ECONOMIC AND SECTOR WORK CARBON SEQUESTRATION IN AGRICULTURAL SOILS M AY 2 0 1 2 REPORT NUMBER: 67395-GLB ECONOMIC AND SECTOR WORK CARBON SEQUESTRATION IN AGRICULTURAL SOILS REPORT NO. 67395-GLB ARD AGRICULTURE AND RURAL DEVELOPMENT © 2012 International Bank for Reconstruction and Development/International Development Association or The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The �ndings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. 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. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Of�ce of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover Photos: Scott Wallace, Tran Thi Hoa, Curt Carnemark, Ami Vitale, and Ray Witlin. CONTENTS III TABLE OF CONTENTS List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v List of Photos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii List of Boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1: Food Security Under a Changing Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2: Carbon Bene�ts Through Climate-Smart Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3: Objectives and Scope of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Chapter 2: Soil Organic Carbon Dynamics and Assessment Methods . . . . . . . . . . . . . . . . . . 5 2.1: Soil Organic Carbon Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2: Carbon Assessment for Land Management Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.3: Techniques of Soil Carbon Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.4: Carbon Assessment in The World Bank’s Sustainable Land Management Portfolio . . . . . . . . . . . . . 17 Chapter 3: Meta-Analyses of Soil Carbon Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.2: Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.3: Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Chapter 4: Ecosystem Simulation Modeling of Soil Carbon Sequestration . . . . . . . . . . . . . . 43 4.1: Model Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.2: Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Chapter 5: Economics of Soil Carbon Sequestration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 5.1: Marginal Abatement Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 5.2: Trade-Offs in Soil Carbon Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 5.3: Implications of the Trade-Offs in Land-Use Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 5.4: Sustainable Land Management Adoption Barriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 5.5: Policy Options for Soil Carbon Sequestration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 EC O N O M I C A N D S E CT OR WORK IV C ONTENTS Appendix A: The Farming Practice Effect, Number of Estimates, and Features in Land Management Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Appendix B: General Scenario Assumptions and Application for World Regions . . . . . . . . . . 67 B.1: Baseline Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 B.2: Global Mitigation Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 B.3: Application to World Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 B.4: Detailed Modeling for Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Appendix C: Global Crop Yields (T ha −1 yr−1) Grouped into 25th, 50th, and 75th Percentile Bins Corresponding to Low, Medium, and High . . . . . . . . . . . . . . . . . . . . . . . 75 Appendix D: Uncertainty Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Appendix E: Assumptions for Deriving the Applicable Mitigation Area for the Land Management Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 E.1: Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 CARBON SEQUESTRATION IN AGRICULTURAL SOILS LI S T O F F I G U R E S V LIST OF FIGURES Figure E1: Abatement Rates of the Land Management Practices (t CO2e Per Hectare Per Year) . . . . . . . . . .xxii Figure E2: Trade-Offs Between Pro�tability and Carbon Sequestration of Sustainable Land Management Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxv Figure E3: Relationship Between Private Bene�ts and Public Costs . . . . . . . . . . . . . . . . . . . . . . . xxvi Figure 1.1: Contribution of Different Sectors to Greenhouse Gas (GHG) Emissions . . . . . . . . . . . . . . . . . 1 Figure 1.2: Proportion of Agricultural Land Derived from Different Land Covers in the Tropics, 1980–2000 . . . . . 2 Figure 2.1: Carbon Stocks in Biomass and Soils. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Figure 2.2: Global Soil Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 2.3: Factors Affecting Soil Carbon Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 3.1: Geographical Distribution of Carbon Sequestration Estimates . . . . . . . . . . . . . . . . . . . . . 21 Figure 3.2: Soil Carbon Sequestration and Precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 3.3: Soil Carbon Sequestration and Temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 3.4: Soil Carbon Sequestration and Soil Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Figure 3.5: Soil Carbon Sequestration and Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 3.6: Soil Carbon Sequestration and Application Levels of Nitrogen Fertilizer (Means and 95 Percent Con�dence Intervals, n = 285) . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 3.7: Soil Carbon Sequestration and Fertilizer Combinations (Means and 95 Percent Con�dence Intervals, n = 285) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 3.8: Mean Soil Carbon Sequestration and Levels of Residue Returned . . . . . . . . . . . . . . . . . . . 30 Figure 3.9: Classi�cation of Tillage Systems Based on Crop Residue Management . . . . . . . . . . . . . . . . 31 Figure 3.10: Mean Soil Carbon Sequestration and Cropping Intensity . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 3.11: Carbon Dioxide Abatement Rates of the Land Management Practices . . . . . . . . . . . . . . . . 39 Figure 4.1: Representation of the RothC Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Figure 4.2: The 12 Strata Used for Ecosystem Simulation Modeling . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure 4.3: Africa Agroecological Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Figure 4.4: A Screen Shot of the Soil Carbon Internet Database . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure 4.5: Cumulative Soil Carbon Loss by 2030 Assuming 15 Percent Residue Retention (t ha−1) under Different Cropping Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 EC O N O M I C A N D S E CT OR WORK VI LIST O F FIGUR ES Figure 4.6: Predicted Cumulative C Sequestration for Different Land Management Practices by 2030 . . . . . . 49 Figure 5.1: The Private Marginal Abatement Cost Curves. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Figure 5.2: Total Private Bene�ts (Blue) and Public Costs (Red) of Land Management Practices (US$, Billion) for the B1 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Figure 5.3: Trade-Offs Between Pro�tability and Carbon Sequestration of Sustainable Land Management Technologies in Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Figure 5.4: Relationship Between Private Bene�ts and Public Costs in Africa . . . . . . . . . . . . . . . . . . . 56 Figure B.1: FAO Land-Use Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 LIST OF PHOTOS Photo E.1: Terracing and Landscape Management in Bhutan . . . . . . . . . . . . . . . . . . . . . . . . . . . .xvii Photo E.2: Crop Residue Management in Irrigated Fields in Indonesia . . . . . . . . . . . . . . . . . . . . . . . xx Photo E.3: Water Management in a Field in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii Photo E.4: Maize Growing under Faidherbia Albida Trees in Tanzania . . . . . . . . . . . . . . . . . . . . . . . xxiv Photo E.5: Crop Harvesting in Mali. The Biomass Is Smaller Compared to that of Agroforestry Systems. . . . . .xxv Photo 3.1: Crop Residue Management in Irrigated Fields in Indonesia . . . . . . . . . . . . . . . . . . . . . . . 29 Photo 3.2: Water Management in a Field in India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Photo 3.3: Maize Growing under Faidherbia Albida Trees in Tanzania . . . . . . . . . . . . . . . . . . . . . . . . 34 Photo 3.4: Crop Harvesting in Mali. The Biomass Is Smaller Compared to that of Agroforestry Systems. . . . . . 40 Photo 5.1: Terracing and Landscape Management in Bhutan . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 CARBON SEQUESTRATION IN AGRICULTURAL SOILS LI S T O F TA B L E S V II LIST OF TABLES Table E1: Carbon Stocks in Vegetation and Top 1 Meter of Soils of World Biomes . . . . . . . . . . . . . . . . .xvii Table E2: Estimates of Erosion-Induced Carbon Emission Across World Regions. . . . . . . . . . . . . . . . . xviii Table E3: Technical Mitigation Potential, Private Bene�ts, and Public Costs of the Land Management Technologies by 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii Table E4: Relative Importance of Different Factors for Adopting Improved Land Management Practices . . . . xxviii Table 1.1: Improvement in Crop Yields Per Ton of Carbon in the Root Zone . . . . . . . . . . . . . . . . . . . . . 3 Table 1.2: Estimated Increase in Grain Crop Production From Land Management Technologies That Sequester Soil Carbon (Million Tons Per Year) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Table 2.1: Carbon Stocks in Vegetation and Top 1 Meter of Soils of World Biomes . . . . . . . . . . . . . . . . . . 6 Table 2.2: Global Carbon Budget (Gt C) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Table 2.3: Forms of Carbon in the Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Table 2.4: Soil Carbon Pool up to 1-Meter Deep for Soil Orders of the World’s Ice-Free Land Surface . . . . . . . . 8 Table 2.5: Estimate of Erosion-Induced Carbon Emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Table 2.6: Comparison of Carbon Assessment for Carbon Mitigation and Noncarbon Mitigation Projects . . . . . 13 Table 2.7: Direct and Indirect Methods of Soil Carbon Assessment. . . . . . . . . . . . . . . . . . . . . . . . . 14 Table 2.8: Characteristics of Emerging In Situ Methods of Soil Carbon Analytical Techniques . . . . . . . . . . . 14 Table 2.9: Comparative Features of Some Carbon Estimation Models . . . . . . . . . . . . . . . . . . . . . . . 15 Table 2.10: Components of Soil Carbon Monitoring at the Regional Scale . . . . . . . . . . . . . . . . . . . . . 16 Table 2.11: Carbon Accounting Systems and Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Table 3.1: Practices That Sequester Carbon in Forest, Grassland, and Cropland . . . . . . . . . . . . . . . . . . 19 Table 3.2: Nutrient Management and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . . . . . . . . 27 Table 3.3: Relative Importance of the Four Domains of Integration on Crop-Livestock Interaction . . . . . . . . . 28 Table 3.4: Tillage, Crop Residue Management, and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . 29 Table 3.5: Crop Rotation and Soil Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . . . . . . . . . . . . . . . . 32 Table 3.6: Water Management and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . . . . . . . . . 34 Table 3.7: Agroforestry and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . . . . . . . . . . . . . 35 Table 3.8: Land-Use Changes and Soil Carbon Sequestration Rates (kg C ha−1 yr−1). . . . . . . . . . . . . . . . . 36 EC O N O M I C A N D S E CT OR WORK VIII LIS T OF TA B LES Table 3.9: Summary of Observed Rates of Soil Carbon Sequestration (kg C ha−1 yr−1) as a Result of Land-Use Changes and Other Practices Relevant to Livestock Management . . . . . . . . . . . . . . 37 Table 3.10: Soil Amendments and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) . . . . . . . . . . . . . . . . 38 Table 4.1: Spatial Datasets Used in the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Table 4.2: Modeled Cumulative Soil Carbon Sequestration Potential by 2030 (Mt C) under Different Land Management Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Table 5.1: Private Savings of Different Technologies Per Ton of Carbon Dioxide Sequestered . . . . . . . . . . . 53 Table 5.2: Public Costs of Different Technologies Per Ton of Carbon Dioxide Sequestered. . . . . . . . . . . . . 53 Table 5.3: Technical Mitigation Potential, Private Bene�ts, and Public Costs of the Land Management Technologies by 2030 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Table 5.4: Relative Importance of Different Factors for Adopting Improved Land Management Practices . . . . . 60 Table 5.5: Interventions for Facilitating Increased Input Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Table B.1: Agricultural Systems and Mitigation Scenario in South America . . . . . . . . . . . . . . . . . . . . . 69 Table B.2: Agricultural Systems and Mitigation Scenario in Central America . . . . . . . . . . . . . . . . . . . . 69 Table B.3: Manure C Inputs for the AEZs in Africa Based on FAOSTAT . . . . . . . . . . . . . . . . . . . . . . . 70 Table B.4: C Inputs for Different Green Manure/Cover Crop Systems. . . . . . . . . . . . . . . . . . . . . . . . 72 Table B.5: C Inputs for Different Agroforestry Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Table D.1: Uncertainty Analyses Using Random Samples from the Mitigation Scenarios. . . . . . . . . . . . . . 82 Table E.1: Estimated Cropland Area in the 2000s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Table E.2: Estimated Cropland and Grassland Area by 2030 (Million Hectare). . . . . . . . . . . . . . . . . . . . 84 LIST OF BOXES Box 2.1: Brief Description of Soil Orders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Box 2.2: Sustainable Land Management Practices Reverse Soil Carbon Loss in Java. . . . . . . . . . . . . . . . 11 Box 5.1: Risk-Related Barriers to Adoption of Soil Carbon Sequestration Activities . . . . . . . . . . . . . . . . . 59 CARBON SEQUESTRATION IN AGRICULTURAL SOILS PR E FA C E IX PREFACE Agriculture’s direct reliance on the natural resource base has always been a de�ning characteristic of the sector. Production relies directly on soil, water, and a variety of biological processes. And it also relies on the climate at the same time that its role in the global carbon cycle makes it a major contributing factor to climate change. Today, more than ever before, we understand not only the signi�cance that climate has for agriculture, but also the enormous signi�cance that agriculture has for the climate. The growing consensus on the need for a climate-smart agriculture emerged largely out of international awareness of the sector’s negative impacts—its ecological footprint. It also grew out of the recognition that conventional forms of agricul- tural production are often unsustainable and deplete or “mine� the natural resources on which production relies over time. Agriculture is the world’s leading source of methane and nitrous oxide emissions, a substantial source of carbon emissions, and the principal driver behind deforestation worldwide. Some 30 percent of global greenhouse gas emissions are attributable to agriculture and deforestation driven by the expansion of crop and livestock production for food, �ber and fuel. More recently, this perspective of agriculture as a source of greenhouse gas emissions and pollution has become more balanced, with a growing understanding of the environmental services the sector can provide if production is well-managed. While agriculture emits a large volume of greenhouse gases, its biomass and especially its soils also sequester carbon out of the atmosphere, and this role as a carbon sink and as a carbon store can be strategically optimized through proven farming techniques and methods that simultaneously reduce emissions. These technical elements of climate-smart agriculture are by now well understood, and in addition to their technical feasibility, they can be highly productive and pro�table. As this document will discuss, this new and more sustainable pattern of agricultural development can make the sector an ac- tive agent in climate change mitigation at the same time that it improves and builds upon the sector’s capacity to adapt to the increasing temperatures and declining precipitation that are already reducing yields of grains and other primary crops in many parts of the vast semi-arid tropics where so many of the poorest reside. This trend is projected to intensify in the coming de- cades and have serious rami�cations for global food security, and for the food security of vulnerable populations in particular. Agricultural production operates under intensifying pressures. Food production will need to effectively double in many devel- oping countries by 2050 to feed a growing and increasingly urban global population. The agriculture systems that supply this food play a pivotal role in these countries’ economies. Agriculture employs up to two-thirds of their workforce and accounts for between 10 and 30 percent of their gross domestic product. Increasing productivity is agriculture’s most pressing priority, but it is not its only priority. Perhaps the most important point conveyed in this document is that the dual roles of agriculture as a source of food security and as a source of environmental services converge in fundamental ways. Too often the relationship between these roles is viewed as a series of painful trade-offs. Yet the same carbon that is sequestered through sustainable practices makes those practices more productive. The carbon that is removed from the atmosphere and captured in soils and plant biomass is the same carbon that makes agricultural soils more fertile, and that leads to higher pro�t margins for producers. Higher carbon content enables the soil to make more water and nutrients available to support crop growth, and increases the resilience of farmland, reducing both the need for fertilizer applications and susceptibility to land degradation. The Intercontinental Panel on Climate Change (IPCC) indicates that carbon sequestration accounts for about 90 percent of global agricultural mitigation potential by 2030. EC O N O M I C A N D S E CT OR WORK X PR EFA C E While technical progress in the area of integrated “landscape� approaches to managing natural and economic resources has been very promising, the adoption of these approaches still faces serious constraints in many developing countries. Among the most important of these constraints are the signi�cant upfront expenditures that many of the newer techniques require. In many of the developing countries in which these techniques would wield some of their most important bene�ts, aware- ness of both the techniques and the bene�ts remain limited. In some settings there is limited capacity to implement them even when people are aware of them. Mobilizing and targeting resources to overcome these constraints has been an important reason the World Bank became determined to get climate-smart agriculture more �rmly onto the agenda of the international dialogue on climate change. It is our hope that this report moves that agenda forward by making the “triple win� of soil carbon sequestration for increased productivity, improved climate resilience, and enhanced mitigation an integral part of that dialogue. Juergen Voegele Director Agriculture and Rural Development Department The World Bank CARBON SEQUESTRATION IN AGRICULTURAL SOILS ACK N O W L E D G ME NT S XI ACKNOWLEDGMENTS The preparation of this report was managed by the Agriculture and Rural Development (ARD) department. Ademola Braimoh wrote the report with meta-analyses and research support from Idowu Oladele, Louis Lebel, and Ijeoma Emenanjo. Matthias Seebauer, Patricia del Valle Pérez, and Katia Obst carried out the ecosystem simulation modeling, while Reza Firuzabadi, Michael Kane, Varuna Somaweera, Dany Jones, Sarah Elizabeth Antos, Katie McWilliams, and Alex Stoicof provided Geographical Information System and Information Technology support. The author is grateful for constructive comments and suggestions from the following peer reviewers: Erick Fernandes, Johannes Woelcke, Yurie Tanimichi Hoberg, Chuck Rice, John Idowu, Ellysar Baroudy, Johannes Heister, Wilhelmus Janssen, Christine Negra, Louis Bockel, Tim Searchinger, Meine van Noordwijk, and Andreas Wilkes. Many others provided inputs and support including Jurgen Voegele, Mark Cackler, Fionna Douglas, Marjory-Anne Bromhead, Patrick Verkoijen, Pai-Yei Whung, Dipti Thapa, Gunnar Larson, Maria Gabitan, Olusola Ikuforiji, Sarian Akibo-Betts, Ramon Yndriago, Kaisa Antikainen, Cicely Spooner, Shunalini Sarkar, and Genalinda Gorospe. This report improves the knowledge base for scaling-up investments in land management technologies that sequester soil carbon for increased productivity under changing climate conditions. EC O N O M I C A N D S E CT OR WORK ABB R E V I AT I O N S X III ABBREVIATIONS AEZ Agroecological Zone HUM humi�ed organic matter BIO microbial biomass INS inelastic neutron scattering CBP Carbon Bene�ts Project IOM inert organic matter CSA climate-smart agriculture IPCC Intercontinental Panel on Climate Change DPM decomposable plant material LIBS laser-induced breakdown spectroscopy EX-ACT Ex Ante Appraisal Carbon-Balance Tool MAC marginal abatement cost FAOSTAT Food and Agriculture Organization of the MMV measurement, monitoring, and veri�cation United Nations NPP net primary productivity GEF Global Environment Facility RPM resistant plant material GHG greenhouse gas SALM Sustainable Agricultural Land Management GIS geographical information system SLM sustainable land management GPS global positioning system UNFCCC UN Framework Convention on Climate ha hectare Change EC O N O M I C A N D S E CT OR WORK EX E C U T I V E S U MMARY XV EXECUTIVE SUMMARY Ensuring food security in a context of growing population and changing climate is arguably the principal challenge of our time. The current human population of 7 billion will increase to more than 9 billion by 2050. Moreover, rising incomes and the increasing proportion of the global population living in urban areas are changing the compo- sition of food demand in fundamental ways. Higher income urban populations have more diverse diets that feature a variety of high-value food sources, such as livestock that are more resource intensive to produce and process. This adds to the chal- lenge of maintaining and preserving the resilience of both natural and agricultural ecosystems. Based on these developments, projections indicate that global food production must increase by 70 percent by 2050. In many African countries, where the challenge is most acute, food production must increase by more than 100 percent—it must effectively double. The onus of this challenge falls on agriculture, which is the sector of the global economy that is most vulnerable to the effects of global warming, such as more variable rainfall and more extreme weather-generated events. At the same time, agriculture and the changes in land-use that are associated with it, are one of the principal contributors to climate change, accounting for one-third of global greenhouse gas (GHG) emissions. Projected increases in demand for food and bioenergy by 2050 have profound implications for the pressure that agriculture wields on forests and other natural ecosystems in the tropics. These ecosystems are vital, both in the role their biomass plays in sequestering carbon and in providing habitat for biodiversity. When they are lost, they become a massive source of GHG emissions. Increasing agricultural productivity, enhancing its resilience to climate change, and reducing the emissions that come from the agriculture sector are therefore triple imperatives that require alternative sets of practices. Climate-smart agriculture (CSA) seeks to increase productivity in an environmentally and socially sustainable way, strengthen farmers’ resilience to climate change, and reduce agriculture’s contribution to climate change by reducing GHG emissions and sequestering carbon. A key element of CSA is sustainable land management (SLM), involving the implementation of land-use systems and management practices that enable humans to maximize the economic and social bene�ts from land while maintaining or enhancing the ecosystem services that land resources provide. Because soil is the basic resource in agricultural and forest land use, it is the central element of most SLM technologies. Soil carbon has a direct correlation with soil quality. It is a major determinant of the soil’s ability to hold and release water and other nutrients that are essential for plants and their root systems to grow. Soil carbon also plays an important role in maintaining the biotic habitats that make land management systems sustainable, resilient, and able to resist degradation. Carbon seques- tration, the process by which atmospheric carbon dioxide is taken up by plants through photosynthesis and stored as carbon in biomass and soils, can help reverse soil fertility loss, limit GHG concentrations in the atmosphere, and reduce the impact of climate change on agricultural ecosystems. The objective of this report is to improve the knowledge base that informs investment decisions in land management tech- nologies that purposefully sequester soil carbon. The �ndings reported are based on three exercises. The �rst was a review of soil carbon dynamics and assessment methods and a meta-analysis of soil carbon sequestration rates in Africa, Asia, and Latin America. The second exercise was to apply an ecosystem simulation modeling technique to predict future carbon storage in global cropland soils. The third consisted of a series of estimations of marginal abatement costs and trade-offs to assess the cost-effectiveness of deploying the land management technologies for climate-smart agriculture. The results EC O N O M I C A N D S E CT OR WORK XV I EX EC UTIV E S UM M A RY reported in this document complement a number of related publications, including empirical lessons from recent project examples and policy briefs that were used as inputs at the Durban Climate Change Conference in November 2011. At least four key messages emerge over the course of this report, and these relate to pro�tability, managing trade-offs, barriers to adoption, and the need for targeted public support. Pro�tability In addition to storing soil carbon, sustainable land management technologies can be bene�cial to farmers because they can increase yields and reduce production costs. Total private pro�ts by the year 2030 are estimated at US$105 billion for Africa, $274 billion for Latin America, and $1.4 trillion for Asia. Maximizing Bene�ts and Managing Trade-Offs Soil carbon sequestration can be maximized by managing trade-offs across space, time, and sectors. Working at the land- scape level is useful for addressing food security and rural livelihood issues and in responding to the impacts of climate change and contributing to its mitigation. Barriers to Adoption and Up-Front Costs The adoption of sustainable land management practices can face a variety of socioeconomic and institutional barriers. These include the need for signi�cant up-front expenditures on the part of poorer farmers, the nonavailability of some inputs in the local markets, lack of information about the potential of improved techniques, and often limited capacity to implement the techniques. Certain techniques associated with sustainable land management can be incompatible with traditional practices. In some instances, the diffusion of new technologies relies on a level of social capital and experience with collective action that farmers simply do not yet have. The Need for Targeted Public Support Without public support for farmers, poor agricultural land management will intensify land degradation, increase farmers’ vul- nerability to the effects of climate change, and lead to the emission of additional GHGs into the atmosphere. The amount of support that governments will need to provide by the year 2030 to enable farmers to implement SLM practices are projected at US$20 billion in Africa, $41 billion in Latin America, and $131 billion in Asia. Mechanisms for Carbon Enhancement in Agro-Ecosystems Sustainable land management delivers carbon bene�ts in three important ways. The �rst is carbon conservation, in which the large volumes of carbon stored in natural forests, grasslands, and wetlands remain stored as carbon stocks. Conserving this terrestrial carbon represents a “least-cost opportunity� in terms of climate change adaptation and mitigation and is essential to increasing the resilience of agricultural ecosystems. The second bene�t is carbon sequestration, in which the growth of agricultural and natural biomass actively removes carbon from the atmosphere and stores it in soil and biomass. The third bene�t delivered by SLM is to reduce the emissions of GHGs that emanate from agricultural production, including those emissions that result from land-use change in which carbon stocks become carbon sources as agricultural production expands into natural ecosystems. SLM practices are alternatives to conventional agriculture in all three of these paths—conservation, sequestration, and reductions in GHG emissions. While it capitalizes more purposefully on the positive impacts of conservation and sequestration, its reversal of agriculture’s negative impacts also presents profound contrast with conventional practices. These conventional agricultural prac- tices include deforestation, the burning of biomass, draining of wetlands, uncontrolled grazing, and plowing and other forms of soil disturbance that release not only carbon dioxide into the atmosphere, but also nitrous oxide and methane—GHGs with extremely high impacts on global warming. Investment in soil quality improvement practices such as erosion control, water management, and judicious application of fertilizers can reduce these emissions directly and increase rates of soil carbon sequestration. The Dynamics of Soil Organic Carbon Different ecosystems store different amounts of carbon depending on their species compositions, soil types, climate, relief, and other biophysical features. (Globally, volumes of carbon are generally measured in gigatonnes [Gt], which is equal to CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY X V II PHOTO E.1: Terracing and Landscape Management in Bhutan Source: Curt Carnemark/World Bank. 1 billion tons, or metric tons in the United States.) The amount of carbon stored in plant biomass ranges from 3 Gt in croplands to 212 Gt in tropical forests (table E1). Soils hold more carbon than plant biomass (or vegetation) and account for 81 percent of the world’s terrestrial carbon stock. Soil carbon stocks also vary by ecosystem, ranging, for instance, from 100 Gt in temper- ate forests to 471 Gt in boreal forests. Boreal ecosystems are a particular concern. Because much of the soil organic carbon stored there is permafrost and wetlands, any large-scale melting caused by global warming will release massive volumes of carbon into the atmosphere. Conservation and protection are therefore widely recognized as major priorities, with the excep- tion of limited areas selected for forest management. TABLE E1: Carbon Stocks in Vegetation and Top 1 Meter of Soils of World Biomes CARBON STOCKS (Gt C) AND PROPORTION IN THE ECOSYSTEM (%) AREA PROPORTION PROPORTION BIOMES (MILLION km2) VEGETATION (%) SOILS (%) TOTAL Tropical forests 17.6 212 49.5 216 50.5 428 Temperate forests 10.4 59 37.1 100 62.9 159 Boreal forests 13.7 88 15.7 471 84.3 559 Tropical savannas 22.5 66 20.0 264 80.0 330 Temperate grasslands 12.5 9 3.0 295 97.0 304 Deserts 45.5 8 4.0 191 96.0 199 Tundra 9.5 6 4.7 121 95.3 127 Wetlands 3.5 15 6.3 225 93.8 240 Croplands 16 3 2.3 128 97.7 131 Total 151.2 466 2,011 2,477 Proportion (%) 19 81 100 Source: Watson, Robert, et al. (2000). EC O N O M I C A N D S E CT OR WORK XV I I I EX EC UTIV E S UM M A RY TABLE E2: Estimates of Erosion-Induced Carbon Emission Across World Regions SOIL CARBON DISPLACED BY EMISSION (20 PERCENT OF GROSS EROSION EROSION (2 TO 3 PERCENT OF DISPLACED SOIL CARBON; REGION (Gt/YEAR) SEDIMENT; Gt C/YEAR) Gt C/YEAR) Africa 38.9 0.8–1.2 0.16–0.24 Asia 74.0 1.5–2.2 0.30–0.44 South America 39.4 0.8–1.2 0.16–0.24 North America 28.1 0.6–0.8 0.12–0.16 Europe 13.1 0.2–0.4 0.04–0.08 Oceania 7.6 0.1–0.2 0.02–0.04 Total 201.1 4.0–6.0 0.8–1.2 Source: Lal, R. (2003). The global carbon cycle describes the transfer of carbon in the earth’s atmosphere, vegetation, soils, and oceans. The two most important anthropogenic processes responsible for the release of carbon dioxide into the atmosphere are the burning of fossil fuels (coal, oil, and natural gas) and land use. Rapidly growing emissions are outpacing the growth in natural sinks (lands and oceans). The ef�ciency of oceans and lands as carbon dioxide sinks has declined over time. These sinks currently remove an average of 55 percent of all anthropogenic carbon dioxide emissions; 50 years ago they removed 60 percent. Soils are critically important in determining global carbon cycle dynamics because they serve as the link between the atmo- sphere, vegetation, and oceans. Globally, the soil carbon pool (also referred to as the pedologic pool) is estimated at 2,500 Gt up to a 2-m depth. Out of this, the soil organic carbon pool comprises 1,550 Gt, while the soil inorganic carbon and elemental pools make up the remaining 950 Gt (Batjes 1996). The soil carbon pool is more than 3 times the size of the atmospheric pool (760 Gt) and about 4.5 times the size of the biotic pool (560 Gt). The soil organic carbon pool represents a dynamic balance between gains and losses. The amount changes over time de- pending on photosynthetic C added and the rate of its decay. Under undisturbed natural conditions, inputs of carbon from litter fall and root biomass are cycled by output through erosion, organic matter decomposition, and leaching. The potential carbon sequestration is controlled primarily by pedological factors that set the physico-chemical maximum limit to storage of carbon in the soil. Such factors include soil texture and clay mineralogy, depth, bulk density, aeration, and proportion of coarse fragments. Attainable carbon sequestration is determined by factors that limit the input of carbon to the soil system. Net primary produc- tivity (NPP)—the rate of photosynthesis minus autotrophic respiration—is the major factor influencing attainable sequestra- tion and is modi�ed by above-ground versus below-ground allocation. Land management practices that increase carbon input through increasing NPP tend to increase the attainable carbon sequestration to nearer to the potential level. Climate has both direct and indirect effects on attainable sequestration. Decomposition rate increases with temperature but decreases with increasingly anaerobic conditions. Actual carbon sequestration is determined by land management factors that reduce carbon storage such as erosion, tillage, residue removal, and drainage. Theoretically, the potential soil carbon sequestration capacity is equivalent to the cumulative historical carbon loss. However, only 50 to 66 percent of this capacity is attainable through the adoption of sustainable land management practices. The current rate of carbon loss due to land-use change (deforestation) and related land-change processes (erosion, tillage operations, biomass burning, excessive fertilizers, residue removal, and drainage of peat lands) is between 0.7 and 2.1 Gt carbon per year. Soil erosion is the major land degradation process that emits soil carbon. Because soil organic matter is concentrated on the soil surface, accelerated soil erosion leads to progressive depletion of soil carbon. The annual rate of soil loss ranges from 7.6 Gt for Oceania to 74.0 Gt for Asia (table E2). This corresponds to carbon emissions ranging from 0.02 to 0.04 Gt per year for Oceania to 0.30 to 0.44 Gt per year for Asia. Globally, 201 Gt of soil is lost to erosion, corresponding to 0.8 to 1.2 Gt of emitted carbon per year. Africa, Asia, and South America emit between 0.60 and 0.92 Gt of carbon per year through soil erosion. Agricultural soils must be prevented from being washed into streams and rivers where the relatively stable soil carbon pools are rapidly oxidized to carbon dioxide. CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY X IX Soil respiration, the flux of microbially and plant-respired carbon dioxide, estimated at 75 to 100 Gt carbon per year, is the next largest terrestrial carbon flux following photosynthesis. Soil respiration is a potentially important mechanism of positive feedback to climate change. A small change in soil respiration can signi�cantly alter the balance of atmospheric carbon dioxide concentration compared to soil carbon stores. Conventional tillage leads to the destruction of soil aggregates, excessive respiration, and soil organic matter decomposition, leading to reduced crop production and decreased resilience of the soil ecosystem. When other factors are at optimum, conservation tillage, use of cover crops (green manure), crop rotations, use of deep-rooted crops, application of manure, and water management can optimize soil respiration in addition to improving soil carbon leading to the triple win of enhanced agricultural productivity, adaptation, and mitigation. Approaches to Soil Carbon Assessment Soil carbon assessment in different parts of the world requires methods that are appropriate to the circumstances. The variety of methods that have been developed and tested for use in different countries raises concerns about their comparability. Ensuring this comparability warrants serious international priority. In the case of carbon projects, credible and cost-effective techniques of monitoring changes in soil carbon still need to be developed. Soil carbon assessment methods can be broadly classi�ed into direct and indirect methods, depending on whether carbon content in soil samples is directly measured or inferred through a proxy variable. The most established type of direct soil carbon assessment entails collecting soil samples in the �eld and analyzing them in the laboratory using combustion techniques. Field sampling is technically challenging, but most of its challenges can be addressed through an appropriate design that accounts for soil spatial variation. The degree and nature of sampling depend on the objectives of the carbon assessment objective, whether, for instance, the assessment is used for national or regional accounting or for a carbon offset project. Each context will require a differing degree of granularity and measurement set to assess uncertainty in the estimates. Direct methods are more precise and accurate but also more time and labor intensive as well as very expensive. Some in situ soil carbon analytical methods are being developed with the objective of offering