.gov.br IBGE 2006. Mapa de fertilidade de solos do Brasil. 1:5.000.000. IBGE.

IBGE. Instituto Brasileiro de Geografia e Estatística. www.ibge.com.br.

IBS - Instituto Brasileiro de Siderurgia. Available at: http://www.ibs.org.br/index. asp. IEA, 2008, International Energy Agency, Energy Technology Perspectives 2008

IEA, 2009, International Energy Agency, http://www.iea.org

IISI - International Iron and Steel Institute. Available at: http://www.worldsteel. org/.

Instituto Brasileiro de Geografia e Estatistica – IBGE. Pesquisa Agrícola Municipal 2007 - PAM. Available at . Accessed November 9 2008a.

Instituto Brasileiro de Geografia e Estatistica – IBGE. Pesquisa Pecuária Municipal 2007 – PPM. Disponível em http://www.sidra.ibge.gov.br/bda/pesquisas/ppm/default.asp. Accessed November 8 2008b.

Instituto Brasileiro de Geografia e Estatistica – IBGE. Produção da Extração Vegetal e da Silvicultura 2007. Available at . Acesso em 9 de setembro de 2008c.

Instituto Nacional de Pesquisas Espaciais – INPE. Projeto PRODES - monitoramento da floresta amazônica brasileira por satélite. Available at . Accessed October 20 2008.

Instituto Nacional de Pesquisas Espaciais – INPE. Relatório de Gestão. São José dos Campos, 2009. Available at . Last access 05/12/2009.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

IBGE – Instituto Brasileiro de Geografia e Estatistica. Censo Agropecuário: 19951996. Rio de Janeiro, 1996. Available at: www.ibge.gov.br

Instituto Nacional de Pesquisas Espaciais – INPE. (2009). Projeto DEGRAD (Mapeamento da Degradação Florestal na Amazônia). Available at < http://www.obt.inpe.br/ degrad/>. Last access 05/12/2009. 270

Instituto Nacional de Pesquisas Espaciais – INPE. (2009). Projeto DEGRAD (Mapeamento da Degradação Florestal na Amazônia). Available at < http://www.obt.inpe.br/ degrad/>. Last access 05/12/2009. Instituto Nacional de Pesquisas Espaciais – INPE. (2009). PRODES (Programa de Monitoramento do Desmatamento na Amazônia). Available at < http://www.obt.inpe. br/prodes/> . Last access 05/12/2009.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Intergovernmental Panel on Climate Change (IPCC). Good practice guidance for land use, land use change and forestry. (2003) Available at

Intergovernmental Panel on Climate Change – IPCC. IPCC Special Report: Land Use, Land-Use Change and Forestry.(2000). Available at: < http://www.ipcc.ch/pdf/special-reports/spm/srl-en.pdf> Accessed 05/11/2009. ISBN:92-3169-114-3.

IPCC. International Panel on Climate Change. Guidlines for National Greenhouse Gas Inventories. Reference Manual, < http://www.ipcc-nggip.iges.or.jp/public/2006gl>. 2006.

IPCC. International Panel on Climate Change. Guidlines for National Greenhouse Gas Inventories. Reference Manual, 1996.

IPCC, 2001, Climate Change 2001: The scientific basis. Contribution of the Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change.

Ivaunaskas, N.M.; R.R. Rodrigues and A.G. Nave. 1999. Fitossociologia de um trecho de floresta estacional semidecidual em Itatinga – SP, Brasil. Scientia Florestalis. 56:8399.

Iverson, L. R.; S. Brown; A. Prasad; H. Mitasova; A.J.E. Gillespie and A.E. Lugo. 1994. Use of GIS for Estimating Potential and Actual Forest Biomass for Continental South and Southeast Asia. Effects of Land-Use Change on Atmospheric CO2 Concentrations. Springer-Verlag.

Jacobson, M. Z. Strong Radiative Heating due to the Mixing State of Black Carbon in Atmospheric Aerosols. Nature, vol. 409, 695-697, 2001.

