WORLD SOIL RESOURCES AND FOOD SECURITY
Advances in Soil Science
WORLD SOIL RESOURCES AND FOOD SECURITY
Edited by
Ra...
61 downloads
742 Views
28MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
WORLD SOIL RESOURCES AND FOOD SECURITY
Advances in Soil Science
WORLD SOIL RESOURCES AND FOOD SECURITY
Edited by
Rattan Lal B. A. Stewart
CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2012 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Version Date: 20110623 International Standard Book Number-13: 978-1-4398-4451-9 (eBook - PDF) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright. com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com
Contents Preface......................................................................................................................vii Editors........................................................................................................................ix Contributors...............................................................................................................xi Chapter 1 Sustainable Management of Soil Resources and Food Security...........1 Rattan Lal and B. A. Stewart Chapter 2 Global Food Situation and Unresolved and Emerging Issues............. 11 Shahla Shapouri, Stacey Rosen, and Summer Allen Chapter 3 World Soil Resources: Opportunities and Challenges........................ 29 H. Eswaran, P. F. Reich, and E. Padmanabhan Chapter 4 Soil Resources and Human Adaptation in Forest and Agricultural Ecosystems in Humid Asia............................................. 53 S. Funakawa, T. Watanabe, A. Kadono, A. Nakao, K. Fujii, and T. Kosaki Chapter 5 Pedogenetic Acidification in Upland Soils under Different Bioclimatic Conditions in Humid Asia............................................. 169 S. Funakawa, T. Watanabe, A. Nakao, K. Fujii, and T. Kosaki Chapter 6 Soil Resources Affecting Food Security and Safety in South Asia.................................................................................................... 271 Tapan J. Purakayastha, Bal Ram Singh, R. P. Narwal, and Promod K. Chhonkar Chapter 7 Formation and Management of Cracking Clay Soils (Vertisols) to Enhance Crop Productivity: Indian Experience........................... 317 D. K. Pal, T. Bhattacharyya, and Suhas P. Wani
v
vi
Contents
Chapter 8 Role of Nuclear and Isotopic Techniques in Sustainable Land Management: Achieving Food Security and Mitigating Impacts of Climate Change............................................................................. 345 Long Nguyen, Felipe Zapata, Rattan Lal, and Gerd Dercon Chapter 9 New Paradigm to Unlock the Potential of Rainfed Agriculture in the Semiarid Tropics..................................................................... 419 Suhas P. Wani, Johan Rockstrom, B. Venkateswarlu, and A. K. Singh Chapter 10 Land Degradation.............................................................................. 471 Freddy O. Nachtergaele, Monica Petri, and Riccardo Biancalani Chapter 11 Where Do We Stand 20 Years after the Assessment of Soil Nutrient Balances in Sub-Saharan Africa?....................................... 499 E. M. A. Smaling, J. P. Lesschen, C. L. van Beek, A. de Jager, J. J. Stoorvogel, N. H. Batjes, and L. O. Fresco Chapter 12 Research Needs for Credible Data on Soil Resources and Degradation....................................................................................... 539 Rattan Lal
Preface World soils, the basis of all terrestrial life, are finite in extent, variable over time and space, and prone to alterations by natural and anthropogenic perturbations. Productive soils of high quality are essential to human well-being, economic and sustainable development, political stability, and ethnic and cultural harmony. Ancient civilizations and cultures (i.e., Mayan, Aztec, Mesopotamian, Indus, Yangtze) arose and thrived on good soils, and survived only as long as soils had the capacity to support them. These and other once thriving civilizations collapsed with declines in the quality of their soils. Even during the twenty-first century, a good soil is the engine of economic development and essential to present and future food security. Yet, the quality of soil resources is threatened by human-induced and natural perturbations. Soil quality is degraded by land misuse and soil mismanagement. Soil resources available for agricultural use are also being diminished for conversion to other uses (e.g., urbanization, industrial, recreational). The Food and Agriculture Organization (FAO) of the United Nations defines soil degradation as a decline in “the current or potential capability of soils to produce goods and services.” Soil use and management produce several goods including food, feed, fiber, fuel, and industrial raw materials. Under both natural and managed ecosystems, soils also generate numerous ecosystem services including moderation of climate, purification and filtration of water, enhancement of biodiversity, an archive of planetary and earth history, a reservoir of germplasm, etc. Whereas soil degradation may lead to reductions in its ability to produce goods and generate services, its capacity to do so also depends on numerous endogenous and exogenous factors. Endogenous factors are those related to soil-forming factors including climate, vegetation, parent material, terrain, and time. The exogenous factors include anthropogenic perturbations including deforestation, biomass burning, drainage, irrigation, and use of inputs (fertilizer, amendments, tillage methods, residues management, vehicular traffic, cropping and farming systems, etc). Exogenous factors may also be natural perturbations such as volcanic eruption, seismic activity, tsunami, etc. The impacts of exogenous factors on the extent and severity are exacerbated by several endogenous factors (i.e., climate, terrain). There exists a vast amount of literature on soil degradation. However, the data reported in the literature are often confusing, contradictory, erratic, and misleading. The problem is confounded by several factors such as: (i) using interchangeably different terminology such as land vs. soil, degradation vs. desertification; (ii) using different methodologies to assess degradation; (iii) using different categories to assess the severity of a degradation process such as light vs. slight, severe vs. strong; (iv) using proxy methods to indirectly assess the extent and severity of degradation; (v) not validating the estimates by ground truthing; (vi) not relating goods produced (e.g., crop yields) to the severity of degradation under different land uses and management; and (vii) not evaluating the masking effects of improved management practices through partial or complete elimination of some soil-related constraints to vii