increased accuracy, precision, and cost-effectiveness over con- ventional ex situ methods. The in situ soil carbon analytical methods include mid-infrared (IR) spectroscopy, near-IR spectros- copy, laser-induced breakdown spectroscopy (LIBS), and inelastic neutron scattering (INS). While LIBS and INS technologies are still in their infancy, IR spectroscopy has proven valuable in developing soil spectral libraries and for rapid characterization of soil properties for soil quality monitoring and other agricultural applications in developed and developing countries. Indirect estimation of soil organic carbon changes over large areas using simulation models has become increasingly im- portant. Indirect methods are needed to �ll knowledge gaps about the biogeochemical processes involved in soil carbon sequestration. One of the more important indirect methods involves the use of simulation models that project changes in soil organic carbon under varying climate, soil, and management conditions. Although simulation models can have limited accuracy, particularly in the context of developing countries in which land resources data are scarce, they are a cost-effective means of estimating GHG emissions in space and time under a wide range of biophysical and agricultural management condi- tions. The data can be particularly useful in scaling-up site-speci�c information to larger scales of magnitude. Monitoring and verifying soil carbon sequestration at the project or regional scale require �ve activities. These include selec- tion of landscape units suitable for monitoring soil carbon changes, development of measurement protocols, use of remote sensing to estimate soil organic carbon controlling parameters, spatially explicit biogeochemical modeling, and scaling-up the results to the entire project area. The selection of landscape monitoring units is based on the responsiveness of the area to land management practices as determined by climate, soil properties, management history, and availability of historical data. Protocols for temporally repeated measurements at �xed locations will generally include strati�cation and selection of sampling sites, sampling depth and volume, measurement of bulk density, laboratory analyses, other ancillary �eld measure- ments, and estimation of the marginal cost of carbon sequestration. Remote sensing can provide information on net primary productivity, leaf area index, tillage practices, crop yields and location, and amounts of crop residues. All of this is critical information used for input into models. Recently, the cellulose absorption index, derived from remote imaging spectroscopy, has been used to infer tillage intensity and residue quantity. These param- eters are fed into biogeochemical models to predict soil carbon sequestration. Scaling-up to larger areas requires integration from a variety of sources including �eld measurements, existing databases, models, geographical information systems, and remote sensing. Multitemporal moderate resolution remote sensing such as the Landsat Thematic Mapper and Moderate EC O N O M I C A N D S E CT OR WORK XX EX EC UTIV E S UM M A RY Resolution Imaging Spectroradiometer can provide information such as land-use and land-cover change, crop rotations, and soil moisture, which can markedly improve our ability to scale-up soil carbon assessments. Monitoring trends in soil carbon over a large geographical area through repeated sampling is, for the most part, restricted to in- dustrialized countries and a handful of developing countries. Examples of national carbon accounting system and tools include Australia’s National Carbon Accounting System; Canada’s National Forest Carbon Monitoring, Accounting, and Reporting System; Indonesia’s National Carbon Accounting System; and New Zealand’s Carbon Accounting System. The Agriculture and Land Use National Greenhouse Gas Inventory Software tool was recently developed by Colorado State University to support countries’ efforts to understand current emission trends and the influence of land-use and manage- ment alternatives on future emissions. The tool can be used to estimate emissions and removals associated with biomass C stocks, soil C stocks, soil nitrous oxide emissions, rice methane emissions, enteric methane emissions, and manure meth- ane and nitrous oxide emissions, as well as non-CO2 GHG emissions from biomass burning. PHOTO E.2: Crop Residue Management in Irrigated Fields in Indonesia Source: Curt Carnemark/World Bank. The Food and Agriculture Organization of the United Nations has developed the Ex Ante Appraisal Carbon-Balance Tool (EX-ACT) to assess GHGs in the agricultural sector. EX-ACT can provide ex ante assessments of the impact of agriculture and related forestry, �sheries, livestock, and water development projects on GHG emissions and carbon sequestration, thereby indicating the overall effects on the carbon balance. A detailed analysis of lessons learned in testing EX-ACT in World Bank agriculture projects can be found in a separate report. The BioCarbon Fund of the World Bank has also developed a methodology to encourage adoption of sustainable land man- agement practices by small-scale farmers in developing countries. The methodology, referred to as Sustainable Agricultural Land Management (SALM), provides a protocol for quantifying carbon emissions and removals and includes guidelines for identifying baseline scenario and assessing additionality in all carbon pools relevant to sustainable land management projects. Factors Affecting Soil Carbon Sequestration Climate signi�cantly influences large-scale patterns of soil carbon sequestration. In this study, irrespective of land manage- ment practices, higher sequestration rates were observed in the wettest locations with annual precipitation above 1,500 mm. CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY XXI There was also a trend to lower sequestration rates in the coolest (mean annual temperature less than 20°C) and warmest (mean annual temperature greater than 30°C) conditions. Sites in warmer and middle temperature regions tend to accumulate soil carbon more rapidly than those in colder regions, while semi-humid areas have higher sequestration rates than their semi-arid counterparts. Soil type is signi�cant to soil carbon sequestration as well. Soils with higher clay content sequester carbon at higher rates. In Africa and Latin America, carbon sequestration rates and variability are highest on inceptisols—relatively young soils that con- stitute about 9 percent of soils in the tropics. In Asia, the highest sequestration rates and variability are observed in oxisols, formed principally in humid tropical zones under rain forest, scrub, or savanna vegetation. Oxisols comprise about 24 percent of tropical land mass and are typically found on old landscapes that have been subject to shifting cultivation for some time. Timing is another factor that warrants careful consideration when introducing improved land management practices that increase carbon sequestration. Most of the potential soil carbon sequestration takes place within the �rst 20 to 30 years of adopting improved land management practices. The patterns of change in sequestration rates are nonlinear and differ between major types of practices. With most practices, the highest rates of sequestration are achieved in the intermediate term, with lower or even negative rates in the short term.1 Greenhouse Gas Mitigation by Sustainable Land Management Technologies The climate bene�ts of sustainable land management technologies are measured by the net rate of carbon sequestration adjusted for emissions associated with the technologies—a measurement referred to as the abatement rate. The emissions associated with the technologies are classi�ed as land emissions and process emissions. Land emissions are the differences between emissions of nitrous oxides and methane by conventional and improved practices. Process emissions are those arising from fuel and energy use. The abatement rate is expressed in tons of carbon dioxide equivalent (t CO2e) per hectare (ha) per year. Increases in productivity from nitrogen fertilizers need to be considered against the increased emission of GHGs from soils as well as the energy-related emissions associated with the fertilizer’s production and transport. In Latin America, the abatement rate of inorganic fertilizer is −0.23 t CO2e per ha per year compared to 0.13 t CO2e per ha per year for Asia and 0.29 t CO2e per ha per year for Africa. The greenhouse mitigation of manure is much higher at about 2.2 to 2.7 t CO2e per ha per year across the regions. No-tillage and residue management generated abatement rates ranging from 0.9 to 3.5 t CO2e per ha per year across the three regions. These rates represent the marginal carbon bene�t of mulching or incorporating residues relative to burning, grazing, and removal of the residues for other uses. Commonly applied residues on croplands include biomass from trees, sugarcane, rice, and other grain crops. Cover crops and crop rotation are key complementary practices for successful implementation of no-tillage. Cover crops improve soil quality by increasing soil organic carbon through their biomass, and they also help in improving soil aggregate stability and protecting the soil from surface runoff. Crop rotation is the deliberate order of speci�c crops sown on the same �eld. The succeeding crop may be of a different species (e.g., maize or sorghum followed by legumes) or a variety from the previous crop, and the planned rotation may be for 2 or more years. GHG abatements of cover crops were 1.7 to 2.4 t CO2e per ha per year, while those of crop rotation were 0.7 to 1.5 t CO2e per ha per year. There is a tendency toward higher carbon sequestration rates in triple cropping systems, although variation is high. Differences in soils, climate, and cropping systems also affect carbon sequestration under crop rotation. Supplemental irrigation and water harvesting are needed to minimize production risks in dry land agriculture. They also sequester carbon in the soil. Improved irrigation generated low to moderately high abatement rates (0.2 to 3.4 t CO2e per ha 1 The World Bank has posted a useful geographical information system tool on the Internet that summarizes the results of a series of ecosystem modeling exercises (see http://www-esd.worldbank.org/SoilCarbonSequestration/). The tool comprises several land man- agement scenarios reflecting situations typically encountered in agricultural projects. The Internet GIS database provides per-hectare estimates of soil carbon sequestration under different land management practices for a period of 20 to 25 years. Information on carbon sequestration potential of a location can be derived by point-and-click or by searching using place names. Users can download data from the database and integrate them with other GIS information to estimate soil carbon stock changes for different agricultural projects. EC O N O M I C A N D S E CT OR WORK XXI I ch ch ro ch e e m ta em im mic tio i pr al no ica n ca -2 0 2 4 6 8 10 12 14 16 ov fe o l fe in l f 0 2 4 6 8 10 12 te er 0 2 4 6 8 10 12 14 16 18 di ed rtili re int r re rtil ns til ve irr ze s e d iz ifi ize r ig r ap idue nsi uce er ro ca r in sify ati pl f ta re tio te tio Source: This study. ro on ica ma y ro d ti Asia ns t n sid n Africa tio na tat ll gr r if at n g io di m ue ve u s as edu y ro ion re of em n rs lch sla ce ta du m en ifi e nd d- tio ce ulc t c s d h Latin America -t gr n no atio no o-p azin cr co -gra es co til n or lant g op ve zin ot ve lag re ati -to r c g he rc e du on -g ro rs ro an im ra ps o m ps co ced pr ss nu v ov m lan pa il a t an st m er ure al- er till ed a d to cro -p n ur en ra e d cin re er ps an in irri ure sid en in nu clu ga w imp em g at ro en ue n te al- de tio er v ts m m ial ns to t n i -p re h e an an cr in arv me u (t CO2e Per Hectare Per Year) pa w ve v ere es os t e nt st age re at e n e ur m er ge ni in s s rcr stin clu lo op g e- en ha ta al in to-f t b de pe pi in rve les clu or te rs tre tre /bar ng e rc tin in de t st cr r e- es rie cr i rs te re op bi opp g op n r f pa c cr es -to ofe ing -p r af far ield st rop opp fo m ur cr lan tiliz a re in e- -to- ing op ta er s to fo -to tio im lley tat g -p r -fo n pr fa ion lan est ov rm ta r ed in ti bi est oc fa g bi on ha ll FIGURE E1: Abatement Rates of the Land Management Practices oc ha r bi ow oc r ha r CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX EC UTIV E S UM M A RY EX E C U T I V E S U MMARY X X III per year). Process and land emissions under irrigation can signi�cantly offset gains from carbon sequestration. Apart from energy-related emissions, a critical issue for soil carbon sequestration activities in irrigated areas is reduced emissions of methane from rice �elds. Mid-season drainage is a viable practice to reduce such emissions. The GHG abatement of water harvesting, the process of concentrating runoff from a larger area for use in a smaller target area, averaged 3.9 to 4.8 t CO2e per ha per year. Terracing and construction of slope barriers on sloping lands for soil and water conservation produced abate- ments of 2.4 to 5.3 t CO2e per ha per year. PHOTO E.3: Water Management in a Field in India Source: Ray Witlin/World Bank. Abatement rates of agroforestry systems, integrated land-use systems combining trees and shrubs with crops and livestock, are fairly high. This is due to the relatively large time-averaged biomass of trees compared to crops. The average abatement rates in t CO2e per ha per year are 7.6 for alley farming (the growing of crops simultaneously in alleys of perennial, preferably leguminous trees or shrubs), 7.5 for tree-crop farming, 8.7 for improved fallow (involving the use of fast-growing trees to ac- celerate soil rehabilitation), 4.6 to 6.3 for intercropping (the growing of crops near existing trees), and 4.3 to 6.7 for croplands where trees are introduced. The impacts of land-use changes on tree-based systems are also relatively large. Conversion of cropland to forest or pasture to plantation resulted in an abatement of 6.7 to 7.5 t CO2e per ha per year, while conversion of cropland to plantation gener- ated an abatement of 5.7 t CO2e per ha per year. Pasture improvement generated an abatement of 3.21 t CO2e per ha per year, whereas conversion of cropland to grassland produced GHG mitigation of 2.6 t CO2e per ha per year. By de�nition, most of the potential impact of changes in agricultural practices on carbon stocks is below ground. However, land-use changes away from cropland to agroforestry or plantations provide more convincing examples where it is useful to think of both above- and below-ground sequestration rates at the same time and possible trade-offs or interactions between them. Application of biochar, on average, resulted in the highest overall GHG abatement rate (10.3 to 15.7 t CO2e per ha per year), but its impact on crop productivity and soil resilience is still uncertain. In general, biochar production should not deplete the soil of the crop residues needed to protect against erosion and increase soil resilience. Decisions to adopt any of the land management practices should not be based solely on their respective climate mitigation bene�ts. Rather, they should be based on whole farm systems analysis that comprehensively assesses the productivity, EC O N O M I C A N D S E CT OR WORK XXI V EX EC UTIV E S UM M A RY PHOTO E.4: Maize Growing under Faidherbia Albida Trees in Tanzania Source: World Agroforestry Centre. on-farm resource use, and environmental load of the system. Farm-scale management decisions, taken within a wider socio- economic context, particularly the influence of public policy and markets, will most likely generate optimum social bene�ts. Pro�tability of Soil Carbon Sequestration In addition to storing soil carbon, sustainable land management technologies can be bene�cial to farmers by increasing yields and reducing production costs. Increases in crop yields derive from the ability of the land management technologies to maintain soil organic matter and biological activity at levels suitable for soil fertility. The pattern of increase in yield, however, varies from crop to crop. The pro�tability of no-tillage systems results mainly from the reduced labor requirement for seedbed preparation and other tillage operations compared to conventional tillage systems. In Zambia, yields have doubled for maize and increased by 60 percent for cotton compared to the conventional tillage system. Farmers also frequently reported signi�cant crop yield increases for maize, sorghum, millet, cotton, and groundnut in agroforestry systems, but relatively high labor inputs are required to reduce competition effects of trees from negatively impacting crop growth. Inorganic fertilizers also show relatively high pro�ts because they provide nutrients that can be readily absorbed by plants. Judicious fertilizer application counters soil nutrient depletion, reduces deforestation and expansion of cultivation to marginal areas, and increases crop yields. Excessive fertilizer use is less environmentally friendly, however, due to nitrous oxide emissions associated with high application rates of nitrogen fertilizers and fossil fuel–based emissions associated with fertilizer production and transportation. Capitalizing on Synergies and Managing Trade-Offs in Soil Carbon Sequestration Synergies occur when there is a positive correlation between carbon sequestration and pro�tability (where pro�tability refers to the net present value of implementing the land management practices). Trade-offs occur when attempts to increase carbon storage reduce pro�ts. Increasing food security under a changing climate requires the analysis and identi�cation of the land management technologies that maximize synergies and minimize trade-offs. A plot of pro�t versus carbon sequestration reveals synergies in two agroforestry systems—intercropping and alley farming (top right quadrant of �gure E2). In �gure E2, land management technologies in the lower right quadrant have high mitigation potentials but are modestly pro�table. Afforestation, improved fallow (including trees in croplands), and establishing barriers across sloping areas tend to take land out of production for a signi�cant period of time. They reduce the amount of land available for cultivation in the short run but can lead to overall increases in productivity and improved resilience in the long run. The time-averaged, above-ground CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY XXV PHOTO E.5: Crop Harvesting in Mali. The Biomass Is Smaller Compared to that of Agroforestry Systems Source: Curt Carnemark/World Bank. FIGURE E2: Trade-Offs Between Pro�tability and Carbon Sequestration of Sustainable Land Management Technologies 1,000 pro�t per tone of carbon dioxide sequestered (US $) No-tillage Inorganic fertilizer Intercropping 100 Manure Alley farming Cover crops Soil amendments Include trees Crop residues Terracing Afforestation Rotation Rotation diversi�cation Tree crop farming 10 intensi�cation Rainwater harvesting Improved Cross slope barriers fallow 1 0 2 4 6 8 10 carbon dioxide sequestered (ton per hectare per year) Source: This study. biomass of crop residues and other technologies in the lower left quadrant of �gure E2 is relatively small compared to that of agroforestry systems. Also, the biomass of crop residues does not accumulate easily, resulting in lower mitigation bene�ts. Judicious fertilizer application increases crop yields and pro�tability. Yields also increase with manure application and ac- cumulation of soil carbon, but with patterns that depend on crop type. Manure is less pro�table than inorganic fertilizer because of the labor costs associated with collecting and processing manure (top left quadrant of �gure E2). The relatively high pro�tability of no-tillage derives primarily from the decrease in production costs after the establishment of the system. EC O N O M I C A N D S E CT OR WORK XXV I EX EC UTIV E S UM M A RY The trade-offs exhibited by the land management technologies have important implications for land-use decision making. Sustainable land management interventions should be planned and implemented in a coordinated manner across space, time, and sectors. Working at the landscape level within an ecosystems approach is useful for addressing food security and rural livelihood issues and in responding to the impacts of climate change and contributing to its mitigation. The landscape approach entails the integrated planning of land, agriculture, forests, �sheries, and water at local, watershed, and regional scales to ensure that synergies are properly captured. The landscape approach provides a framework for the better manage- ment of ecosystem services, such as agricultural productivity, carbon storage, freshwater cycling, biodiversity protection, and pollination. Public Costs of Soil Carbon Sequestration Public cost refers to government support toward the implementation of land management practices. They include invest- ments in seeds and seedlings, input subsidies, extension services, and other administrative costs. The pattern of public sup- port is as crucial as the amount of support for full realization of productivity, adaptation, and mitigation bene�ts in agriculture. Public support that focuses on research, investments in improved land management, and land tenure rather than on input support is generally more effective, bene�ts more farmers, and is more sustainable in the long run. Technologies that involve signi�cant change in land use (such as afforestation and improved fallows) and landscape alteration (such as terracing and cross-slope barriers) incur high public costs but generate low private bene�ts (lower right quadrant of �gure E3). The low pro�ts suggest that farmers may be reluctant to privately invest in these technologies. Strong public involvement in these technologies is required given their relatively high mitigation potentials. Crop residues, cover crops, crop rotation, and rainwater harvesting with lower pro�ts and also manure and no tillage that generate relatively higher pro�ts require minimal government support (lower left and upper left quadrants of �gure E3, respectively). These technologies generally have low mitigation potentials. The relatively high public cost of inorganic fertilizer (top right quadrant, �gure E3) reflects the use of subsidies in spurring farmers’ access to the technology. Fertilizer subsidies are associated with high �scal costs, dif�cult targeting, and crowding out of commercial sales. Thus, fertil- izer subsidies are appropriate in situations when the economic bene�ts clearly exceed costs, the subsidies help achieve social rather than economic objectives, and the support helps improve targeting through market-smart subsidies while providing impetus for private sector input development. Examples of market-smart subsidies include demonstration packs, vouchers, matching grants, and loan guarantees. FIGURE E3: Relationship Between Private Bene�ts and Public Costs 1000 private bene�t (per tonne of carbon dioxide sequestered) No-tillage Inorganic fertilizer Intercropping 100 Manure Alley farming Cover crops Terracing Include trees Afforestation Crop residues Tree crop farming 10 Crop rotation Improved fallow Rainwater harvesting Cross slope barriers 1 0 3 5 8 10 13 public cost ($ per tonne of carbon dioxide sequestered) Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY X X V II The overall biophysical mitigation, potential savings, and the costs of soil carbon sequestration by 2030 depend on the emission scenarios influenced by a wide range of driving forces from demographic to social and economic developments. The total mitigation potential varies from 2.3 Gt CO2-eq for Latin America to 7.0 Gt CO2-eq for Asia (table E3). Total private pro�ts range from US$105 billion in Africa to $1.4 trillion in Asia, while total public costs range from US$20 billion in Africa to $160 billion in Asia. Barriers to the Adoption of Sustainable Land Management Practices Despite the fact that improved land management technologies generate private and public bene�ts, their adoption faces many socioeconomic and institutional barriers: Most of the land management technologies require signi�cant up-front ex- penditure that poor farmers cannot afford; the nonavailability of inputs in the local markets can be a signi�cant obstacle; lack of information on the potentials of alternative techniques of farming and limited capacity is a major constraint in many developing countries; when technologies are inconsistent with community rules and traditional practices, their adoption is often resisted; and willingness and ability to work together is crucial for many technologies such as improved irrigation and communal pastures. The absence of collective action will hinder successful uptake, diffusion, and impact of such land management technologies. Factors affecting adoption tend to be more speci�c to the land management technologies. Table E4 suggests that lack of credit and inputs and land tenure problems are by far the most important factors for adoption across the range of technologies. However, improved availability of inputs is a necessary but insuf�cient condition for adoption of land management practices. Better market prices for crops and other agricultural produce are crucial. Secure land rights is a precondition for climate-smart agriculture as it provides incentive for local communities to manage land more sustainably. Ill-de�ned land ownership may inhibit sustainable land management changes. TABLE E3: Technical Mitigation Potential, Private Bene�ts, and Public Costs of the Land Management Technologies by 2030 TECHNICAL POTENTIAL PRIVATE BENEFITS PUBLIC COSTS SCENARIO (MILLION TONS CO2-eq) (US$, BILLION ) (US$, BILLION ) Africa B1 3,448 105.4 19.6 A1b 3,505 108.6 19.7 B2 3,678 111.4 20.8 A2 3,926 120.9 22.3 Asia B1 5,977 1,224.5 131.3 A1b 6,388 1,259.3 143.6 B2 7,007 1,368.1 159.7 A2 6,678 1,310.8 150.4 Latin America B1 2,321 273.8 40.8 A1b 2,425 279.4 42.9 B2 2,538 288.8 44.3 A2 3,097 319.4 55.1 Source: This study. Notes: B1 = a world more integrated and more ecologically friendly; A1b = a world more integrated with a balanced emphasis on all energy sources; B2 = a world more divided but more ecologically friendly; A2 = a world more divided and independently operating self-reliant nations. EC O N O M I C A N D S E CT OR WORK XXV I I I EX EC UTIV E S UM M A RY TABLE E4: Relative Importance of Different Factors for Adopting Improved Land Management Practices LAND MANAGEMENT INPUTS/ MARKET TRAINING/ LAND TECHNOLOGY CREDITS ACCESS EDUCATION TENURE RESEARCH INFRASTRUCTURE Inorganic fertilizer *** ** ** ** * ** Manure ** ** * ** * ** Conservation agriculture ** ** *** ** ** * Rainwater harvesting ** ** ** *** ** ** Cross-slope barriers ** * ** ** ** * Improved fallows ** * * *** ** * Grazing management *** *** ** *** ** * Source: Synthesized from Liniger et al. 2011. Liniger, H. P., Mekdaschi Studer, R., Hauert, C., and Gurtner, M. 2011. Sustainable Land Management in Practice—Guidelines and Best Practices for Sub- Saharan Africa. World Overview of Conservation Approaches and Technologies and Food and Agriculture Organization of the United Nations. Key * = Low importance, ** = Moderate importance; *** = High importance. Behavioral change through education and extension services is required to enable change-over to improved land manage- ment technologies. For instance, conservation agriculture, the farming system involving no-tillage, residue management, and use of cover crops is highly knowledge intensive, requiring training and practical experience of those promoting its adoption. Learning hubs, regional platforms, scienti�c research, south-south knowledge exchange, and technical support mechanisms may increase innovation and facilitate adoption of improved land management technologies. The knowledge base of land management practices at the local level can be also improved through careful targeting of capacity development programs. Policy Implications Private bene�ts that drive land-use decisions often fall short of social costs; thus, carbon sequestration may not reach the optimal level from a social point of view unless some mechanisms exist to encourage farmers. Some public policies that can potentially incentivize carbon sequestration include the following options. 1. Strengthen the capacity of governments to implement climate-smart agriculture. Countries must be prepared to access new and additional �nance. There is a need to build the technical and institutional capacity of government ministries to implement climate-smart agriculture programs. Existing national policies, strategies, and investment plans should be strengthened to form the basis for scaling-up investments for climate-smart agriculture. Readiness for carbon sequestration and climate-smart agriculture can be achieved through improved extension services and training in relevant land management technologies for different locales. 2. Global cooperative agreement. Given the tremendous signi�cance that agriculture has for the global climate, prog- ress in incorporating it into the UN Framework Convention on Climate Change (UNFCCC) has been slower than many people hoped for. While the negative impacts of agricultural production in terms of land-use change and GHG emis- sions were reasonably well covered by the convention, the real and potential contributions the sector can and does make in terms of sequestering carbon in agricultural biomass and soils were for the most part omitted. Redressing this omission promises to foster a more balanced perspective in which food security is not necessarily at odds with climate change adaptation and mitigation (an unworkable conflict in which longer term environmental concerns are virtually guaranteed to universally lose out politically to the more immediate concern of food supply). A more practi- cal and thorough picture makes it possible for agriculture to be rewarded for its positive environmental impacts and to be an integral part of the solution as well as part of the problem. This is vitally important because agriculture needs to be fully incorporated into adaptation and mitigation strategies. As a result, the international community has recognized the importance of integrating agriculture into the ongoing negotiations on the international climate change regime. At the 17th Conference of Parties to the UNFCCC in Durban, South Africa, in November 2011, the parties asked the UNFCCC Subsidiary Body for Scienti�c and Technological Advice to explore the possibility of a formal work program on agriculture. CARBON SEQUESTRATION IN AGRICULTURAL SOILS EX E C U T I V E S U MMARY X X IX 3. Boost �nancial support for early action. A blend of public, private, and development �nance will be required to scale-up improved land management practices. Integrating sources of climate �nance with those that support food security may be one of the most promising ways to deliver to climate-smart agriculture the resources it requires. For technologies that generate signi�cant private returns, grant funding or loans may be more suitable to overcom- ing adoption barriers. For technologies such as conservation agriculture that require speci�c machinery inputs and signi�cant up-front costs, payment for an ecosystem services scheme could be used to support farmers and break the adoption barrier. There is also the potential for carbon �nance to support farmers during the initial period before the trees in agroforestry systems generate an economic return. 4. Raise the level of national investment in agriculture. While this may appear a tall order in countries with severe bud- get constraints, �nite public resources can be more selectively targeted using the criteria given above—prioritizing technologies that generate no short-term returns and those that most effectively address the barriers that prevent prospective adopters from moving forward. In some cases, relatively affordable technologies that generate quick and demonstrable bene�ts may warrant priority and potentially establish some of the channels through which more sophisticated technologies are dispersed in the future. Nationally owned climate-smart agricultural policies and action frameworks will increase the adoption of sustainable land management practices. However, public investment is only one sphere, and involving the private sector in climate-smart agriculture and sustainable land management is the other. 5. Create enabling environments for private sector participation. Introducing policies and incentives that provide an enabling environment for private sector investment can increase overall investment. This private investment can be targeted to some degree as well, particularly when government priorities translate clearly into business opportunities and certain areas of investment are looked upon favorably by public of�cials and institutions. Public investment can also be used to leverage private investment in areas such as research and development, establishing tree planta- tions, and developing improved seeds and seedlings. Particular attention should go to encouraging private �nancial service providers to tailor instruments that enable farmers who adopt SLM practices to overcome the barriers de- scribed above. Bundling agricultural credit and insurance together and providing different forms of risk management such as index-based weather insurance or weather derivatives are areas of private investment that can be encour- aged through public policy and public-private partnerships. EC O N O M I C A N D S E CT OR WORK C H A P T E R 1 — I N T RODUCT ION 1 Chapter 1: INTRODUCTION 1.1 FOOD SECURITY UNDER A CHANGING Agriculture is highly vulnerable to climate change and needs CLIMATE to adapt to changing climate conditions. Under optimistic Ensuring food security under changing climate conditions is lower end projections of temperature rise, climate change one of the major challenges of our era. There are about 925 may reduce crop yields by 10 to 20 percent (Jones and million food-insecure people in the world—about 16 percent Thornton 2009), while increased incidence of droughts and of the population in developing countries. Global population floods may lead to a sharp increase in prices of some of the will increase from 7 billion currently to over 9 billion people main food crops by the 2050s. Climate change will also im- by 2050, creating a demand for a more diverse diet that re- pact agriculture through effects on pests and disease. The quires additional resources to produce. Competition for land, interactions between ecosystems and climate change are water, and energy will intensify in an attempt to meet the complex, and the full implications in terms of productivity need for food, fuel, and �ber and will contribute to economic and food security are uncertain (Gornall et al. 2010) development and poverty reduction. Over this period, global- The agriculture sector has a pivotal role to play in mitigating ization may further expose the food system to the vagaries greenhouse gas (GHG) emissions. Agriculture and land-use of economic and political forces. Various projections suggest change currently account for about one-third of total emis- that global food requirements must increase by 70 to 100 sions (�gure 1.1). Agriculture is the primary driver of defores- percent by 2050 (Burney, Davis, and Lobell 2010), in addition tation in many developing countries. The net increase in agri- to maintaining and, where possible, enhancing the resilience cultural land during the 1980s and 1990s was more than 100 of natural ecosystems. million ha across the tropics. About 55 percent of the new FIGURE 1.1: Contribution of Different Sectors to Greenhouse Gas Emissions Buildings 8% Manure Forestry mgt, 7% 17% Rice production, 11% Industry 19% Agriculture Biomass Nitrous oxide 14% burning, 12% from soils, 38% Energy Waste 3% 26% Enteric fermentation, Transport 32% 13% Source: IPCC 2007; Smith et al. 2008. EC O N O M I C A N D S E CT OR WORK 2 CH A PTER 1 — INTR OD UC TION FIGURE 1.2: Proportion of Agricultural Land Derived From Different Land Covers in the Tropics, 1980–2000 100% 90% 80% 70% 60% Water Plantations 50% Shrubs 40% Disturbed forests 30% Intact forests 20% 10% 0% ics ica ca ia ica a ca ia ric As As fri fri op er er Af lA tA m Am st h -Tr ut st Ea lA ra es n So Ea nt h Pa h- ra W ut Ce ut nt So So Ce Source: Redrawn from Gibbs et al. (2010). agricultural land in the tropics came at the expense of intact Soil is central to most SLM technologies because it is the forests, while another 28 percent came from the conversion basic resource for land use. It supports all the terrestrial eco- of degraded forests (Gibbs et al. 2010; �gure 1.2). Projected systems that cycle much of the atmospheric and terrestrial increases in demand for food and bioenergy by 2050 may carbon. It also provides the biogeochemical linkage between further increase pressure on forests in the tropics with pro- other major carbon reservoirs, namely the biosphere, at- found implications for an increase in GHG emissions. Even mosphere, and hydrosphere. Soil carbon is held within the if emissions in all other sectors were eliminated by 2050, soil, primarily in association with its organic constituent. Soil growth in agricultural emissions under a business-as-usual carbon has a strong correlation with soil quality, de�ned as world with a near doubling in food production would perpetu- the ability of soils to function in natural and managed ecosys- ate climate change. tems. Soil carbon influences �ve major functions of the soil (Larson and Pierce 1991), namely the ability to 1.2 CARBON BENEFITS THROUGH accept, hold, and release nutrients; CLIMATE-SMART AGRICULTURE accept, hold, and release water both for plants and for The triple imperatives of increasing productivity, reducing surface and groundwater recharge; emissions, and enhancing resilience to climate change call promote and sustain root growth; for alternative approaches to practicing agriculture. Climate- maintain suitable biotic habitat; and smart agriculture (CSA) seeks to increase productivity in an respond to management and resist degradation. environmentally and socially sustainable way, strengthen farmers’ resilience to climate change, and reduce agriculture’s Increasing soil organic carbon can reverse soil fertility deteri- contribution to climate change by reducing GHG emissions oration, the fundamental cause of declining crop productivity and increasing soil carbon storage. One of the key elements in developing countries. Table 1.1 indicates the potential in- of CSA is sustainable land management (SLM) involving crease in crop yields from increasing the soil organic carbon the implementation