Jank, M.S. 2008. An overview of the Brazilian sugarcane industry. Presentation given at the Better Sugarcane Initiative General Assembly, São Paulo, SP, November 13, 2008. Jantalia, C; D. Resck, D; B. Alves; L. Zotarelli; S. Urquiaga and R. Boddey. Tillage effect on C stocks of a clayey Oxisol under a soybean-based crop rotation in the Brazilian Cerrado region. Soil & Tillage Research, v. 95, p. 97-109, 2007

Jantalia, C.P.; L. Vilela; B.J.R. Alves; R.M. Boddey and S. Urquaia. Influencia de pastagens e sistemas de produção de grãos no estoque de carbono e nitrogênio em um Latossolo Vermelho. Seropédica: EMBRAPA Agrobiologia. 2006a, 50 p. (Boletim de Pesquisa e Desenvolvimento, 11).

Jantalia, C.P.; L. Zotarelli; H.P. dos Santos; E. Torres; S. Urquiaga; R.M. Boddey and 271 B.J.R. Alves. Em Busca da Mitigação da Produção de Óxido Nitroso em Sistemas Agrícolas: Avaliação de Práticas Usadas na Produção de Grãos no suldo País. In: Bruno J. R. Alves; Segundo Urguiaga; Celso Aita; Robert Michael Boddey; Claudia P. Jantalia; Flávio A. O. Camargo. (Org.). Manejo de Sistemas Agrícolas: Impacto no Sequestro de C e nas Emissões de Gases de Efeito Estufa. 1 ed. Porto Alegre: Genesis, 2006b, v. , p. 81-108. Lægreid, M.; Bøckman, O.C.; Kaarstad, O. Agriculture, fertilizers and the environment. Wallingford: CABI, 1999, 320p. José, F. J. A., 2009, comunicação pessoal

Jones, P.G. and A.Gladkov. 2003. FloraMap. A Computer Tool for Predicting the distribution of Plants and Other Organisms in the Wild.Version 1.02. Centro Internacional de Agricultura Tropical: Cali, Colômbia.

Juvenal, Thais Linhares and René Luiz Grion Mattos. O setor florestal no Brasil e a importância do reflorestamento. Rio de Janeiro: BNDES, 2002. Disponível em: www. bndes.gov.br

Kaufman, Y. J. Remote Sensing of Direct and Indirect Aerosol Forcing. Aerosol Forcing of Climate. Ed. by R. J. Charlson and J. Heintzenberg, John Wiley & Sons Ltd., 1995. Koren, I.; Y. Kaufman; L.A. Remer and J.V.Martins. Measurement of the Effect of Amazon Smoke on Inhibition of Cloud Formation, Science, 2004, 303, 1342-1345.

Lamas, M. P. Manejo cultural do algodoeiro nas condições do Cerrado. In: O Agronegócio do Algodoeiro no Brasil. (N.E.M. Beltrão & D.M.P. Azevedo, eds). Volume 2, 2ª. Edição. Brasília, DF, EMBRAPA Informação Tecnológica, 2008. pp. 621 - 648.

Laurance, W.F.; P.M. Fearnside; S.G. Laurance; P. Delamonica; T.E. Lovejoy; J.M.R. Merona; J.Q. Chambers and C. Gascona. 1999. Relationship between soils and Amazon forest biomass: a landscape-scale study. Forest Ecology and Management 118:127-138 Leal, M.R.L.V, Uso do bagaço e palha para produção de etanol e energia, F.O. Licht Ethanol Production Workshop, São Paulo, SP, 23 de março de 2009, Leite, Nelson Barbosa. O setor florestal no Brasil. In: A QUESTÃO FLORESTAL E O

Desenvolvimento, 8-9 jul. 2003, Rio de Janeiro. BNDES, Rio de Janeiro, 2003. Seminário. Available at: www.bndes.gov.br.

Lima, A.M.N.; I.R. Silva; J.C.L. Neves, R.F.Novais; N.F.Barros; E.S. Mendonca; M.S.M.Demolinari and F.P. Leite. Frações da matéria orgânica do solo após três décadas de cultivo de eucalipto no Vale do Rio Doce-MG. Revista Brasileira de Ciência do Solo. vol.32, n.3, pp. 1053-1063. 2008.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Jantalia, C.P.; D.V.S. Resck; B.J.R. Alves; L. Zotarelli; S. Urquiaga; R.M. Boddey. Tillage effect on C stocks of a clayey Oxisol under a soybean-based crop rotation in the Brazilian Cerrado region. Soil & Tillage Research 95, 97–109. 2007.