viii
Preface
agronomic production. Thus, there is a strong need for creating credible data based on the use of standardized methodology and terminology, ground truthing and validation, assessing impacts on agronomic production under a range of management scenarios, and evaluating social and economic impacts on the community. This volume is specifically dedicated to soil resources of the world in terms of their availability and status in the context of the growing demands of increasing world population and rising expectations of living standards. It comprises invited chapters contributed by renowned scientists in their specific fields of expertise including global food situations by Shahla Shapouri, world soil resources by Hari Eswaran and colleagues, soil resources for humid Asia and their acidification by Funakawa and colleagues, soil resources of South Asia by Tapan Purakayastha and others, properties and management of vertisols by D. Pal and colleagues, use of radio isotopic techniques in soil management by Long Nguyen and his colleagues at IAEA, the potential of rainfed agriculture in the semiarid tropics by S. Wani and group, the status of land degradation by F. Nachtergaele and colleagues, and the nutrient balance in sub-Saharan Africa by E. Smaling and others. The volume is a useful reference source for those interested in the state of the soils of the world in relation to food security and environmental quality. Rattan Lal B. A. Stewart
Editors Rattan Lal is a distinguished professor of soil physics in the School of Environment and Natural Resources and director of the Carbon Management and Sequestration Center, Food, Agricultural, and Environmental Sciences/Ohio Agriculture Research and Development Center, at The Ohio State University. Before joining Ohio State in 1987, he was a soil physicist for 18 years at the International Institute of Tropical Agriculture, Ibadan, Nigeria. In Africa, Professor Lal conducted long-term experiments on land use, watershed management, soil erosion processes as influenced by rainfall characteristics, soil properties, methods of deforestation, soil-tillage and crop-residue management, cropping systems including cover crops and agroforestry, and mixed/relay cropping methods. He also assessed the impact of soil erosion on crop yields and related erosion-induced changes in soil properties to crop growth and yields. Since joining Ohio State University in 1987, he has continued research on erosion-induced changes in soil quality and developed a new project on soils and climate change. He has demonstrated that accelerated soil erosion is a major factor affecting emission of carbon from the soil to the atmosphere. Soil-erosion control and adoption of conservation-effective measures can lead to carbon sequestration and mitigation of the greenhouse effect. Other research interests include soil compaction, conservation tillage, mine soil reclamation, water table management, and sustainable use of soil and water resources of the tropics for enhancing food security. Professor Lal is a fellow of the Soil Science Society of America, American Society of Agronomy, Third World Academy of Sciences, American Association for the Advancement of Sciences, Soil and Waste Conservation Society, and Indian Academy of Agricultural Sciences. He is a recipient of the International Soil Science Award of the Soil Science Society of America, the Hugh Hammond Bennett Award of the Soil and Water Conservation Society, the 2005 Borlaug Award, and the 2009 Swaminathan Award. He also received an honorary degree of Doctor of Science from Punjab Agricultural University, India, from the Norwegian University of Life Sciences, Aas, Norway, and the Alecu Russo Balti State University in Moldova. He is a past president of the World Association of the Soil and Water Conservation, the International Soil Tillage Research Organization, and the Soil Science Society of America. He is a member of the United States National Committee on Soil Science of the National Academy of Sciences (1998 to 2002, 2007 to present). He has served on the Panel of Sustainable Agriculture and the Environment in the Humid Tropics of the National Academy of Sciences. He has authored and coauthored around 1500 research papers. He has also written 15 books and edited or coedited another 48. B. A. Stewart is a distinguished professor of soil science at the West Texas A&M University, Canyon, Texas. He is also the director of the Dryland Agriculture Institute, and a former director of the United States Department of Agriculture Conservation and Production Laboratory at Bushland, Texas; past president of the Soil Science Society of America; and member of the 1990–1993 Committee on Long-Range Soil ix
x
Editors
and Water Policy, National Research Council, National Academy of Sciences. He is a fellow of the Soil Science Society of America, the American Society of Agronomy, the Soil and Water Conservation Society, a recipient of the United States Department of Agriculture Superior Service Award, a recipient of the Hugh Hammond Bennett Award of the Soil and Water Conservation Society, and an honorary member of the International Union of Soil Sciences in 2008. Dr. Stewart is very supportive of education and research on dryland agriculture. The B.A. and Jane Anne Stewart Dryland Agriculture Scholarship Fund was established at West Texas A&M University in 1994 to provide scholarships for undergraduate and graduate students with a demonstrated interest in dryland agriculture.