of land-use systems and management pool in the root zone by 1 ton C/ha/yr through SLM technolo- practices that enable humans to maximize the economic and gies. The overall increase in grain productivity in Africa, Asia, social bene�ts from land while maintaining or enhancing the and Latin America due to such increase in soil organic carbon ecosystem services from land resources. is estimated at 24 to 40 million tons per year (table 1.2). CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 1 — I N T RODUCT ION 3 TABLE 1.1: Improvement in Crop Yields per Furthermore, the removal of crop residues and cattle manure Ton of Carbon in the Root Zone for fuel leads creates a negative carbon budget and must be CROP POTENTIAL YIELD INCREASE (kg/ha) prevented. Maize 200–400 Sustainable land management practices are an alternative to Wheat 20–70 several conventional agricultural practices that lead to emis- Soybean 20–30 sions of GHG from the soil to the atmosphere. These con- Cowpea 5–10 ventional practices include biomass burning (that releases Rice 10–50 carbon dioxide, methane, and nitrous oxide), plowing and soil Millet 50–60 disturbance (carbon dioxide), deforestation (carbon dioxide, Source: Lal (2011). methane, and nitrous oxide), draining of wetlands (carbon dioxide and nitrous oxide), and uncontrolled grazing (carbon dioxide and nitrous oxide). Emission of these gases from Soil carbon also enhances resilience to climate variability agricultural ecosystems is increased through subsistence ag- and change by improving soil structure and stability, reducing ricultural practices that do not invest in soil quality improve- soil erosion, improving aeration and water-holding capacity, ment practices such as erosion control, water management, reducing the impacts of drought, improving soil biodiversity, and application of fertilizers and other amendments (World and increasing nutrient use ef�ciency. Bank 2010). Sustainable land management provides carbon bene�ts Soil carbon sequestration is the process by which atmospheric through three key processes, namely carbon conservation, carbon dioxide is taken up by plants through photosynthesis reduced emissions, and carbon sequestration. Many natural and stored as carbon in biomass and soils. It entails replen- land systems such as native forests, grasslands, and wet- ishing lost carbon and adding new carbon (organic inputs) lands have relatively high carbon stocks. Conserving this ter- beyond original levels. Historically, agricultural soils have lost restrial carbon pool accumulated over millennia should be a more than 50 Gt (1 Gt = 1 billion tons) of carbon. Some of this major priority, as it offers the greatest least-cost opportunity carbon, however, can be recaptured through sustainable land for climate mitigation and ecosystem resilience. Zero toler- management practices. For instance, new technologies such ance for soil erosion is indispensable for soil carbon conser- as deeper-rooted crops and pasture grasses can enhance vation. Removal of the vegetation cover aggravates losses by original soil carbon up to a given equilibrium. The use of crop soil erosion and increases the rate of decomposition due to residues as mulch, intercropping food crops with trees, and changes in soil moisture and temperature regimes. Because integrated nutrient and water management also sequester soil organic matter is concentrated on the soil surface, ac- carbon in the soil. By adopting improved land management celerated soil erosion leads to progressive depletion of soil practices to increase soil carbon, farmers can increase crop carbon. Agricultural soils should be prevented from being yields, reduce rural poverty, limit GHG concentrations in the washed to streams and rivers where the relatively stable atmosphere, and reduce the impact of climate change on soil C pools are rapidly oxidized to carbon dioxide (Lal 2003). agricultural ecosystems. TABLE 1.2: Estimated Increase in Grain Crop Production from Land Management Technologies That Sequester Soil Carbon (Million Tons/Year) CROP AFRICA ASIA LATIN AMERICA TOTAL Maize 0.8–1.3 4.1–8.2 4.5–6.9 9.4–16.4 Wheat 0.2–0.4 2.9–4.9 0.5–0.6 3.6–5.9 Rice 0.1–0.2 4.1–6.9 0.2–0.3 4.7–7.4 Sorghum 1.7–2.6 1.3–1.8 0.4–0.6 3.4–5.0 Millet 0.6–1.0 0.4–0.7 0.01–0.01 1.0–1.8 Beans 0.1–0.2 0.4–0.7 0.3–0.5 0.8–1.4 Soybean 0.02–0.03 0.3–0.5 0.7–1.2 1.0–1.7 Total 3.5–5.7 13.5–23.7 6.6–10.1 23.6–39.5 Source: Lal (2003). EC O N O M I C A N D S E CT OR WORK 4 CHAPTER 1 — INTRODUCTION 1.3 OBJECTIVES AND SCOPE OF THE REPORT Gibbs, H. K., Ruesch, A. S., Foley, J. A., Ramankutty, N., Achard, ., . F and Holmgren, P 2010. “Pathways of Agricultural Expansion The purpose of this report is to improve the knowledge Across the Tropics: Implications for Forest Resources. � base for facilitating investments in land management tech- Proceedings of the National Academy of Sciences 107 (38): nologies that sequester soil organic carbon. While there are 16732–16737 . many studies on soil carbon sequestration, there is no single Gornall, J., Betts, R., Burke, E., Clark, R., Camp, J., Willett, K., unifying volume that synthesizes knowledge on the impact and Wiltshire, A. 2010. “Implications of Climate Change for of different land management practices on soil carbon se- Agricultural Productivity in the Early Twenty-First Century. � Philosophical Tranactions of the Royal Society B 365: 2973–2989. questration rates across the world.2 A meta-analysis was car- doi:10.1098/rstb.2010.0158. ried out to provide soil carbon sequestration rates in Africa, Asia, and Latin America. This is one important element in Guo, L., and Gifford, R. 2002. “Soil Carbon Stocks and Land-Use � Change: A Meta Analysis. Global Change Biology 8: 345–360. decision-making for sustainable agricultural intensification, agro-ecosystems resilience, and comprehensive assess- . Intercontinental Panel on Climate Change (IPCC). 2007 Climate Change 2007: Synthesis Report—Summary for Policymakers. ments of greenhouse mitigation potentials of SLM practices. Fourth Assessment Report. Furthermore, the ecosystem simulation modeling technique was used to predict future carbon storage in global cropland ., . Jones, P and Thornton, P 2009. “Croppers to Livestock Keepers: � Livelihood Transitions to 2050 in Africa Due to Climate Change. soils. Last, marginal abatement cost curves and trade-off Environmental Science and Policy 12: 427–437 . graphs were used to assess the cost-effectiveness of the technologies in carbon sequestration. � Lal, R. 2003. “Soil Erosion and the Global Carbon Budget. Environment International 29: 437–450. The remainder of the report is organized as follows. Chapter � Lal, R. 2011. “Sequestering Carbon in Soils of Agroecosystems. 2 provides a brief review of soil organic carbon dynamics and Food Policy 36: S33–S39. the methods for soil carbon assessment. The chapter con- . Larson, W. E., and Pierce, F J. 1991. “Conservation and cludes with brief information on carbon assessment in The � Enhancement of Soil Quality. In Evaluation for Sustainable World Bank’s sustainable land management projects portfo- Land Management in the Developing World, Vol. 2: Technical Papers, ed. J. Dumanski, E. Pushparajah, M. Latham, and R. lio. Chapter 3 reports the increase in soil carbon for selected Myers. Bangkok, Thailand: International Board for Research and sustainable land management practices in Africa, Asia, and Management. IBSRAM Proceedings No. 12 (2): 175–203. Latin America. Chapter 4 reports the estimates from ecosys- . Ogle, S. M., Bredit, F J., and Pasutian, K. 2005. “Agricultural tem simulation, while Chapter 5 concludes with the benefits Management Impacts on Soil Organic Carbon Storage Under and costs of adopting carbon sequestering practices and a Moist and Dry Climatic Conditions of Temperate and Tropical discussion of policy options to support climate-smart agricul- � Regions. Biogeochemistry 72: 87–121. ture in developing countries. The report will provide a broad ., Smith, P Martino, D., Cai, Z., et al. 2008. “Greenhouse Gas perspective to natural resource managers and other profes- � Mitigation in Agriculture. Philosophical Transactions of the sionals involved in scaling up CSA. Royal Society B 363: 789–813. Watson, Robert et al., ed. 2000. Land Use, Land-Use Change, REFERENCES and Forestry. Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Burney, J. A., Davis, S. J., and Lobell, D. B. 2010. “Greenhouse Gas Mitigation by Agricultural Intensification. Proceedings of the � World Bank. 2010. Sustainable Land Management for Mitigation of National Academy of Sciences 107 (26): 12052–12057 . and Adaptation to Climate Change. Environment Department. 2 Major exceptions are Guo and Gifford (2002) and Ogle et al. (2005), but the sequestration rates in these papers are highly variable and not specific to local conditions. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 5 Chapter 2: SOIL ORGANIC CARBON DYNAMICS AND ASSESSMENT METHODS 2.1 SOIL ORGANIC CARBON DYNAMICS two most important anthropogenic processes responsible Different ecosystem types store different amounts of car- for the release of carbon dioxide into the atmosphere are bon depending on their species compositions, soil types, burning of fossil fuels (coal, oil, and natural gas) and land use climate, relief, and other biophysical features (�gure 2.1). Of (table 2.2). Emissions from land-use change are about 1.5 Gt C the estimated over 150 million km2 of terrestrial ecosystems per year, largely determined by tropical deforestation that area, forests account for more than 40 million km2 (about exacerbates soil erosion and organic matter decomposition. 28 percent). Savannahs and grasslands both cover about 23 The underlying driving factors of tropical deforestation are percent, while croplands occupy about 11 percent (table 2.1). highly interconnected and include poverty, policy and institu- Among the biomes, vegetation carbon stocks range from 3 tional failures, population growth, and the attendant demand Gt for croplands to 212 Gt for tropical forests, while soil car- for natural resources, urban expansion, and international bon stocks range from 100 Gt for temperate forests to 471 trade. Gt for boreal forests. The tundra biome, covering an area of less than 10 million km2, has the highest density of carbon Rapidly growing emissions are outpacing the growth in storage. Soils generally hold more carbon than vegetation natural sinks. The ef�ciency of oceans and lands as carbon across biomes and account for 81 percent of terrestrial car- dioxide sinks has declined over the years. Currently, natural bon stock at the global level. sinks remove an average of 55 percent of all anthropogenic carbon dioxide emissions, which is slightly lower than 60 The global carbon cycle describes the transfer of carbon in percent they removed some 50 years ago (Global Carbon the earth’s atmosphere, vegetation, soils, and oceans. The Project 2009). FIGURE 2.1: Carbon Stocks in Biomass and Soils Carbon storage in terrestrial ecosystems (Tonnes per ha) 0 to 10 10 to 20 20 to 50 50 to 100 100 to 150 150 to 200 200 to 300 300 to 400 400 to 500 More than 500 Source: Ruesch and Gibbs, 2008; IGBP-DIS, 2000. Source: UNEP/GRID, http://www.grida.no/publications/rr/natural-�x/page/3724.aspx. EC O N O M I C A N D S E CT OR WORK 6 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S TABLE 2.1: Carbon Stocks in Vegetation and Top 1 Meter of Soils of World Biomes CARBON STOCKS (Gt C) AND PROPORTION IN THE ECOSYSTEM (%) AREA BIOMES (MILLION km2) VEGETATION PROPORTION (%) SOILS PROPORTION (%) TOTAL Tropical forests 17.6 212 49.5 216 50.5 428 Temperate forests 10.4 59 37.1 100 62.9 159 Boreal forests 13.7 88 15.7 471 84.3 559 Tropical savannas 22.5 66 20.0 264 80.0 330 Temperate grasslands 12.5 9 3.0 295 97.0 304 Deserts 45.5 8 4.0 191 96.0 199 Tundra 9.5 6 4.7 121 95.3 127 Wetlands 3.5 15 6.3 225 93.8 240 Croplands 16 3 2.3 128 97.7 131 Total 151.2 466 2,011 2,477 Proportion (%) 19 81 100 Source: Based on Watson et al. (2000) and Ravindranath and Ostwald (2008). TABLE 2.2: Global Carbon Budget (Gt C) SOURCE 1980s 1990s 2000–2008 Atmospheric increase 3.3 ± 0.1 3.2 ± 0.1 4.1 ± 0.1 Fossil fuel emissions 5.4 ± 0.3 6.4 ± 0.4 7.2 ± 0.3 Net ocean-to-atmosphere flux −1.8 ± 0.8 −2.2 ± 0.4 −2.3 ± 0.5 Net land-to-atmosphere flux −0.3 ± 0.9 −1.0 ± 0.6 −1.3 ± 0.7 Partitioned as: Land-use change flux 1.4 ± 1.0 1.6 ± 0.7 1.4 ± 0.7 Residual land sink −1.7 ± 1.7 −2.6 ± 0.9 −2.7 ± 1.0 Sources: IPCC (2007) and the Global Carbon Project (2009). Soils are critically important in determining global carbon Through photosynthesis, plants reduce carbon from its oxi- cycle dynamics because they serve as the link between the dized form to organic forms (net primary productivity; NPP) atmosphere, vegetation, and oceans. Globally, the soil car- useful for growth and energy storage. Over time, the C �xed bon pool (also referred to as the pedologic pool) is estimated in the atmosphere becomes soil carbon through the process at 2,500 Gt up to 2 meters deep. Out of this, the soil organic of above- and below-ground decomposition of materials, carbon pool comprises 1,550 Gt, while the soil inorganic release of sap exudates from plant roots into the soil, and carbon and elemental pools make up the remaining 950 Gt root die-off. Breeding crop plants with deeper and bushy root (Batjes 1996). The soil carbon pool is more than three times ecosystems could simultaneously sequester more carbon, the size of the atmospheric pool (760 Gt) and about 4.5 times improve soil structure, improve water and nutrient retention, the size of the biotic pool (560 Gt). and increase crop yields (Kell 2011). The elemental and inorganic forms of soil carbon primarily Different fractions or soil organic carbon pools have different result from mineral weathering and are less responsive to functions within the soil system. Crop residues are readily land management than soil organic carbon (table 2.3). Soil broken down and serve as substrates to soil microorgan- organic carbon is a complex mixture of organic compounds isms. Particulate organic carbon is broken down relatively composed of decomposing plant tissue, microbial organ- quickly but more slowly than other crop residues and is isms, and carbon bound to soil minerals. These compounds important for soil structure, energy for biological processes, originate from the photosynthetic activities of plants. and provision of nutrients for plants. A more stable fraction, CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 7 TABLE 2.3: Forms of Carbon in the Soil FORMS SOURCES Elemental Geologic materials (e.g., graphite and coal) Incomplete combustion of organic materials (e.g., charcoal, graphite, and soot) Dispersion of these carbon forms during mining Inorganic Geologic or soil parent materials, usually as carbonates—that is, calcite, CaCO3 dolomite, CaMg(CO3)2 and, to some extent, siderite (Fe CO3) Agricultural inputs such as liming can also introduce calcite and dolomite into the soil. Organic Plant and animal materials at various stages of decomposition ranging from crop residues with size of 2 mm or more Plant debris, also referred to as particulate organic carbon, with size between 0.05 and 2 mm humus, highly decomposed materials less than 0.05 mm that are dominated by molecules attached to soil minerals Source: Synthesized from Schumacher (2002). humus, can be classi�ed into two depending on the level stable humus complexes can remain in the soil for centuries of decomposability: The �rst is active humus that is still or millennia. subject to further decomposition, and the other is passive humus (or recalcitrant carbon), the highly stable, insoluble At the global level, the soil organic carbon pool is concen- form that is not subject to further decomposition. Active trated in �ve major soil orders: histosols, inceptisols, enti- humus is an excellent source of plant nutrients (nitrates sols, al�sols, and oxisols. In the tropics, the largest amount and phosphates), while passive humus is important for soil of soil organic carbon is found in oxisols, histosols, ultisols, physical structure, water retention, and tilth. Some very and inceptisols (�gure 2.2, table 2.4, and box 2.1). FIGURE 2.2: Global Soil Regions Robinson projection scale 1:130,000,000 Soil orders Alfisols Entisols Inceptisols Spodosols Rocky land Andisols Gelisols Mollisols Ultisols Shifting sand Aridisols Histosols Oxisols Vertisols Ice/glacier Source: United States Department of Agriculture. EC O N O M I C A N D S E CT OR WORK 8 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S TABLE 2.4: Soil Carbon Pool up to 1-M Deep for Soil Orders of the World’s Ice-Free Land Surface GLOBAL LAND AREA TROPICAL LAND AREA SOIL SOIL ORGANIC ORGANIC EXTENT PROPORTION CARBON PROPORTION EXTENT PROPORTION CARBON PROPORTION SOIL ORDER (1000 km2) (%) POOL (Gt) (%) (1000 km2) (%) POOL (Gt) (%) Al�sols 18,283 13.5 127 8.1 6,411 12.9 30 5.9 Andisols 2,552 1.9 78 4.9 1,683 3.4 47 9.3 Aridisols 31,743 23.5 110 7 9,117 18.4 29 5.7 Entisols 14,921 11 148 9.4 3,256 6.6 19 3.8 Histosols 1,745 1.3 357 22.7 286 0.6 100 19.8 Inceptisols 21,580 16 352 22.3 4,565 9.2 60 11.9 Mollisols 5,480 4.1 72 4.6 234 0.5 2 0.4 Oxisols 11,772 8.7 119 7.6 11,512 23.2 119 23.5 Spodosols 4,878 3.6 71 4.5 40 0.1 2 0.4 Ultisols 11,330 8.4 105 6.7 9,018 18.2 85 16.8 Vertisols 3,287 2.4 19 1.2 2,189 4.4 11 2.2 Others 7,644 5.7 18 1.1 1,358 2.7 2 0.4 Total 135,215 100 1,576 100 49,669 100 506 100 Source: Eswaran et al. (1993). The soil organic carbon pool represents a dynamic balance erosion, tillage, residue removal, and drainage. Theoretically, between gains and losses. The amount changes over time the potential soil carbon sequestration capacity is equivalent depending on photosynthetic C added and the rate of its de- to the cumulative historical carbon loss. However, only 50 cay. Under undisturbed natural conditions, inputs of carbon to 66 percent of this capacity is attainable through the adop- from litter fall and root biomass are cycled by output through tion of sustainable land management practices (Lal 2004; erosion, organic matter decomposition, and leaching. box 2.2). The potential carbon sequestration is controlled primarily by The current rate of carbon loss due to land-use change (defor- pedological factors that set the physico-chemical maximum estation) and related land change processes (erosion, tillage limit to storage of carbon in the soil. Such factors include operations, biomass burning, excessive fertilizers, residue soil texture and clay mineralogy, depth, bulk density, aera- removal, and drainage of peat lands) is between 0.7 and 2.1 tion, and proportion of coarse fragments (�gure 2.3). The Gt carbon per year (table 2.2). This is more than 50 percent attainable carbon sequestration is set by factors that limit of the carbon absorbed by land. The conversion of natural the input of carbon to the soil system. NPP—the rate of vegetation to agricultural ecosystems leads to a depletion photosynthesis minus autotrophic respiration—is the major of the soil organic carbon pool by as much as 60 percent factor influencing attainable sequestration and is modi�ed by in the temperate regions and by 75 percent or more in the above-ground versus below-ground allocation. Land manage- tropics (box 2.1). The degree of loss is higher in soils that are ment practices that increase carbon input through increas- susceptible to accelerated erosion and other soil degradation ing NPP tend to increase the attainable level to nearer the processes. Soil erosion is the major land degradation process potential level. Climate has both direct and indirect effects on that emits soil carbon. The annual soil losses in Africa, South attainable sequestration. The decomposition rate increases America, and Asia are estimated at 39 to 74 Gt, correspond- with temperature but decreases with increasingly anaerobic ing to carbon emissions of 0.16 to 0.44 Gt per year (table 2.5). conditions. The actual carbon sequestration is determined by Globally, soil erosion accounts for up to 1.2 Gt of C emitted land management factors that reduce carbon storage such as to the atmosphere each year. This is more than 57 percent CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 9 BOX 2.1: Brief Description of Soil Orders Al�sols: Formed primarily under forest or mixed vegeta- Inceptisols: Exhibiting modest soil weathering and horizon tive cover, al�sols result from weathering processes that development, inceptisols are formed on recent geomor- leach clay minerals from the surface to the subhorizon. phic surfaces in semi-arid to humid environments. Included in this category are partially developed soils of the Sahel re- Andisols: Common in cool areas with moderate to high gion of West Africa, some soils of the riverine floodplains precipitations, andisols result from weathering process- of the Ganges and Brahmaputra Rivers in Bangladesh and es that generate minerals with little orderly crystalline India, and the floodplains of Southeast Asia. structure (volcanic glass) and usually have high nutrient- and water-holding capacity. Mollisols: Formed under moderate to pronounced sea- sonal moisture de�cits, mollisols are grassland soils Aridisols: Formed under arid climates, the lack of mois- with dark-colored surface horizons, relatively high or- ture markedly restricts the intensity of weathering and ganic matter, and high base saturation. development of aridisols. The paucity of vegetation also leads to low organic matter content. Oxisols: Dominated by low activity minerals, oxisols are highly weathered soils of tropical and subtropical re- Entisols: Occurring in areas of recently deposited parent gions. They are found on stable landscapes, have low materials or areas where erosion or deposition rates ex- natural fertility, and low capacity to retain fertilizer and ceed the rate of soil development, entisols are charac- soil amendments. terized with little or no horizon development. They occur in many environments such as on steep slopes, flood Spodosols: Commonly occurring in areas of coarse-tex- plains, or sand dunes. tured deposits of humid regions, spodosols have devel- oped from weathering processes that strip organic matter Gelisols: Found mostly in very cold areas under the influ- and iron and aluminum oxides from the surface to the sub- ence of glaciation, gelisols are characterized by perma- soil. Spodosols tend to be acidic and are inherently infertile. frost within 2 m of the soil surface. High amounts of soil organic matter accumulate in the upper layer, mak- Ultisols: Formed from fairly intense weathering and ing most gelisols black or dark brown in color. Gelisols leaching that results in clay accumulation at the subsoil, are not highly fertile because nutrients are very easily ultisols are typically acidic with most nutrients concen- leached above the permafrost. trated in the topsoil. They have moderately low capacity to retain fertilizer and soil amendments. Histosols: Formed in decomposed organic materials that accumulate faster than they decay, histosols Vertisols: Dominated by high content of swelling and have a high content of organic matter and no perma- shrinking clay minerals, vertisols typically form from highly frost. Most histosols are saturated all the year round. basic rocks in climates that are seasonally humid or subject They are commonly called peats, bogs, mucks, or to erratic floods, droughts, or impeded drainage. Vertisols moors. tend to be high in natural fertility, but they are dif�cult to till. Source: Modi�ed from United States Department of Agriculture. of the emission through land-use change and underscores to the atmosphere, estimated at 75 to 100 Gt C per year the need for carbon conservation through zero tolerance for is the next largest terrestrial carbon flux (Raich and Potter soil erosion. 1995). It is about 60 times the annual contribution of land-use change and about 11 times that of fossil fuel to atmospheric Each year, the terrestrial carbon pool assimilates 120 Gt C emissions. Thus, a small change in soil respiration can sig- from the atmosphere in the form of gross primary productiv- ni�cantly alter the balance of atmospheric carbon dioxide ity (or photosynthesis). Soil respiration, the flux of microbially concentration compared to soil carbon stores. Soil respira- and plant-respired carbon dioxide (CO2) from the soil surface tion is regulated by several factors including temperature, EC O N O M I C A N D S E CT OR WORK 10 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S FIGURE 2.3: Factors Affecting Soil Carbon Sequestration – Mineralogy/content – Depth Potential Defining factors: – Stoniness – Bulk density – Aeration carbon sequestration situation – NPP and allocation Attainable Limiting factors: – Climate (direct) – Climate (via NPP) SOC-increasing measures – Erosion – Tillage Actual Reducing factors: – Residue removal – Disrupted biology – Drainage SOC-protecting measures SOC level (t1/2 ≥ 10 years) ton ha–1 Source: Redrawn from Ingram and Fernandes (2001). TABLE 2.5: Estimate of Erosion-Induced Carbon Emission GROSS EROSION SOIL CARBON DISPLACED BY EROSION EMISSION (20 PERCENT OF DISPLACED CONTINENT (X 109 Mg/YEAR) (2 TO 3 PERCENT OF SEDIMENT; Gt C/YEAR) SOIL CARBON; Gt C/YEAR) Africa 38.9 0.8–1.2 0.16–0.24 Asia 74.0 1.5–2.2 0.30–0.44 South America 39.4 0.8–1.2 0.16–0.24 North America 28.1 0.6–0.8 0.12–0.16 Europe 13.1 0.2–0.4 0.04–0.08 Oceania 7.6 0.1–0.2 0.02–0.04 Total 201.1 4.0–6.0 0.8–1.2 Source: Adapted from Lal (2003). moisture, vegetation type, nitrogen content, and level of to speed up their consumption of plant residues and other aeration of the soil. organic matter. Variations in temperature are signi�cantly and positively correlated with changes in global soil res- Climate change is positively correlated with increasing rate piration (Bond-Lamberty and Thompson 2010). In 2008, of soil respiration. Higher temperatures trigger microbes the global soil respiration reached roughly 98 Gt, about 10 CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 11 BOX 2.2: Sustainable Land Management Practices Reverse Soil Carbon Loss in Java Research in the tropics has demonstrated the decline forests to cropland. From the Japanese occupation of soil organic carbon by as much as 60 percent after in 1942, throughout its independence years, and until conversion of forest to cropland. However, sustainable the early 1960s, Indonesia faced a serious problem of land management practices can accumulate soil organ- food scarcity. The Green Revolution of the 1960s saw ic carbon, reverse chronic soil degradation, improve soil Java producing close to two-thirds of the country’s rice. quality, and enhance ecosystem services supply from As a result, between 1960 and 1970, soil organic carbon the soil. Some soil scientists have recently used legacy markedly declined by 62 percent of its natural condition. soil survey data to capture the long-term trend of soil or- ganic carbon in Java in Indonesia (Minasny et al. 2010). Since the late 1960s, soil organic C has increased With an estimated population density of 1,026 persons slightly as a result of the government extension km−2, Java is undoubtedly the most densely populated program to disseminate new agricultural production and the most intensively cultivated island in Indonesia. knowledge among farmers, including the use of high- An analysis over the period from 1930 to 2010 revealed yielding varieties and chemical inputs. The increased that human activities are more important than environ- biomass and the return of crop residues, green com- mental factors in explaining soil organic carbon trend. post, and animal manure application were mostly The median soil organic carbon stock in the topsoil responsible for the increase in soil organic carbon dropped from 20.4 t ha−1 between 1930 and 1940 to 7.3 stock. By the 1990s, soil organic carbon stock had t ha−1 between 1960 and 1970 (see �gure below). This risen to about 11 t ha−1 as there was also a large inter- huge drop was mostly due to the high conversion of for- est in organic farming in Java. Further intensi�cation ests and natural vegetation into plantations and subse- has resulted in improved environmental awareness, quently to food crops. During the Dutch colonial period, increased likelihood of adoption of sustainable land most land development was for plantations such as tea, management practices, increased soil carbon se- rubber, and coffee. Between 1930 and 1950, decline questration, and increased resilience of the agricul- in soil carbon stock was primarily due to conversion of tural system. 25 20 15 t ha–1 10 5 0 1930–1940 1940–1950 1950–1960 1960–1970 1970–1980 1980–1990 1990–2000 2000–2010 Source: Minasny, B., Sulaeman, Y., and McBrateney A.B. 2010. Is soil carbon disappearing? The dynamics of soil organic carbon in Java. Global Change Biology 17:1917–1924. EC O N O M I C A N D S E CT OR WORK 12 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S times more carbon than humans release into the atmo- Typically, different countries adapt the Intergovernmental sphere each year. Soil respiration increased 0.1 Gt C per Panel on Climate Change (IPCC) guideline for national GHG year between 1989 and 2008. A rise in temperature by 2°C by using sampling methods, measurement techniques, and is estimated to release an additional 10 Gt C per year to models tailored to their particular circumstances. the atmosphere through soil respiration (Friedlingstein et al. 2003). Carbon assessment for land management projects can be either purposely for climate mitigation or for nonclimate Tillage operations can signi�cantly affect soil respiration. mitigation. Mitigation projects involve estimation of veri�- Conventional tillage leads to the destruction of soil aggre- able changes in carbon stocks over a given period in the de- gates, excessive respiration, and soil organic matter decom- �ned project area and require methods for estimating car- position, leading to reduced crop production and decreased bon stocks and changes for the baseline scenario (without resilience of the soil ecosystem. Excessive application of the project) and the project. Carbon assessment for land large amounts of nitrogenous fertilizer can markedly in- management projects not principally designed for climate crease root biomass and stimulate soil respiration rates. change mitigation is carried out for a number of reasons: When other factors are at optimum, conservation tillage, use of cover crops (green manure), crop rotations, use of The need to assess the carbon footprint of the opera- deep-rooted crops, application of manure, and water man- tional work of funding agencies (see section 2.4). agement can optimize soil respiration in addition to improv- Changes in soil carbon over the lifetime of a project ing soil carbon. are an indicator of the success of SLM intervention. Changes in soil carbon stocks can help track changes in regulating, supporting, and provisioning ecosystem services. 2.2 CARBON ASSESSMENT FOR LAND MANAGEMENT PROJECTS Interest in bene�ting from carbon �nance, though this is hardly a prime objective. Carbon assessment entails the estimation of stocks and fluxes of carbon from different land-use systems in a given The key differences for carbon assessment for the two types area over a period of time. The assessment covers four bio- of projects are summarized in table 2.6. mass pools—above ground, below ground, dead wood, and litter—and the soil organic carbon pool. The assessment can be undertaken either at national or project level. 2.3 TECHNIQUES OF SOIL CARBON Signatory parties of the UN Framework Convention on ASSESSMENT Climate Change (UNFCCC) are required to prepare national Methods to assess above-ground biomass are more ad- GHG inventories on a periodic basis and report them to vanced than for soil carbon. The three major methods for the body. Annex I or industrialized countries are required above-ground carbon assessment include the following to estimate and report emissions and removals annually, (Gibbs et al. 2007): while non-Annex I or developing countries only need to report every 3 to 5 years. The key steps involved are as 1. Biome averages involving the estimation of follows: average forest carbon stocks for broad forest categories based on a variety of input data sources, 1. Estimating the area under a given land-use category 2. Forest inventory that relates tree diameters or in a given year and the area under each category volume to forest carbon stocks using allometric subjected to land-use change relationships, and 2. Estimating the stocks of carbon in each pool at 3. Use of optical, radar, or laser remote-sensing data the beginning and end of the period to calculate integrated with allometry and ground measurements. net emissions or removal (stock difference approach) Soil carbon assessment in different parts of the world re- 3. Estimating the gain in carbon stock for each pool due quires methods that are appropriate to the circumstances. to accumulation or losses and calculating the differ- Many different methods have been tested in a number of ence between gains and losses as net emissions or countries, but effort is required to ensure that the methods removal (a gain-loss approach). are comparable. Furthermore, for carbon projects, credible CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 13 TABLE 2.6: Comparison of Carbon Assessment for Carbon Mitigation and Non-Carbon-Mitigation Projects NONCARBON-MITIGATION PROJECTS (SUSTAINABLE LAND MANAGEMENT INCLUDING PROJECT PHASE CARBON MITIGATION PROJECTS FOREST, GRASSLAND, CROPLAND MANAGEMENT) Conceptualization Primary focus: carbon mitigation and carbon credits—global Primary focus: forest and biodiversity conservation, watershed environmental bene�t protection, and livelihoods enhancement Secondary focus: soil and biodiversity conservation Cobene�ts: carbon mitigation is implicit though often not men- tioned in proposal Proposal development Clear historical records of the past vegetation and soil carbon Historical vegetation status not so critical to project eligibility status are required Project boundary needed for estimating environmental and Project boundarya impacted by project activities needs clear socioeconomic bene�ts restricted to project area de�nition Baseline economic bene�ts, soil fertility, and biodiversity need Estimation of baseline carbon stocks is crucial as well as to be clearly identi�ed. Also, well-de�ned plan is required for rigorous plan for monitoring carbon stock changes monitoring of local environmental and socioeconomic impacts Project review and appraisal Baseline and project scenario carbon monitoring methods are Monitoring plan for local environmental and socioeconomic critical bene�ts is important Implementation Activities aimed at maximizing carbon bene�ts, followed by other Activities are aimed at maximizing biomass production, crop cobene�ts yields, biodiversity conservation, and livelihood improvement Monitoring and evaluation Approved methodologiesb are crucial. Additionality must be Project-speci�c methodology is used demonstrated Additionality of local environmental and socioeconomic bene�ts All the relevant carbon pools must be considered are critical Large transaction cost likely for carbon inventory and monitoring Soil carbon critical for land development projects due to effects on agricultural sustainability Moderate transaction cost for monitoring Source: Modi�ed from Ravindranath and Ostwald (2008). a Project boundary refers to the physical boundary of the land area delineated either with a geographical information system or a global positioning system and the greenhouse gas boundary that includes all fluxes of all gases affected by project activity. b Carbon assessment methodologies are the blueprints to design, verify, and operate carbon projects. They document the protocol for quantifying carbon emissions and removals and include guidelines for identifying baseline scenario and assessing additionality in all carbon pools relevant to the project. and cost-effective techniques of monitoring changes in soil In Finland, Makipaa et al. (2008) observed that organic layer carbon are required. carbon measurements cost 520 per plot if 10 samples are analyzed. The precision obtained with such sampling cor- Soil carbon assessment methods can be broadly classi�ed responds to detection of soil carbon change greater than into direct and indirect methods depending on whether car- 860 g C m−2. At the national level, two measurements for a bon content in soil samples is directly measured or inferred minimum of 3,000 plots are needed to detect an expected through a proxy variable (table 2.7). Most assessments typi- change of 11 g C m−2 yr−1 in the organic layer of upland forest cally involve a combination of these techniques. Each of the soils at 10-year sampling intervals. One round of measure- methods depicted in table 2.7 has unique constraints related ment was estimated to cost about 4 million, corresponding to costs, inadequacies, geographic scope, and sampling de- to 8 percent of the value of the annual sequestration of about sign requirements and associated levels of bias or uncertainty. 