Lima, M.A. Relatório Técnico Final do Projeto “Emissão de metano em sistemas de produção de arroz irrigado: quantificação e análise”, período 2000 a 2004. EMBRAPA Meio-Ambiente: Jaguariúna. 2009. 272

Longo, K. M.; S.R. Freitas; M. Silva Dias; P. Silva Dias. Numerical modeling of the biomassburning aerosol direct radiative effects on the thermodynamics structure of the atmosphere and convective precipitation. In: International Conference on Southern Hemisphere Meteorology and Oceanography (ICSHMO), 8., 2006, Foz do Iguaçu. Proceedings... São José dos Campos: INPE, 2006. p. 283-289. CD-ROM. ISBN 85-17-000234.

Longo, K. M.; S.R. Freitas, et al. The Coupled Aerosol and Tracer Transport model to the Brazilian developments on the Regional Atmospheric Modeling System. Part 2: Model sensitivity to the biomass burning inventories. Atmospheric Chemistry and Physics. Discussion Paper, 2007, 8571-8595 p.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Macedo, I.C., M.R.L.V. Leal and J.E.A.R da Silva. 2004, Balanço das emissões de gases do efeito estufa na produção e no uso do etanol no Brasil, Relatório submetido à Secretaria Estadual do Meio Ambiente do Estado de São Paulo, Abril de 2004, 37 pg. Macedo, I.C., J.E.A. Seabra and J.E.A.R. da Silva. 2008, Greenhouse gas emissions in the production and use of ethanol from sugarcane in Brazil: The 2005/2006 averages and the prediction for 2020, Biomass & Bioenergy, 2008. Macedo, M.C.M. 1995. Pastagens no ecossistema cerrados. In: SIMPÓSIO SOBRE PASTAGENS NOS ECOSSISTEMAS BRASILEIROS: pesquisas para o desenvolvimento sustentável, 1995, Brasília. Anais. Brasília: SBZ, p.28-62.

Macedo, M.C.M., and A.H.Zimmer. Sistema de Pasto-Lavoura e seus efeitos na produtividade agropecuária. In: SIMPÓSIO SOBRE ECOSSISTEMA DE PASTAGENS. 1993. Jaboticabal,SP. Anais. Jaboticabal,1993.p.217-245.

Machado, R. B., M.B. Ramos Neto; P.G.P. Pereira; E.F. Caldas, D.A. Gonçalves; N.S. Santos; K. Tabor and M. Steininger (2004). Estimativas de perda da área do Cerrado brasileiro. [technical report]. Brasília-DF, Brasil: Conservation International.

Maddock, J.; M. dos Santos, M. and K. Prata. Nitrous oxide emission from soil of the Mata Atlantica, Rio de Janeiro State, Brazil, J. Geophys. Res., 106(D19), 23055-23060. 2001. Mattoso, M.J.; J.C. Cruz and I.A. Pereira Filho. Custo de produção em plantio direto. Informe Agropecuário, v. 22, p. 109-116. 2001.

Malhi, Y.J., et al. Climate Change deforestation and the faith of the Amazon. Science, 319: 169-172, 2008.

MAPA – Ministério da Agricultura, Pecuária e Abastecimento. Disponível em: http:// www.agricultura.gov.br Marengo, J.A., et al. The drought of Amazonia in 2005. J. Clim., 21, 495-516, 2008.

Margulis, S. Causas do desmatamento na Amazônia brasileira. 1ªed. Brasília: Banco Mundial, 2003.

Martins, O.S. 2004. Determinação do potencial de remoção de carbono na recuperação de matas ciliares na região de São Carlos, SP. Tese de Doutorado. Programa de PósGraduação em Ecologia e Recursos Naturais. Universidade Federal de São Carlos. 136 p.

Masera, O.R. et al. 2003. Modeling carbon sequestration in afforestation, agroforestry and forest management projects: the CO2FIX V.2 approach. Ecological Modelling 273 164:177–199.