Contributors Summer Allen Economic Research Service United States Department of Agriculture Washington, D.C. N. H. Batjes ISRIC—World Soil Information Wageningen, the Netherlands T. Bhattacharyya Division of Soil Resource Studies National Bureau of Soil Survey and Land Use Planning Nagpur, India Riccardo Biancalani Land and Water Division Food and Agriculture Organization of the United Nations Rome, Italy Promod K. Chhonkar Division of Soil Science and Agricultural Chemistry Indian Agricultural Research Institute New Delhi, India A. de Jager North and West Africa Division International Fertilizer Development Center Accra, Ghana Gerd Dercon International Atomic Energy Agency Vienna, Austria
H. Eswaran Natural Resources Conservation Service United States Department of Agriculture Washington, D.C. L. O. Fresco University of Amsterdam Amsterdam, the Netherlands K. Fujii Graduate School of Agriculture Kyoto University Kyoto, Japan S. Funakawa Graduate School of Agriculture Kyoto University Kyoto, Japan A. Kadono Graduate School of Urban Environmental Sciences Tokyo Metropolitan University Tokyo, Japan T. Kosaki Graduate School of Urban Environmental Sciences Tokyo Metropolitan University Tokyo, Japan Rattan Lal Carbon Management and Sequestration Center The Ohio State University Columbus, Ohio
xi
xii
Contributors
J. P. Lesschen Alterra Wageningen University and Research Centre Wageningen, the Netherlands
Tapan J. Purakayastha Division of Soil Science and Agricultural Chemistry Indian Agricultural Research Institute New Delhi, India
Freddy O. Nachtergaele Land and Water Division Food and Agriculture Organization of the United Nations Rome, Italy
P. F. Reich Natural Resources Conservation Service United States Department of Agriculture Washington, D.C.
A. Nakao Department of Radioecology Institute for Environmental Sciences Aomori-Ken, Japan R. P. Narwal Chaudhary Charan Singh Haryana Agricultural University Haryana, India Long Nguyen International Atomic Energy Agency Vienna, Austria E. Padmanabhan Petroleum and Geoscience Department University of Technology Petronas Perak, Malaysia D. K. Pal Division of Soil Resource Studies National Bureau of Soil Survey and Land Use Planning Nagpur, India Monica Petri Land and Water Division Food and Agriculture Organization of the United Nations Rome, Italy
Johan Rockstrom Stockholm Environment Institute Stockholm, Sweden Stacey Rosen Economic Research Service United States Department of Agriculture Washington, D.C. Shahla Shapouri Economic Research Service United States Department of Agriculture Washington, D.C. A. K. Singh Indian Council of Agricultural Research New Delhi, India Bal Ram Singh Department of Plant and Environmental Sciences Norwegian University of Life Sciences Ås, Norway
xiii
Contributors
E. M. A. Smaling Royal Tropical Institute (KIT) Amsterdam, the Netherlands and University of Twente Faculty of Geo-Information Science and Earth Observation (ITC) Enschede, the Netherlands B. A. Stewart Dryland Agriculture Institute West Texas A&M University Canyon, Texas J. J. Stoorvogel Land Dynamics Group Wageningen University and Research Centre Wageningen, the Netherlands C. L. van Beek Alterra Wageningen University and Research Centre Wageningen, the Netherlands
B. Venkateswarlu Central Research Institute for Dryland Agriculture Andhra Pradesh, India Suhas P. Wani International Crops Research Institute for the Semi-Arid Tropics Andhra Pradesh, India T. Watanabe Graduate School of Agriculture Kyoto University Kyoto, Japan Felipe Zapata International Atomic Energy Agency Vienna, Austria
Management 1 Sustainable of Soil Resources and Food Security Rattan Lal and B. A. Stewart CONTENTS 1.1 Introduction....................................................................................................... 1 1.2 Soil and Climatic Effects on Agricultural Production......................................2 1.3 Enhancing Soil Resilience through Sustainable Management.......................... 3 1.4 Per Capita Arable Land Area and Energy Use..................................................5 1.5 Diet Preferences.................................................................................................6 1.6 Carbon Sequestration in Terrestrial Ecosystems............................................... 7 1.7 Soil Degradation................................................................................................8 1.8 Conclusions........................................................................................................8 Acronyms.................................................................................................................... 9 References................................................................................................................... 9
1.1 INTRODUCTION The world population is projected to increase from about 7 billion in 2011 to 9.2 billion in 2050. The current rate of increase is about 6 million/month, with almost all growth occurring in developing countries where natural resources are already under great stress. The Green Revolution technology led to the doubling of food production between 1950 and 2010, with only a 10% increase in the area under production [FAO 2010]. However, meeting the food demand of the growing population, rising standards of living, and changes in diet preferences will necessitate an additional 70% increase in production between 2010 and 2050 [Burney et al. 2010]. Yet yields in the Indo-Gangetic plains and other regions have stagnated or declined over the past two decades and may be jeopardized further by climate change and increases in the frequency and intensity of extreme events [Semonov 2009], especially in sub-Saharan Africa [IIED 2010]. Grain yields of wheat [Semenov 2009] and rice [Wassmann et al. 2009] are sensitive to high temperatures. The problem of food insecurity is also exacerbated by increases in the severity and extent of soil degradation. This is especially true because of declines in the soil structure and hydrological properties in conjunction with reductions in the quantity and quality of soil organic carbon (SOC) content caused by a widespread use of extractive farming practices (i.e., indiscriminate residue removal, excessive grazing, the use of animal dung as household fuel 1
2
World Soil Resources and Food Security
rather than as manure, and a negative nutrient budget). Over and above the biophysical constraints being exacerbated by a changing climate and an increase in the frequency of extreme events, there are also issues related to the human dimensions. To the resource-poor and small-size land holders of the tropics and subtropics, neither the essential inputs are available (e.g., improved seeds and fertilizers and new equipment such as no-till seed drills and soil testing facilities), nor affordable. One, these inputs are prohibitively expensive. Two, farmers are not sure about their effectiveness, especially under conditions of uncertain rains, frequent droughts, and a high incidence of weeds and other pests. This volume is especially focused on the soil resources of the world and the challenge of their management on a sustainable basis. This introductory chapter describes some of the soil-related constraints on enhancing agricultural production on smallholder farms of the tropics and subtropics, and outlines mitigative and adaptive strategies for changing climates and declining qualities of soil and water resources.