3 million tCO2 of Finland’s upland forest soils. Strategies to The most established type of direct soil carbon assessment reduce the cost of soil carbon monitoring include lengthening entails collecting soil samples in the �eld and analyzing them the sampling interval, increasing the ef�ciency of sampling in the laboratory by combustion techniques. Field sampling is through strati�cation, pooled sampling, use of in situ analyti- technically challenging, but it can be addressed with appropri- cal methods, and the use of biogeochemical models. ate design that accounts for soil spatial variation. The degree and nature of sampling depend on the carbon assessment Several in situ soil carbon analytical methods are being de- objective, whether for national or regional accounting or for veloped with the objective of offering increased accuracy, carbon offset project. Each context will require a differing precision, and cost-effectiveness over conventional ex situ degree of granularity and measurement set to assess un- methods. A comparison of these techniques is provided in certainty in the estimates. The direct method, though more table 2.8. Most of the in situ techniques are still in their in- precise and accurate, is quite laborious and very expensive. fancy. The exception is infrared spectroscopy currently being EC O N O M I C A N D S E CT OR WORK 14 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S TABLE 2.7: Direct and Indirect Methods of Soil Carbon Assessment DIRECT METHODS INDIRECT METHODS 1. Field sampling and laboratory measurements using dry combustion or Accounting techniques wet combustion Strati�ed accounting with database Remote sensing to infer factors determining above-ground carbon inputs 2. Eddy covariance; flux tower measurements Biogeochemical/ecosystem simulation modeling to understand below-ground biological processes, for example, RothC Century DNDC PROCOMAP CO2FIX 3. Emerging technologies for in situ determination Laser-Induced Breakdown Spectroscopy Inelastic Neutron Scattering (still being assessed for improved reli- ability for measurement) Near-infrared and mid-infrared spectroscopy Source: Modi�ed from Post et al. (2001). TABLE 2.8: Characteristics of Emerging In Situ Methods of Soil Carbon Analytical Techniques TYPE OF PENETRATION SAMPLED RADIATION DEPTH VOLUME DIRECT METHOD PROCESS MEASURED (CM) (CM3) ADVANTAGES DISADVANTAGES Mid-infrared Molecular/diffuse Infrared 1 10 In situ–based Costs are prohibitive spectroscopy reflectance measurement of on per project basis carbon. Better than near infrared in distinguishing soil organic from inorganic carbon. Near-infrared Molecular/diffuse Near infrared 0.2 1 Rapid, low cost, Less accurate than spectroscopy reflectance in situ method mid-infrared in pre- dicting soil organic carbon Laser-induced break- Atomic/ Visible 0.1 0.1 Very fast—provides Interference with iron down spectroscopy plasma-induced total soil carbon compounds around emission measurements in 248 nm wavelength; seconds; capable of currently, the spectrally resolving technology cannot several elements directly distinguish apart from carbon soil inorganic from organic carbon Inelastic neutron Nuclear/neutron- Gamma rays 30 100,000 Large footprint of The technology is scattering induced nuclear about 2 m2 and still at its infancy reactions sampling depth and needs to be calibrated for wide variety of soil types, and scanner must be adapted to capture large areas Source: Adapted from Chatterjee and Lal (2009). CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 15 used to develop a spectral library for soils of the world.3 The Monitoring and verifying soil carbon sequestration at the spectral library provides a valuable resource for rapid char- project or regional scale require �ve components (Post et al. acterization of soil properties for soil quality monitoring and 1999). These include the selection of landscape units suit- other agricultural applications. able for monitoring soil carbon changes, development of measurement protocols, application of remote sensing to Indirect estimation of soil organic carbon changes over large estimate soil organic carbon controlling parameters, spatially areas using simulation models is increasingly important to explicit biogeochemical modeling, and scaling-up the results �ll knowledge gaps about the biogeochemical processes to the entire project area (table 2.9). of soil carbon sequestration. Simulation models describe changes in soil organic carbon under varying climate, soil, Monitoring the trends in soil carbon over a large geographi- and management conditions. Though the models could have cal area through repeated sampling is mainly restricted to limited accuracy, they are particularly useful in the context of developed and few developing countries. Examples of national developing countries where land resources data are scarce. carbon accounting system and tools are presented in table 2.11. Models provide a cost-effective means of estimating GHG emissions in space and time under a wide range of biophysi- Progress is being made in developing and testing cost-effec- cal and agricultural management conditions, and they are tive soil carbon monitoring methods. The Global Environment particularly useful for up-scaling site-speci�c information to Facility (GEF) in collaboration with other partners is currently the regional level. Table 2.9 compares the features of some implementing the Carbon Bene�ts Project (CBP) to develop of the biogeochemical models commonly used for soil car- standardized, cost-effective methods of quantifying the bon assessment. carbon bene�ts of sustainable land management projects.4 TABLE 2.9: Comparative Features of Some Carbon Estimation Models MODEL FEATURES KEY INPUTS KEY OUTPUTS CENTURY Simulates long-term dynamics of carbon, Monthly mean maximum and minimum air Total carbon, soil water dynamics, commercial nitrogen, phosphorus, and sulfur for different temperature and total precipitation; plant N, crop yield, total dry matter, and carbon in plant ecosystems P, and S content; soil texture; atmospheric and residue soil nitrogen inputs; and initial soil carbon, nitrogen, phosphorus, and sulfur levels CO2FIX Simulates carbon dynamics of single/multiple Simulation length, maximum biomass in stand, Carbon stocks and fluxes, total biomass and species, forests, and agroforestry systems carbon content, wood density, initial carbon, soil carbon, above- and below-ground biomass, yield tables, precipitation, temperature, and deadwood, and litter and soil organic carbon length of growing period production RothC Estimation of turnover of organic carbon in Clay, monthly rainfall, monthly open pan evapo- Total organic carbon content and carbon topsoil ration, average monthly mean air temperature, content in microbial biomass and an estimate of the organic input PROCOMAP Equilibrium model for estimating carbon stocks Activity data, planting rate, vegetation carbon Biomass and soil carbon stock, incremental stocks, rotation period, and mean annual incre- carbon stocks, and cost-effectiveness ment in biomass and soil indicators DNDC DeNitri�cation-DeComposition is used for Plant growth data, soil clay, bulk density, pH, Total carbon, total nitrogen, soil water dynam- predicting crop growth, soil temperature and air temperature, rainfall, atmospheric nitrogen ics, biomass carbon, carbon dioxide, crop yield, moisture regimes, carbon sequestration, decomposition rate, crop rotation timing and carbon input into soil, fluxes of gases including nitrogen leaching, and emissions of nitrous type, inorganic fertilizer timing, amount and N2O, nitric oxide NO, NH3, and methane CH4 oxide (N2O), nitric oxide (NO), dinitrogen (N2), type, irrigation timing and amount, residue ammonia (NH3), methane (CH4), and carbon incorporation timing and amount, and tillage dioxide (CO2) timing and type Source: This study. 3 World Agroforestry Centre, ISRIC-World Soil Information: A glob- ally distributed soil spectral library: visible near-infrared diffuse reflectance spectra, http://www.africasoils.net/sites/default/ �les/ICRAF-ISRICSoilVNIRSpectralLibrary.pdf. 4 http://www.unep.org/climatechange/carbon-bene�ts/ EC O N O M I C A N D S E CT OR WORK 16 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S TABLE 2.10: Components of Soil Carbon Monitoring at the Regional Scale COMPONENTS DESCRIPTION Selection of landscape units The selection will depend on responsiveness of the area to land management practices as determined by climate, soil properties, management history, and availability of historical data. Participation of local agronomists, farmer organizations, and other stakeholders can be of help in selecting pilot areas and the extent to which the results can be extrapolated over the region. Development of protocol Changes in soil carbon can generally be estimated as changes in stocks (from direct measurement) or fluxes (using eddy covariance methods) (see table 2.7). Protocols for temporally repeated measurements at �xed locations will generally include strati�cation and selection of sampling sites, sampling depth and volume, measurement of bulk density, laboratory analyses, other ancillary �eld measurements, and estimation of the marginal cost of carbon sequestration. Application of remote sensing Remote sensing can provide information on net primary productivity, leaf area index, tillage practices, crop yields, and location and amount of crop residue for input into models. Recently, cellulose absorption index derived from remote imaging spectroscopy has been used to infer tillage intensity and residue quantity (Serbin et al. 2009) Biogeochemical modeling Models are used to determine soil carbon changes over large areas because satellites cannot sense below-ground biological processes. Models are useful for understanding soil properties–land management interactions and for predicting soil carbon sequestration. They can simulate full ecosystem–level carbon balance, multiple land uses, or several land management practices Up-scaling Scaling-up to large areas requires integration from a variety of sources including �eld measurements, existing database, models, geographical information system, and remote sensing. Multitemporal moderate resolution remote sensing such as Landsat Thematic Mapper and Moderate Resolution Imaging Spectroradiometer can provide information such as land-use and land cover change, crop rotations, and soil moisture that can markedly improve up-scaling of soil carbon assessment. Source: Synthesized from Post et al. 1999. TABLE 2.11: Carbon Accounting Systems and Tools NAME DESCRIPTION AND INTERNET LOCATION Australia’s National Carbon NCAS estimates emissions through a system that combines satellite images to monitor land use and land-use change across Accounting System (NCAS) Australia that are updated annually; monthly maps of climate information, such as rainfall, temperature, and humidity; maps of soil type and soil carbon; databases containing information on plant species, land management, and changes in land manage- ment over time; and ecosystem modeling—the Full Carbon Accounting Model. http://www.climatechange.gov.au/government/initiatives/national-carbon-accounting.aspx National Forest Carbon Monitoring, NFCMARS is designed to estimate past changes in forest carbon stocks and to predict, based on scenarios of future distur- Accounting and Reporting System, bance rates and management actions, changes in carbon stocks in the next two to three decades. Canada (NFCMARS) http://carbon.cfs.nrcan.gc.ca/index_e.html Agriculture and Land Use National The program supports countries’ efforts to understand current emission trends and the influence of land-use and management Greenhouse Gas Inventory Software alternatives on future emissions. It can be used to estimate emissions and removals associated with biomass C stocks, soil C (Colorado State University, United stocks, soil nitrous oxide emissions, rice methane emissions, enteric methane emissions, manure methane, and nitrous oxide States) emissions, as well as non-CO2 GHG emissions from biomass burning. The software accommodates Tier 1 and 2 methods as de- �ned by the Intercontinental Panel on Climate Change. It allows compilers to integrate global information system spatial data along with national statistics on agriculture and forestry and is designed to produce a consistent and complete representation of land use for inventory assessment. http://www.nrel.colostate.edu/projects/ghgtool/software.php National Carbon Accounting System Provides monitoring capabilities for greenhouse gas (GHG) emissions/sinks to establish a credible reference emission level. of Indonesia The three major activities linked are the remote sensing program, the modeling and measurement program for GHG accounting and reporting, and the data program. http://www.dpi.inpe.br/geoforest/pdf/group2/04%20-%20National%20carbon%20accounting%20system%20of%20Indonesia.pdf New Zealand’s Carbon Accounting The National Carbon Accounting System for New Zealand’s indigenous forest, shrub land, and soils was developed for the System Ministry of the Environment by Landcare Research and Scion. It monitors forest de�nition, land-use change, forest inventory and modeling, and reporting methods. http://www.joanneum.at/carboinvent/workshop/1000_Peter_Stephens_ver_�nal.pdf Forest Vegetation Simulator, United The Forest Vegetation Simulator (FVS) is a family of forest growth simulation models. The basic FVS model structure has been States calibrated to unique geographic areas to produce individual FVS variants. Since its initial development in 1973, it has become a system of highly integrated analytical tools. http://www.fs.fed.us/fmsc/fvs/description/index.shtml Source: This study. The new suite of tools estimate and model carbon and other Management Evaluation Tool (http://www.cometvr. GHG flows under present and alternative management and colostate.edu/), and the GEF Soil Organic Carbon System measures. They also monitor changes in carbon under speci- that approximates national- and subnational-scale soil carbon �ed land use and management. The CBP comprises a na- stock variations in developing countries using RothC and tional GHG inventory tool; the Agriculture and Land Use Tool Century models (table 2.9). (table 2.11), Voluntary Reporting of Greenhouse Gases-Carbon CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 2 — S O I L ORGANIC CARBON DYNAMICS A ND A SS ES S MENT METH OD S 17 2.4 CARBON ASSESSMENT IN THE farmers in developing countries (http://www.v-c-s.org/sites/ WORLD BANK’S SUSTAINABLE LAND v-c-s.org/files/VM0017%20SALM%20Methodolgy%20 MANAGEMENT PORTFOLIO v1.0.pdf). The methodology, referred to as Sustainable Carbon Assessment Using the Ex Ante Appraisal Agricultural Land Management (SALM), provides protocol Carbon-Balance Tool for quantifying carbon emissions and removals and includes The World Bank is increasingly looking to assess the carbon guidelines for identifying baseline scenario and assessing footprint of its operational work across sectors. Emphasis is additionality in all carbon pools relevant to sustainable land placed on cost-effective approaches that do not add exces- management projects. SALM is applicable to projects that sively to the burden of project management. The Ex-Ante introduce sustainable land management practices into crop- Appraisal Carbon-Balance Tool (EX-ACT; http://www.fao. lands subject to conditions such that soil organic carbon org/tc/exact/en/) has been developed with this objective in would remain constant or decrease with time in the absence mind. of the project. The methodology currently being applied in the �rst African soil carbon project allows small-holder EX-ACT can provide ex ante assessments of the impact of farmers in Kenya to access the carbon market and receive agriculture and related forestry, �sheries, livestock, and water additional carbon revenue streams through the adoption of development projects on GHG emissions and carbon seques- productivity-enhancing practices and technologies. tration, thereby indicating the overall effects on the carbon balance. EX-ACT was developed following the IPCC guideline for national GHG inventory (IPCC 2006), supplemented by other existing methodologies and reviews of default coef�- REFERENCES cients. It is easy to use in the context of program formulation; Batjes, N. H. 1996. “Total Carbon and Nitrogen in the Soils of the it is cost-effective and requires a minimum amount of data. It World.� European Journal of Soil Science 47: 151–163. also has resources (linked tables and maps) that can assist in Bond-Lamberty, B., and Thomson, A. 2010. “Temperature-Associated gathering the information necessary to run the model. While Increases in the Global Soil Respiration Record.� Nature 464: 579–582. EX-ACT primarily works at the project level, it can easily be up-scaled at the program/sector level. Chatterjee, A., and Lal, R. 2009. “On Farm Assessment of Tillage Impact on Soil Carbon and Associated Soil Quality Parameters.� Carbon assessment in EX-ACT is implemented in the follow- Soil Tillage Research. 104: 270–277. ing three steps: Friedlingstein, P., Dufresne, J., and Cox, P. 2003. “How Positive is the Feedback Between Climate Change and the Global Carbon General description of the project (geographic area, cli- Cycle?� Tellus 55B: 692–700. mate and soil characteristics, and duration of the project) Eswaran, H., van den Berg, E., Reich, P. 1993. ``Organic Carbon in Identi�cation of changes in land use and technologies Soils of the World,’’ Soil Science Society of America Journal 57, 192–194. foreseen by project components using speci�c “mod- ules� (deforestation, forest degradation, afforestation/ Gibbs, H. K., Brown, S., Niles, J. O., and Foley, J. A. 2007. “Monitoring and Estimating Tropical Forest Carbon Stocks: Making REDD a reforestation, annual/perennial crops, rice cultivation, Reality.� Environmental Research Letters 2: 045023. grasslands, livestock, inputs, organic soils, and energy) Global Carbon Project. 2009. Policy Brief November 2009. No. 10. Computation of carbon balance with or without the project using IPCC default values and, when available, Ingram, J. S. I., and Fernandes, E. C. M. 2001. “Managing Carbon Sequestration in Soils: Concepts and Terminology.� Agriculture, ad hoc coef�cients. Ecosystems and Environment 87: 111–117. A detailed analysis of lessons learned in testing EX-ACT in IPCC. 1996. Guidelines for National Greenhouse Gas Inventories Vol. World Bank agriculture projects can be found in World Bank 1–3. Intergovernmental Panel on Climate Change, London. (2012). IPCC. 2006. Guidelines for National Greenhouse Gas Inventories Vol. 4: Agriculture, Land Use, and Forestry. http://www.ipcc-nggip. iges.or.jp. Sustainable Agricultural Land Management IPCC. 2007. Climate Change 2007: Synthesis Report—Summary for Methodology Policymakers. Fourth Assessment Report. The BioCarbon Fund of the World Bank has recently devel- Kell, D.B. 2011. “Breeding Crop Plants With Deep Roots: Their Role oped a carbon accounting methodology to encourage adop- in Sustainable Carbon, Nutrient, and Water Sequestration.� Ann. tion of sustainable land management practices by small-scale Bot. 108 (3): 407–418. EC O N O M I C A N D S E CT OR WORK 18 CHAPTER 2 — S OIL OR GA NIC CA R B ON D YNA MIC S A ND A SS ES S MENT M ETH OD S Lal, R. 2003. “Soil Erosion and the Global Carbon Budget.� Raich, J., and Potter, C. 1995. “Global Patterns of Carbon Dioxide Environment International 29: 437–450. Emissions From Soils.� Global Biogeochemical Cycles 9: 23–36.0 Lal, R. 2004. “Carbon Emission From Farm Operations.� Environment Ravindranath, N. H., and M. Ostwald. 2008. Carbon Inventory International 30: 981–990. Methods Handbook for Greenhouse Gas Inventory, Carbon Mitigation and Roundwood Production Projects. Advances in Makipaa, R., Hakkinen, M., Muukkonen, P., et al. 2008. “The Costs Global Change Research. Springer-Verlag, Berlin. of Monitoring Changes in Forest Soil Carbon Stocks.� Boreal Environmental Research 13: 120–130. U.S. Department of Agriculture, Natural Resources Conservation Service. “Global Soil Regions.� http://soils.usda.gov/use/world- Post, W. M., Izaurralde, R. C., Mann, L. K., and Bliss, N. soils/mapindex/orders.jpg. 1999. “Monitoring and Veri�cation of Soil Organic Carbon Sequestration.� In Proceedings of the Symposium on Carbon Watson, Robert et al., ed. 2000. Land Use, Land-Use Change, Sequestration in Soils Science, Monitoring and Beyond, ed. N.J. and Forestry. Intergovernmental Panel on Climate Change. Rosenberg, R.C. Izaurralde, and E.L. Malone, Columbus, OH: Cambridge University Press, Cambridge. Batelle Press. World Bank. 2012. Carbon Foot-Printing of ARD Projects: Testing Post, W. M., Izaurralde, R. C., Mann, L. K., and Bliss, N. 2001. the Ex-Ante Carbon Balance Appraisal Tool. World Bank, “Monitoring and Verifying Changes of Organic Carbon in Soil.� Washington, DC. Climatic Change 51 (1): 73–99. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 19 Chapter 3: META-ANALYSES OF SOIL CARBON SEQUESTRATION 3.1 INTRODUCTION livestock and manure management. The impacts of changes A range of practices has been suggested as important to in agricultural practices on soil carbon stocks such as chang- soil carbon sequestration and thus of potential relevance es to crop rotation or reduced grazing are usually more subtle to increasing crop yield, increasing the resilience of agro- than those brought about by more dramatic changes in land ecosystems, and mitigating GHG emissions (table 3.1). use such as conversion of cropland to forest or grassland to Mitigation of GHG in agriculture can involve several practices tree crops. such as avoiding the conversion of native forests and grass- This chapter documents the evidence from the published lit- lands to croplands; enhancing removal of carbon from the erature on the impacts of agricultural land management prac- atmosphere through a range of soil and water management tices and agricultural land-use changes on soil carbon seques- practices including crop diversi�cation; restoration of barren, tration in Africa, Asia, and Latin America. The main emphasis abandoned, or seriously degraded agricultural lands; and is on obtaining better estimates of soil carbon sequestration TABLE 3.1: Practices That Sequester Carbon in Forest, Grassland, and Cropland FOREST GRASSLAND CROPLAND Protection of existing forests—Avoided deforesta- Improved grassland management—Optimize stocking No or reduced tillage—Reduces the accelerated tion preserves existing soil C stocks and prevents rates to reduce land degradation, depletion of soil decomposition of organic matter associated with emissions associated with biomass burning and soil organic carbon, and methane emissions through intensive (conventional or traditional) tillage exposure by land clearing enteric fermentation Reforestation—Increasing tree density in degraded Introduction of improved pasture species and legumes Mulching/residue management—Improves soil mois- forests increases carbon accumulation to increase above- and below-ground biomass produc- ture, prevents soil erosion, and increases soil organic tion and soil organic carbon accumulation matter when incorporated into the soil; crop residues also prevent loss of carbon from the soil system Afforestation—Establishment of new forests on Application of inorganic fertilizers and manure to Application of inorganic fertilizers and manure to nonforest land (cropland, grassland, or degraded stimulate biomass production—Chemical fertilizers stimulate biomass production—Chemical fertilizers lands) increases carbon stock through the increase are, however, less environmentally friendly due to are, however, less environmentally friendly due to in above-ground biomass as well as greater organic nitrous oxide (N2O) emissions associated with N fertil- nitrous oxide (N2O) emissions associated with N fertil- materials input for soil decomposition izers, the greenhouse cost of fertilizer production, and izers, the greenhouse cost of fertilizer production, and emissions associated with transport of fertilizers emissions associated with transport of fertilizers Water management to increase productivity, but this Use of cover crops/green manure increases the has to be put in the perspective of emissions associ- biomass returned to the soil and thus increases soil ated with the process of irrigation carbon stock Introduction of earthworms to improve aeration and Use of improved crop varieties—Improved crop aid organic matter decomposition in the soil pro�le varieties help to sequester carbon in the soil through increased above- and below-ground biomass production Establishment of pasture on degraded land reintro- Agroforestry/tree-crop farming—Introduction of fruit duces large amounts of organic matter into the soil trees, orchards, and woodlots into croplands helps to store more carbon, optimize water use, diversify production, and increase income Introduction of improved crop varieties Application of biochar and other soil amendments Source: This study. EC O N O M I C A N D S E CT OR WORK 20 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION rates. This is one important element in making comprehen- Analysis sive assessments of the impact of soil quality on agricultural Most studies reported concentrations of carbon in soil sustainability and greenhouse mitigation potentials. samples (Cc in g kg−1). These were converted to volumes and then areas to calculate stocks (Cs in kg−1 ha−1) and sequestra- tion rates (kg ha−1 yr−1) using bulk density (BD, in g cm−3) and 3.2 METHODS sample soil depth (D, in cm): Searches and Data Sources Searches were carried out using online database and Cs = BD x Cc x D x 10,000 search tools, including ProQuest, Scopus, Sciencedirect, In a few studies, value was given in terms of percent soil SpringerLink, Wiley Science Library, and Google Scholar with organic matter. In these cases, concentrations of Cc (g kg−1) an emphasis on key terms such as soil organic matter, or- were calculated as ganic matter, soil organic carbon, soil carbon, carbon seques- tration, soil sequestration, and soil properties, in combination Cc = 0.58 x OM% x 10 with geographical descriptors (e.g., countries and continents) and terms for particular agricultural practices. In some cases, only a single value, either initial or average across treatments, was provided for bulk density. In these Inclusion-Exclusion Criteria cases, that value was assumed to apply to all treatments. If For soil fertility and surface management effects that are no bulk density information was provided in a paper (or other commonly studied in agricultural science, only studies of at reports about the same study cited by that paper), then bulk least 3 years duration were included. A major effort was made density was estimated using known pedotransfer functions to collect data from as many long-term studies as possible. (that is, simple regression equations) developed for that region Almost all studies adopted formal experimental designs, or extracted from the International Soil Reference Information setting up control and treatments. The variations applied in Center–derived soil properties database (www.isric.org). the treatments accounted for the different levels of carbon Effect sizes and importance of contextual variables (e.g., added to the soil. In a few cases where paired designs were temperature, precipitation, duration, and soil type) were employed, logical contrasts were made with appropriate con- summarized by means and 95 percent con�dence intervals trols using �nal values of stocks under each treatment. for the mean. Associations of the context variables with car- Experimental study designs are rare for land-use change bon sequestration were assessed by grouping observations effects. Most adopted nonexperimental designs such as into a few classes so that nonlinear patterns could be clearly chronosequence where adjacent plots of different ages were identi�ed. Geographical distribution of datasets is shown in compared, paired studies where adjacent plots of different �gure 3.1, while the characteristics of the estimates with land uses and similar ages were compared, or repeated sam- respect to duration of study, soil sampling depth, and experi- ples where same plot was measured over time. Only studies mental design are shown in Appendix 3.1. of at least 4 years duration were included, and where repeat- ed measures were made, sequestration rates for the longest time interval were taken. A major reason for excluding papers 3.3 RESULTS with data on different land uses was dif�culty in assuming Contextual Factors and Soil Carbon Sequestration particular sites could be taken as a reasonable control. Climate Climate signi�cantly influences large-scale patterns of soil Effect Sizes carbon sequestration. In this study, higher sequestration The effect of a land management practice was estimated by rates were observed in the wettest locations (�gure 3.2). comparing the �nal level of soil carbon stock in one treatment There was also a trend to lower sequestration rates in the with that practice and an appropriate control. Thus, all soil coolest and warmest conditions (�gure 3.3). Sites in warmer carbon sequestration rates are estimates of effect size—the and middle temperature regions tended to accumulate soil difference with respect to a control—and thus represent the carbon more rapidly than those in colder regions, whereas marginal bene�t of adopting that practice. Effect sizes were semi-humid areas had higher average rates than their semi- estimated for all logical contrasts with suf�cient information arid counterparts. Potter et al. (2007) explored interactions provided in a paper. with residue management practices in maize �elds at six CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 21 FIGURE 3.1: Geographical Distribution of Carbon Sequestration Estimates Carbon sequestration number of estimates 1–10 11–30 31–50 51–100 101–200 201–630 Source: This study. sites across a wide range of annual temperature regimes in soil pro�les give some indication of humus, clay minerals, or Mexico and discovered that as temperature increased, more metal oxides accumulating in their layers. In Asia, the highest crop residues needed to be retained to increase levels of soil sequestration rates and variability were observed on oxisols, organic carbon. An increase in soil temperature exacerbated formed principally in humid tropical zones under rain forest, the rate of mineralization, leading to a decrease in the soil scrub, or savanna vegetation on flat to gently sloping up- organic carbon pool (SOC). However, decomposition by- lands. Oxisols are typically found on old landscapes that have products at higher temperatures may be more recalcitrant been subject to shifting cultivation for several years. than those at lower temperatures. Duration Soils Longer term studies on average have resulted in lower Soil type, especially those with a higher clay content, leads to sequestration rates, as would be expected from saturation higher carbon sequestration rates. However, obtaining com- (�gure 3.5). Most of the potential soil carbon sequestration parable data is dif�cult as not all studies provide suf�cient takes place within the �rst 20 to 30 years. The pattern of information on soil properties, and those that do use differ- change in sequestration rates is nonlinear and differs be- ent soil classi�cation schemes at different levels of detail. As tween major groups of practices, with the highest rates at a �rst-level analysis, the reported soil types were reclassi�ed intermediate times and low or even negative rates in the into major soil orders of the U.S. Department of Agriculture short term. classi�cation system (�gure 3.4). Carbon sequestration rates were highest and also highly variable on inceptisols in Africa Nutrient Management and Latin America. Inceptisols are relatively young soils Fertilizer use sequesters carbon by stimulating biomass pro- characterized by having only the weakest appearance of ho- duction. Judicious fertilizer application also counters nutrient rizons, or layers, produced by soil-forming factors. Inceptisol depletion, reduces deforestation and expansion of cultivation EC O N O M I C A N D S E CT OR WORK 22 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION FIGURE 3.2: Soil Carbon Sequestration and FIGURE 3.3: Soil Carbon Sequestration and Precipitation Temperature Africa Africa 3,000 3,000 carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 2,500 2,500 2,000 2,000 1,500 1,500 1,000 1,000 500 500 0 0 <500 500–1,000 1,001–1,500 <1,500 <20 20–25 25–30 >30 mean annual temperature (°C) annual precipitation (mm) Asia Asia carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 1,250 800 1,000 600 750 400 500 250 200 < 10 10 – 20 20 – 30 30+ <500 500–1,000 1,001–1,500 1,500+ annual precipitation (mm) mean annual tempeature (°C) Latin America Latin America 800 1,200 carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 1,000 600 800 600 400 400 200 200 0 –200 <500 500–1,000 1,000–1,500 1,500+ <15 15.20 20.25 25+ annual precipitation (mm) mean annual temperature (°C) Source: This study. Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 23 FIGURE 3.4: Soil Carbon Sequestration and Soil Order to marginal areas, and increases crop yields. Strategies to promote nutrient use ef�ciency include the following: Africa Vertisols Adjusting application rates based on assessment of crop needs Ultisols Minimizing losses by synchronizing the application of nutrients with plant uptake Oxisols soil type Correcting placement to make the nutrients more Mollisols accessible to crop roots (microfertilization and Inceptisols microdosing) Using controlled-release forms of fertilizer that delay Andisols its availability for plant uptake and use after Alfisols application Using nitri�cation inhibitors that hold-up microbial 0 1,000 2,000 3,000 4,000 processes leading to nitrous oxide formation carbon sequestration (kg C/ha/yr) The average effect size of applying fertilizer was an additional 124 kg C ha−1 yr−1 sequestered for Latin America, 222 kg C ha−1 yr−1 for Asia, and 264 kg C ha−1 yr−1 for Africa (table 3.2). The ma- Asia jority of studies have designs focused on the influence of dif- Vertisols ferent levels of nitrogen and, in some cases, the combination Ultisols of fertilizer with locally available manure sources. Aggregating across locations and cropping systems there was no signi�cant Oxisols association between level of N applied across annual cropping Mollisols soil type cycles and carbon sequestration rates (�gure 3.6). Inceptisols Entisols Across the full dataset, studied average sequestration Aridisols rates with NPK - Nitrogen, Phosphorus and Potassium Andisols compound fertilizers N = Nitrogen, P = Phosphorus, K = Potassium were signi�cantly higher than other combi- Alfisols nations (Figure 3.7). Within individual experiments, some 0 1,000 2,000 3,000 4,000 studies show that integrated management of N, P, and K carbon sequestration (kg C/ha/yr) fertilizers is important to maintaining or increasing soil car- bon and nitrogen and thus soil fertility. Alvarez’s (2005a) analysis of a global dataset indicated that Latin America for every additional tonne of nitrogen fertilizer applied, two more tonnes of soil organic carbon were stored in fertilized Vertisols than unfertilized plots. Soil organic carbon levels clearly in- Ultisols creased under nitrogen fertilization only when crop residues Oxisols were returned to the soil. Another meta-analysis at the global level concluded that addition of nitrogen fertilizer resulted, on soil type Mollisols Inceptisols average, in a 3.5 percent increase in soil carbon in agricultural ecosystems (Lu et al. 2011). Entisols Andisols Biofertilizers are an essential component of organic farming. They contain living microorganisms that colonize the rhizo- Alfisols sphere and promote plant growth by increasing the supply –2,000 –1,000 0 1,000 2,000 3,000 4,000 of nutrients through nitrogen �xation or enhancing the avail- carbon sequestration (kg C/ha/yr) ability of primary nutrients to the host plant by solubilizing Source: This study. phosphorus and other nutrients. The microorganisms in EC O N O M I C A N D S E CT OR WORK 24 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION FIGURE 3.5: Soil Carbon Sequestration and Time Africa 1,600 4,500 Tillage and residue carbon sequestration (kg C/ha/yr) 1,400 management 4,000 All carbon sequestration (kg C/ha/yr) 1,200 3,500 1,000 3,000 2,500 800 2,000 600 1,500 400 1,000 200 500 0 <5 5–10 11–20 0 <5 5–10 11–20 >30 duration of study (years) duration of study (years) 4,500 3,000 Land-use change Nutrient 4,000 manangement carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 2,500 3,500 3,000 2,000 2,500 1,500 2,000 1,000 1,500 1,000 500 500 0 0 <5 5–10 11–20 <5 5–10 11–20 20–30 >30 duration of study (years) duration of study (years) Asia Tillage and residue management 1,250 All 600 carbon sequestration (kg C/ha/yr) 1,000 400 750 200 500 250 0 <5 5–10 10–20 20–30 30+ <5 5–10 10–20 20–30 30+ duration of study (years) duration of study (years) CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 25 1,000 2,500 Land-use change Nutrient management 95% CI carbon sequestration (kg C/ha/yr) 800 2,000 600 1,500 400 1,000 200 500 0 0 <5 5–10 10–20 20–30 30+ <5 5–10 10–20 20–30 30+ duration of study (years) duration of study (years) Latin America 1,200 All 1,500 Tillage and residue management 1,000 carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 1,000 800 600 500 400 0 200 0 –500 <5 5–10 10–20 20–30 30+ <5 5–10 10–20 20–30 30+ duration of study (Years) duration of study (Years) Nutrient 1,500 Land-use change 750 management 1,000 carbon sequestration (kg C/ha/yr) carbon sequestration (kg C/ha/yr) 500 500 250 0 0 –500 –1,000 –250 –1,500 0 <5 5–10 10–20 20–30 30+ <5 5–10 10–20 20–30 30+ duration of study (years) duration of study (years) Source: This study. EC O N O M I C A N D S E CT OR WORK 26 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION FIGURE 3.6: Soil Carbon Sequestration and Application Levels of Nitrogen Fertilizer (Means and 95 Percent Con�dence Intervals, n = 285) 600 carbon sequestration (kg C/ha/yr) 400 200 0 <100 100–200 200–300 300+ N fertilizer (kg N/ha/yr) Source: This study. Note: N = Nitrogen. FIGURE 3.7: Soil Carbon Sequestration and Fertilizer Combinations (Means and 95 Percent Con�dence Intervals, n = 285) 500 400 carbon sequestration (kg C/ha/yr) 300 200 100 0 N NP PK NK NPK fertilizer mix Source: This study. Note: N = Nitrogen only; NP = Nitrogen and Phosphorus only; PK = Phosphorus and Potassium only; NK = Nitrogen and Potassium only; NPK = combination of Nitrogen, Phosphorus and Potassium. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 27 TABLE 3.2: Nutrient Management and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 PERCENT UPPER 95 PERCENT CONFIDENCE CONFIDENCE INTERVAL INTERVAL NUMBER OF PRACTICE MEAN OF MEAN OF MEAN ESTIMATES Africa Chemical fertilizer 264 169 359 30 Manure 325 224 427 30 Asia Chemical fertilizer 222 157 288 297 Manure 465 374 556 146 Biofertilizer 1,459 −42 2,960 3 Latin America Chemical fertilizer 124 −15 262 74 Manure 455 23 887 25 Source: This study. biofertilizers restore the soil’s natural nutrient cycle and help also increased with manure application and accumulation of in building soil organic matter. Biofertilizers are more envi- soil carbon but with patterns that depend on crop. In China, ronmentally friendly and cost-effective relative to chemical yields of maize and maize-wheat systems increased over fertilizers. Three studies reviewed indicate that biofertilizers the longer term, while in rice-based systems, the gains hap- sequestered about 1.4 t C ha−1 yr−1. pened in the �rst few years and were not followed by further yield improvements (Zhang et al. 2009). Manure application to agricultural soils can reduce nitrous oxide emissions by displacing N fertilizer use. Methane emis- One major constraint is the availability of manure and labor sions can also be minimized by displacing anaerobic storage costs associated with collecting and processing it. Studies options with aerobic decomposition. These bene�ts have in Nepal (Acharya et al. 2007) and Thailand (Matsumoto, already been recognized in efforts to divert organic waste Paisancharoen, and Hakamata 2008) have pointed to trends from land�lls. Pattey, Trzcinski, and Desjardins (2005) found of declining livestock numbers and speculated on impacts of that compared to untreated manure storage, composting this on manure application practices. The impact of manure reduced total GHG emissions (CH4 and N2O) by 31 to 78 on soil carbon sequestration is best realized in farming sys- percent, depending on carbon-to-nitrogen ratio, moisture tems that integrate crops and livestock. Crop-livestock inte- content, and aeration status. gration can occur in space, time, management, or ownership domains. The agronomic and economic justi�cation for the The impact of composting on emissions post-land applica- integration is based on the exchange of four main types of tion is of further interest. Fronning, Thelen, and Min (2008) resources: crop residues, manure, animal power, and �nan- examined GHG fluxes following land application of solid beef cial income (Sumberg 2003). The spatial domain integration manure and composted dairy manure over a 3-year period. is based on the idea that crops and livestock activities can be Net CH4 flux was minimal (less than 0.01 t CO2e ha−1 yr−1), colocated with the level of integration increasing as the scale while untreated manure application generated higher N2O becomes smaller. Close spatial integration is required for emissions than did compost (0.9 versus 0.7 t CO2e ha−1 yr−1). crop-livestock interactions involving crop residues, manure, However, these land emission impacts were small when and animal power (table 3.3). At large distances (scale), eco- compared to soil C sequestration rates, which were 1.8 nomic movement of crop residues, manure, and livestock is times greater for compost than for manure, suggesting that markedly curtailed, hindering interaction. the organic matter stabilization during composting reduces post-application respiration losses. Integration in the temporal domain connotes that crop and livestock production can take place simultaneously (in paral- Manure sequestered more carbon than fertilizer, yielding 61 lel) or can be temporally segregated (in sequence). Temporal kg C ha−1 yr−1 more in Africa, 243 C ha−1 yr−1 more in Asia, and integration can only occur after some form of spatial integra- 331 kg C ha−1 yr−1 more in Latin America (table 3.2). Yields tion has taken place, and the latter is important given the EC O N O M I C A N D S E CT OR WORK 28 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION TABLE 3.3: Relative Importance of the Four Domains of Integration on Crop-Livestock Interaction SPACE TIME OWNERSHIP MANAGEMENT Crop residue *** ** * ** Manure *** * * ** Animal power *** *** * * Financial income * ** *** * Source: Adapted from Sumberg (2003). Note: * = little importance; ** = some importance, *** = much importance. seasonality of feed and water for livestock. Integration in the resistant to decomposition. High-quality residues of perennial ownership domain underscores the fact that a given crop- legumes are generally the most effective in supplying nitrogen livestock combination can be owned by the same or a differ- in the short to medium term, while low-quality residues tend to ent entity, thereby promoting control and secure access to immobilize nitrogen. As large carbon losses occur under very resources. However, the formal and informal links between wet conditions, the best results are obtained when residues crop and livestock producers for accessing manure, crop are applied shortly before the beginning of the rainy season. residue, or power implies that though desired, integration in the ownership domain is not required for bene�cial crop- One of the main barriers to the use of crop residues and livestock interaction. mulch for soil fertility management is the numerous compet- ing uses for feed, fodder, thatch, and biofuel. Crop-livestock Last, integration in the management domain implies that integration can minimize the trade-off in the use of residues management of crop and livestock production may or may for feed (see the section of this report on nutrient manage- not be in the hands of the same entity and that manage- ment on page 21). Controlled grazing and the establishment ment may not necessarily coincide with ownership of both of plots of permanent forages for direct grazing can also crops and livestock. While ownership may increase the ef- reduce conflicts between soil organic matter accumulation �ciency of the bene�cial effects of interaction, integration in and grazing needs. Zero grazing involving the con�nement the management domain is not a prerequisite for successful of livestock in a stall and developing a cut-and-carry fodder bene�cial crop-livestock interaction (Sumberg 2003). system can make for more residue retention on the �eld, but it requires more labor. The establishment of bioenergy plants to meet the demand for biofuel may also help. In Crop Residue Management and Tillage general, desirable results will be achieved if integrated food- Crop residues are an important renewable resource for agro- feed-energy systems are tailored to speci�c local conditions. ecosystems. Crop residue management influences soil resil- Examples include intercropping maize and pigeon pea, and ience, agronomic productivity, and GHG emissions by using cookstoves for rural dwellers in Malawi and using agro- forestry systems for food, income generation, and bioetha- aiding nutrient cycling; nol production in Mozambique. The effects of crop residues intercepting raindrops, thereby allowing water to gen- and mulches on carbon sequestration are highest in Latin tly percolate into the soil; America and lowest in Africa (table 3.4). In Latin America, lowering soil evaporation; most studies looked speci�cally at the effects on soil carbon increasing aggregation of soil particles; and sequestration of mulching or incorporating residues rela- reducing run-off and erosion. tive to burning. Others were on the effects of sugar cane residues on sequestration, while others looked at the effects The quality and quantity of residues markedly influence the of grazing crop residues on soil carbon sequestration. Apart amount of carbon sequestered (�gure 3.8). The quantity of from biomass from trees, use of straw from rice and other residue produced is a function of the cropland area and ag- crop residues was found to be prevalent in Asia. ronomic practices, including tillage method. Cereals are two to three times better than legumes at sequestering carbon. Tillage, the agricultural preparation of the soil for planting, Cereals also have higher concentrations of lignin that are has three primary purposes: to facilitate seed germination CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 29 PHOTO 3.1: Crop Residue Management in Irrigated Fields in Indonesia Source: Curt Carnemark/World Bank. TABLE 3.4: Tillage, Crop Residue Management, and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 PERCENT UPPER 95 PERCENT CONFIDENCE CONFIDENCE NUMBER OF PRACTICE MEAN INTERVAL OF MEAN INTERVAL OF MEAN ESTIMATES Africa Crop residues 374 292 457 46 Mulches 377 159 595 6 Cover crops 406 298 515 24 No-tillage 370 322 418 108 Asia Crop residues 450 379 521 189 Mulches 565 371 759 53 Cover crops 414 233 594 38 No-tillage 224 97 351 48 Latin America Crop residues 948 638 1,258 56 Mulches 748 262 1,108 16 Cover crops 314 108 520 33 No-tillage 535 431 639 249 Source: This study. EC O N O M I C A N D S E CT OR WORK 30 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION FIGURE 3.8: Mean Soil Carbon Sequestration and Levels of Residue Returned 1,000 800 carbon sequestration (kg C/ha/yr) 600 400 200 0 <3 3–5 5–8 8+ residue application (t/ha/yr) Source: This study. Notes: n = 165; the 95 percent con�dence intervals are shown as whiskers. by creating a smooth, uniform soil surface for planting; to minimum soil disturbance through mechanical tillage, incorporate fertilizer, manures and crop residues into the soil; permanent soil cover through residue management, and and to control weeds. Depending on the amount of residue crop rotation and diversi�cation using legumes and left on the soil surface, tillage systems can be broadly clas- green manure or cover crops (�gure 3.2). si�ed into conventional, reduced, and conservation tillage (�gure 3.9). The conventional method, more appropriately In this study, carbon sequestration under conservation tillage referred to as intensive tillage, entails motorized multiple ranged from 224 kg ha−1 yr−1 for Asia to 535 kg ha−1 yr−1 for farm operations with mold board, disk, plow, and harrow Latin America (table 3.4). for seedbed preparation. Conventional tillage should not be confused with traditional tillage techniques involving manual Most of the conservation tillage systems are large-scale or animal-drawn operations. Conventional tillage leaves the farms in the United States, Canada, Brazil, Argentina, and least residue on the soil surface. While plowing loosens and Australia. Africa lags behind with only about 500,000 ha aerates the topsoil and facilitates seedling establishment, it under conservation agriculture. Recent experience in can lead to many unfavorable effects including soil compac- Zambia—conservation agriculture with trees—suggests that tion, destruction of soil aggregates, decrease in in�ltration the system holds promise for replenishing soil fertility and rate, increase in soil erosion and loss of nutrients, increase improving productivity and rural livelihoods. Conservation in evaporation loss, and reduction in soil organic matter. agriculture in Zambia entails (1) dry season land preparation Conservation tillage systems leave the most crop residues using minimum tillage methods and utilizing �xed planting on the surface, and they are the precursor to conservation stations (small shallow basins), (2) retention of crop residue agriculture, the holistic agricultural production system that in- from previous harvests in the �eld or use of other mulches tegrates management of soil, water, and biological resources from the tree component (Faidherbia albida) or other cover (Liniger et al. 2011). Conservation agriculture is based on crops, and (3) rotation of grains with legumes in the �eld. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 31 FIGURE 3.9: Classi�cation of Tillage Systems Based on Crop Residue Management Tillage Systems Conventional tillage Reduced tillage Conservation tillage Less than 15% residue cover 15% to 30% residue cover on Greater than 30% crop on the soil (< 500 kg ha-1 crop the soil (500 to 1,000 kg ha-1 residue on the soil (>1,000 kg residue equivalent) crop residue equivalent) ha-1 crop residue equivalent) No tillage Strip tillage Mulch tillage Ridge tillage Rotational tillage Also called zero It uses minimum Soil is disturbed prior Ridge tillage is similar This is a system in tillage; aims for 100% tillage by combining to planting and crop to the traditional which different soil cover and does the soil aeration residues are systems with planting tillage methods are not involve soil benefits of incorporated using on preformed ridges, used to establish disturbance through conventional tillage chisels, sweeps, and hills or bunds that different crops tillage. Helps to with the soil- field cultivators. provides warmer during a crop increase soil protecting It also includes conditions for plant rotation sequence. moisture, organic advantages of no planting operations growth. The ridges For instance, a matter, nutrients, tillage. Only the soil such as hoe drills and are formed from crop rotational system and crop yields; less area containing seed air seeders. The residues that are left may include both tillage reduces labor rows are tilled. Strip primary purpose of on the soil. The mulch tillage and no costs. Higher soil tillage allows for a mulch tillage is to system is mainly used tillage, or rotation moisture also helps better seedbed and increase soil organic on poorly drained from annual to to increase cropping for nutrients to be matter and tilth, soils. The rows are perennials. The soils intensity rather than better adapted to the reduce erosion, maintained in the are tilled at specific leave the field fallow. plant’s needs. Soil improve water use same location each intervals e.g. at the To avoid yield erosion on strip tilled efficiency and reduce season, while the top introduction of a depression during farms is much lower energy use. Residue of the ridges are crop in the rotation transition period (3 to than for conventional management is used leveled off at sequence, or every 5 years), fertilizer tillage. Strip tillage in conjunction with planting. Like other other year depending should be applied in requires higher crop rotation, cover conservation tillage on cropping slightly higher precision planters crops, and systems, the residue sequence. quantities. As an compared to no adjustment of cover prevents soil option to increased tillage system. Also, planting density. erosion, conserves herbicide use, cover appropriate cover soil moisture and crops can be used to crop mix for weed helps increase soil control weeds and suppression is organic matter. supply nutrients. essential. Source: This study. Over 180,000 farmers used this system at the end of 2010, Cover Crops and this �gure was projected to rise to 250,000 farmers by In this study, the practice of growing cover crops in situ 2011, representing some 30 percent of the population of was distinguished from mulches and crop residues of main small-scale farmers in Zambia. The tree component provides harvested crops. Cover crops help to improve soil macro- mulch and nutrients. By eliminating the need for laborious nutrients and micronutrients and are termed green manure land preparation, farmers adopting the system have been because of their ability to enhance soil fertility. Green manure better able to plant close to the onset of the rains. Using crops are commonly leguminous crops with high nitrogen conservation agriculture, yields have doubled for maize and content. Examples include cowpea, groundnut and mucuna. increased by 60 percent for cotton compared to conventional Cover crops can also improve soil quality by increasing soil tillage system. organic carbon through their biomass, and they also help in EC O N O M I C A N D S E CT OR WORK 32 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION improving soil aggregate stability, protecting the soil from The apparent lower level for double compared to single or surface runoff and suppressing weeds. triple cropping may reflect differences in soils, climate, and cropping systems rather than effects of cropping intensity. Crop Rotation Rotation diversi�cation is different in Africa compared to Latin Crop rotation is a key complementary practice for successful America. In Africa, the traditional element of crop rotation is implementation of no-tillage. Crop rotation is the deliberate the replenishment of nitrogen through the use of legumes in order of speci�c crops sown on the same �eld. The suc- sequence with other crops. In the Sahel, a typical cropping ceeding crop may be of a different species (e.g., grain crops sequence is millet/sorghum, followed by maize, groundnuts, followed by legumes) or variety from the previous crop, and cowpea, sesame, cassava, yams, and tree legumes, while in the planned rotation may be for 2 or more years. Rotating to Ethiopia, the sequence is usually maize/barley, followed by a different crop such as cowpea or soybean usually results sorghum, millet, and tef. in higher grain yields when compared to continuous cropping of maize. Other bene�ts of crop rotation include improved Water Management soil fertility, increased soil water management, reduced soil Improved water productivity in agriculture is achieved by erosion, and reduced pest and diseases. The recommended reducing water loss, harvesting water, managing excess wa- crop rotation strategies include ter, and maximizing water storage. Rainwater harvesting is particularly important for rain-fed agriculture in arid and semi- producing large amounts of biomass and residue for arid regions. The practice aims at minimizing the effects soil protection and incorporation in the soil, of seasonal variations in water availability due to droughts maintaining a continuous sequence of living vegetation, and dry periods and enhancing the reliability of agricultural including perennial crops in the rotation, and production. Conveyance and distribution ef�ciency are also diversifying the rotation to include nitrogen-�xing important measures in irrigation. Terracing on steep slopes legumes. and cross-slope barriers helps in reducing surface runoff. Improved irrigation sequestered carbon the most, while ter- Two variants of crop rotation observed in the review are rota- racing sequestered the least (table 3.6). tion intensi�cation and diversi�cation (table 3.5). Intensifying rotation means replacing a fallow with another crop, while Agroforestry diversifying rotation implies altering cropping sequences Agroforestry is an integrated land-use system combining within or across years while keeping the same number of trees and shrubs with crops and livestock. Agroforestry crops in the rotation. maintains soil organic matter and biological activity at levels There is a tendency toward higher sequestration rates in suitable for soil fertility. It also contributes to agro-ecosystem triple cropping systems, but variation is high (�gure 3.10). resilience by controlling runoff and soil erosion, thereby TABLE 3.5: Crop Rotation and Soil Sequestration Rates (kg C ha−1 yr−1) LOWER 95 PERCENT UPPER 95 PERCENT CONFIDENCE CONFIDENCE INTERVAL INTERVAL NUMBER OF PRACTICE MEAN OF MEAN OF MEAN ESTIMATES Africa Diversify rotation 378 306 451 49 Intensify rotation 342 277 407 55 Asia Intensify rotation 345 87 604 43 Latin America Intensify rotation 331 165 496 25 Diversify rotation 136 −62 334 43 Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 33 PHOTO 3.2: Water Management in a Field in India Source: Ray Witlin/World Bank. FIGURE 3.10: Mean Soil Carbon Sequestration and Cropping Intensity 1,500 carbon sequestration (kg C/ha/yr) 1,250 1,000 750 500 250 Single Double Triple of more cropping intensity Source: This study. Note: The 95 percent con�dence intervals are shown as whiskers (n = 536). EC O N O M I C A N D S E CT OR WORK 34 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION TABLE 3.6: Water Management and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 PERCENT UPPER 95 PERCENT CONFIDENCE CONFIDENCE NUMBER OF PRACTICE MEAN INTERVAL OF MEAN INTERVAL OF MEAN ESTIMATES Africa Rainwater harvesting 839 556 1,122 33 Cross-slope barriers 1,193 581 1,805 22 Terracing 421 276 566 15 Asia Rainwater harvesting 1,086 405 1,767 4 Improved irrigation 1,428 477 2,379 10 Latin America Improved irrigation 571 −59 1,201 34 Source: This study. reducing losses of water and nutrients. The shade provided has the special feature of reversed leaf phenology, a char- by the trees helps in moderating microclimate and reducing acteristic that makes it dormant and sheds its leaves during crops and livestock stress and helps to improve crop yields. the early rainy season and leafs out at the onset of the dry One of the most promising fertilizer tree species is Faidherbia season. This makes Faidherbia compatible with food crop albida, an Acacia species native to Africa and the Middle production because it does not compete for light, nutrients, East. Faidherbia is widespread throughout Africa, thrives on and water. Farmers have frequently reported signi�cant a range of soils, and occurs in different ecosystems ranging crop yield increases for maize, sorghum, millet, cotton, and from dry lands to wet tropical climates. It �xes nitrogen and groundnut when grown in proximity to Faidherbia. PHOTO 3.3: Maize Growing under Faidherbia Albida Trees in Tanzania Source: World Agroforestry Centre. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 35 Improved fallow involves the use of fast-growing trees to shade. In addition to C sequestration in biomass and soil, accelerate the process of soil rehabilitation and thereby tropical plantations are needed for timber and, more im- shorten the length of fallow sequester carbon the most portantly, as fuel for cooking. Thus, the area under tropical (about 2.5 t ha−1 yr−1). Nitrogen-�xing plants are normally plantations has increased drastically since the 1960s from used because they are generally sturdy, easy to establish, 7 Million hectares (Mha) in 1965 to 21 Mha in 1980, 43 Mha deep rooted, drought tolerant, and �x atmospheric nitro- in 1990, and 187 Mha in 2000. gen. The improved fallow trees and shrubs are left in the �eld for several months or years. During the fallow period, Intercropping examines the effects of crops on soils where the plants accumulate nitrogen from the atmosphere and there are trees, as opposed to the effects of including trees deep layers of the soil, while leaf litter protects the soil where there are crops. The responses over time vary in dif- from erosion, enriches the soil with nutrients, and helps ferent studies and may be affected by biomass harvesting. to conserve moisture. When the trees are removed after Competition with crops is an important trade-off. Although in- fallow, their roots remaining in the soil gradually decom- cluding the nitrogen-�xing tree Dalbergia sisso leads to more pose, releasing additional nutrients to the subsequent accumulation of organic carbon in the soil, the incorporation crops. Examples of species used for improved fallow in- of more trees reduces spacing between crops, and shading clude pigeon pea, sesban, sun hemp, Gliricidia sepium, and of crops by trees may reduce crop yields. The highest effects Tephrosia vogelii. recorded in Latin America for intercropping were 1.1 t ha−1 yr−1, while the highest effects for trees recorded in Africa was The average soil carbon sequestration rate of tree-crop 1.2 t ha−1 yr−1 (table 3.7). farming is approximately 1.4 t C ha−1 yr−1.The estimates covered cocoa in Ghana and Cameroon, coffee in Burkina Land-Use Changes Faso, indigenous fruit trees in South Africa, oil palm in Cote The review captured diverse categories of land-use changes d’Ivoire, exotic tree species in Ethiopia, rubber plantation in Asia and Latin America compared to Africa (table 3.8). in Nigeria and Ghana, and cashew and teak plantation Replacing annual crops with perennials increased soil carbon in Nigeria. Cocoa planted at low plant density and under sequestration on average by 1 t C ha−1 yr−1 in Asia and by 0.5 shade stores more carbon per unit area of soil than an t C ha−1 yr−1 in Latin America. In virtually all cases, the switch equivalent area of cocoa planted at high density without was to perennial grasses used as fodder for livestock. On TABLE 3.7: Agroforestry and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 UPPER 95 PERCENT PERCENT CONFIDENCE CONFIDENCE INTERVAL OF INTERVAL OF NUMBER OF PRACTICE MEAN MEAN MEAN ESTIMATES Africa Include trees in �eld 1,204 798 1,610 125 Intercropping 629 162 1,421 14 Alley farming 1,458 869 2,047 46 Tree-crop farming 1,359 755 1,964 44 Improved fallow 2,413 1,886 2,941 71 Asia Include trees in �eld 562 220 904 58 Intercropping 803 65 1,541 17 Latin America Include trees in �eld 1,065 270 1,860 43 Diversify trees 1,365 516 2,213 6 Intercropping 1,089 116 2,063 7 Source: This study. EC O N O M I C A N D S E CT OR WORK 36 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION TABLE 3.8: Land-Use Changes and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 UPPER 95 PERCENT PERCENT CONFIDENCE CONFIDENCE INTERVAL OF INTERVAL OF NUMBER OF PRACTICE MEAN MEAN MEAN ESTIMATES Africa Crop-to-forest 1,163 619 1,706 37 Pasture improvement 799 469 1,129 32 Asia Crop-to-forest 932 554 1,309 60 Crop-to-plantation 878 662 1,094 158 Crop-to–grassland 302 −36 640 35 Exclusion or reduction in grazing 502 126 877 39 Restoration of wetlands 471 1 Annual-to-perennial 1,004 615 1,392 36 Intensive vegetables and specialty 2,580 1,226 3,933 56 crops Latin America Crop-to-forest 528 −80 1,135 59 Pasture-to-forest 362 −32 756 62 Crop-to-plantation 893 299 1,488 14 Pasture-to-plantation 1,169 315 2,024 53 Grassland–to-plantation −406 −842 32 32 Exclusion or reduction in grazing 172 −393 737 30 Crop-to-pasture 1,116 −32 2,265 7 Annual-to-perennial 526 239 812 13 Pasture improvement 1,687 825 2,549 13 Source: This study. average, conversion of cultivated lands to secondary forests In Latin America, the conversion of native grasslands includ- sequestered more than 1 t C ha−1 yr−1 in Africa. The Global ing savannahs, which are frequently grazed, to plantations, Partnership on Forest Landscape Restoration estimates that on average, resulted in a net loss of soil carbon of 0.4 t C ha−1 over 400 Mha of degraded forest landscapes offer oppor- yr−1 (table 3.8). This is in sharp contrast to �ndings for con- tunities for restoring or enhancing the functionality mosaic version of pastures to forest or plantation. Converting grass- landscapes of forest, agriculture, and other land uses in the lands to plantations in the Pampas region results in acidi�- continent. In any afforestation project, emphasis should be cation of soils (Jobbagy and Jackson 2003), an impact also placed on maximizing the use of available land by planting observed in some studies of savannas in Brazil (Lilienfein high-yielding tree species. The species may be similar or et al. 2000). Other studies have suggested that grassland mixed in a manner that will generate the highest yield and soils may not accumulate carbon once forested and that biodiversity. The growing of plantations on former agricul- some humid soils may even lose carbon (Paruelo et al. 2010). tural land sequestered on average an additional 0.9 t C ha−1 yr−1 in Asia and Latin America—a value comparable to that for The highest soil carbon sequestration rate for land-use secondary forests. However, more C is sequestered when change observed in this review was for intensive veg- the former land use is pasture (about 1.2 t C ha−1 yr−1 in Latin etable production in Asia (2.6 t C ha−1 yr−1). One green- America). The establishment of pasture on cultivated land house system in Taiwan had 26 crops in 4 years with sequesters 1.1 t C ha−1 yr−1. high inputs of fertilizers and manures (Chang, Chung, and CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 37 Wang 2008). But other systems included farms with or- uses the appropriate mix of grass or legume species for ganic production using no pesticides or chemical fertilizers pasture, manages stocking rates, encourages more uniform (Ge et al. 2010c). Although intensi�ed cultivation in green- use of paddocks, and adjusts the timing of grazing. Pasture houses produced the highest average rates of soil carbon improvement sequestered 0.8 and 1.7 t C ha−1 yr−1 in Africa sequestration, the differences from estimates from �eld and Latin America, respectively. or agroforestry settings were not statistically signi�cant. One repeated-sampling design study in India, for example, Livestock grazing is relevant to many different land-use and documented the consequences of intensive cultivation of agricultural practices. This study looks at livestock manage- high-value medicinals and aromatics in an agroforestry set- ment practices from several perspectives, recognizing that ting (Sujatha et al. 2011), while another looked at growing there is not always a clear boundary between categories vanilla orchids under different organic manure and mulch of effects. These different practices are summarized in combinations in an agroforestry setting (Sujatha and Bhat table 3.9 to aid comparison. 2010). Fertigation, the inclusion of liquid fertilizers as part of a drip irrigation system, has been experimented with for Biochar and Other Soil Amendments high-value crops in arecanut agroforestry systems (Bhat Of the soil amendments studied, biochar sequestered car- and Sujatha 2009). Increases in soil organic carbon under bon the most (table 3.10). Biochar is produced by pyrolysis, these high-input systems are likewise rapid (Bhat and the thermal decomposition of biomass under limited oxygen Sujatha 2009). supply and at temperatures below 700°C. Biochar is a key ingredient in the formation of anthropogenic Amazonian Grazing management, the control of animal grazing to sus- dark earth (soils). Its application has gained recognition in tain productivity and ensure continuous supply of forages to the last few years for both climate change mitigation and animals, sequesters about 0.5 and 0.2 t C ha−1 yr−1 in Asia and soil improvement. The climate mitigation bene�t of biochar Latin America, respectively. Grazing management helps to lies in the fact that it decomposes more slowly and stabilizes maintain a healthy and productive pasture; biomass carbon. Application of biochar also leads to avoided increase water use ef�ciency by increasing in�ltration emissions of nitrous oxide and methane. As a soil amend- and reducing runoff; ment, biochar reduce soil and nutrient losses in runoff, thereby main- adds nutrients and improves uptake of applied taining soil physical and chemical quality; and fertilizers, maintain higher amounts of soil organic matter and increases water holding capacity of the soil, rapid cycling of nutrients. increases microbial biomass and activity, and Grazing management and pasture improvement should be increases mycorrhizal abundance linked to enhanced integrated for optimal bene�ts. An ef�cient grazing system agronomic ef�ciency and yield. TABLE 3.9: Summary of Observed Rates of Soil Carbon Sequestration (kg C ha−1 yr−1) as a Result of Land-Use Changes and Other Practices Relevant to Livestock Management GRASS TREES LATIN PRACTICE EFFECT PLANTED PLANTED GRAZING AFRICA ASIA AMERICA Pasture improvement (perennial, productive Y N Y 799 1687 grasses) Pasture to plantation Y Y (N) 1169 Include trees (silvopasture) (N) Y Y 1167 Pasture to forest Y N (N) 362 Excluding grazing (N) N N 502 172 Grassland to plantation N Y (N) −406 Crop to grassland Y N Y 302 Source: This study. Note: Letters in parentheses indicate typical but not absolute conditions. EC O N O M I C A N D S E CT OR WORK 38 CH A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION TABLE 3.10: Soil Amendments and Soil Carbon Sequestration Rates (kg C ha−1 yr−1) LOWER 95 PERCENT UPPER 95 PERCENT CONFIDENCE CONFIDENCE NUMBER OF PRACTICE MEAN INTERVAL OF MEAN INTERVAL OF MEAN ESTIMATES Africa Biochar 2,303 1,219 3,387 11 a Soil amendment 569 299 839 15 Asia Biochar 3,818 747 6,889 6 Sulfur 425 106 743 5 Lime 39 1 Zinc 53 1 Latin America Biochar 3,237 1,079 5,395 8 Lime 114 −287 516 9 Source: This study. a Ash, sawdust, cocoa husk, rice bran. Biochar can remain resident in the soil approximately 10 to GHGs from soils as well as energy-related emissions. It 1,000 times longer than the residence time of most soil or- should be noted that as much as 70 to 75 percent of fossil ganic matter. However, research results on biochar’s effect fuel use in the agricultural sector in the tropics is for the on some soil properties are not consistent. No signi�cant production and use of chemical fertilizers (Vlek, Rodriquez- increase in nutrient-holding capacity was observed after Kuhl, and Sommer 2004). Fertilizers may make no net con- the addition of biochar to a coastal plain soil (Novak et al. tribution to mitigation of climate change if the CO2 emitted 2009). Other studies have also indicated an adverse effect to produce and transport them exceeds the soil storage of biochar application on earthworm survival, possibly due bene�t (Schlesinger, 2010). Shang et al. (2010) calculated to increases in soil pH. In general, the use of biochar should full GHG budgets over a 3-year period in a long-term ex- ensure that crop residues and mulch needed for soil protec- periment on fertilization in a double rice-cropping system in tion are not removed from the �eld. China. They found fertilizer plots sequestered 470 kg C ha−1 yr−1 more carbon in soil than controls but that long-term fer- tilization increased CH4 emissions during flooded rice and Net Climate Change Mitigation Bene�ts of the Land increased N2O emissions from drained soils at other times. Management Practices They estimated a net impact of 4.1 t CO2e ha−1 yr−1 above Estimates of the net climate change mitigation benefits unfertilized controls although in terms of emissions per of the agricultural land management practices are sum- unit yield fertilization was still bene�cial. Shang et al. noted marized in figure 3.11. The estimates were derived by that mixtures of inorganic fertilizer and chemical fertilizers converting carbon sequestration rates from this study to increased net annual greenhouse warming potential even carbon dioxide equivalent by multiplying by 3.67 and also further, to as much as 13.5 t CO2e ha−1 yr−1 above unfertil- by accounting for land and process emissions. Land emis- ized controls. sions are the differences between emissions of nitrous oxides and methane expressed in CO2 equivalents by A modeling study for Indian rice and wheat suggested that conventional and improved practices, while process emis- increased irrigation and fertilizer application would increase sions are those arising from fuel and energy use (Eagle the carbon ef�ciency ratio even as net emissions rise (Bhatia et al., 2010). et al. 2010). At the same time, intensi�cation of agricultural production (using more fertilizers) on better lands may make Net Mitigation Bene�ts for Nutrient Management less suitable land available for conversion to grasslands and Increases in productivity from nitrogen fertilizers and irriga- forests with high soil carbon sequestration potential (Vlek, tion need to be considered against increased emission of Rodriquez-Kuhl, and Sommer 2004). Reducing wasteful CARBON SEQUESTRATION IN AGRICULTURAL SOILS ch ch ro ch e e m ta em im mic tio i pr al no ica n ca -2 0 2 4 6 8 10 12 14 16 ov fe o l fe in l f 0 2 4 6 8 10 12 te er 0 2 4 6 8 10 12 14 16 18 di ed rtili re int r re rtil ns til ve irr ze sid en d ize ifi ize r ig r ap ue si uce r ro ca r in sify ati pl f ta re tio EC O N O M I C A N D S E CT OR WORK te o tio Source: This study. ica ma y ro d ti Asia ns rot n n sid n Africa tio na tat ll di m ue gr re ify atio as du ro n n g io ve u s sla ce ta re of em n rs lch du m en ifi e nd d- tio ce ulc t c s Latin America -t gr n d- he no atio no o-p azin cr co gra s co til n or lant g op ve zin ot ve lag re ati -to r c g he rc e du on -g ro rs ro an im ra ps o m ps co ced pr ss nu v ov m lan pa il a t an st m er ure al- er till ed a d to cro -p n ur en ra e d cin re er ps an in irri ure sid en in nua clud gat w imp em g ue n te l- i at ro en er v ts m m ial ns to e t on ive -pe ree h e an an w cr in arv me pa u at ve ren s os t e nt e st age re ur m er ge ni in s s rcr stin clu lo op g e- en ha ta al b C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION in to-f t in rve les de pe pi clu or te rs tre tre /bar ng e rc tin in de t st cr r e- es rie cr i rs te re op bi opp g op n r f pa c cr es -to ofe ing -p r af far ield st rop opp fo m ur cr lan tiliz a re in e- -to- ing op ta er s to fo -to tio im lley tat g -p r -fo n pr fa ion lan est ov rm ta r ed in ti bi est oc fa g bi on ha ll oc ha r bi ow oc r ha r FIGURE 3.11: Carbon Dioxide Abatement Rates of the Land Management Practices 39 40 C H A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION fertilizer use by ensuring that applied rates do not exceed Kobayashi, and Shindo (2007), for example, found 43 percent crop requirements is an important mitigation strategy. lower CH4 emissions in no-till rice. The GHG mitigation bene�ts of residue management also Net Mitigation Bene�ts for Residue Management require consideration of processes apart from soil carbon and Tillage sequestration. Returning straw to �elds rather than burning The net GHG mitigation potential of residue management it helps avoid emissions associated with producing synthetic has been assessed in a few instances. Key constraints in- fertilizer as well as CH4 and N2O emissions from burning clude controlling methane emission from rice paddies. The (Lu et al. 2010). Improved management of compost pro- net potential of straw return (rather than burning) in China cesses and mulches can reduce non-CO2 emissions (Zeman, was assessed using a GHG budget model by Lu and et al. Depken, and Rich 2002). (2010). They found that across 10 provinces, straw return Net Mitigation Bene�ts of Intensi�cation increased net GHG emissions; in the other provinces, the and Water Management total net mitigation potential at soil saturation was equivalent Very intensive systems such as vegetable production under to just 1.7 percent of the fossil fuel emission budget in China greenhouses can sequester a lot of carbon in the soil, but for 2003. they obviously depend a lot on high levels of inputs as well. The life cycle analysis by Koga, Sawamoto, and Tsuruta Wang et al. (2011) made one of the few full carbon budgets (2006) of conventional and reduced tillage in intensive crop- for a greenhouse system. Their analyses suggest that green- ping systems in Hokkaido, Japan, suggested that soil-derived houses are a net sink of 1,210 and 1,230 kg C ha−1 yr−1 in CO2 emissions accounted for 64 to 76 percent of total GHG temperate and subtropical areas, respectively. The conver- emissions, emphasizing the importance of soil management sion from conventional agriculture enhances carbon sink practices. Adoption of reduced till in these systems was ex- potential as much as 8 times in temperate and 1.3 times in pected to reduce total GHG emissions by 4 to 18 percent for tropical areas. The mitigation potential of improved irrigation various crops as a result of slower decomposition rates and is almost offset by land and process emissions, but cross- fuel saving for plowing. The experimental study by Harada, slopes/barriers achieve moderate mitigation impact. PHOTO 3.4: Crop Harvesting in Mali. The Biomass Is Smaller Compared to that of Agroforestry Systems Source: Curt Carnemark/World Bank. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 3 — M E TA- ANALYSE S OF SOIL CARBON SEQUES TR ATION 41 A critical issue for soil carbon sequestration activities across Babu, Y., Li, C., Frolking, S., Naya, D., Adhya, T. 2006. “Field Validation humid parts of Asia is how to reduce emissions of CH4 from of DNDC Model for Methane and Nitrous Oxide Emissions From Rice-Based Production Systems of India.� Nutrient Cycling in rice �elds. There is a very large scienti�c literature on factors Agroecosystems 74: 157–174. influencing emissions and management options (e.g., Babu et al., 2006; Li et al. 2006, Minamikawa and Sakai 2005, Bhat, R., and Sujatha, S. 2009. “Soil Fertility and Nutrient Uptake by Arecanut (Areca catechu L.) as Affected by Level and Frequency Pathak 2010, Wassmann et al. 2000, and Zheng et al. 2007). of Fertigation in a Laterite Soil.