McKinsey and Company, 2009, Pathway to a low-carbon economy for Brazil. Report. 46 p.

Ministério de Ciência e Tecnologia - MCT. Primeiro inventário brasileiro de emissões antrópicas de GEE: emissões e remoções de dióxido de carbono por conversão de florestas e abandono de terras cultivadas. Brasília, 2004.

Miranda, E. E. (Coord.). 2005 . Brasil em Relevo. Campinas: EMBRAPA Monitoramento por Satélite,

Moraes, R.J.; MM.G.L. Azevedo; F.M.A. Delmanto. 2005. As leis federais mais importantes de proteção ao meio ambiente comentadas. Rio de Janeiro: Renovar.

Nepstad, D. et al. 2007. Manejo e recuperação de mata ciliar em regiões florestais da Amazônia. Série Boas Práticas. Livro 1. Instituto de Pesquisa Ambiental da Amazônia.

Mosier, A. R.; E. Wassmann; L. Verchot; J. King and C. Palm. Methane and nitrogen oxide fluxes in tropical agricultural soils:sources, sinks and mechanisms. Environment, Development and Sustainability, v. 6, p. 11-49, 2004. Nakicenovic, N.; A. Grübler and A. McDonald, 1998, Global energy perspectives, Cambridge University Press, Cambridge, UK, 1998.

Nepstad, D. et al. The Costs and Benefits of Reducing Carbon Emissions from Deforestation and Forest Degradation in the Brazilian Amazon. Woods Hole: WHRC, 2007. v. 1. 26 p. Nepstad, D. et al. Globalization of the Amazon Soy and Beef Industries: Opportunities for Conservation. In: Conservation Biology, v. 6 n. 20, p. 1595-1603, 2006.

NIPE, 2005, Núcleo Interdisciplinar de Planejamento Energético da UNICAMP, Estudo sobre as possibilidades e impactos da produção de grandes quantidades de etanol visando a substituição parcial de gasolina no mundo – Fase 1, Relatório para o Centro de Gestão e Estudos Estratégicos (CGEE), Coordenado pelo Dr. Rogério Cezar de Cerqueira Leite, Dezembro de 2005. NIPE, 2007, Núcleo Interdisciplinar de Planejamento Energético da UNICAMP, Estudo sobre as possibilidades e impactos da produção de grandes quantidades de etanol visando a substituição parcial de gasolina no mundo – Fase 2, Relatório para o Centro de Gestão e Estudos Estratégicos (CGEE), Coordenado pelo Dr. Rogério Cezar de Cerqueira Leite, Março de 2007. Nobre, C. A., P. J. Sellers and J. Shukla. 1991. Amazonian deforestation and regional climate change. J. Climate 4, 957-988.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Ministério do Meio Ambiente - MMA. Plana Amazônia Sustentável. Brasília, 2008.

Nobre, et al. Amazônia: Fonte ou Sumidouro de Carbono? Causas e dinâmica do desmatamento na Amazônia. Brasília: Ministério do Meio Ambiente, 2001, 436 p.

274

NREL, 2002, Lignocellulosic Biomass to Ethanol Process Design and Economics Utilizing Co-Current Dilute Acis Prehydrolysis and Ezymatic Hydrolysis for Corn Stover, Report NREL/TP-510-32438, prepared by Adem, A., Ruth, M., Ibsen, K., Jechura, J., Neeves, Sheeran, J. and Wallace from NREL and Montague, L. Slayton, A. and Lukas, J.from Harris Group.

Nucleo Interdisciplinar de Planejamento Energetico NIPE; Centro de Gestao e Estudos Estrategicos - CGEE. Estudo sobre as possibilidades e impactos da produção de grandes quantidades de etanol visando à substituição parcial de gasolina no mundo Relatório final. 320p. Dezembro 2005.

Olivério, J.L. 2007. As Usinas de Açúcar e Álcool – Estado da Arte da Tecnologia Industrial, Congresso Internacional de Agroenergia e Biocombustíveis, Teresina, PI, 13 d3 Julho de 2007.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Pimentel, D. CRC Handbook of energy utilization in agriculture. Boca Raton: CRC Press. 1980, 475p.