1.2 S OIL AND CLIMATIC EFFECTS ON AGRICULTURAL PRODUCTION There is a strong interaction between soil degradation and extreme events (e.g., drought, inundation), with positive feedback from reducing the agronomic or net primary productivity (NPP). Declines in productivity are accentuated by reductions in the use-efficiency of nutrients and water, both inherent and applied. These interactive effects lead to declines in both plant- and animal-based outputs (Table 1.1). Specific reductions in plant-based outputs are those related to declines in yields of crops (cereals, grain legumes, roots and tubers, oil seeds, fruits and vegetables), fuel [crop residues, wood, charcoal, biomass from switchgrass (Pennistum vergatum), etc.], feed (forages, crop residues, fodder trees), and fiber [cotton (Gossipium hirsutum)]. Because of close interactions between plants and animals, a decline in soil quality also affects animal health and productivity. Healthy soils produce healthy plants, raise healthy animals, improve human health, and vice versa. Consequently, TABLE 1.1 Agricultural Products Affected by Soil Resource Degradation, Climate Change, and Related Extreme Events Production 1.
Food
2.
Fuel
3.
Feed
4. 5.
Fiber Raw materials
Plant-Based Cereals, grain legumes, roots and tubers, oil seeds, vegetables, fruits Wood, charcoal, biomass, residues, sugarcane, oils Forages, fodder trees and shrubs, rangelands, seed cakes Cotton, hemp, banana, agave Bioeconomy, timber, resins, medicines, pharmaceuticals, oils
Animal-Based Milk, poultry, fish, meat Dung, animal fat, residues Fishmeal, meat and poultry by-products Soil, wool, animal hairs Fish oil, animal products
Sustainable Management of Soil Resources and Food Security
3
a decline in soil quality coupled with an increase in the frequency of extreme events reduce the output of animal-based products such as milk, meat, poultry, wool, etc. (Table 1.1).
1.3 E NHANCING SOIL RESILIENCE THROUGH SUSTAINABLE MANAGEMENT No technological panacea exists for the global challenges of sustainable use of soil and water resources [The Royal Society 2009]. This is especially true because the constraints on agronomic production differ among climates, soils, and geographical regions. Nonetheless, the strategy is to reverse the degradation trend, restore degraded soils, and enhance resilience to any adverse effects of climate change. Conversion to restorative land use and adoption of recommended management practices (RMPs) would improve soil quality and adaptation to climate change (e.g., onset of rains, changes in soil temperature, moisture regime, and growing season duration). Adoption of RMPs can mitigate climate change [UNFCCC 2008; Lobell and Burney 2010], and is also essential to advancing global food security [Gregory et al. 2005; Lobell et al. 2008]. Furthermore, agricultural RMPs, in addition to enhancing natural processes, are more cost-effective than engineering techniques for carbon capture and storage [McKinsey & Company 2009]. In this context, agricultural intensification in developing countries is an important solution to climate change [World Bank 2009]. Land uses and RMPs that enhance adaptation to climate change also have mitigative effects through (i) sequestering carbon in soils and biota, thereby enhancing the ecosystem carbon pools, and (ii) reducing emissions through conversion of plow-till systems to no-till (NT) or conservation agriculture (CA) systems and enhanced use-efficiency of energy-based inputs (e.g., fertilizers, irrigation). Some RMPs for improving soil quality and enhancing soil resilience are outlined in Figure 1.1. Important among these are (i) integrated nutrient management (INM) for balancing budgets for macronutrients (N, P, K, Ca, Mg) and micronutrients (Zn, Cu, Fe, Se), (ii) improving the chemical properties of soil, such as soil reaction through liming, while also enhancing cation exchange capacity (CEC), (iii) enhancing aggregation and improving soil tilth, (iv) increasing the soil organic carbon (SOC) pool and its quality, along with improvements in the activity and species diversity of soil fauna (microbial biomass carbon, earthworm activity), and (v) moderating soil temperature so that it is within the optimal range (25°C–30°C) in the root zone. Soil-specific RMPs to achieve these goals include CA, cover cropping, manuring, biochar application, and those RMPs that enhance earthworm and other biota activity (Figure 1.1). Similar to soil management, there are also RMPs for improving the use-efficiency of water resources (Figure 1.1). Important among these are (i) conserving water in the root zone by increasing infiltration capacity, decreasing surface runoff, and reducing evaporation, (ii) harvesting and recycling water through supplemental irrigation by using microirrigation, drip subsurface irrigation (DSI), fertigation, and even condensation irrigation (CI), (iii) using genetically modified crops with deep root systems and built-in mechanisms for drought avoidance and tolerance, and (iv) using an aerobic rice culture that saves water and improves water use-efficiency
4