� Agricultural Water Management For example, midseason drainage is a viable practice in some 96: 445–456. locations in India to reduce CH4 emissions (Babu et al. 2006). Bhatia, A., Pathak, H., Aggarwal, P. K., Jain, N. 2010. “Trade-Off Increasing fertilizer use increased both yields and CH4 emis- Between Productivity Enhancement and Global Warming sions. Other more complex water management strategies Potential of Rice and Wheat in India.� Nutrient Cycling in have been proposed and demonstrated to reduce CH4 emis- Agroecosystems 86, 413–424. sions (Minamikawa and Sakai 2007). Chang, E., Chung, R., Wang, F. 2008. “Effect of Different Types of Organic Fertilizers on the Chemical Properties and Enzymatic Net Mitigation Bene�ts for Agroforestry and Activities of an Oxisol Under Intensive Cultivation of Vegetables Land-Use Changes for 4 years.� Soil Science & Plant Nutrition 54: 587–599. The impacts of land-use changes to trees are positive and Eagle, A., Henry, L., Olander, L., Haugen-Kozrya, K., Millar, N., and large. The effects of some practices such as excluding grazers Robertson, G.P. 2010. Greenhouse Gas Mitigation Potential of from rangelands or grasslands are, however, fairly small. Most Agricultural Land Management in the United States: A Synthesis of the Literature. Technical Working Group on Agricultural Greenhoused of the potential impacts of changes in agricultural practices Gases (T-AGG) Report. Nicholas Institute, Duke University, Durham. on carbon stocks are by de�nition below ground. The time- averaged above-ground biomass of crops is small and does Fronning, E., Thelen, D., and Min, H. 2008. “Use of Manure, Compost, and Cover Crops to Supplant Crop Residue Carbon in not accumulate easily. Land-use changes away from cropping, Corn Stover Removed Cropping Systems.� Agronomy Journal such as to agroforestry or plantations, provide more compel- 100 (6): 1703–1710. ling examples where it is useful to think of both above- and Ge, T., Nie, S., Hong, Y., Wu, J., Xiao, H., Tong, C., and Iwasaki, below-ground sequestration rates at the same time and pos- K. 2010c. “Soluble Organic Nitrogen Pools in Greenhouse and sible trade-offs or interactions between them. One estimate Open Field Horticultural Soils Under Organic and Conventional for humid tropics globally suggested that tree-based agrofor- Management: A Case Study.� European Journal of Soil Biology estry systems could sequester 70 Mg C ha−1 in vegetation and 46: 371–374. up to 25 Mg C ha−1 in the topsoil (Mutuo et al. 2005). Harada, H., Kobayashi, H., Shindo, H. 2007. “Reduction in Greenhouse Gas Emissions by No-Tilling Rice Cultivation Net Mitigation Bene�ts for Biochar Application in Hachirogata Polder, Northern Japan: Life-cycle Inventory Analysis.� Soil Science and Plant Nutrition 53 (5): 668–677. Applications of biochar or charcoal, on average, resulted in higher overall GHG mitigation potential than other practice Jobbagy, E., and Jackson, R. 2003. “Patterns and Mechanisms of changes reviewed in this study, but precise estimates were Soil Acidi�cation in the Conversion of Grasslands to Forests.� Biogeochemistry 64: 205–229. not possible given the paucity of data. These �ndings are consistent with reviews that suggest potential value of bio- Koga, N., Sawamoto, T., Tsuruta, H. 2006. “Life Cycle Inventory- char for improving soil conditions and increasing sequestra- Based Analysis of Greenhouse Gas Emissions From Arable Land Farming Systems in Hokkaido, Northern Japan.� Soil Science tion of GHGs (Lehmann et al., 2006; Sohi et al., 2010). and Plant Nutrition 52 (4): 564–574. Lehmann, J., Gaunt, J., and Rondon., M. 2006. “Biochar REFERENCES Sequestration in Terrestrial Ecosystems—A Review.� Mitigation Acharya, G. P., McDonald, M., Tripathi, B., Gardner, R., and and Adaptation Strategies for Global Change 11: 403–427. Mawdesley, K. 2007. “Nutrient Losses From Rain-Fed Bench Li, C., Salas, W., DeAngelo, B., Rose, S. 2006. “Assessing Terraced Cultivation Systems in High Rainfall Areas of the Mid- Alternatives for Mitigating Net Greenhouse Gas Emissions Hills of Nepal.� Land Degradation & Development 18: 486–499. and Increasing Yields from Rice Production in China Over the Alvarez, R. 2005a. “A Review of Nitrogen Fertilizer and Conservation Next Twenty Years.� Journal of Environmental Quality 35 (4): Tillage Effects on Soil Organic Carbon Storage.� Soil Use and 1554. Management 21: 38–52. Lilienfein, J., Wilcke, W., Ayarza, M., Vilela, L., Lima, S., and Zech, Alvarez, R. 2005b. “A Review of Nitrogen Fertilizer and Conservation W. 2000. “Soil Acidi�cation in Pinus Caribaea Forests on Tillage Effects on Soil Organic Carbon Storage.� Soil Use and Brazilian Savanna Oxisols.� Forest Ecology and Management Management 21: 38–52. 128: 145–157. EC O N O M I C A N D S E CT OR WORK 42 C H A PTER 3 — META -A NA LY S ES OF SOIL CA R B ON SEQU ES TR ATION Liniger, H. P., Mekdaschi Studer, R., Hauert, C. and M. Gurtner. Schlesinger, W. H., 2010. “On Fertilizer-Induced Soil Carbon 2011. Sustainable Land Management in Practice - Guidelines Sequestration in China’s Croplands.� Global Change Biology 16: and Best Practices for Sub Saharan Africa. World Overview of 849–850. Conservation Approaches and Technologies (WOCAT) and Food Shang, Q., Yang, X., Gao, C., Wu, P., Liu, J., Xu, Y., Shen, Q., and Agriculture Organization of the United Nations (FAO). Zou, J., and Guo, S., 2010. “Net Annual Global Warming Lu, F., Wang, X., Han, B., Ouyang, Z., Duan, X., and Zheng, H. 2010. Potential and Greenhouse Gas Intensity in Chinese Double “Net Mitigation Potential of Straw Return to Chinese Cropland: Rice-Cropping Systems: A 3-Year Field Measurement in Long- Estimation With a Full Greenhouse Gas Budget Model.� Term Fertilizer Experiments.� Global Change Biology, doi: Ecological Applications 20 (3): 634–647. 10.1111/j.1365-2486. Lu, M., Zhou, X., Luo, Y., Yang, Y., Fang, C., Chen, J., and Li, B. 2011. Sohi S., Krull, E., Lopez-Capel, E., and Bol, R. 2010. “A Review of “Minor Stimulation of Soil Carbon Storage by Nitrogen Addition: Biochar and Its Use and Function in Soil.� Advances in Agronomy A Meta-Analysis.� Agriculture, Ecosystems, and Environment 105:47–82. 140: 234–244. Sujatha, S., and Bhat, R. 2010. “Response of Vanilla (Vanilla Planifolia A.) Intercropped in Arecanut to Irrigation and Nutrition in Humid Matsumoto, N., Paisancharoen, K., and Hakamata, T. 2008. “Carbon Tropics of India.� Agricultural Water Management 97: 988–994. Balance in Maize Fields Under Cattle Manure Application and No- Tillage Cultivation in Northeast Thailand.� Soil Science and Plant Sujatha, S., Bhat, R., Kannan, C., and Balasimha, D. 2011. “Impact Nutrition 54: 277–288. of Intercropping of Medicinal and Aromatic Plants With Organic Farming Approach on Resource Use Ef�ciency in Arecanut Minamikawa, K., and Sakai, H. 2005. “The effect of Water (Areca catechu L.) Plantations in India.� Industrial Crops and Management Based on Soil Redox Potential on Methane Products 33: 78–83. Emission From Two Kinds of Paddy Soils in Japan.� Agriculture, Ecosystems, and Environment 107: 397–407. Sumberg, J. 2003. “Toward a Disaggregated View of Crop–Livestock Integration in Western Africa.� Land Use Policy 20: 253–264. Mutuo, P., Cadisch, G., Albrecht, A., Palm, C., and Verchot, L. 2005. “Potential of Agroforestry for Carbon Sequestration and Vlek, P., Rodriquez-Kuhl, G., and Sommer, R., 2004. “Energy Use and Mitigation of Greenhouse Gas Emissions From Soils in the CO2 Production in Tropical Agriculture and Means and Strategies Tropics.� Nutrient Cycling in Agroecosystems 71: 43–54. for Reduction or Mitigation.� Environment, Development, and Sustainability 6: 213–233. Novak, J. M., Busscher, W. J., Laird, D. L., Ahmedna, M., Watts, D. Wang, Y., Xu, H., Wu, X., Zhu, Y., Gu, B., Niu, X., Liu, A., Peng, C., Ge, W., and Niandou, M. A. S. 2009. “Impact of Biochar Amendment Y., Chang, J., 2011. Quanti�cation of net carbon flux from plastic on Fertility of a Southeastern Coastal Plain Soil.� Soil Science greenhouse vegetable cultivation: A full carbon cycle analysis. 174: 105–112. Environmental Pollution 159, 1427–1434. Paruelo, J. M., Pineiro, G., Baldi, G., Baeza, S., Lezama, F., Altesor, Wassmann, R., Lantin, R., Neue, H., Buendia, L., Corton, T., and A., and Oesterheld, M. 2010. “Carbon Stocks and Fluxes in Lu, Y. 2000. “Characterization of Methane Emissions From Rangelands of the Rio de la Plata Basin.� Rangeland Ecology & Rice Fields in Asia III: Mitigation Options and Future Research Management 63: 94–108. Needs.� Nutrient Cycling in Agroecosystemes 58: 23–36. Pathak, H., 2010. “Mitigating Greenhouse Gas and Nitrogen Loss Zeman, C., Depken, D., and Rich, M., 2002. “Research on How the With Improved Fertilizer Management in Rice: Quanti�cation and Composting Process Impacts Greenhouse Gas Emissions and Economic Assessment.� Nutrient Cycling in Agroecosystemes Global Warming.� Compost Science & Utilization 10 (1) : 72. 87: 443–454. Zhang, H., Xu, M., and Zhang, F. 2009. “Long-Term Effects of Manure Pattey, E., Trzcinski, M. K., and Desjardins, R. L. 2005. “Quantifying Application on Grain Yield Under Different Cropping Systems and the Reduction of Greenhouse Gas Emissions as a Result of Ecological Conditions in China.� Journal of Agricultural Science Composting Dairy and Beef Cattle Manure.� Nutrient Cycling in 147 (1): 31. Agroecosystems 72: 173–187. Zheng, J., Zhang, X., Li, L., Zhang, P., and Pan, G. 2007. “Effect of Potter, K., Velazquez-Garcia, J., Scopel, E., and Torbert, H. 2007. Long-Term Fertilization on C Mineralization and Production of “Residue Removal and Climatic Effects on Soil Carbon Content CH4 and CO2 Under Anaerobic Incubation From Bulk Samples of No-Till Soils.� Journal of Soil and Water Conservation 62: and Particle Size Fractions of a Typical Paddy Soil.� Agriculture, 110–114. Ecosystems, and Environment 120: 129–138. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 4 — E C O SYST E M SIMUL AT ION MODE L ING OF S OIL CA R B ON SEQUES TR ATION 43 Chapter 4: ECOSYSTEM SIMULATION MODELING OF SOIL CARBON SEQUESTRATION 4.1 MODEL DESCRIPTION which is highly resistant to microbial decomposition (�gure The RothC model (Coleman and Jenkinson 2008) was used 4.1). Both DPM and RPM decompose to form CO2, BIO, and to project the amount of soil carbon sequestered by dif- HUM. The proportion that is converted to CO2 and to BIO ferent land management practices up to 2035. The RothC plus HUM is primarily determined by the clay content of the model describes the fate of organic inputs entering the soil soil. Subsequent further decomposition of the BIO and HUM environment, the undergoing decomposition within the soil produces more CO2, BIO, and HUM. biomass to form a number of carbon pools, and the release One of the main advantages of the RothC model is its re- of CO2. The pools have different susceptibilities to decom- quirement of a few, easily obtainable inputs to estimate soil position, ranging from highly labile to inert materials. The carbon. The required inputs are monthly rainfall, monthly pools include easily decomposable plant material (DPM), open pan evaporation, average monthly mean air tempera- resistant plant material (RPM), microbial biomass (BIO), hu- ture (in degrees Celsius), clay content of the soil, and an mi�ed organic matter (HUM), and inert organic matter (IOM), estimate of the decomposability of the incoming organic FIGURE 4.1: Representation of the RothC Model Organic inputs IOM RPM DPM HUM BIO CO2 HUM BIO CO2 HUM BIO CO2 Source: This study. Note: DPM = decomposable plant material, RPM = resistant plant material, BIO = microbial biomass, HUM = humi�ed organic matter, IOM = inert organic matter. EC O N O M I C A N D S E CT OR WORK 44 CHAP T E R 4 — EC OSY S TEM SIMULATION MOD ELING OF S OIL CA R B ON SEQU ES TR ATION material referred to as the DPM/RPM ratio (Coleman and decomposition rates (BIO plus HUM) pools formed during Jenkinson 2008). The model has been validated across the decomposition using the following exponential equation: agro-ecological zones of the world and has been used for many subnational and national GHG inventories. x = 1.67(1.85 + 1.60e−0.0786%clay), [2] The amount of carbon (Y) that decomposes from an active where x is the ratio CO2/(BIO + HUM) and BIO and HUM are pool in a given month can be represented by an exponential the corresponding biomass and humic pools formed initially decay function of the form as incoming plant materials. Y = Y0(1 − e−abckt), [1] The global soil carbon mitigation potential due to the adop- tion of sustainable land management practices was modeled where Y0 is the initial amount of carbon in the particular pool, to a depth of 30 cm using the following relationship: a is the rate-modifying factor for temperature, b is the rate- modifying factor for soil moisture, c is the rate-modifying fac- Cs = A × f, [3] tor for soil cover, k is the yearly decomposition rate constant for that particular compartment, and t = ¹�¹² is to scale k into where Cs is the change in soil organic carbon as a result of monthly values. adoption, A is the activity data or land area (in ha) where a given sustainable land management practice was adopted, Equations for calculating each of these factors can be found and f the emission factor is the sequestered carbon in t C in Coleman and Jenkinson (2008). While the above factors ha−1 yr−1. The activity data (global cropland area) were de- contribute exponentially to the soil carbon remaining at the rived from available spatial datasets (table 4.1). The harvest- end of each month, others related to the input of carbon such ed areas of eight major crops (barley, maize, millet, pulses, as crop yields, root biomass, and the proportion of carbon in rice, sorghum, soybean, and wheat) occupying more than plant residues are linearly related to the amount of carbon 70 percent of the global agricultural area were estimated decomposing. The RothC model also adjusts for clay content within a geographical information system and used for by altering the partitioning between evolved CO2 and soil C modeling. TABLE 4.1: Spatial Datasets Used in the Study DATA PURPOSE REFERENCES Clay content, initial soil carbon RothC model parameterization Harmonized World Soil Database v 1.1: FAO/IIASA/ISRIC/ISSCAS/JRC, 2009. content Harmonized World Soil Database (version 1.1). FAO, Rome, Italy and IIASA, Laxenburg, Austria. http://www.iiasa.ac.at/Research/LUC/External-World-soil- database/HTML/index.html Temperature and precipitation RothC model parameterization FAOCLIM 2; World-wide Agro Climatic Data Base; Food and Agriculture Organization of the United Nations; Environment and Natural Resources Service— Agrometerology Group Crop calendar RothC model parameterization, FAO Crop Calendar—a crop production information tool for decisionmaking (FAO modeling 2010): http://www.fao.org/agriculture/seed/cropcalendar/welcome.do Direct manure/composted manure Carbon input for modeling Global Fertilizer and Manure Application Rates; Land Use and the Global input data Environment, Department of Geography, McGill University; Food and Agriculture Organization of the United Nations livestock data for Africa, 2009: http://faostat.fao.org/site/569/default.aspx#ancor Harvested area (ha) of selected crops Carbon input for modeling Harvested area and yields of selected crops; Harvested area and Yields of 175 and crop yield (t/ha/year) data crops (M3-Crops Data); Navin Ramankutty; Land Use and the Global Environment, Department of Geography, McGill University Carbon input data for agroforestry and From several published literature See references cover crops Land-use systems Additional data used to estimate land FAO Land Use Systems area for which a given technology is http://www.fao.org/geonetwork/srv/en/main.home applicable http://www.fao.org/geonetwork/srv/en/metadata.show Sustainability and the Global Strati�cation of Africa Center for Sustainability and the Global Environment, University of Wisconsin Environment Global Agro-Ecological Zones Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 4 — E C O SYST E M SIMUL AT ION MODE L ING OF S OIL CA R B ON SEQUES TR ATION 45 Using cluster analysis, the global cropland extent was The typical land management practices associated strati�ed into mapping units based on temperature, precipi- with the cropping systems. For example, most of the tation, and clay content. This resulted in 12 distinct clusters farming systems in North America already leave the (strata) within eight regions (Africa, Asia, Central America, residues on the �eld, while in Africa, a common prac- Europe, North America, Oceania, Russia, and South America) tice is to burn or remove the residues from the �eld. (Figure 4.2). Crop yields and manure were converted into Documented impact of agricultural land management organic residues as model inputs using IPCC standard equa- practices on carbon sequestration (see Chapter 3). tions (IPCC, 2006). A standard DPM/RPM ratio of 1.44 was As a result, residue management, manure management, set for modeling sustainable land management scenarios tillage management, agroforestry, and integrated fertility except for agroforestry, where a ratio of 0.25 was assumed. management were modeled. A detailed description of the The speci�c organic inputs for the land management practice baseline and mitigation scenarios is provided in Appendix 4.1. being modeled were set on a monthly basis using crop calen- The study also took advantage of the recently released crop dars speci�c for each stratum. For each stratum and region, calendar for Africa (http://www.fao.org/agriculture/seed/ the most dominant cropping systems were identi�ed from cropcalendar/welcome.do) to model carbon sequestration the literature. A summary of the farming systems is given in under various levels of organic inputs for Africa. Africa was Appendix 4.1. classi�ed into four agroecological zones using procedures The choice of the suitable mitigation scenarios for each world similar to the global cropland extent (Figure 4.3). The land region was guided by the following baseline considerations: management practices include integrated residue and ma- nure management; agroforestry systems including perennial The most dominant cropping systems in a speci�c crops; land rehabilitation; coppice and improved fallow; and region. For instance, mixed smallholder farming cropping systems involving mucuna, cowpea, and groundnut systems are the dominant system in Africa, cropping as cover crops. To account for trade-offs between mulch systems in South Asia are dominated by rice, and residues and livestock and fuel biomass, different fractions maize-soybeans systems are dominant in several of retained residues (i.e., 25 percent, 50 percent, and 75 per- parts of South America. cent) were modeled. A detailed description of the scenarios for Africa is provided in Appendix 4.2. FIGURE 4.2: The 12 Strata Used for Ecosystem Simulation Modeling Source: This study. EC O N O M I C A N D S E CT OR WORK 46 CHAP T E R 4 — EC OSY S TEM SIMULATION MOD ELING OF S OIL CA R B ON SEQU ES TR ATION FIGURE 4.3: Africa Agroecological Zone Uncertainties in model parameters were estimated fol- worldbank.org/SoilCarbonSequestration/. The tool includes lowing the adoption of the Sustainable Agricultural Land over 4,000 land management scenarios carefully chosen to Management (SALM) methodology (http://www.v-c-s.org/ reflect situations typically encountered in agricultural proj- sites/v-c-s.org/files/SALM%20Methodolgy%20V5%20 ects. The Internet GIS database provides per-ha estimates 2011_02%20-14_accepted%20SCS.pdf). The procedures of soil carbon sequestration under different land manage- are provided in Appendix 4.3. ment practices for a period of 20 to 25 years (�gure 4.4). Information on carbon sequestration potential of a location can be derived by point-and-click or by searching using 4.2 RESULTS place name. Users can download data from the Internet Soil Carbon Sequestration Internet Tool database and integrate with other GIS information to es- Modeling results are summarized in an Internet geo- timate soil carbon stock changes for different agricultural graphical information system (GIS) tool at http://www.esd. projects. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 4 — E C O SYST E M SIMUL AT ION MODE L ING OF S OIL CA R B ON SEQUES TR ATION 47 FIGURE 4.4: A Screen Shot of the Soil Carbon Internet Database Source: http://www.esd.worldbank.org/SoilCarbonSequestration/. Soil Carbon Loss Under Low Input Baseline Scenario Agroforestry by far has the highest sequestration potentials The predicted cumulative C loss by 2030 varies for different for all world regions. The time-averaged above-ground bio- cropping systems and regions of the world. The loss is high- mass of trees is relatively large compared to crops. est for Russia under wheat, rice, pulses, and barley (35 to Carbon sequestration potential of the land management 40 t C ha−1) where the drive to exploit minerals and other natu- practices is in the order of agroforestry > cover crops > ral resources has spread agriculture to unproductive soils and manure > crop residues > no-tillage. The highest emphasis low fertilizer use has led to a sharp decrease in soil fertility. should be placed on agroforestry systems because of the Middle America is predicted to experience the next highest diverse bene�ts they provide including compatibility of some loss due to depletion of crop residues in virtually all its crop- tree species with crops and livestock production, increased ping systems (25 to 37 t C ha−1). The highest cumulative C loss income through production of indigenous fruit trees, and suit- under the low input scenario occurs under rice and pulses for ability of certain tree species for bioenergy. Agroforestry is Africa (20 t C ha−1), under pulses for South America (26 t C ha−1), also vital for the restoration of marginal and degraded lands. and under millet for Europe (23 t C ha−1). The cumulative C loss is around 15 to 20 t C ha−1 for all cropping systems in Asia. Carbon Sequestration Maps Soil Carbon Sequestration Under Different Land Figure 4.6 reveals differences in the predicted spatial pat- Management Practices tern of carbon sequestration for the land management prac- Carbon sequestration through residue management depends tices. High sequestration rates are generally observed in the much on the land area devoted to a given crop (table 4.2). Guinea savannah areas in Africa for most of the practices. Based on the assumption of 50 percent residue retention, cu- The highest cumulative sequestration for green manure (6 mulative carbon sequestration by 2030 varies from 0.5 Million to 10 t C ha−1) are predicted for Europe and North America, tons (Mt) C for soybean to 37 Mt C for maize (�gure 4.5). while the highest for maize residue (7 to 12 t C ha−1) are In Asia, the sequestered carbon varies from 10 Mt for millet predicted for Asia. The spatial patterns of composted and to 517 Mt for rice. The lowest amount of sequestered car- direct manure are similar because both models are based bon from cover crops was recorded for Middle America (15 on frequency of livestock. Composted manure sequesters Mt), while the highest was recorded for Asia (1 Gigaton). The slightly higher than direct manure (0.04 to 14 t C ha−1 versus highest sequestration potentials for direct and composted 0.02 to 13 t C ha−1). No-tillage sequesters least (0.08 to 1.3 manure (550 and 587 Mt, respectively) were observed for t C ha−1); its estimates markedly suffer from lack of good North America, while Russia has the least (less than 0.2 Mt). resolution spatial data of no-tillage adopting areas. EC O N O M I C A N D S E CT OR WORK 48 CHAP T E R 4 — EC OSY S TEM SIMULATION MOD ELING OF S OIL CA R B ON SEQU ES TR ATION FIGURE 4.5: Cumulative Soil Carbon Loss by 2030 Assuming 15 Percent Residue Retention (t ha−1) Under Different Cropping Systems Middle North South Soybean Wheat Middle South North America America Europe America Asia Africa Oceania Russia America America America Europe Asia Africa Oceania 0 0 –5 –5 –10 –10 –15 –15 –20 –20 –25 –25 –30 –35 –30 –40 –35 –45 –40 Middle South North Sorghum Rice Middle South North America America America Asia Africa Oceania Russia America Africa Europe America Asia America Oceania 0 0 –5 –5 –10 –10 –15 –15 –20 –25 –20 –30 –25 –35 –30 –40 Millet Europe Asia North America Africa Middle South North Maize 0 America Africa America Asia Europe America Oceania 0 –5 –5 –10 –10 –15 –20 –15 –25 –20 –30 –35 –25 –40 Pulses Middle North South Barley Middle South North Russia America America AmericaEurope Africa Asia Oceania Russia America AmericaEurope America Asia Africa Oceania 0 0 –5 –5 –10 –10 –15 –15 –20 –20 –25 –25 –30 –30 –35 –35 –40 –40 –45 –45 Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 4 — E C O SYST E M SIMUL AT ION MODE L ING OF S OIL CA R B ON SEQUES TR ATION 49 FIGURE 4.6: Predicted Cumulative C Sequestration for Different Land Management Practices by 2030 Agroforestry Green Manuring 2030 2030 Tons per Hectare Tons per Hectare 2.563 – 3.915 3,014 – 3,552 3.916 – 6.479 3,553 – 4,341 6.480 – 9.338 4,342 – 5,438 9.339 – 12.859 5,439 – 6,743 12.860 – 23.976 6,744 – 9,799 0 1,500 3,000 6,000 9,000 12,000 0 1,500 3,000 6,000 9,000 12,000 Kilometers Kilometers Non Tillage Composted Manure Direct Manuring Residue Management, Maize Source: This study. EC O N O M I C A N D S E CT OR WORK 50 CHAP T E R 4 — EC OSY S TEM SIMULATION MOD ELING OF S OIL CA R B ON SEQU ES TR ATION TABLE 4.2: Modeled Cumulative Soil Carbon Sequestration Potential by 2030 (Mt C) Under Different Land Management Practices MIDDLE NORTH SOUTH AFRICA ASIA EUROPE AMERICA AMERICA OCEANIA RUSSIA AMERICA Residue management Barley 7.898 16.359 Maize 37.281 209.574 Millet 10.657 9.993 Pulses 13.664 32.995 Rice 17.771 516.843 1.637 3.279 Sorghum 21.494 11.562 Soybean 0.524 35.120 Wheat 34.504 360.966 No-tillage 20.763 33.209 33.128 0.700 0.557 7.131 Cover crops 513.237 1009.402 772.082 14.727 632.415 136.495 Direct manure 400.101 203.703 23.556 2.252 549.558 1.740 0.080 20.098 Composted manure 427.890 478.064 57.106 5.460 586.731 4.218 0.193 48.721 Agroforestry 1309.511 2416.434 803.907 18.608 727.361 81.229 19.868 210.233 Source: This study. REFERENCES Coleman, K., and Jenkinson, D. S.2008 ROTHC-26.3, A Model for Intergovernmental Panel on Climate Change. 2006. “Volume 4. the Turnover of Carbon in Soil: Model Description and Windows Agriculture, Forestry, and Other Land Uses,� in 2006 IPCC Users Guide. Rothamsted Research, Harpenden, UK. Available Guidelines for National Greenhouse Gas Inventories, ed. H. online at http://www.rothamsted.ac.uk/aen/carbon/mod26_3_ S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe. win.pdf. Institute for Global Environmental Strategies, Japan. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 51 Chapter 5: ECONOMICS OF SOIL CARBON SEQUESTRATION 5.1 MARGINAL ABATEMENT COSTS included investments in seeds and seedlings, input subsi- Sustainable land management technologies can generally dies, extension services, and other administrative costs. The be deployed at varying costs, creating the need to evaluate cost-bene�t flows were discounted to present value to cal- their cost-effectiveness. Such analysis helps in identifying culate NPV using a discount rate of 9 percent. The adoption potential mitigation pathways for a given context. The cost- period was assumed to be 25 years; the time carbon seques- effectiveness of the land management practices in mitigat- tration reaches saturation for most of the land management ing climate change has been evaluated using the marginal technologies. abatement cost (MAC) curve. The MAC curve analysis was The abatement rates of the land management practices a quantitative assessment of all possible costs and bene�ts (�gure 3.11) were used to scale-up for each continent by that would accrue if the various management practices were multiplying by the suitable areas for each practice within a implemented. A MAC curve depicts the relationship be- continent in 2030. The assumptions for estimating the suit- tween the cost-effectiveness of different land management able areas for the four IPCC special reports on emission sce- practices vis-à-vis the amount of GHG abated. The MAC is narios are described in Appendix 5.1. Efforts were made to plotted on the y-axis and GHG abated on the x-axis, with the avoid double counting as some of the practices are mutually land management practices ranked against the MAC from exclusive. the lowest to the highest. Moving along the curve from left to right worsens the cost-effectiveness of the mitigation Figure 5.1 shows the MACs for Africa, Asia, and Latin measures. The width of the column is the amount of GHG America. The shapes of the curve are similar across scenar- mitigated by the land management practice, while the area ios, so only the curves for the A1b scenario are presented. of each column equals the cost or bene�t of adopting the practice. The MAC curve can also be used for cost-bene�t All the land management practices are pro�table to the farm- analysis by comparing the unit mitigation cost with the er, but to varying degrees (table 5.1). The marginal bene�t of shadow price of carbon or the cost of purchasing emissions no-tillage is greater than US$100 per tonne of carbon dioxide allowance. Negative MACs indicate that a land management mitigated for the three regions. Alley farming and intercrop- practice is self-�nancing (that is, it both reduces emissions ping also yield relatively high pro�ts in Africa. With the excep- and saves money), while positive MACs imply that the land tion of Asia, the marginal bene�t of residues for the regions is management practice reduces emissions at a cost and thus modest (less than US$50). Table 5.1 also reveals the inherent requires judgment against the cost of inaction. trade-off between the pro�tability of the land management practices and their mitigation potentials. Afforestation and In this study, private and public marginal abatement costs pasture establishment on degraded land with relatively high were computed. For the private MACs, all possible costs mitigation potentials are modestly pro�table. This suggests and bene�ts that would accrue to the farmers were valued that farmers may be reluctant to privately implement land at market prices the farmers are likely to face in switching rehabilitation. On the other hand, manure and fertilizer with to the practices. The public costs, on the other hand, refer modest mitigation potential yielded relatively high pro�ts. to government support toward the implementation of land management practices. Without public support to farm- The public costs of all the land management practices ers, poor agricultural land management will intensify land are lower than US$20 per ton of GHG mitigated in Africa. degradation, increase farmers’ vulnerability, and contribute Afforestation and grassland rehabilitation cost governments additional GHGs in the atmosphere. Computed public costs more than $20 per ton of GHG mitigated in Asia and Latin EC O N O M I C A N D S E CT OR WORK 52 C H A PTER 5 — EC ONOMIC S OF SOIL CA R B ON SEQU ES TR ATION FIGURE 5.1: The Private Marginal Abatement Cost Curves CO2 abated (Mt yr –1) Mt CO2 abated 0 500 1,000 1,500 0 500 1,000 1,500 2,000 0 0 marginal abatement cost ($/t CO2) –25 Include trees –100 Slope/Barriers marginal abatement cost ($/t CO2) –50 Residues –200 –75 Water Cover crops Tree-crop management Grazing Improved fallow –100 Improved Rotation management Manure pastures Alley Intensification –300 Afforestation Manure cropping –125 Intercropping Rainwater harvesting –150 Pasture Inorganic Rotation –400 established fertilizer diversification degraded land –175 Other soil –500 Biochar –200 ammendments –225 Fertilizer –600 –250 No tillage scenario A1b-Africa cropland scenario A1B-Africa grassland CO2 abated (Mt yr –1) CO2 abated (Mt yr–1) 0 1,000 2,000 3,000 4,000 0 500 1,000 1,500 2,000 0 0 –200 –100 marginal abatement cost ($/t CO2) Intensive marginal abatement cost ($/t CO2) vegetables –200 –400 Biochar –300 Pasture on –600 Improved degraded irrigation land Crop to Crop to Alley Manure –400 forest grassland –800 farming Organic soil –500 Intensify Crop to restoration Improved –1,000 rotation plantation No tillage irrigation Annual to –600 Biofertilizer perennial Inorganic –1,200 fertilizer Cover crops –700 Reduced Manure Rainwater grazing harvesting –1,400 –800 Residues –900 scenario A1B-Asia cropland Fertilizer scenario A1b-Asia grassland Rainwater harvesting cummulative CO2 abated (Mt yr–1) 0 500 1,000 1,500 cummulative CO2 abated (Mt yr–1) 0 0 100 200 300 400 500 600 700 800 0 marginal abatement cost ($/t CO2) Cover crops –100 marginal abatement cost ($/t CO2) Residue Pasture to management –200 forest Crop to forest Grassland to Pasture plantation improvement Diversify trees –300 –800 Inculde-trees Rainwater Rainwater Pasture to harvesting harvesting plantation –400 Manure Application of Annual to Intensify manure rotation perennial –500 Biochar Intercropping –600 Pasture on degraded –1,600 land Diversify Reduced –700 rotation grazing No tillage Improved irrigation scenario A1b-Latin America cropland Improved irrigation scenario A1b–Latin America grassland Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 53 TABLE 5.1: Private Savings of Different Technologies Per Ton of Carbon Dioxide Sequestered LESS THAN US$50 US$51 TO $100 MORE THAN US$100 Africa Cover crops, residues, other soil amendments, Manure No tillage, biochar, inorganic fertilizer, intercrop- terracing, afforestation, tree crop farming, rotation, ping, alley farming rainwater harvesting, cross-slope barriers, pasture improvement, grazing management, pasture on degraded lands Asia Intensify rotation, cover crops, crop-to-plantation, Include trees, organic soil Residues, rainwater harvesting, inorganic fertilizer, afforestation, annual-to-perennial grass, restoration, biofertilizer, no-tillage, manure, biochar pasture on degraded land, grazing management, improved irrigation crop-to–grassland Latin America Diversify trees, pasture-to-forest, cover crops, Pasture on degraded land No-tillage, improved irrigation, diversify rotation, afforestation, pasture- to-plantation, intercropping, biochar, intensify rotation, rainwater harvesting, manure, include trees, residues, annual-to-perenni- grassland–to-plantation, grazing management al grass, pasture improvement Source: This study. America (table 5.2). Intensive vegetable production, biofer- The total mitigation potential varies from 2.3 Gt CO2-eq for tilizer application, and organic soil restoration also display Latin America to 7.0 Gt CO2-eq for Asia (table 5.3). Total relatively high costs in Asia, while in Latin America, the land private pro�ts range from US$105 billion in Africa to $1.4 tril- management practices with the largest costs are mainly lion in Asia, while total public costs range from $20 billion in those associated with trees. Africa to $160 billion in Asia. Figure 5.2 indicates that all the land management practices generate bene�ts to the farmers, but at varying costs to the 5.2 TRADE-OFFS IN SOIL CARBON public. Private bene�ts that motivate decisions often fall SEQUESTRATION short of social costs, with the implication that in the absence Trade-off is inherent in the attempt to achieve the triple wins of countervailing policies, GHGs from poor land management of food security, increased resilience, and reduced GHG will continue to accumulate in the atmosphere. The total emissions. For instance, attempts to increase soil carbon cost for afforestation was highest for Africa (US$2.8 billion), storage through afforestation may reduce productivity (prof- Asia (US$16.7 billion), and Latin America (US$5.5 billion), itability), as afforestation tends to take land out of production while the lowest total public cost was for terracing in Africa for a signi�cant period of time. Conversely, intercropping, the (US$18.7 million), inorganic fertilizer in Asia (US$154.7 mil- growing of crops near existing trees, provides synergy be- lion), and rotation diversi�cation in Latin America (US$30.1 tween pro�tability and increased soil carbon sequestration. million). TABLE 5.2: Public Costs of Different Technologies Per Ton of Carbon Dioxide Sequestered LESS THAN US$10 US$10 TO $20 MORE THAN US$20 Africa Tree crop farming, rainwater harvesting, Biochar, inorganic fertilizer, intercropping, no-tillage, manure, cover crops, rotation alley farming, include trees, afforestation, intensi�cation, rotation diversi�cation, pasture establishment on degraded land residues, terracing, slope barriers, improved fallows, other soil amendments, improved pastures Asia Residues, rainwater harvesting No-tillage, manure, improved irrigation, Inorganic fertilizer, alley farming, biochar, intensify rotation, cover crops, crops- include trees, organic soil restoration, to-plantation, grazing management, biofertilizer, afforestation, intensive veg- cropland–to–grassland etables, annual-to-perennial grass, pasture establishment on degraded land Latin America No-tillage, diversify rotation, intensify Improved irrigation, biochar Grassland–to-plantation, diversify trees, rotation, rainwater harvesting, cover crops, pasture to forest, afforestation, pasture to manure, residues, grazing management plantation, intercropping, include trees Source: This study. EC O N O M I C A N D S E CT OR WORK 54 C H A PTER 5 — EC ONOMIC S OF SOIL CA R B ON SEQU ES TR ATION FIGURE 5.2: Total Private Bene�ts (Orange) and Public Costs (Green) of Land Management Practices (US$, Billion) for the B1 Scenario other soil amendments rotation intensification rotation diversification rainwater harvesting tree crop farming inorgnic fertilizer improved fallow slope barriers intercropping include trees afforestation alley frming cover crops no tillage terracing residues manure 4 2 0 –2 –4 –6 –8 –10 –12 –14 Africa –16 organic soil restoration intensive vegetables rainwater harvesting improved irrigation crop-to-plantation intensify rotation inorgnic fertilizer include-trees afforestation alley frming cover crops biofertilizer no tillage residues manure biochar 50 0 –50 –100 –150 –200 –250 –300 –350 Asia –400 grassland to plantation pasture to plantation rainwater harvesting improved irrigation intensify rotation diversify rotation pasture to forest diversify trees intercropping include-trees afforestation cover crops no tillage manure residue biochar 20 0 –20 –40 –60 –80 –100 –120 Latin America –140 –160 Source: This study. Notes: The public costs for Africa were adapted from a World Bank study on Nigeria’s Agricultural, Forest, and Other Land Use sectors where public support for agriculture is 3 percent. The public costs for Asia and Latin America were assumed to increase proportionately to the state support for agriculture for China (8 percent) and