Plano Anual de Outorga Florestal – PAOF 2009. Brasília: MMA/ Serviço Florestal Brasileiro, 2008.111f.

Presidencia da Republica. Decreto nº 6.527 de 1º de agosto de 2008. Available at . Last access on 05/09/2009.

PNMC - Plano Nacional de Mudanças sobre o Clima (versão para consulta pública). Available at: . Last access on 05/13/2009.

Rochette, P. No-till only increases N2O emissions in poorly-aerated soils. Soil & Tillage Research, v. 101, p. 97–100, 2008.

Rodrigues, J.E.R. 2005. Sistema Nacional de Unidades de Conservação. São Paulo: Ed. Revista dos Tribunais.

Rodrigues, et al. Dinamica EGO, uma plataforma para modelagem de sistemas ambientais. In: Simpósio Brasileiro de Sensoriamento Remoto, 13. Instituto Nacional de Pesquisas Espaciais. 2007.

Saatchi, S.S.; R.A. Houghton; R.C.S. Alvalá; J.V. Soares and Y. Yu 2007. Distribution of Aboveground Live Biomass in the Amazon Basin. Global Change Biology. Accepted article online: 24-Jan-2007

Saatchi, S.S; R.A. Houghton; R.C. Dos Santos; J.V. Soares; Y. Yu. 2007. Distribution of aboveground live biomass in the Amazon Basin. Global Change Biology, 13, 816–837.

Sampaio, G.; C. Nobre; M. Costa; P. Satyamurty; B.S. Soares-Filho; M. Cardoso. 2007. Regional climate change over eastern Amazonai caused by pasture and soybean cropland expansion. Geophysical Research Letters 34. doi:10.102.

Sanchez, P.A. Properties and management of soils in the tropics. New York: John Wiley & Sons, 1976. 618p.

Saturnino, H. M.; J.N. Landers. O Meio Ambiente e o Plantio Direto. Goiânia: APDC, 1997. 116 p.

SEMAD – Secretaria de Estado de Meio Ambiente e Desenvolvimento Sustentável. Disponível em: http://www.semad.mg.gov.br/

Silva, R.R., et al. Regional impacts of future land-cover changes on the Amazon Basin 275 wet season climate. J. Clim., 21 (6), 1153-1170. SINDIFER - Sindicato das Indústrias de Ferro de Minas Gerais. Disponível em: http://www.sindifer.com.br/inst.html.

Smith, K.A. and F. Conen. Impacts of land management of fluxes of trace greenhouse gases. Soil Use and Management, 20:255-263. 2004.

Smith, P.; D. Martino; Z. Cai; D. Gwary; H. Janzen; P. Kumar; B. McCarl; S. Ogle. F. O’Mara; C. Rice; B. Scholes and O. Sirotenko. 2007: Agriculture. In Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (eds)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Soares, J.J.; D.W. Silva and M.I.S. Lima. 2003. Current state and projection of the probable original vegetation of the São Carlos region of São Paulo State, Brazil. Brazilian Journal of Biology. 63(3):527-536.

Soares, L.H.B.; B.J.R. Alves; S. Urquiaga; R.M. Boddey. Mitigação dos gases de efeito estufa pelo uso de etanol de cana-de-açúcar produzido no Brasil. EMBRAPA Agrobiologia. Circulat Técnica no. 27. Seropedica, RJ. 14 p. 2009.

Soares-Filho, et al. Cenários de desmatamento para a Amazônia. Estudos Avançados, São Paulo, v.19, n.54, p.138-152, 2005. Soares-Filho, et al. Nexos entre as dimensões socioeconômicas e o desmatamento: a caminho de um modelo integrado. In: Batistela, M. et al (org.): Amazônia. Natureza e Sociedade em Transformação. Edusp. São Paulo, 2008.

Soares-Filho, et al. Redução das emissões de carbono do desmatamento no Brasil: o papel do programa Áreas Protegidas da Amazônia (ARPA). (em revisão), 2008b.

Soares-Filho, et al. Reduction of carbon emissions associated with deforestation in brazil: the role of the amazon region protected areas program (ARPA). Brasília: IPAM, 2008. v. 1. 32 p. Soares-Filho, et al. (2009). Projeto Dinamica EGO. Disponível em último acesso em 13/05/2009.