World Soil Resources and Food Security
Soil management Nutrients: macro, micro (INM, slow release, nano-enhanced fertilizers) Reaction: liming, leaching Structure: conservation agriculture Temperature: optimize Biota: macro, micro Precision farming: soil-specific management Amendments: manure, biochar
Water management
Soil water retention Infiltration capacity and storage Runoff control Evaporation reduction Supplemental irrigation (micro, DSI, CI) Water harvesting and recycling Plastic mulch Aerobic rice
Farming systems management
1. Crops: rotations, complex systems, perennial culture, GM crops 2. Trees: species management, GM varieties, super-CO2 adsorbing trees, albedo management 3. Animals: species management, methane management
Ecosystem management Silvo-pastoral systems Agro-silvo-pastoral systems Eco-efficiency approach
FIGURE 1.1 Agricultural intensification to enhance soil and ecosystem resilience for mitigation of and adaptation to climate change and related extreme events. CDSI, drip subirrigation; CI, condensation irrigation; INM, integrated nutrient management.
[Bouman et al. 2007]. There exists a strong interaction between soil temperature and the use-efficiency of soil water. Practiced in conjunction with the retention of crop residue mulch and INM needed for complex rotations, CA can moderate surface soil temperature while reducing losses (by runoff and evaporation) and improving use-efficiency of stored and applied water (Figure 1.1). These strategies and the relevant policy interventions have been described elsewhere [The Royal Society 2009; FAO 2010]. Adoption of improved farming systems is integral to the sustainable management of soil and water resources, and to the mitigation of, and adaptation to, climate change. The goal is to integrate the cultivation of crops with the growth of suitable trees and the raising of livestock. Important options include the perennial culture of wheat (Triticum aestivum) and barley (Hordeum vulgare) [Glover et al. 2010], complex rotations (cereals–legumes–forages/hay), agroforestry, silvopastoral systems, and agrosilvopastoral systems (Figure 1.1). Agroforestry systems enhance, stabilize, and sustain production under small-holder agriculture [Zomer et al. 2009; Sileshi et al. 2008]. These farming systems increase land use intensity and have high NPP per unit area, unit time, and unit consumption of inherent and applied resources (e.g., nutrients, water, energy). While stabilizing productivity and restoring soil quality, complex systems also minimize the incidence of pests. Amudavi et al. [2007] show the effectiveness of a push–pull system of integrated pest management (IPM) for
Sustainable Management of Soil Resources and Food Security
5
maize (Zea mays L) cultivation in East Africa by intercropping maize with desmodium (Desmodium uncinatum), which suppresses striga (Striga hermonthica) and growing Napier grass (Pennisetum purpureum) that is more attractive to moths of the maize stalk borer (Busseola fusca) [Cook et al. 2007; Hassanali et al. 2008]. Complex systems involving intercropping with perennial shrubs can also improve drought tolerance through hydraulic lift [Caldwell and Richards 1989]. Complex farming systems may also enhance the use-efficiency of fertilizers and reduce the rate of application. Future demand for nitrogenous fertilizers may require 2% of the total global energy use by 2050 as the Haber–Bosch process is highly energy-intensive [Glendining et al. 2009]. Complex systems based on reduced cultivation may also promote arbiscular mychorrhizal (AM) associations that enhance plant nutrient uptake and agronomic sustainability [Leigh et al. 2009].
1.4 PER CAPITA ARABLE LAND AREA AND ENERGY USE The increase in population, from 7 billion in 2011 to 9.2 billion in 2050, will occur almost entirely in developing countries. The per capita arable land area—which in most countries of South and East Asia is <0.15 ha (Table 1.2), and as low as 0.04 ha in densely populated countries such as Bangladesh—is declining as the population is increasing. Thus, all basic necessities (food, feed, fiber, fuel, and the raw materials listed in Table 1.1) have to be met from this ever dwindling and fragile resource. Similar to arable land area, fresh renewable water resources are also scarce and in heavy demand for competing uses such as industry and urbanization [Molden 2007]. With arable land and fresh water supplies scarce, resource-poor farmers are unable to invest in or use energy-based inputs. For example, oil consumption per unit of
TABLE 1.2 Arable Land and Oil Consumption in Selected Countries Country USA China Germany Brazil Indonesia India Nigeria Philippines Kenya
Per Capita Arable Land (ha)
Oil Consumption per Hectare of Arable Land (L/ha)
0.568 0.136 0.146 0.343 1.147 1.143 1.220 0.119 0.135
2.73 0.311 2.78 0.231 0.407 0.060 0.256 0.348 0.077
Source: Winslow, M.D., and Ortiz, R., The International Dimensions of the American Society of Agronomy: Past and Future, Madison, WI: ASA, 2010. Note: The oil consumption refers to use in mechanized farm operations and does not include fertilizer equivalent.