Brazil (6 percent), respectively. Synergies and trade-offs analyses can therefore help in quan- pro�tability and between private bene�ts and public costs. tifying the extent of “triple wins� of different land manage- The analysis was limited to the Africa dataset, as the graphs ment technologies. Synergies and trade-offs in CSA affect for other regions exhibit similar patterns leading to the same decision making at various levels ranging from the household conclusions. to the policy levels. Figure 5.3 reveals synergies between pro�tability and miti- In this study, trade-off was analyzed by using two-dimen- gation in two agroforestry systems: intercropping and alley sional graphs to depict relationships between carbon and farming (top right quadrant of �gure 5.3). Intercropping is CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 55 TABLE 5.3: Technical Mitigation Potential, Private Bene�ts, and Public Costs of the Land Management Technologies by 2030 TECHNICAL POTENTIAL PRIVATE BENEFITS PUBLIC COSTS SCENARIO (MILLION TONS CO2-eq) (US$, BILLION ) (US$, BILLION ) Africa B1 3,448 105.4 19.6 A1b 3,505 108.6 19.7 B2 3,678 111.4 20.8 A2 3,926 120.9 22.3 Asia B1 5,977 1,224.5 131.3 A1b 6,388 1,259.3 143.6 B2 7,007 1,368.1 159.7 A2 6,678 1,310.8 150.4 Latin America B1 2,321 273.8 40.8 A1b 2,425 279.4 42.9 B2 2,538 288.8 44.3 A2 3,097 319.4 55.1 Source: This study. Notes: B1 = a world more integrated and more ecologically friendly; A1b = a world more integrated with a balanced emphasis on all energy sources; B2 = a world more divided but more ecologically friendly; A2 = a world more divided and independently operating self-reliant nations. FIGURE 5.3: Trade-Offs Between Pro�tability and Carbon Sequestration of Sustainable Land Management Technologies in Africa 1000 pro�t per tone of carbon dioxide sequestered (US $) No-tillage Inorganic fertilizer Intercropping 100 Manure Alley farming Cover crops Soil amendments Include trees Crop residues Terracing Afforestation Rotation Rotation diversi�cation Tree crop farming 10 intensi�cation Rainwater harvesting Improved Cross slope barriers fallow 1 0 2 4 6 8 10 carbon dioxide sequestered (ton per hectare per year) Source: This study. EC O N O M I C A N D S E CT OR WORK 56 C H A PTER 5 — EC ONOMIC S OF SOIL CA R B ON SEQU ES TR ATION growing crops near existing trees, whereas alley farming is conservation Manure is less pro�table than inorganic fertil- growing crops simultaneously in alleys of perennials, prefer- izer because of the labor costs associated with collecting and ably leguminous trees or shrubs. Both are important strate- processing manure (top left quadrant of �gure 5.3). Manure gies for increased productivity and resilience of the farming also has quite low nutrient content relative to inorganic fertil- system. Land management technologies in the lower right izers, so a large amount needs to be applied on relatively quadrant of �gure 5.3 have high carbon sequestration rates small �elds. This explains why manure works well for small- but are modestly pro�table. Afforestation, improved fallow scale intensive and high-value vegetable gardening. Manure involving the use of fast-growing trees to accelerate soil re- systems are also associated with high methane emissions. habilitation, including trees in croplands, and establishing bar- The relatively high pro�tability of no-tillage derives primarily riers across sloping areas, tends to take land out of produc- from the decrease in production costs after establishment of tion for a signi�cant period of time. It reduces the amount of the system. land available for cultivation in the short run, but can lead to The relationship between public costs and private bene�ts overall increases in productivity and stability in the long run. of the land management technologies is shown in �gure 5.4. The time-averaged, above-ground biomass of crop residues Public cost refers to government support toward the imple- and other technologies in the lower left quadrant of �gure 5.3 mentation of land management practices. They include is relatively small compared to that of agroforestry systems. investments in seeds and seedlings, input subsidies, exten- Also, the biomass of crop residues does not accumulate eas- sion services, and other administrative costs. The pattern of ily, resulting in lower mitigation bene�ts. public support is as crucial as the amount of support for full realization of productivity, mitigation, and adaptation bene�ts Judicious fertilizer application counters soil nutrient deple- in agriculture. Public support that focuses on research, invest- tion, reduces deforestation and expansion of cultivation to ments in improved land management, and land tenure rather marginal areas, and increases crop yields. Reversing devel- than on input support are generally more effective, bene�t oping countries’ (especially Africa’s) soil productivity declines more farmers, and are more sustainable in the long run. cannot be adequately addressed without increased fertilizer use. Farmers apply 9 kg/ha of fertilizer in Africa compared to Technologies that involve signi�cant change in land-use (af- 86 kg/ha in Latin America, 104 kg/ha in South Asia, and 142 forestation and improved fallows) and landscape alteration kg/ha in Southeast Asia (Kelly 2006). Yields also increase with (terracing and cross-slope barriers) incur high public costs manure application and accumulation of soil carbon, but with but generate low private bene�ts (lower right quadrant of patterns that depend on crop type. Manure plays a crucial �gure 5.4). The low pro�ts suggest that farmers may be re- role in improving fertilizer use ef�ciency and soil moisture luctant to privately invest in these technologies. Strong public FIGURE 5.4: Relationship between Private Bene�ts and Public Costs in Africa 1,000 private bene�t (per tonne of carbon dioxide sequestered) No-tillage Inorganic fertilizer Intercropping 100 Manure Alley farming Cover crops Terracing Include trees Afforestation Crop residues Tree crop farming 10 Crop rotation Improved fallow Rainwater harvesting Cross slope barriers 1 0 3 5 8 10 13 public cost ($ per tonne of carbon dioxide sequestered) Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 57 involvement in these technologies is justi�able given their rel- making. Sustainable land management interventions should atively high mitigation potentials. Crop residues, cover crops, be planned and implemented in a coordinated manner crop rotation, and rainwater harvesting with lower pro�ts and across space, time, and sectors. Working at the landscape also manure and no tillage that generate relatively higher prof- level with an ecosystems approach is useful for addressing its require minimal government support (lower left and upper food security and rural livelihood issues and in responding to left quadrants of �gure 5.4, respectively). These technologies the impacts of climate change and contributing to its mitiga- generally have low mitigation potentials. The relatively high tion. The landscape level is the scale at which many eco- public cost of inorganic fertilizer (top right quadrant, �gure system processes operate and at which interactions among 5.4) reflects the use of subsidies in spurring farmers’ access agriculture, environment, and development objectives are to the technology. Fertilizer subsidy is, however, associated mediated. It entails the integrated planning of land, agricul- with high �scal costs, dif�cult targeting, and crowding out of ture, forests, �sheries, and water at local, watershed, and commercial sales. Thus, fertilizer subsidies are appropriate regional scales to ensure synergies are properly captured. in situations when the economic bene�ts exceed costs, the The landscape approach provides a framework for the bet- subsidies help achieve social rather than economic objec- ter management of ecosystem services, such as agricultural tives, and the support helps improve targeting through mar- productivity, carbon storage, fresh water cycling, biodiversity ket-smart subsidies while providing impetus for private sec- protection, and pollination. It allows trade-offs to be explic- tor input development. Examples of market-smart subsidies itly quanti�ed and addressed through negotiated solutions include demonstration packs, vouchers, matching grants, and among various stakeholders. loan guarantees (Agwe, Morris, and Fernandes 2007). Two examples taken from World Bank (2011c) illustrate the ef�cacy of the landscape approach. The �rst example is the silvopastoral farming systems of Costa Rica and Nicaragua. 5.3 IMPLICATIONS OF THE TRADE-OFFS After several years of intensive grazing in Costa Rica and IN LAND-USE DECISIONS Nicaragua, pastures were degraded, erosion was accelerat- The trade-offs exhibited by the land management tech- ing, and livestock productivity was falling. To address these nologies have important implications for land-use decision challenges, a pilot project introduced silvopastoral techniques PHOTO 5.1: Terracing and Landscape Management in Bhutan Source: Curt Carnemark/World Bank. EC O N O M I C A N D S E CT OR WORK 58 C H A PTER 5 — EC ONOMIC S OF SOIL CA R B ON SEQU ES TR ATION to 265 farms on 12,000 ha between 2001 and 2007. A can be a signi�cant barrier in situations where farmers might payment scheme for environmental services—carbon se- want to invest in a technique. Third, lack of information on questration and biodiversity conservation—was introduced the potentials of alternative techniques of farming and limited as an additional income stream for livestock production. capacity is a major constraint in many developing countries. Silvopastoral techniques were used to transform degraded Fourth, when technologies are inconsistent with community lands with monocultures of one grass species into more rules and traditional practices, their adoption will most likely complex agroforestry systems of different tree species, live encounter the resistance of the people. Last, willingness and fences, riparian forests, and trees dispersed in pastures. the ability to work together are crucial for many technologies The techniques have been shown to enhance biodiversity, such as improved irrigation and communal pastures. The ab- sequester carbon, and reduce methane emissions. Results sence of collective action will hinder successful uptake, dif- showed a typical win-win situation: An annual sequestration fusion, and impact of these land management technologies. of 1.5 Mt of CO2-equivalent was accompanied with increases of 22 percent in milk production, 38 percent in stocking rate, Factors affecting adoption tend to be more speci�c to the and 60 percent in farm income. The methane emission per land management technologies. Table 5.4 suggests that product kilogram decreased, while biodiversity (measured by lack of credit and inputs and land tenure problems are by the number of bird species and water quality) increased. far the most important factors for adoption across the range of technologies. However, improved availability of inputs is The last example is one of the world’s largest erosion control a necessary but insuf�cient condition for adoption of land programs in China. Revegetation has successfully restored management practices. Better market prices for crops and the devastated Loess Plateau to sustainable agricultural other agricultural produce are crucial. Secure land rights is production, improving the livelihoods of 2.5 million people a precondition for climate-smart agriculture as it provides in- and securing food supplies in an area where food was some- centive for local communities to manage land more sustain- times scarce in the past. The project encouraged natural ably. Ill-de�ned land ownership may inhibit sustainable land regeneration of grasslands, trees, and shrubs on previously management changes. cultivated sloping lands. Replanting and a grazing restriction allowed the perennial vegetation cover to increase from 17 Behavioral change through education is required to enable to 34 percent between 1999 and 2004, sustaining soil fertility changeover to improved land management technologies. For and enhancing carbon sequestration. Together with terrac- instance, conservation agriculture, the farming system involv- ing, these measures not only increased average yields, but ing no-tillage, residue management, and use of cover crops, also signi�cantly lowered their variability. Agricultural produc- is highly knowledge intensive, requiring training and practi- tion has changed from generating a narrow range of food cal experience of those promoting its adoption. Learning and low-value grain commodities to high-value products. As hubs, regional platforms, scienti�c research, south-south a result, the evolution of farm and family incomes has shown knowledge exchange, and technical support mechanisms a steady increase. It is estimated that as many as 20 mil- may increase innovation and facilitate adoption of improved lion people have bene�ted from the replication of the Loess land management technologies. The knowledge base of land Plateau approach throughout China. management practices at the local level can also be improved through careful targeting of capacity development programs. 5.4 SUSTAINABLE LAND MANAGEMENT Table 5.5 summarizes possible demand- and supply-side in- ADOPTION BARRIERS terventions for facilitating the adoption of sustainable land management inputs. It is unlikely that any of these interven- Despite the fact that improved land management technolo- tions alone will be effective in increasing input use. Careful gies generate private bene�ts, their adoption faces many selection of combinations of demand- and supply-side mea- socioeconomic and institutional barriers. The commonly sures will allow the demand and supply to grow, leading to cited risk-related barriers to adoption of carbon sequester- the emergence of viable private sector–led input markets. ing technologies in agriculture are permanence, leakage, and additionality (box 5.1). Beyond these, there are a number of other implementation constraints. 5.5 POLICY OPTIONS FOR SOIL CARBON SEQUESTRATION First, most of the land management technologies require signi�cant up-front expenditure that poor farmers cannot af- Private bene�ts that drive land-use decisions often fall short ford. Second, the nonavailability of inputs in the local markets of social costs; thus, carbon sequestration may not reach an CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 59 BOX 5.1: Risk-Related Barriers to Adoption of Soil Carbon Sequestration Activities • Permanence: Permanence refers to the secure of leakage. Macroeconomic policies induce retention of newly sequestered carbon. Carbon changes in market conditions and prices, sequestration only removes carbon from the which in turn influence farmers’ land-use atmosphere until the maximum capacity of the and management practices. While most occur- ecosystem is reached, which may be about rence of leakage has a negative effect on project 25 years for most land management practices. bene�ts, positive leakage spillover effects Storage of carbon in soils is relatively volatile that lead to reduction in emissions outside and subject to re-emission into the atmosphere the project boundary can occur. This could be in a subsequent change in land management. as a result of technology transfer or changes The risk of nonpermanence is lower when the in market conditions that stimulate mitigation adoption of soil carbon sequestration practices activities. also leads to more pro�table farming systems. • Additionality: The concept implies that in order Note that not all agricultural mitigation options to attract compensation, emissions reduc- are transient. Substitution of fossil fuels by tion must be in addition to what would have bioenergy is a permanent mitigation option, and occurred under the business-as-usual scenario. reduction in nitrous oxide and methane emis- Additionality is usually calculated as postproject sions are nonsaturating. carbon stocks less the forward-looking baseline, • Leakage: Leakage occurs when a project dis- less deduction for leakage and risk of reversal, places greenhouse gas emissions outside its and less emission generated by the project boundary. For instance, control of grazing in an (Fynn et al. 2010). area might force herders to move their animals Permanence, leakage, and additionality can be ad- to another location. Economic adjustment to dressed through temporary crediting, ex ante discount- meet market demand is the underlying driver ing, and comprehensive accounting (Murray et al. 2007). TEMPORARY CREDITING EX ANTE DISCOUNTING COMPREHENSIVE ACCOUNTING Description Balances debits and credits for �nite Accounts for the possibility of future Balances debits and credits as they oc- periods with provision for reversal loss by reducing the amount of credit cur in the course of the project. These at the onset based on the expectation can be based on stock change or aver- of reversal age stock change during the period Environmental rigor Rigorous as temporary credits must be Credits may not equal debits for a given This achieves consistency as long as replaced when they expire project; as such, ex ante discounting the system is monitored perpetually may lead to underdebiting or overdebit- ing of ex post reversal. Feasibility of Enables up-front payment; book balanc- Relatively easy to impose discounts on Boosts attractiveness of investment by implementation ing at the end of the project is also credits if amounts of reversal can be allowing credits to be earned as soon possible reasonably projected as they are generated by the project; however, perpetual accounting may hinder balancing the books at the end of a �nite-life project Transaction costs Measurement, monitoring, and veri�ca- MMV are not necessary; rather, credits MMV are carried out into perpetuity tion (MMV) and contract renewal costs are reduced by formula, not observed need to be borne by the project changes in carbon Source: Table synthesized from Murray et al. 2007. optimal level from a social point of view unless some mecha- 1. Strengthen the capacity of governments to imple- nisms exist to encourage farmers. Some public policies that ment climate-smart agriculture. Countries must can potentially incentivize carbon sequestration include the be prepared to access new and additional �nance. following: There is a need to build the technical and institutional EC O N O M I C A N D S E CT OR WORK 60 C H A PTER 5 — EC ONOMIC S OF SOIL CA R B ON SEQU ES TR ATION TABLE 5.4: Relative Importance of Different Factors for Adopting Improved Land Management Practices LAND MANAGEMENT INPUTS/ MARKET TRAINING/ LAND TECHNOLOGY CREDITS ACCESS EDUCATION TENURE RESEARCH INFRASTRUCTURE Inorganic fertilizer *** ** ** ** * ** Manure ** ** * ** * ** Conservation agriculture ** ** *** ** ** * Rainwater harvesting ** ** ** *** ** ** Cross-slope barriers ** * ** ** ** * Improved fallows ** * * *** ** * Grazing management *** *** ** *** ** * Source: Synthesized from Liniger et al. 2011. Key * = Low importance, ** = Moderate importance; *** = High importance. TABLE 5.5: Interventions for Facilitating Increased Input Use DEMAND-SIDE INTERVENTIONS SUPPLY-SIDE INTERVENTIONS Strengthen soil-crop research and extension Reduce input sourcing costs Support to public agencies Lowering trade barriers to increase national and regional market size Public-private partnership On-farm trials and demonstrations Improve farmers’ ability to purchase inputs Reduce distribution costs Improve access to credits Improve road and rail infrastructure to lower transport costs Phased and incremental use (e.g., small bags for fertilizers) Implement laws that enables farmers to use risk-free collaterals for loans Provide farmers with risk management tools Strengthen business �nance and risk management Improved weather forecasting, weather-indexed crop insurance Use credit guarantee and innovative insurance schemes Improved quality and dissemination of market information Improve supply chain coordination mechanisms Public and private sector information systems easily accessible to farmers Product grades and standards Market information systems to reduce information costs Protecting farmers against low and volatile output prices Investment in measures to reduce production variability such as drought-tolerant crops, deep-rooted crops, irrigation, and storage systems Empowering farmers by supporting producer organizations Investment in rural education Training farmers in organizational management Improving the resource base so that input use is more pro�table Investment in soil and water management and irrigation infrastructure Source: Modi�ed from Agwe, Morris, and Fernandes (2007). capacity of government ministries to implement progress in incorporating it into the UNFCCC has climate-smart agriculture programs. Existing national been slower than many people hoped for. Although policies, strategies, and investment plans should be the negative impacts of agricultural production in strengthened to form the basis for scaling-up invest- terms of land-use change and GHG emissions were ments for climate-smart agriculture. Readiness for reasonably well covered by the convention, the real carbon sequestration and climate-smart agriculture and potential contributions the sector can and does can be achieved through improved extension services make in terms of sequestering carbon in agricultural and training in relevant land management technolo- biomass and soils were for the most part omitted. gies for different locales. Redressing this omission promises to foster a more 2. Global cooperative agreement. Given the tremendous balanced perspective in which food security is not signi�cance that agriculture has for the global climate, necessarily at odds with climate change adaptation CARBON SEQUESTRATION IN AGRICULTURAL SOILS C H A P T E R 5 — E C O NOMICS OF SOIL CARBON SE QUESTR ATION 61 and mitigation (an unworkable conflict in which longer agricultural policies and action frameworks will term environmental concerns are virtually guaranteed increase the adoption of sustainable land manage- to universally lose out politically to the more immedi- ment practices. However, public investment is only ate concern of food supply). A more practical and one sphere, involving the private sector in climate- thorough picture makes it possible for agriculture to smart agriculture and sustainable land management be rewarded for its positive environmental impacts, is the other. and to be an integral part of “the solution� as well 5. Create enabling environments for private sector as part of “the problem.� This is vitally important be- participation. Introducing policies and incentives that cause agriculture needs to be fully incorporated into provide an enabling environment for private sector adaptation and mitigation strategies. As a result, the investment can increase overall investment. This pri- international community has recognized the impor- vate investment can be targeted to some degree as tance of integrating agriculture into the ongoing ne- well, particularly when government priorities translate gotiations on the international climate change regime. clearly into business opportunities and certain areas At the 17th Conference of Parties to the UNFCCC in of investment are looked upon favorably by public of- Durban, South Africa, in November, 2011, the parties �cials and institutions. Public investment can also be asked the UNFCCC Subsidiary Body for Scienti�c and used to leverage private investment in areas such as Technological Advice to explore the possibility of a research and development, establishing tree planta- formal work program on agriculture. tions, and in developing improved seeds and seed- 3. Boost �nancial support for early action. A blend of lings. Particular attention should go to encouraging public, private, and development �nance will be private �nancial service providers to tailor instruments required to scale-up improved land management that enable farmers who adopt SLM practices to practices. Integrating sources of climate �nance with overcome the barriers described above. Bundling ag- those that support food security may be one of the ricultural credit and insurance together and providing most promising ways to deliver climate-smart different forms of risk management, such as index- agriculture with the resources it requires. For based weather insurance or weather derivatives, are technologies that generate signi�cant private areas of private investment that can be encouraged returns, grant funding or loans may be more through public policy and public-private partnerships. suitable to overcoming adoption barriers. For tech- nologies such as conservation agriculture that REFERENCES require speci�c machinery inputs, and signi�cant Agwe, J., Morris, M., and Fernandes, E. 2007. “Africa’s Growing Soil up-front costs, payment for ecosystem services Fertility Crisis: What Role for Fertilizer?� World Bank Agriculture scheme could be used to support farmers and and Rural Development Note Issue 21 (May): 4 pp. break the adoption barrier. There is also the Fynn, A. J., Alvarez, P., Brown, J. R., George, M. R., Kustin, C., Laca, potential for carbon �nance to support farmers E. A., Old�eld, J. T., Schohr, T., Neely, C. L., and Wong, C. P. 2010. during the initial period before the trees in Soil Carbon Sequestration in U.S. Rangelands: Issues Paper for agroforestry systems generate an economic return. Protocol Development. Environmental Defense Fund, New York. 4. Raise the level of national investment in agriculture. Kelly, V. A. 2006. Factors Affecting Demand for Fertilizer in Sub- While this may appear a tall order in countries with Saharan Africa. Agriculture and Rural Development Discussion Paper 23. World Bank, Washington, DC. severe budget constraints, �nite public resources can be more selectively targeted using the criteria Liniger, H. P., Mekdaschi Studer, R., Hauert, C., and Gurtner, M. 2011. Sustainable Land Management in Practice—Guidelines given above—prioritizing technologies that generate and Best Practices for Sub-Saharan Africa. World Overview no short-term returns and those that most effec- of Conservation Approaches and Technologies and Food and tively address the barriers that prevent prospective Agriculture Organization of the United Nations. adopters from moving forward. In some cases, Murray, B. C., Sohngen, B. L., and Ross, M. T. 2007. “Economic relatively affordable technologies that generate quick Consequences of Consideration of Permanence, Leakage, and and demonstrable bene�ts may warrant priority and Additionality for Soil Carbon Sequestration Projects.� Climatic potentially establish some of the channels through Change 80: 127–143. which more sophisticated technologies are dispersed World Bank. 2011. Climate-Smart Agriculture—A Call to Action. in the future. Nationally owned climate-smart Washington, DC. EC O N O M I C A N D S E CT OR WORK AP P E N D I X A — FA RMING P RACT ICE E F F E CT, NUMB ER OF ESTIMATES , A ND FEATUR E IN LA ND MA NA GEMENT 63 Appendix A: FARMING PRACTICE EFFECT, NUMBER OF ESTIMATES, AND FEATURE IN LAND MANAGEMENT PRACTICES Africa LAND MANAGEMENT NUMBER OF MEAN EXPERIMENTAL DESIGN PRACTICES ESTIMATES SUBTOTALS MEAN DURATION DEPTH (%) Nutrient management 60 8.3 15 100 Chemical fertilizer 30 Animal manure 30 Tillage and residue 184 4 15 100 management No tillage 108 Residues 46 Mulches 6 Cover crops 24 Agroforestry 185 6 22 100 Trees/forest 125 Intercropping 14 Alley farming 46 Tree-crop farming 44 44 2.2 18.4 Land-use changes 103 3 20 100 Afforestation 16 Grazing pasture 32 Cropping intensity 55 Soil management 187 4.5 10 100 Crop rotation 49 Improved fallow 71 Natural fallow 68 Water management 56 2.5 12 100 Water/rain harvesting 33 Slope/barriers 22 Terracing 15 Others 100 Biochar 11 11 1.8 7.4 Soil amendment 15 15 1.8 10 EC O N O M I C A N D S E CT OR WORK 64 AP P E NDIX A — FARMING P RACT ICE EFFECT, NUMB ER OF ESTIMATES , A ND FEATUR E IN LA ND MA NA GEM ENT Asia NUMBER OF MEAN DURATION EXPERIMENTAL PRACTICES ESTIMATES SUBTOTALS (yr) MEAN DEPTH (cm) DESIGN (%) Nutrient management 443 17.0 26 97 Application of fertilizer 297 Application of manure 146 Tillage and residue 328 9.3 20 99 management Reduced or no till 48 Return of crop residues 189 to �eld Application of mulches 53 Cover crops 38 Agroforestry 75 8.3 27 64 Inclusion of trees 58 Intercropping 17 Intensi�cation 150 14.4 49 34 Intensive vegetables 57 Annual-to-perennial 36 Intensify rotation 43 Improved irrigation 10 Rain harvest 4 Land-use change 292 18.5 29 5 Crop-to-forest 60 Crop-to-plantation 158 Crop-to-grassland 35 Reduced grazing 39 Other amendments 25 8.6 18 100 and practices Biochar 6 Bio-inoculant 3 Gypsum 8 Sulfur 5 Lime 2 Zinc 1 TOTAL 14.5 29 68 CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X A — FA RMING P RACT ICE E F F E CT, NUMB ER OF ESTIMATES , A ND FEATUR E IN LA ND MA NA GEMENT 65 Latin America NUMBER OF MEAN DURATION EXPERIMENTAL PRACTICES ESTIMATES SUBTOTALS (yr) MEAN DEPTH (cm) DESIGN (%) Nutrient management 99 9.7 17.2 92 Application of fertilizer 74 Application of manure 25 Tillage and residue 364 8.9 21.8 90 management Reduce or no till 249 Return of crop residues 56 to �eld Application of mulches 16 Cover crops 33 Graze residues 10 Agroforestry 56 8.1 24.3 61 Inclusion of trees 43 Diversify trees 6 Intercropping 7 Intensi�cation 138 15.5 33.0 64 Intensify rotation 25 Diversify rotation 43 Improved irrigation 34 Improved pasture 15 Improved fallow 8 Annual-to-perennial 13 Land-use change 257 19.0 38.5 5 Pasture-to-forest 62 Crop-to-forest 59 Pasture-to-plantation 53 Grassland-to-plantation 32 Crop-to-plantation 14 Crop-to-pasture 7 Reduced or excluded 30 grazing Other amendments 17 5.2 29.1 82 Biochar 8 Lime 9 TOTAL 931 12.6 28.5 61 Source: This study. EC O N O M I C A N D S E CT OR WORK AP P E N D I X B — G E NE RAL SCE NARIO ASSUMP T IONS A ND A PPLICATION FOR WORLD REGIONS 67 Appendix B: GENERAL SCENARIO ASSUMPTIONS AND APPLICATION FOR WORLD REGIONS B.1 BASELINE SCENARIO The average fresh yield was converted to amount of residues Using the initial soil carbon stocks (in t C/ha) from the produced on the basis of IPCC equations (IPCC 2006). Harmonized World Soil Database, the models were run in Crop yields were grouped into three bins representing the reverse mode to estimate 25th percentile, the 50th percentile, and the 75th percentile initial carbon mass of decomposable plant material of the yields of a speci�c crop in one stratum to assess the (DPM), opportunity of adapting the residue management to local situations. initial carbon mass of resistant plant material (RPM), initial carbon mass of fast decomposing biomass Further, it served as a proxy to consider increase of yields (BIO-F), over time due to improved management practices including initial carbon mass of slow decomposing biomass the increase in application of inorganic fertilizer (integrated (BIO-S), nutrient management). For instance, a farmer in a spe- initial carbon mass of humi�ed organic matter (HUM), ci�c stratum whose current maize yield is within the 25th and percentile may be able to increase the yields to within initial carbon mass of soil. the 75th percentile due to increased inorganic fertilizer application. All models were run to equilibrium state increasing the organic inputs in 0.1 t C steps until the initial carbon stock Two scenarios were considered with regard to increased represented the equilibrium of the speci�c soil in each productivity as a result of integrated nutrient management stratum. The required addition of organic inputs to the soil practices: A shift from low productivity to medium produc- varied greatly depending on climate parameters and the clay tivity (25 percentile to 50 percentile of crop yields in a spe- content of the soil. However, the inputs were in line with ci�c stratum) and a shift from medium to high productivity observations made by Young (1997), who estimated plant (50 percentile to 75 percentile of crop yields in a speci�c biomass requirements to maintain soil organic matter range stratum). The crop yields for each stratum are presented in between 3.5, 7, and 14 t d.m. per ha per year for semi-arid, Appendix C. subhumid, and humid ecosystems, respectively. Manure Management For each stratum, one low organic input baseline scenario Generally, manure management can be classi�ed into direct was modeled for each crop and crop area, respectively, as- manure application and application of composted manure. suming a conventional management of 15 percent of resi- Similar to the procedure for residue calculation, the raw dues left on the ground after harvesting. manure and composted manure model inputs in tC/ha were estimated by applying IPCC factors to the average amount of farm animals per ha (IPCC 2006). The global data estimated B.2 GLOBAL MITIGATION SCENARIOS manure application in kg per ha of nitrogen. Therefore, the C Residue and Integrated Nutrient Management input per ha for each kg N was calculated based on Food and This scenario implies additional residue inputs due to crop Agriculture Organization of the United Nations (FAOSTAT) management improvement. The calculation of residues in- numbers of cattle, sheep, goats, pigs, and poultry for each puts from the crops was based on the global crop yield data. region. EC O N O M I C A N D S E CT OR WORK 68 AP P E NDIX B — GENERA L SC ENA RIO A SS UMPTIONS A ND A PPLIC ATION FOR WORLD R EGIONS The amount of manure/composted manure represents the from trees to the soil, either as litter or through pruning and amount of potential manure production and not the amount mulching. of manure actually spread on the �eld in the baseline. For each stratum, the average manure/composted manure B.3 APPLICATION TO WORLD REGIONS production was calculated for its use in the RothC model. Improved manure and composted manure application are Africa considered mitigation opportunities for all climatic regions. In Africa, crop yields have remained stagnant for decades due to continuous depletion of soil organic matter over time from unsustainable practices. To reverse this situation, sus- Green Manure/Cover Crops tainable practices such as cover cropping, water harvesting, Green manure is a type of cover crop grown to add organic agroforestry, and water and nutrient management to improve matter and nutrients to the soil. On average, such crops yield soil carbon sequestration, increase yields, and enhance resil- around 4 t dry matter ha−1 yr−1. The above-ground biomass ience to climate change need to be adopted. was converted into t C ha−1 yr−1 using the IPCC equations for N-�xing forage, non-N-�xing forage, and grass, and then Agroforestry options that produce high-value crops and ad- computing the average value for RothC modeling. Based on ditional sources of farm revenues offer additional mitigation this average conversion, the input value for the model was bene�ts. In general, the best package of practices for soil 1.44 t C ha−1, of which 0.43 C ha−1 was allocated as above- carbon sequestration for the region consists of a combina- ground input and 1.01 C ha−1 as below-ground input. tion of manure application, fertilizer application, and residue management. Agroforestry and Improved Fallow There is a signi�cant trade-off between residues on the Agroforestry, including improved fallow, was considered a �eld versus residues used for livestock feeding. Therefore, mitigation potential for all climate regions. 50 percent residue retention was assumed, and the remain- ing 50 percent was assumed to be removed as animal feed. Based on the literature research, the input value for the All mapping units (see page 45) in Africa were considered RothC model concerning improved fallow is found to be simi- for the modeling. A residue management scenario of 50 per- lar to that of other agroforestry systems. Take for instance, cent of available residues per ha was modeled for each of the following: the main crops in Africa. In addition, manure management (direct and composting), green manuring, and agroforestry In Zambia, improved fallows in maize systems with were modeled for each cluster. several nitrogen-�xing tree species (both coppiced and noncoppiced) resulted in above-ground carbon inputs Asia of 2.8 tC/ha on average (Kaonga and Coleman 2008). In many areas, the most dominant farming system is inten- In Asia, the introduction of the mungbean (Vigna sive wetland rice cultivation with or without irrigation. Rice is radiata) as a grain legume in the short fallow of the grown in the wet season under dry land farming. In the dry wheat-rice system produced a total biomass of season, a second crop of rice (where irrigated) or another 4.5 t d.m./ha (Yaqub et al. 2010). less water–demanding crop (legumes and coarse grains) is In Mexico, the use of several varieties of mucuna grown. Apart from rice, mixed smallholder farming systems (Mucuna pruriens) in rotation with maize produced on are dominant (soybean, maize, wheat, and roots and tubers) average 6.8 t d.m./ha (Eilittä et al. 2003). with currently low input (organic and inorganic), apart from Southeast Asian countries. Compared to green manure/cover crops, a robust average input value was used based on input values from various Like Africa, there exists a signi�cant trade-off between resi- studies for tropical and temperate climate regions, covering dues on the �eld versus residues used for livestock feeding. a wide range of different agroforestry practices such as al- All mapping units in Asia were considered for the modeling. ley cropping, trees on cropland, and so forth (Oelbermann, A residue management scenario of 50 percent of available Voroney, and Gordon 2005a, 2005b; Gama-Rodrigues and residues per ha was modeled for each of the main crops in Antonio 2011). Based on this, the input value for tropical and Asia. In addition, manure management (direct and compost- temperate agroforestry systems averaged 2.3 tC/ha and 1.06 ing), green manuring, and agroforestry were modeled for tC/ha, respectively. These values represent the organic input each of the strata. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X B — G E NE RAL SCE NARIO ASSUMP T IONS A ND A PPLICATION FOR WORLD REGIONS 69 TABLE B.1: Agricultural Systems and Mitigation Scenario in South America WHEAT MAPPING UNIT/ SOYBEAN SOYBEAN BEANS STRATUM MAIZE MAIZE MAIZE RICE AGROFORESTRY 2 4 6 8 9 12 Cover crop Mitigation options Cover crop Cover crop no-tillage Cover crop residue management Source: This study. South America Compared to South America, cover crops and residue man- Several agricultural systems exist in South America. The agement of rice-based systems were identi�ed as mitigation agricultural systems found in the mapping units/stratum are options. displayed in Table B.1. North America Common land management practices in South America in- clude rotational wheat/soybean and fallow systems, maize The dominant crops in mapping unit 1 are barley, wheat, soy- and soybean systems with residue management, and tillage. bean, maize, and pulses. Those in zone 3 are barley, wheat, The use of a cover crop during the fallow period was iden- soybean, maize, pulses, and sorghum. In zone 7, wheat, soy- ti�ed as a promising mitigation opportunity. No-tillage was bean, and maize predominate. Crops are mostly cultivated identi�ed as another mitigation option. Residue manage- during the summer with bare fallow during the winter. In ment in rice systems was modeled. No-tillage in soybean/ recent years, no-tillage has been adopted by many produc- maize systems was modeled and applied to the area where ers, but there are still opportunities to increase its use. No- soybeans and maize are grown. Cover crop was modeled tillage was considered a mitigation option in 50 percent of and applied to the total area of crops for which green manur- the cropped area. Green manure using winter cover crops ing is practiced. during the fallow period was also modeled. Central America Oceania The most dominant agricultural systems in Central America The main crops are wheat, barley, and pulses cultivated as are sorghum, beans/maize, rice, and agroforestry (table B.2). winter crops and usually in rotations. No-tillage is used in TABLE B.2: Agricultural Systems and Mitigation Scenario in Central America MAPPING UNIT/ BEANS STRATUM SORGHUM MAIZE RICE AGROFORESTRY 2 4 6 9 10 12 Cover crop Mitigation options Cover crop Residue management Source: This study. EC O N O M I C A N D S E CT OR WORK 70 AP P E NDIX B — GENERA L SC ENA RIO A SS UMPTIONS A ND A PPLIC ATION FOR WORLD R EGIONS approximately 50 percent of the cropped area. Residue is �ed above. The average fresh yield (for instance maize) was commonly left on the �eld or incorporated (around 75 per- converted to amount of residues using IPCC Guidelines. cent of the cropped area). Residues are sometimes burnt just before sowing. Crop yields were grouped into three bins representing the 25th percentile, the 50th percentile, and the 75th percentile of the yields of a speci�c crop in one stratum to assess the Russia opportunity of adapting the residue management to local The main crops are wheat and barley cultivated as summer situations. To account for possible trade-offs between reten- crops with bare fallow during the rest of the year. Tillage is tion of residues in the �eld and residues needed as livestock frequently used. No-tillage was modeled with the average feed, different fractions of residues applied in the �eld were value for organic inputs for the two main crops used. Winter modeled (25 percent, 50 percent, and 75 percent). Each crop cover crops during the fallow period for green manure were was modeled separately, but if there was more than one also modeled. cropping season, each season was modeled separately (e.g., maize 1s and maize 2s). Europe The main crops are wheat and barley in winter and maize Manure Management in summer. Cover crops and no-tillage techniques are rarely Generally, manure management can be classi�ed into direct used (around 1 percent). The scenarios modeled include use manure application and application of composted manure. of no-tillage. For each climate zone, the average value of in- Similar to the procedure for residue calculation, the raw ma- puts for the main crops was used. For mapping units 2, 7, nure and composted manure model inputs in t C/ha were and 8, no-tillage was assumed to be suitable on 35 percent estimated by applying IPCC factors to the average amount of the cropped area. Cover crops during the fallow period of farm animals per ha (IPCC 2006). The C input per ha was for green manure were also modeled. For each zone, the calculated based on FAOSTAT numbers of cattle, sheep, average value for summer and winter cover crops was used. goats, pigs, and poultry for each country of each region. For each mapping unit, the average manure/composted manure B.4 DETAILED MODELING FOR AFRICA production was calculated for its use in the RothC model (table B.3). Residue and Integrated Nutrient Management This scenario implies additional residue inputs due to crop Direct manure and composted manure application were management improvement. The calculation of residues in- modeled in combination with different fractions of crop resi- puts from the crops was based on the crop yield data identi- dues (25 percent, 50 percent, and 75 percent) left in the �eld. TABLE B.3: Manure C Inputs for the Agroecological Zones (AEZs) in Africa Based on FAOSTAT DIRECT MANURE COMPOSTED MANURE MAPPING UNIT/AEZ t C/hA/APPL. t C/hA/APPL. 1 0.031 0.075 2 0.035 0.085 3 0.032 0.078 4 0.029 0.070 5 0.030 0.073 6 0.017 0.042 7 0.046 0.110 8 0.019 0.046 9 0.023 0.057 10 0.025 0.061 11 0.056 0.136 12 0.096 0.232 Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X B — G E NE RAL SCE NARIO ASSUMP T IONS A ND A PPLICATION FOR WORLD REGIONS 71 FIGURE B.1: FAO Land-Use Map 10 Herbaceous-mod. intensive pastoralism FAO Land use system 11 Herbaceous-intensive pastoralism Undefined 13 Rainfed agriculture (subsistence/commercial) 1 Forestry-no use/not managed (Natural) 14 Agro-pastoralism mod. intensive 2 Forestry-protected areas 15 Agro-pastoralism intensive 4 Forestry-pastoralism moderate or higher 16 Agro-pastoralism mod. intensive or higher with large scale irrigation 5 Forestry-pastoralism moderate or higher with scattered plantations 17 Agriculture-large scale irrigation (>25% pixel size) 6 Forestry-scattered plantations 18 Agriculture-protected areas 7 Herbaceous-no use/not managed (Natural) 19 Urban areas 8 Herbaceous-protected areas 20 Wetlands-no use/not managed (Natural) 9 Herbaceous-extensive pastoralism 21 Wetlands-protected areas Source: FAO and World Bank. Green Manure/Cover Crops (GMCCS) were used for the modeling. The activity data were Based on a study by Barahona (2004), the largest share of the crop areas of maize and sorghum. GMCCs worldwide is from Africa (51 percent) with maize Groundnuts + maize and groundnuts + sorghum: This sce- cropping systems being the most dominant (66 percent). nario assumes that groundnuts are intercropped with Other main crops include cassava and sorghum. The most maize and sorghum. The input values are the strata- frequently used GMCCs are Mucuna sp., Cowpeas, pigeon speci�c combinations of crop residues of groundnuts peas, and groundnuts. The following GMCCs scenarios are in addition to the residues of maize and sorghum, considered for the modeling: respectively. Only the mean values of residues were used for the modeling. The activity data are the crop Mucuna sp.: An input value of 3.27 t C/ha/year was areas of maize and sorghum. The input values are used for the modeling in all strata based on Kaizzi shown in table B.4. et al. (2006) and Anthofer (2005). The activity data for this scenario are potentially the area of all crops. Agroforestry, Improved Fallow, and Land Rehabilitation Cowpea + maize and cowpea + sorghum: This scenario Five different agroforestry mitigation scenarios were consid- assumes that cowpeas are predominantly inter- ered in this study. cropped with maize and sorghum. The input values are the strata-speci�c combinations of crop residues The General Agroforestry Mitigation Scenario of cowpeas in addition to the residues of maize and This scenario can be seen as representative for all agrofor- sorghum, respectively. Only mean values of residues estry systems on cropland. The input values were calculated EC O N O M I C A N D S E CT OR WORK 72 AP P E NDIX B — GENERA L SC ENA RIO A SS UMPTIONS A ND A PPLIC ATION FOR WORLD R EGIONS TABLE B.4: C Inputs for Different Green Manure/Cover Crop Systems COWPEA + GROUNDNUTS + MAPPING UNIT MUCUNA COWPEA + MAIZE SORGHUM GROUNDNUTS + SORGHUM (AFRICA) (tC/ha/YEAR) (tC/ha/YEAR) (tC/ha/YEAR) MAIZE (tC/ha/YEAR) (tC/ha/YEAR) AEZ 01, Clay 25 3.27 1.45 1.24 1.83 2.51 AEZ 01, Clay 50 3.27 2.12 1.58 2.37 3.15 AEZ 01, Clay 75 3.27 1.73 1.40 2.13 2.87 AEZ 02, Clay 25 3.27 1.51 1.34 1.85 2.59 AEZ 02, Clay 50 3.27 1.61 1.40 1.94 2.71 AEZ 02, Clay 75 3.27 1.61 1.35 1.99 2.73 AEZ 03, Clay 25 3.27 1.66 1.43 1.93 2.71 AEZ 03, Clay 50 3.27 1.76 1.46 2.02 2.81 AEZ 03, Clay 75 3.27 1.62 1.40 1.89 2.65 AEZ 04, Clay 25 3.27 1.63 1.44 1.88 2.69 AEZ 04, Clay 50 3.27 1.79 1.50 2.03 2.84 AEZ 04, Clay 75 3.27 1.73 1.44 1.98 2.80 AEZ 05, Clay 25 3.27 1.70 1.52 1.89 2.67 AEZ 05, Clay 50 3.27 1.86 1.57 2.06 2.87 AEZ 05, Clay 75 3.27 1.77 1.57 1.91 2.73 AEZ 06, Clay 25 3.27 1.78 1.55 1.95 2.72 AEZ 06, Clay 50 3.27 1.78 1.58 1.87 2.61 AEZ 06, Clay 75 3.27 1.81 1.62 1.92 2.71 AEZ 07, Clay 25, N 3.27 3.76 2.93 3.27 4.52 AEZ 07, Clay 25, S 3.27 1.78 1.42 2.10 2.85 AEZ 07, Clay 50, N 3.27 4.58 3.17 4.25 5.74 AEZ 07, Clay 50, S 3.27 2.53 1.88 3.12 4.33 AEZ 07, Clay 75, N 3.27 3.81 3.20 3.51 5.03 AEZ 07, Clay 75, S 3.27 2.55 1.84 3.26 4.42 AEZ 08, Clay 25, N 3.27 0.00 0.00 2.14 2.99 AEZ 08, Clay 25, S 3.27 1.55 1.40 1.72 2.44 AEZ 08, Clay 50, N 3.27 0.00 0.00 1.91 2.79 AEZ 08, Clay 50, S 3.27 2.19 1.83 2.59 3.74 AEZ 08, Clay 75, N 3.27 0.00 0.00 1.78 2.64 AEZ 08, Clay 75, S 3.27 1.88 1.56 2.08 2.96 AEZ 09, Clay 25, N 3.27 0.00 0.00 2.11 3.06 AEZ 09, Clay 25, S 3.27 1.72 1.50 1.87 2.71 AEZ 09, Clay 50, N 3.27 0.00 0.00 2.15 3.09 AEZ 09, Clay 50, S 3.27 2.19 1.57 2.49 3.41 AEZ 09, Clay 75, N 3.27 0.00 0.00 1.94 2.91 AEZ 09, Clay 75, S 3.27 2.08 1.62 2.34 3.28 AEZ 10, Clay 25, N 3.27 0.00 0.00 2.62 3.47 AEZ 10, Clay 25, S 3.27 2.15 1.91 2.45 3.71 AEZ 10, Clay 50, N 3.27 0.00 0.00 2.23 3.02 AEZ 10, Clay 50, S 3.27 2.49 1.76 2.94 4.10 AEZ 10, Clay 75, N 3.27 0.00 0.00 2.36 3.18 AEZ 10, Clay 75, S 3.27 2.12 1.71 2.50 3.62 AEZ 11, Clay 25, N 3.27 0.00 0.00 2.68 3.62 AEZ 11, Clay 25, S 3.27 2.66 1.87 3.08 4.28 AEZ 11, Clay 50, N 3.27 0.00 0.00 2.70 3.66 AEZ 11, Clay 50, S 3.27 2.65 1.62 3.04 3.97 AEZ 11, Clay 75, N 3.27 0.00 0.00 2.81 3.82 AEZ 11, Clay 75, S 3.27 2.19 1.90 2.36 3.58 AEZ 12, Clay 25 3.27 0.00 0.00 2.35 3.25 AEZ 12, Clay 50 3.27 0.00 0.00 2.29 3.23 AEZ 12, Clay 75 3.27 0.00 0.00 2.58 3.38 Source: This study. Note: AEZ = Agroecological Zone. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X B — G E NE RAL SCE NARIO ASSUMP T IONS A ND A PPLICATION FOR WORLD REGIONS 73 TABLE B.5: C Inputs for Different Agroforestry Systems COFFEE AND COCOA COPPICED AGROFORESTRY SHADE TREE LEGUME IMPROVED IMPROVED LAND MAPPING UNIT GENERAL SYSTEMS FALLOW + MAIZE FALLOW + MAIZE REHABILITATION (AFRICA) (tC/ha/YEAR) (tC/ha/YEAR) (tC/ha/YEAR) (tC/ha/YEAR) (tC/ha/YEAR) AEZ 01, Clay 25 1.58 8.19 2.85 4.35 3.843 AEZ 01, Clay 50 1.58 8.19 3.27 4.77 3.843 AEZ 01, Clay 75 1.58 8.19 3.03 4.53 3.843 AEZ 02, Clay 25 5.17 8.19 2.86 4.36 3.843 AEZ 02, Clay 50 5.17 8.19 2.94 4.44 3.843 AEZ 02, Clay 75 5.17 8.19 2.95 4.45 3.843 AEZ 03, Clay 25 5.17 8.19 2.96 4.46 3.843 AEZ 03, Clay 50 5.17 8.19 3.05 4.55 3.843 AEZ 03, Clay 75 5.17 8.19 2.93 4.43 3.843 AEZ 04, Clay 25 5.17 8.19 2.95 4.45 3.843 AEZ 04, Clay 50 5.17 8.19 3.05 4.55 3.843 AEZ 04, Clay 75 5.17 8.19 3.06 4.56 3.843 AEZ 05, Clay 25 5.17 8.19 2.91 4.41 3.843 AEZ 05, Clay 50 5.17 8.19 3.04 4.54 3.843 AEZ 05, Clay 75 5.17 8.19 2.97 4.47 3.843 AEZ 06, Clay 25 5.17 8.19 2.94 4.44 3.843 AEZ 06, Clay 50 5.17 8.19 2.89 4.39 3.843 AEZ 06, Clay 75 5.17 8.19 2.92 4.42 3.843 AEZ 07, Clay 25, N 1.58 8.19 4.03 5.53 3.843 AEZ 07, Clay 25, S 1.58 8.19 3.05 4.55 3.843 AEZ 07, Clay 50, N 1.58 8.19 4.85 6.35 3.843 AEZ 07, Clay 50, S 1.58 8.19 3.80 5.30 3.843 AEZ 07, Clay 75, N 1.58 8.19 4.08 5.58 3.843 AEZ 07, Clay 75, S 1.58 8.19 3.83 5.33 3.843 AEZ 08, Clay 25, N 1.58 8.19 3.09 4.59 3.843 AEZ 08, Clay 25, S 5.17 8.19 2.82 4.32 3.843 AEZ 08, Clay 50, N 1.58 8.19 2.95 4.45 3.843 AEZ 08, Clay 50, S 5.17 8.19 3.46 4.96 3.843 AEZ 08, Clay 75, N 1.58 8.19 2.81 4.31 3.843 AEZ 08, Clay 75, S 5.17 8.19 3.15 4.65 3.843 AEZ 09, Clay 25, N 1.58 8.19 3.06 4.56 3.843 AEZ 09, Clay 25, S 5.17 8.19 3.01 4.51 3.843 AEZ 09, Clay 50, N 1.58 8.19 3.14 4.64 3.843 AEZ 09, Clay 50, S 5.17 8.19 3.48 4.98 3.843 AEZ 09, Clay 75, N 5.17 8.19 2.92 4.42 3.843 AEZ 09, Clay 75, S 5.17 8.19 3.35 4.85 3.843 AEZ 10, Clay 25, N 5.17 8.19 3.44 4.94 3.843 AEZ 10, Clay 25, S 5.17 8.19 3.46 4.96 3.843 AEZ 10, Clay 50, N 5.17 8.19 3.08 4.58 3.843 AEZ 10, Clay 50, S 5.17 8.19 3.83 5.33 3.843 AEZ 10, Clay 75, N 5.17 8.19 3.22 4.72 3.843 AEZ 10, Clay 75, S 5.17 8.19 3.47 4.97 3.843 AEZ 11, Clay 25, N 5.17 8.19 3.72 5.22 3.843 AEZ 11, Clay 25, S 5.17 8.19 3.94 5.44 3.843 AEZ 11, Clay 50, N 5.17 8.19 3.75 5.25 3.843 AEZ 11, Clay 50, S 5.17 8.19 3.91 5.41 3.843 AEZ 11, Clay 75, N 5.17 8.19 3.86 5.36 3.843 AEZ 11, Clay 75, S 5.17 8.19 3.46 4.96 3.843 AEZ 12, Clay 25 5.17 8.19 3.40 4.90 3.843 AEZ 12, Clay 50 5.17 8.19 3.33 4.83 3.843 AEZ 12, Clay 75 5.17 8.19 3.62 5.12 3.843 Source: This study. Note: AEZ = Agroecological Zone. EC O N O M I C A N D S E CT OR WORK 74 AP P E NDIX B — GENERA L SC ENA RIO A SS UMPTIONS A ND A PPLIC ATION FOR WORLD R EGIONS as mean values of more than 30 different systems taking Coppiced improved fallow + maize: This system into account different climate regions (humid, subhumid, and assumes a 3-year fallow period and a 7-year trees semi-arid) (see Schroeder 1995; Oelbermann, Voroney, and and maize period. The tree species considered are Kass 2005a, 2005b; and Lemma et al., 2006). It is modeled Gliricidia sepium, Callliandra callothyrsus, and Senna for all areas classi�ed as FAO Land Use System 13-18 (i.e., siamea. rain-fed agriculture, agro-pastoralism, and irrigated agricul- ture) (see Figure B.1). The input values from the trees are mean values of the dif- ferent tree species. In addition, the mapping unit-speci�c Perennial Crop—Tree Systems maize residues were included as organic inputs in this sys- This scenario considered two cash crops: coffee and cocoa. It tem (mean residues of maize). The input values represent assumes combinations of improved perennial crop manage- the mean annual input values over the whole system (fallow ment (pruning and mulching) and the introduction of shade and cropping period). These two mitigation scenarios were trees (see Szott, Palm, and Sanchez 1991; Szott, Fernandez, modeled for all maize areas. and Sanchez 1991; van Noordwijk et al. 2002; and Dossa et Land Rehabilitation al., 2008). Land degradation may be de�ned as the long-term loss of Improved Fallows + Maize ecosystem function and productivity caused by disturbances Improved fallows, in which leguminous trees and coppiced from which land cannot recover unaided. Due to the non- trees and shrubs are grown in association with crops, can availability of reliable spatial data of degraded lands in Africa, sequester substantial amounts of C in plants and the soil. the mitigation potential was applied to the FAO Land Use Following a study by Kaonga and Coleman (2008), two im- Systems 7-11 (herbaceous land-use systems in Figure B.1). proved fallow scenarios were modeled in association with The input value is based on organic inputs from tree-domi- maize. nated fallow systems (Szott et al. 1994). The C inputs for the �ve agroforestry systems are shown in table B.5. Legume improved fallow + maize: This system assumes a 2-year fallow period and a 1-year trees REFERENCE and maize period. The tree species considered are Young, A. 1997. Agroforestry for soil management. CAB International, Tephrosia vogelli, Cajanus cajan, and Sesbania sesban. Wallingford, UK CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X C — G L OBAL CROP YIE L DS ( T HA –1 YR –1 ) GR OUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE BINS 75 Appendix C: GLOBAL CROP YIELDS (T HA –1 YR –1) GROUPED INTO 25TH, 50TH, AND 75TH PERCENTILE BINS CORRESPONDING TO LOW, MEDIUM, AND HIGH BARLEY STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH (Leer) 1.45 2.72 3.99 4North America 3.16 4.13 5.11 10South America 1.97 2.42 2.86 4South America 0.69 0.70 0.70 11Asia 0.99 1.58 2.17 5Asia 0.92 1.37 1.83 11Oceania 2.55 2.61 2.66 5Europe 2.38 2.70 3.02 12South America 0.61 1.29 1.98 6Africa 0.97 1.21 1.45 1Asia 1.26 1.52 1.78 6Asia 1.18 1.89 2.59 1North America 2.50 2.70 2.90 6North America 1.15 2.08 3.02 1Russia 1.73 1.73 1.73 6South America 0.97 1.36 1.75 2Africa 0.56 1.22 1.87 7Africa 0.71 1.03 1.35 2Asia 1.35 1.98 2.61 7Asia 2.19 2.79 3.38 2Europe 1.92 2.42 2.92 7Europe 3.98 4.97 5.96 2North America 3.42 4.60 5.77 7North America 3.18 3.60 4.02 2Oceania 1.79 2.02 2.24 7Oceania 3.26 4.29 5.31 2South America 2.47 2.52 2.56 7Russia 1.58 1.69 1.81 3Africa 0.63 0.99 1.36 8Asia 2.14 2.70 3.26 3Asia 1.33 1.92 2.52 8Europe 2.83 3.60 4.37 3Europe 1.63 2.06 2.49 8North America 3.50 3.80 4.10 3North America 2.21 2.78 3.35 8Oceania 3.47 4.45 5.44 3Oceania 2.41 2.55 2.69 8South America 1.81 2.04 2.26 3Russia 1.63 1.75 1.88 9Africa 1.33 1.51 1.68 3South America 1.06 1.15 1.23 9Asia 1.10 1.16 1.22 4Africa 0.81 1.11 1.41 9Middle America 0.86 1.35 1.83 4Asia 0.93 1.10 1.27 9South America 0.56 0.69 0.83 EC O N O M I C A N D S E CT OR WORK 76 AP P E NDIX C — GL OBAL CROP YIELD S (T H A –1 Y R –1 ) GROUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE B INS BEANS STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 10Africa 0.49 0.65 0.81 4Africa 0.23 0.37 0.51 10Asia 0.81 0.98 1.15 4Asia 0.48 0.57 0.66 10Middle America 0.39 0.44 0.50 4Middle America 0.70 0.78 0.86 10South America 0.94 1.24 1.53 4North America 0.86 1.20 1.54 11Africa 0.67 0.86 1.04 4South America 0.17 0.32 0.47 11Asia 0.75 0.98 1.21 5Asia 1.19 1.37 1.54 11South America 1.05 1.11 1.17 6Africa 0.50 0.67 0.84 12Africa 0.49 0.68 0.87 6Asia 0.71 0.92 1.13 12Asia 0.77 0.82 0.87 6Middle 0.55 0.69 0.82 12Middle America 0.67 0.78 0.89 America 12South America 0.67 0.84 1.02 6North America 0.43 0.65 0.86 1Asia 1.41 1.47 1.53 6South America 0.70 0.94 1.18 1North America 1.60 1.76 1.92 7Africa 0.42 0.54 0.67 1Russia 1.46 1.46 1.46 7Asia 0.99 1.22 1.45 2Africa 0.45 0.93 1.40 7Europe 3.15 3.86 4.56 2Asia 0.52 0.64 0.76 7North America 1.56 1.71 1.86 2Europe 0.54 0.87 1.20 7Oceania 0.92 1.35 1.79 2North America 0.76 1.28 1.79 7Russia 0.81 0.86 0.92 2Oceania 1.03 1.13 1.22 8Asia 0.73 0.99 1.25 2South America 0.83 1.16 1.48 8Europe 2.19 3.39 4.60 3Africa 0.58 0.83 1.08 8Oceania 1.36 2.02 2.67 3Asia 1.15 1.36 1.57 8South America 0.73 0.96 1.19 3Europe 0.88 1.24 1.60 9Africa 0.39 0.54 0.68 3North America 1.44 1.78 2.12 9Asia 0.61 0.71 0.80 3Oceania 1.04 1.09 1.13 9Middle America 0.62 0.74 0.87 3Russia 1.45 1.46 1.47 9North America 0.47 0.55 0.64 3South America 0.93 1.03 1.12 9South America 0.36 0.72 1.08 CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X C — G L OBAL CROP YIE L DS ( T HA –1 YR –1 ) GR OUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE BINS 77 MAIZE STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 10Africa 0.72 0.89 1.06 4Asia 1.15 1.65 2.14 10Asia 1.85 2.35 2.86 4Middle America 0.78 1.16 1.54 10Middle 0.84 1.00 1.16 4North America 1.01 1.88 2.75 America 4South America 0.68 1.38 2.08 10South America 1.88 2.98 4.08 5Asia 3.89 4.44 4.99 11Africa 1.94 3.69 5.43 6Africa 0.84 1.60 2.35 11Asia 1.88 2.59 3.29 6Asia 2.08 2.87 3.66 11South America 3.28 4.20 5.12 6Middle America 1.36 1.70 2.03 12Africa 0.66 0.75 0.85 6North America 1.23 2.27 3.31 12Asia 1.52 2.11 2.70 6South America 1.64 2.50 3.35 12Middle 1.26 1.84 2.42 America 7Africa 1.55 2.54 3.54 12South America 2.02 2.68 3.35 7Asia 3.26 4.11 4.97 1Asia 4.85 5.81 6.78 7Europe 5.44 7.10 8.76 1North America 6.30 7.17 8.05 7North America 6.42 7.53 8.65 2Africa 1.04 2.53 4.02 7Russia 2.73 3.23 3.73 2Asia 0.95 1.83 2.70 8Africa 3.86 5.16 6.45 2Europe 6.12 7.96 9.80 8Asia 2.19 2.96 3.73 2North America 3.56 6.34 9.12 8Europe 4.46 6.25 8.04 2Oceania 3.73 5.79 7.86 8North America 5.40 6.38 7.37 2South America 4.29 5.59 6.89 8Oceania 6.54 8.28 10.02 3Africa 1.42 2.54 3.66 8South America 3.33 4.58 5.83 3Asia 3.82 4.80 5.77 9Africa 0.92 1.22 1.52 3Europe 2.61 3.82 5.02 9Asia 1.78 2.49 3.21 3North America 6.44 7.96 9.49 9Middle America 1.22 1.64 2.05 3South America 2.32 3.88 5.44 9North America 0.93 1.45 1.96 4Africa 0.83 1.15 1.47 9South America 1.56 2.48 3.41 MILLET STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 11Asia 1.06 1.18 1.30 4Africa 0.34 0.56 0.79 1Asia 0.88 1.21 1.54 4Asia 0.53 0.82 1.11 2Africa 0.34 0.54 0.75 5Asia 1.41 1.50 1.58 2Asia 0.39 0.67 0.95 6Africa 0.62 0.94 1.26 2North America 1.05 1.12 1.20 6Asia 0.55 0.72 0.89 3Africa 0.57 0.59 0.62 7Asia 1.43 1.71 1.99 3Asia 1.67 1.81 1.96 8Asia 0.62 0.90 1.18 3Europe 0.79 0.88 0.97 9Africa 0.80 1.03 1.26 3North America 1.23 1.43 1.62 9Asia 0.66 0.92 1.18 EC O N O M I C A N D S E CT OR WORK 78 AP P E NDIX C — GL OBAL CROP YIELD S (T H A –1 Y R –1 ) GROUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE B INS RICE SORGHUM STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 10Africa 1.75 2.17 2.59 10South America 3.13 3.20 3.27 10Asia 3.29 3.83 4.37 11Africa 0.73 1.43 2.12 10Middle America 2.40 3.22 4.05 11Asia 0.78 1.90 3.01 10South America 2.97 4.70 6.44 11Oceania 2.91 3.18 3.44 11Asia 2.83 4.25 5.68 12South America 2.37 2.85 3.34 12Africa 1.33 1.62 1.92 1Asia 3.93 4.12 4.31 12Asia 2.58 3.14 3.70 2Africa 0.59 1.69 2.80 12Middle America 2.31 3.42 4.53 2Asia 0.45 0.62 0.80 12South America 2.79 3.74 4.68 2Middle America 0.70 0.70 0.70 1Asia 5.73 6.77 7.82 2North America 2.45 3.10 3.75 2Africa 4.15 6.39 8.63 2Oceania 2.18 2.69 3.21 2Asia 2.30 3.05 3.81 2South America 4.26 4.55 4.84 2Europe 6.23 6.69 7.15 3Africa 1.02 1.67 2.33 2North America 7.54 8.29 9.04 3Asia 3.08 3.53 3.98 2Oceania 8.78 8.93 9.07 3North America 3.16 3.59 4.01 2South America 4.33 5.24 6.15 3Oceania 2.23 2.91 3.59 3Asia 3.44 4.91 6.38 4Africa 0.51 0.74 0.96 3Europe 3.43 4.50 5.57 4Asia 0.58 0.80 1.02 3North America 9.22 9.27 9.33 4Middle America 0.70 0.70 0.70 3Russia 3.14 3.14 3.14 4North America 1.61 2.64 3.67 3South America 5.46 6.18 6.90 4South America 0.64 1.10 1.56 4Africa 0.84 1.56 2.28 5Asia 3.81 3.81 3.81 4Asia 1.62 2.54 3.46 6Africa 0.89 1.24 1.60 4Middle America 2.66 3.42 4.18 6Asia 0.48 0.82 1.16 4North America 4.79 5.71 6.64 6Middle America 0.55 1.13 1.71 4South America 2.60 3.71 4.82 6North America 4.49 5.52 6.54 5Asia 4.45 4.90 5.36 6South America 1.60 2.16 2.72 6Africa 1.36 2.04 2.72 7Asia 3.22 3.70 4.19 6Asia 3.75 4.99 6.23 7North America 4.48 4.95 5.42 6Middle America 3.66 4.37 5.09 7Oceania 3.26 3.52 3.78 6North America 5.39 5.89 6.38 8Asia 0.38 0.77 1.17 6South America 3.57 4.94 6.30 8North America 4.20 4.68 5.17 7Asia 5.17 6.19 7.21 8Oceania 2.24 2.78 3.31 7Europe 5.97 6.34 6.70 8South America 4.13 4.54 4.96 7North America 6.35 6.57 6.79 9Africa 0.81 1.02 1.23 7Russia 3.32 4.05 4.78 9Asia 0.71 0.95 1.18 8Asia 4.26 5.45 6.63 9Middle America 0.94 1.54 2.15 8Europe 6.01 6.08 6.14 9North America 1.45 2.07 2.69 8North America 5.74 6.26 6.77 9South America 1.75 2.13 2.51 8South America 4.28 4.97 5.67 9Africa 1.19 1.64 2.09 9Asia 2.38 3.09 3.79 9Middle America 2.65 3.39 4.13 9South America 1.43 2.54 3.66 CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X C — G L OBAL CROP YIE L DS ( T HA –1 YR –1 ) GR OUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE BINS 79 SOYBEANS STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 10Asia 1.07 1.17 1.27 4North America 0.93 1.06 1.19 10South America 2.63 2.79 2.95 4South America 1.63 1.77 1.91 11Asia 0.96 1.28 1.59 5Asia 1.14 1.45 1.76 11South America 2.40 2.44 2.48 6Africa 0.47 0.51 0.55 12Asia 1.08 1.16 1.25 6Asia 0.98 1.49 2.01 12South America 2.35 2.55 2.76 6South America 2.19 2.46 2.72 1Asia 1.55 1.95 2.35 7Asia 1.40 1.82 2.24 1North America 1.80 2.04 2.27 7Europe 2.37 2.86 3.34 2Africa 1.06 1.81 2.55 7North America 2.13 2.47 2.82 2Asia 0.59 0.77 0.95 7Russia 1.06 1.21 1.36 2Europe 2.23 2.56 2.88 8Asia 1.13 1.60 2.07 2North America 1.84 2.11 2.39 8Europe 2.61 3.06 3.50 2Oceania 1.02 1.02 1.02 8North America 1.53 1.77 2.02 2South America 2.35 2.45 2.55 8Oceania 1.75 1.86 1.96 3Asia 1.09 1.50 1.90 8South America 2.18 2.41 2.63 3Europe 1.46 1.98 2.50 9Africa 0.61 0.75 0.89 3North America 2.21 2.52 2.83 9Asia 1.10 1.34 1.58 3South America 2.33 2.34 2.35 9Middle America 1.69 1.83 1.97 4Africa 0.70 0.97 1.25 9North America 1.31 1.31 1.31 4Asia 0.74 1.00 1.26 9South America 2.04 2.29 2.55 EC O N O M I C A N D S E CT OR WORK 80 AP P E NDIX C — GL OBAL CROP YIELD S (T H A –1 Y R –1 ) GROUPED INTO 25TH , 50TH , A ND 75TH PER C ENTILE B INS WHEAT STRATUM LOW MEDIUM HIGH STRATUM LOW MEDIUM HIGH 10Asia 1.56 1.74 1.92 4South America 0.68 0.84 1.01 10South America 1.95 2.22 2.49 5Asia 1.00 1.77 2.53 11Africa 1.38 2.28 3.18 5Europe 2.66 3.01 3.36 11Asia 1.41 1.87 2.33 5Russia 0.87 1.09 1.31 11Oceania 2.10 2.27 2.44 6Africa 1.05 1.33 1.61 11South America 2.19 2.30 2.42 6Asia 1.68 2.23 2.77 12South America 1.48 1.71 1.95 6North America 2.05 3.30 4.55 1Asia 0.92 1.48 2.04 6South America 1.05 1.48 1.91 1North America 1.98 2.20 2.42 7Africa 0.83 1.23 1.64 1Russia 1.06 1.26 1.46 7Asia 2.71 3.66 4.61 2Africa 0.97 2.08 3.19 7Europe 4.33 5.70 7.08 2Asia 1.72 2.36 3.00 7North America 3.30 3.80 4.31 2Europe 1.85 2.41 2.96 7Oceania 2.64 4.01 5.39 2North America 1.76 2.92 4.08 7Russia 2.61 3.07 3.52 2Oceania 1.55 1.85 2.15 7South America 4.08 4.31 4.54 2South America 2.10 2.63 3.16 8Africa 6.11 6.31 6.51 3Africa 0.96 1.72 2.48 8Asia 1.84 2.55 3.25 3Asia 1.58 2.67 3.76 8Europe 2.70 3.51 4.32 3Europe 1.70 2.30 2.91 8North America 2.96 3.29 3.62 3North America 1.82 2.60 3.38 8Oceania 2.67 4.27 5.87 3Oceania 1.69 2.03 2.36 8South America 1.83 2.20 2.57 3Russia 1.02 1.30 1.58 9Africa 1.21 1.48 1.75 3South America 1.24 2.11 2.99 9Asia 1.56 1.92 2.29 4Africa 0.81 1.74 2.66 9Middle America 1.67 1.68 1.69 4Asia 1.13 1.58 2.03 9North America 1.70 1.85 1.99 4North America 4.46 5.05 5.63 9South America 1.19 1.52 1.86 Source: This study. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X D — U NCE RTAINT Y ANALYSIS 81 Appendix D: UNCERTAINTY ANALYSIS Uncertainty in the RothC soil carbon modeling was estimat- 2. The minimum Pmin and maximum Pmax values of the ed following the adoption of Sustainable Agricultural Land con�dence interval for the mean of the parameters Management methodology. A precision of 15 percent at the X p were estimated as 95-percent con�dence level was chosen as the criterion for reliability. Pmin = X p − 1.96 × SE p The analysis calculates the soil model response using the Pmax = X p + 1.96 × SE p model input parameters with the upper and lower con�- dence levels. The range of model responses demonstrates where the sensitivity of the soil modeling. The input parameters for which the uncertainty was estimated were minimum and Pmin is the minimum value of the parameter at the maximum monthly temperatures, monthly precipitation, and 95 percent con�dence interval, clay content in percent of the soil. Uncertainty analysis took Pmax is the maximum value of the parameter at the place in two steps: 95 percent con�dence interval, 1. For each mapping unit, the mean values and the SEp is the standard error in the mean of parameter standard deviation for the three parameters were p in year t, and calculated. Thereafter, the standard error in the mean 1.96 is the value of the cumulative normal distribu- was estimated using tion at the 95-percent con�dence interval. ∂p SE p = , np Twenty repetitions were selected randomly among the different scenarios and years for which SOC change where values were modeled (table D.1). For each of these 20 data points, two separate models were done with the SEp is the standard error in the mean of parameter minimum and maximum values as model inputs. Carbon se- p in year t, questration rates using the minimum and maximum values ∂p is the standard deviation of the parameter of the input parameters are given by PRSmin, t and PRSmax, t, p in year t, and respectively. np is the number of samples used to calculate the mean and standard deviation of parameter p. The uncertainty (UNC) in the output model was �nally calcu- lated as In this case, np represents the total number of data points of a parameter used in this analysis for each | PRSmax, t − PRSmin, t | mapping. UNCt = . 2 × PRSt EC O N O M I C A N D S E CT OR WORK 82 A PPEND IX D — UNCERTA INTY A NA LY S IS TABLE D.1: Uncertainty Analyses Using Random Samples from the Mitigation Scenarios SCENARIO RESIDUE MITIGATION MAPPING UNIT YIELD BIN CROP FRACTION SCENARIO YEAR UNC t AEZ01-50 50% Maize 1s 50% Residue 2012 2.3% AEZ02-50 25% Millet 50% Residue 2014 2.1% AEZ03-50 50% Sorghum 1s 50% Residue 2019 1.4% AEZ04-50 50% Maize 1s 75% Residue + compost 2019 2.0% AEZ04-50 75% Rice 2s 25% Residue + compost 2029 1.5% AEZ06-50 50% Maize 2s 50% Residue 2025 1.5% AEZ06-50 50% Sorghum 1s 75% Residue 2025 0.9% AEZ06-50 50% Sorghum 2s 15% Baseline 2021 1.8% AEZ06-50 75% Sorghum 2s 75% Residue + compost 2030 0.9% AEZ08-50 N 75% Maize 15% Baseline 2026 6.7% AEZ09-50 N 75% Maize 25% Residue 2016 9.3% AEZ09-50 S 75% Maize 25% Residue 2032 2.2% AEZ10-50 S 25% Maize 75% Residue 2020 3.9% AEZ10-50 S 50% Maize 75% Residue 2027 6.9% AEZ10-50 S 75% Maize 25% Residue + manure 2027 3.9% AEZ11-50 N 25% Barley 75% Residue 2034 10.7% AEZ11-50 N 50% Wheat 15% Baseline 2035 7.5% AEZ11-50 N 50% Wheat 25% Residue 2017 10.8% AEZ12-50 75% Maize 75% Residue + compost 2029 25.9% AEZ07-50 S Mucuna Cover crop 2024 0.6% Average: 5.1% Source: This study. Note: UNC = Uncertainty , AEZ = Agroecological Zone. The uncertainty ranges from below 1 percent to 26 percent with an average value of 5.1 percent. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X E — A S S UMP T IONS F OR DE RIVING T HE A PPLICA B LE MITIGATION A REA FOR TH E LA ND MA NA GEMENT 83 Appendix E: ASSUMPTIONS FOR DERIVING THE APPLICABLE MITIGATION AREA FOR THE LAND MANAGEMENT PRACTICES E.1 AFRICA Asia Cropland Cropland The difference between current (table E.1) and projected The difference between current (table E.1) and the projected cropland area (table E.2) under each of the four IPCC sce- cropland area (table E.2) under each of the four IPCC sce- narios was allocated to land-use and agroforestry-related narios was allocated to land-use and agroforestry-related land management practices in equal proportion. Tree-crop land management practices in equal proportion. Organic farming was projected to increase by the same proportion soil restoration was applied to the estimated degraded peat the entire cropland area for the continent increased for A1B land area of 13 million ha for each scenario. The abatement and A2 (the more economic focus scenarios), but by just 15 rate for organic soil restoration was taken from Smith et al. percent under the two other scenarios that are more environ- (2008). The estimated irrigable area in Asia is 270 million mentally focused. ha. Two-thirds of this was applied to improved irrigation and rainwater harvesting in equal proportions. Land devoted to No-tillage was assumed to cover 2 percent of cropland area intensive vegetables was assumed to increase by 15 percent in the B1 and A1B scenarios, but only 1 percent in the re- under the B1 and B2 scenarios, while it was assumed to in- maining two scenarios. About 3.6 million ha were estimated crease by the same proportion for total cropland area under as having erosion hazard in Africa. Terracing and sloping bar- the remaining two scenarios. riers were applied to 75 percent of this land area in equal proportion. Sustainable biochar application was assumed for Current land area under biofertilizer is 29 million ha. This was 15 percent of current tree crop area for B1 and B2 scenarios, assumed to increase by 5 percent under scenarios A2 and B2 and 15 percent of projected tree crop area for the remaining and by 6 percent under the remaining two scenarios. Biochar two scenarios. was applied to only 15 percent of the applicable area for each agroforestry-related practice for each scenario. No-tillage is Rainwater harvesting was assumed applicable to 19.5 million currently practiced on 3 percent of land in Asia, and it was ha under each scenario; about 7 percent increase in potential assumed to increase to 7 percent of current land area by irrigable area for 1990. The remaining cropland area under 2030 for all scenarios. The remaining cropland area under each scenario was distributed evenly among inorganic fer- each scenario was distributed evenly among inorganic fertil- tilizer, manure, cover crops, rotation diversi�cation, rotation izer, manure, cover crops, rotation intensi�cation, and crop intensi�cation, and crop residue application. residue application. TABLE E.1: Estimated Cropland Area in the 2000s Grassland MILLION ha The respective abatement rates for fertilizer, manure, im- proved pastures, pasture establishment on degraded land, Africa 165.8 and rainwater harvesting were each applied to one-sixth of Asia 497.4 projected grassland area for each of the four IPCC scenarios Latin America 110.3 for the continent by 2030 (table E.2). Source: Based on Monfreda et al. (2008). EC O N O M I C A N D S E CT OR WORK 84 A P P E NDIX E — ASSUMP T IONS F OR DER IVING TH E A PPLIC A B LE MITIGATION A REA FOR TH E LA ND MA NA GEM ENT TABLE E.2: Estimated Cropland and Grassland Area by 2030 (Million ha) B1 A1B B2 A2 CROPLAND GRASSLAND CROPLAND GRASSLAND CROPLAND GRASSLAND CROPLAND GRASSLAND Africa 279.7 813.7 271.2 826.4 321.3 863.5 325.4 891.9 Asia 799.3 656.9 847.3 678.7 943.4 701.7 871.6 714.4 Latin America 351.0 282.1 361.8 295.6 358.2 384.0 414.3 426.0 Source: Based on Smith et al. (2008). Grassland Dossa, E. L., Fernandes, E. C. M., Reid W. S., Ezui K. 2008. “Above- and Below-Ground Biomass. Nutrient and Carbon Stock The respective abatement rates for fertilizer, manure, pas- Contrasting an Open Grown and a Shade Coffee Plantation.� ture establishment on degraded land, grazing management, Agroforestry Systems 72: 103–115. rainwater harvesting, improved irrigation, cropland–to-pas- Eilittä, M., Carsky, R. J., Mureithi, J., Szabo, N., Bressani, R., ture, and annual-to-perennial grass were each applied to one- Myhrman, R., Sandoval, C., Muinga, R., Carew, L. B., Capo-chichi, eighth of projected grassland area for each of the four IPCC L. J. A., and Teixeira, A. “Future Agenda for Mucuna Research scenarios on the continent by 2030 (table E.2). and Promotion.� Tropical and Subtropical Agroecosystems 1 (2-3): 329–343. Latin America Gama-Rodrigues, Antonio C. 2011. “Soil Organic Matter, Nutrient Cycling and Biological Nitrogen-Fixation in Agroforestry Systems. Cropland Agroforestry Systems 81 (3): 191–193. The difference between current (table E.1) and projected Intergovernmental Panel on Climate Change. 2006. 2006 IPCC cropland area (table E.2) under each of the four IPCC scenarios Guidelines for National Greenhouse Gas Inventories, ed. was allocated to land-use and agroforestry-related land man- H. S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe. agement practices in equal proportion. No-tillage is currently Institute for Global Environmental Strategies, Japan. practiced on 50 Mha. This was assumed to increase to 66 Mha Kaonga, M. L., and Coleman, K. 2008. “Modelling Soil Organic under each scenario by 2030. Sustainable biochar production Carbon Turnover in Improved Fallows in Eastern Zambia Using was assumed applicable to 30 percent of the agroforestry area the RothC-26.3 Model.� Forest Ecology and Management 256 (5): 1160–1166. in each scenario. The potential irrigable area is 77.8 Mha. Two- thirds of this was applied to improved irrigation and rainwater Kaizzi, C. K., Ssali, H., Vlek, P.L. G. 2006. “Differential Use and harvesting in equal proportions. The remaining cropland area Bene�ts of Velvet Bean (Mucuna Pruriens Var. Utilis) and N Fertilizers in Maize Production in Contrasting Agro-Ecological under each scenario was distributed evenly among manure, Zones of E. Uganda.� Agricultural Systems 88: 44–60. residue, cover crops, rotation intensi�cation, and diversi�cation. Lemma, B., Kleja, D. B., Nilsson, I., Olsson, M. 2006. “Soil Carbon Grassland Sequestration Under Different Exotic Tree Species in the Southwestern Highlands of Ethiopia.� Geoderma 136: 886–898. The respective abatement rates for manure, pasture estab- lishment on degraded land, grazing management, rainwater Monfreda, C., Ramankutty, N., and Foley, J. A. 2008. “Farming the Planet: 2. Geographic Distribution of Crop Areas, Yields, harvesting, improved irrigation, cropland–to-pasture, and an- Physiological Types, and Net Primary Production in the Year nual-to-perennial grass were each applied to one-seventh of 2000.� Global Biogeochemical Cycles 22: GB1022. projected grassland area for each of the four IPCC scenarios Oelbermann, M., Voroney, R. P., Kass, D. C. L., et al. 2005a. “Above- continents by 2030 (table E.2). and Below-Ground Carbon Inputs in 19-, 10-, and 4-year-old Costa Rican Alley Cropping Systems.� Agriculture, Ecosystems, REFERENCES and Environment 105 (1-2): 163–172. Anthofer J., and Kroschel J. 2005. “Above-Ground Biomass, Oelbermann, M., Voroney, R. P., Gordon, A. M. 2005b. “Carbon Nutrients, and Persistence of an Early and Late Maturing Mucuna Sequestration in Tropical and Temperate Agroforestry Systems: Variety in the Forest-Savannah Transitional Tone of Ghana.� A Review With Examples From Costa Rica and Southern Canada.� Agriculture, Ecosystems, and Environment 110: 59–77. Agriculture, Ecosystems, and Environment 104 (3): 359–377. Barahona, M. F., Janssen, M. 2004. Green Manure/Cover Crop Smith, P., Martino, D., Cai, Z., et al. 2008. “Greenhouse Gas Systems of Smallholder Farmers, ed. M. Eilittä et al. Kluwer Mitigation in Agriculture. Philosophical Transactions of the Royal Academic Publishers, Dordrecht, The Netherlands. Society B 363: 789–813. CARBON SEQUESTRATION IN AGRICULTURAL SOILS AP P E N D I X E — A S S UMP T IONS F OR DE RIVING T HE A PPLICA B LE MITIGATION A REA FOR TH E LA ND MA NA GEMENT 85 Szott, L. T., E. C. M. Fernandez & P. A. Sanchez. 1991. Soil plant interaction van Noordwijk, M., Rahayu, S., Hairiah, K., Wulan, Y. C., Farida, A., in agroforestry system. Forest Ecology Management 45: 127–152. and Verbist, B. 2002. “Carbon Stock Assessment for a Forest- to-Coffee Conversion Landscape in Sumber-Jaya (Lampunh, Szott, L. T., Palm, C. A., and Sanchez, P. A. 1991. “Agroforestry in Arid Indonesia).� Science in China (Series C) 45: 76–86. Soils of the Humid Tropics.� Advances in Agronomy 45: 275–301. Yaqub, M., Mahmood, T., Akhtar, M., Iqbal, M. M., and Ali, S. 2010. Szott, L. T., Palm, C. A., and Davey, C. B. 1994. “Biomass and Litter “Induction of Mungbean [Vigna Radiata (L.) Wilczek] as a Grain Accumulation Under Managed and Natural Tropical Fallows.� Legume in the Annual Rice-Wheat Double Cropping System.� Forest Ecology and Management 67: 177–190. Pakistan Journal of Botany 42 (5): 3125–3135. EC O N O M I C A N D S E CT OR WORK A G R I C U L T U R E A N D R U R A L D E V E L O P M E N T Agriculture and Rural Development (ARD) 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-477-1000 Internet: www.worldbank.org/ard