Sousa, E.L., 2008, A importância do etanol e da cogeração na atual matriz energética brasileira, Apresentação feita em reunião do Tribunal de Contas da União, Brasília, Df, 12 de novembro de 2008.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Sisti, C. P. J.; H.P. dos Santos; B.J.R. Alves; S. Urquiaga and R.M Boddey. Change in carbon and nitrogen stocks in soil under 13 years of conventional or zero tillage in Southern Brazil. Biology and Fertility of Soils, Amsterdam, v. 76, p. 39-58, 2004.

Teixeira, A. M. et al. Modeling Landscape dynamics in the Atlantic Rainforest domain: Implications for conservation. Forest Ecology and Management 257, 1219– 1230, 2009. 276

UNH, 2007. R-Hydronet: A Regional, Electronic Hydrometeorological Data Network for South America, Central America, And The Caribbean. Water Systems Analysis Group, Universidade de New Hampshire. Unruh, G. C. Understanding carbon lock-in. Energy Policy, 28 (12): 817-830, 2000.

Urquiaga, S.; B.J.R. Alves; R.M. Boddey. Produção de biocombustíveis - A questão do balanço energético. Revista de Política Agrícola, Brasília, DF, v. AnoXIV, n. 5, p. 42-46, 2005. USDOE, 2005, US Department of Energy, Clean Cities Initiative Fact Sheet, April 2005, 2 pg.

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Valverde, Sebastião Renato. Plantações de Eucalipto no Brasil. Revista da Madeira, no 07, 18 de Setembro de 2007.

van den Wall Bake, J.D. 2006. Cane as key in Brazilian ethanol industry, Master Thesis, Utrecht University, 2006.

Veloso, H.P.; A.L.R. Rangel Filho and J.C.A. Lima. 1991. Classificação da vegetação brasileira adaptada a um sistema universal. São Paulo: Fundação Instituto Brasileiro de Geografia e Estatística – IBGE. 123 p.

Vose, R.S., R.L. Schmoyer; P.M. Steurer; T.C. Peterson; R. Heim; T.R. Karl and J.K. Eischeid. 1992. The Global Historical Climatology Network: Long-term monthly temperature, precipitation, sea level pressure, and station pressure data. Oak Ridge National Laboratory, Carbon Dioxide Information Analysis Center.

Walko, R.; L. Band; J. Baron; F. Kittel; R. Lammers; T. Lee; D. Ojima; R. Pielke; C. Taylor; C.Tague; C.Tremback and P.Vidale. Coupled Atmosphere-Biophysics-Hydrology Models for Environmental Modeling. Journal Applied Meteorology, 2000. 39: (6) 931-944 p.

Walter, Arnaldo; Frank Rosillo-Calle; Paulo Dolian; Eric Piacente and Kamyla Borges Cunha. Perspectives on fuel ethanol consumption and trade. Biomass and bioenegy, 23 (2008) p.730-748. Available at www.elsevier.com/locate/biombioe. Accessed on June 8, 2008.

Walter, A.S.; F. Rosillo-Calle; P. Dolzan; E. Piacente and K.B. da Cunha. 2008, Perspectives on ethanol consumption and trade, Biomass & Bioenergy, 2008.

Walter, A.S.; P. Dolzan; O. Quilodrán; J. Garcia; C. da Silva, F. Piacente and A. Segerstedt. 2008, A sustainability analysis of the Brazilian ethanol, Report to the UK Embassy in Brasilia.

Wassmann, R.; R.S. Lantin; H.U. Neue; L.V. Buendia; T.M. Corton and Y. Lu. Characterization of Methane Emissions from Rice Fields in Asia. III. Mitigation Options and Future Research Needs Nutrient Cycling in Agroecosystems 58: 23–36, 2000.

Weck, J. 1970. An improved CVP-index for the delimitation of the potential productivity zones of forest lands of India, Indian Forester. 96:565-572.

Whitmore, T.C. 1984. Tropical Rain Forests of the Far East, 2nd edition, Oxford Science Publications, Clarendon Press, Oxford. WMO 1996. Climatological Normals (Clino) for the Period 1961-1990.