6
World Soil Resources and Food Security
TABLE 1.3 Trends of Wheat Yield in India and China Wheat Yield (Mg/ha) Year
China
India
Ratio (China to India)
2000 2001 2002 2003 2004 2005 2006 2007
3.74 3.81 3.78 3.93 4.25 4.28 4.55 4.64
2.80 2.70 2.80 2.60 2.70 2.60 2.60 2.70
1.34 1.41 1.35 1.51 1.57 1.65 1.75 1.72
Source: Rajaram, S., The International Dimensions of the American Society of Agronomy: Past and Future, Madison, WI: ASA, 2010.
arable land ranges from >2.7 L/ha in Germany to <0.06 L/ha in India (Table 1.2). The low oil consumption is indicative of the lack of, or low intensity of, mechanized farming operations, such as plowing and, harvesting, and the low use of fertilizers and other chemicals. The data in Table 1.3 show trends in the wheat yield in China and India during the 2000s. China’s yield is increasing with time because of the high fertilizer use, up to 600 kg/ha, while India’s yield is stagnating because of soil degradation and nutrient mining. The data in Tables 1.2 and 1.3 raise an important question. Is it possible to increase and sustain production from arable lands managed by resource-poor farmers by reducing external inputs such as fertilizers, amendments, water, and energy? Is recycling a viable option on a national scale where soils are already depleted of their nutrient reserves? Are there viable alternatives to external inputs of fertilizers, pesticides, irrigation, and mechanized farming operations in view of the present and future population? If the answer is yes, where and under what conditions?
1.5 DIET PREFERENCES In addition to a decline in per capita arable land area, reductions in fresh water resources, and rising demands for energy, changes in the dietary preferences of the populations of emerging economies such as China and India must also be considered when seeking to enhance food production. The data in Table 1.4 show gradual increases in meat consumption in developing countries. The ratio of meat consumption in developed vs. developing countries, as computed in terms of kcal/capita/day, is decreasing with time. The ratio was 3.35 in 1990–1992, 2.48 in 1996–1998 and 2.26 in 2003–2005 (Table 1.4). Meat-based diets have a drastically higher footprint for carbon, water, energy, and land area. A shift in the favor of meat-based diets would jeopardize natural resources that are unequally distributed among geographical
7
Sustainable Management of Soil Resources and Food Security
TABLE 1.4 Trends in Meat Consumption in Developed and Developing Countries between 1990 and 2005 Meat Consumption (kcal/capita/day) Produce Beef
Pork
Chicken
Total
Year
Developed Countries (A)
Developing Countries (B)
Ratio A:B
1990–1992 1996–1998 2003–2005 1990–1992 1996–1998 2003–2005 1990–1992 1996–1998 2003–2005 1990–1992 1996–1998 2003–2005
124 102 99 193 178 189 68 73 94 385 357 382
29 32 34 71 89 105 15 23 30 115 144 169
4.27 3.19 2.91 2.72 2.00 1.8 4.53 3.17 3.13 3.35 2.48 2.26
Source: Hawkes, C., and Ruel, M., The International Dimensions of the American Society of Agronomy: Past and Future, Madison, WI: ASA, 2010.
regions, that are dwindling because of conversion to nonagricultural uses such as urbanization, industrial use, and brink-making, and that are prone to degradation because of land misuse, soil mismanagement, and harsh climates.
1.6 CARBON SEQUESTRATION IN TERRESTRIAL ECOSYSTEMS In addition to advancing food security, agricultural intensification through adoption of RMPs has potential for carbon sequestration in soils and biota [Burney et al. 2010; Bellasen et al. 2010; FAO 2010]. Furthermore, land-based carbon sinks are highly cost-effective [McKinsey & Company 2009]. There are numerous cobenefits, especially in relation to increases in biodiversity, improvements in water quality, and increases in agronomic productivity. Because agricultural intensification also creates a positive ecosystem carbon budget, sustainable development is being linked to climate change mitigation [World Bank 2009]. Payments to land managers and farmers for ecosystem services rendered are being adopted to promote agricultural intensification through adoption of RMPs. Paying for ecosystem services (PESs) is a better strategy than paying subsidies. In this context, subsidies in sub-Saharan Africa (SSA) have proven ineffective, creating dependency, exacerbating corruption, and dampening innovation and initiative. The technical potential of carbon sequestration in terrestrial ecosystems, such as soils, biota, wetlands, and peat soils, is about 4 Gt of carbon/yr [Lal 2010]. With payments of $25/t of carbon, the strategy of PES can generate another income stream for farmers. With agroforestry and other complex farming systems, an
8
World Soil Resources and Food Security
average rate of carbon sequestration of 2 t/ha in soils and biota can generate income of $50/ha. This income is important for resource-poor farmers as an initial step toward poverty alleviation and achievement of the U.N. Millennium Development Goals.
1.7 SOIL DEGRADATION The risks of soil degradation, especially those caused by accelerated erosion and secondary salinization, must be considered when addressing the need to increase food production. While the risks of decline in soil quality due to land misuse and soil mismanagement are serious, the available statistics on the extent and severity are often not credible. Further, the data on the estimates of soil degradation by different processes such as erosion, salinization, and nutrient depletion are not related to agronomic production on national, regional, or global scales. The plot-scale data on the yield decline in relation to specific degradation processes are available for some soil-specific situations. However, these data have not been appropriately scaled up through valid scaling techniques, pedotransfer functions, or geostatistical methods (Bellassen et al. 2010). Therefore, a lack of credible data on the extent of degradation (on national, regional, or global scales) and on the impact of agronomic production under different management scenarios have caused misinterpretations, misunderstandings, confusion, and contradictions. Long-term and soil-based studies are needed under diverse climatic conditions to quantify the impact of degradation processes on agronomic production and the use-efficiency of inputs under a range of farming systems.
1.8 CONCLUSIONS Soil resources of the world are finite, unequally distributed among geographical regions, and prone to degradation by land misuse and soil mismanagement. The per capita agricultural land area is decreasing because of increases in population, conversion to urban and industrial uses, and degradation by a range of human-induced processes. Yet food production must be increased, despite the uncertain climate and risks of drought, weeds, pests, and pathogens. In addition to the required increases in food production, dietary changes in emerging economies, with preference given to more animal-based diets, must also be considered. Thus, the strategy is to: • Improve the data on soil resources including the extent, quality, type, and severity of degradation and the resultant effects on agronomic yield • Indentify recommended management practices that enhance and sustain productivity, strengthen soil resilience against climate change, and improve the use-efficiency of inputs • Implement long-term restoration projects of degraded soils to enhance ecosystem services • Enhance productivity from existing agricultural land, thus minimizing the need to convert new lands from natural ecosystems such as tropical rainforests, savannahs, and steppe to croplands and grazing lands
Sustainable Management of Soil Resources and Food Security
9
ACRONYMS AM CA CEC CI DSI ha INM NPP NT PESs RMP SOC SSA
arbiscular mychorrhizal conservation agriculture cation exchange capacity condensation irrigation drip subsurface irrigation hectare integrated nutrient management net primary productivity no till payments for ecosystem services recommended management practices soil organic carbon sub-Saharan Africa
REFERENCES Amudavi, D., Z. Khan, and J. Pickett. 2007. Enhancing the push–pull strategy. Leisa Magazine 23:8–10. Bellassen, V., R.J. Manlay, J.-P. Chèry, V. Gitz, A. Tourè, M. Bernoux, and J.-L. Chotte. 2010. Multi-criteria spatialzation of soil organic carbon sequestration potential from agricultural intensification in Senegal. Climatic Change 98:213–243. Bouman, B.A.M., R.M. Lampayan, and T.P. Toung. 2007. Water Management in Irrigated Rice: Coping with Water Scarcity. Los Baños, Philippines: IRRI. Burney, J.A., S.J. Davis, and D.B. Lobell. 2010. Greenhouse gas mitigation by agricultural intensification. P. Natl. Acad. Sci. USA 107:12052–12057. Caldwell, M.H., and J.H. Richards. 1989. Hydraulic lift: Water efflux from upper roots improves effectiveness of water uptake by deep roots. Oecologia 79:1–5. Cook, S.M., Z.R. Khan, and J.A. Pickett. 2007. The use of push–pull strategies in integrated pest management. Annu. Rev. Entomol. 52:375–400. Food and Agricultural Organization (FAO). 2010. Challenges and opportunities for carbon sequestration in grassland systems. Integrated Crop Management, vol. 9-2010. Rome: FAO. Glendining, M.J., A.G. Dailey, A.G. Williams, F.K. Van Evert, K.W.T. Goulding, and A.P. Whitmore. 2009. Is it possible to increase the sustainability of arable and ruminant agriculture by reducing inputs? Agr. Syst. 99:117–125. Glover, J.D., J.P. Ragnold, L.W. Bell, J. Borevitz, E.C. Brummer, C.M. Cox, T.S. Cox, T.E. Crews, S.W. Culman, L.R. DeHaan, et al. 2010. Increased food and ecosystem security via perennial grains. Science 328:1638–1639. Gregory, P. J., S.I. Ingram, and M. Brklacich. 2005. Climate change and food security. Philos. Trans. R. Soc. B 360:2139–2148. Hassanali, A., H. Herren, Z.R. Khan, J.A. Pickett, and C.M. Woodcock. 2007. Integrated pest management: The push–pull approach for controlling insects, pests and weeds of cereals, and its potential for other agricultural systems including animal husbandry. Philos. Trans. R. Soc. B 363:611–621. Hawkes, C., and M. Ruel. 2010. Changing global diets: Implications for agriculture. In The International Dimensions of the American Society of Agronomy: Past and Future, B. Payne and J. Ryan, eds., 57–65. Madison, WI: ASA.