Xavier, R.P. Contribuição da fixação biológica de nitrogênio na produção sustentável de cana-de-açúcar. Tese de Doutorado, Programa de Pós-Graduação em Fitotecnia, Uni- 277 versidade Federal Rural do Rio de Janeiro. Seropédica: UFRRJ, 114 pp., 2006. Xiang P; H. Huang; M. Huang; D.X. Gan; Y. Zhou and Z. Fu. Studies on Technique of Reducing Methane Emission in a Rice-Duck Ecological System and the Evaluation of Its Economic Significance Agricultural Sciences in China, 5: 758-766 2006.

Legislation:

BRASIL. Presidência da República. Lei nº 4771, 15 de Setembro, 1965. Disponível em: www.planalto.gov.br. Acessado em: 26 de Agosto, 2005.

BRASIL. Presidência da República. Lei nº 5106, 2 de Setembro, 1966. Disponível em: www.planalto.gov.br. Acessado em: 26 de Agosto, 2005.

BRASIL. Assembléia Nacional Constituinte. Constituição da República Federativa do Brasil, 5 de Outubro de, 1988. Disponível em: www.planalto.gov.br. Acessado em: 29 de Agosto, 2005. BRASIL. Presidência da República. Decreto nº 97.628, 10 de Abril, 1989.

BRASIL. Governo do Estado de Minas Gerais. Lei nº 10.561, 27 de Dezembro, 1991. Disponível em: www.ief.mg.gov.br. Acessado em: 09 de Setembro, 2005. Constituição Federal do Brasil

http://www.planalto.gov.br/ccivil_03/constituicao/constitui%C3%A7ao.htm Lei Federal nº 4.771/1965 – Código Florestal

http://www.planalto.gov.br/ccivil_03/LEIS/L4771.htm

Lei Estadual de Mudanças Climáticas nº. 3.135 de 04 de junho de 2007. Manaus: Assembléia Legislativa. Disponível em < http://www.florestavivaamazonas.org.br/download/

Lei_est_n_3135_de_050607.pdf >

Lei Federal nº 6.938/1981 – Lei da Política Nacional do Meio Ambiente http://www.planalto.gov.br/ccivil_03/Leis/L6938.htm

Lei Federal nº 9.985/2000 – Sistema Nacional de Unidades de Conservação (SNUC)

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

Zinn, Y.L.; R. Lal; D.V.S. Resck. Changes in soil organic carbon stocks under agriculture in Brazil. Soil & Tillage Research, v. 84, p. 28-40, 2005.

http://www.planalto.gov.br/ccivil_03/Leis/L9985.htm 278

Lei Federal nº 11.284/2006 – Lei de Gestão de Florestas Públicas

http://www.planalto.gov.br/CCIVIL/_Ato2004-2006/2006/Lei/L11284.htm Lei Federal nº 11.428/2006 – Lei da Mata Atlântica

http://www.planalto.gov.br/ccivil_03/_Ato2004-2006/2006/Lei/L11428.htm Websites and electronic documents consulted:

http://www.cgee.org.br/publicacoes/MudancaDoClima.php http://www.acr.org.br/arquivos2/14.pdf

Technical Synthesis Report | Land Use, Land-Use Change, and Forestry

http://www.cepea.esalq.usp.br/economiaambiental/files/2005/04abr.pdf http://www.mct.gov.br

http://www.sectma.pb.gov.br/comites/rbma/pdf/cad3.pdf http://www.cnpma.embrapa.br

http://www.mct.gov.br/upd_blob/0012/12919.pdf

http://unfccc.int/meetings/cop_13/items/4049.php

http://www.bndes.gov.br/programas/agropecuarios/pronaf.asp http://www.bndes.gov.br/programas/agropecuarios/propflora.asp http://www.integracao.gov.br/fundos/fundos_constitucionais/index.asp http://www.mma.gov.br/sitio/index.php?ido=conteudo.monta&idEstrutura=51 http://www.mma.gov.br/sitio/index.php?ido=conteudo.monta&idEstrutura=5

WORLD BANK