10
World Soil Resources and Food Security
International Institute for Environment and Development (IIED). 2010. The Impacts of Climate Change on Food Security in Africa: A Synthesis of Policy Issues for Europe. London: IIED. Lal, R. 2010. Managing soils and ecosystems for mitigating anthropogenic carbon emissions and advancing global food security. Bioscience 60:708–721. Leigh, J., A. Hodge, and A.H. Fitter. 2009. Arbuscular mycorrhizal fungi can transfer substantial amounts of nitrogen to their host plant from organic material. New Phytol. 181:199–207. Lobell, D., and J. Burney. 2010. Greenhouse gas mitigation by agricultural intensification. Proc. Natl. Acad. Sci. USA 107(26):12052–12057. Lobell, D.B., M.B. Burke, C. Tebaldi, M.D. Mastrandrea, W.P. Falcon, and R.L. Naylor. 2008. Prioritizing climate change adaptation needs for food security in 2030. Science 319:607–610. McKinsey & Company. 2009. Pathways to a low carbon economy. The Global Greenhouse Gas Abatement Cost Curve, version 2. New York: McKinsey & Company. Molden, D., ed. 2007. Water for Food, Water for Life. London: IWMI Earthscan. Rajaram, S. 2010. Challenges in wheat research and development. In The International Dimensions of the American Society of Agronomy: Past and Future, B. Payne and J. Ryan, eds., 39–47. Madison, WI: ASA. The Royal Society. 2009. Reaping the Benefits: Science and the Sustainable Intensification of Global Agriculture. RS Policy Document RD 1608. London: The Royal Society. Semenov, M. 2009. Impacts of climate change on wheat in England and Wales. J. R. Soc. Interface 6:343–350. Sileshi, G., F.K. Akinnifesi, O.C. Ajayi, and F. Place. 2008. Meta-analysis of maize yield response to woody and herbaceous legumes in the sub-Saharan Africa. Plant Soil 307:1–19. United Nations Framework Convention on Climate Change (UNFCCC). 2008. Challenges and Opportunities for Mitigation in the Agricultural Sector. Technical Paper F CC/ TP/2008/8.2008. Wassmann, R.S.,V.K. Jagdish, S. Hauer., A. Ismail, E. Redona, R. Serraj, R.K. Singh, G. Howell, H. Pathak, and K. Sumfleth. 2009. Climate change affecting rice production: The physiological and agronomic basis for possible adaptation. Adv. Agron. 101:59–122. Winslow, M.D., and R. Ortiz. 2010. Biofuels: Risks, opportunities and dilemmas for international agriculture. In The International Dimensions of the American Society of Agronomy: Past and Future, B. Payne and J. Ryan, eds., 99–105. Madison, WI: ASA. World Bank. 2009. World Development Report 2010. Development and Climate Change. Washington, DC: World Bank. Zomer, R.J., A. Trabucco, R. Coe, and F. Place. 2009. Trees on Farm: Analysis of Global Extent and Geographical Patterns of Agroforestry. ICRAF Working Paper 89. Nairobi, Kenya: World Agroforestry Center.
Food Situation 2 Global and Unresolved and Emerging Issues Shahla Shapouri, Stacey Rosen, and Summer Allen CONTENTS 2.1 2.2 2.3 2.4 2.5 2.6
Introduction..................................................................................................... 11 Population Growth and Disparity.................................................................... 12 Income Growth and Disparity......................................................................... 12 Production Growth and Disparity................................................................... 13 Food Consumption and Composition.............................................................. 17 Unresolved and Emerging Issues..................................................................... 21 2.6.1 Persistent Food Insecurity................................................................... 21 2.7 Growing Demand for Biofuels......................................................................... 22 2.8 Natural Resources and Climate Change.......................................................... 23 2.9 Access to New Technology to Enhance Production Growth...........................24 2.10 Rising Obesity Rates.......................................................................................25 2.11 Looking Ahead................................................................................................26 References................................................................................................................. 27
2.1 INTRODUCTION The global food situation depends on the interactions between consumption and production that are, in turn, influenced by population and income growth, technological change, geography, environment, and social and cultural factors. The key factors currently driving per capita food consumption are income levels and food prices, while the key factor driving food composition is the geographical setting (e.g., rural vs. urban areas). Therefore, any changes in income, prices, and/or geographical location are normally accompanied by changes in food access, dietary composition, and food security. Growth in food production can result from expanding the agricultural land base and/or intensifying production per unit of land. Given the economic and environmental constraints on cropland expansion, however, the bulk of increased food production will need to come from higher yields on existing cropland. This yield growth depends on the quality and availability of the required inputs, including natural resources such as land and water. Noting the importance of the multiple factors that influence global food security, this chapter will discuss the dynamics of population, income, production, and 11
12
World Soil Resources and Food Security
food consumption and composition, spatially and temporally. These dynamics are affected by the unresolved and emerging issues of chronic food insecurity, rising obesity, increased demand for biofuels, technology adoption, resource scarcity, and climate change. These components will be discussed in more detail.
2.2 POPULATION GROWTH AND DISPARITY During the twentieth century, population grew dramatically compared to historical rates. Between 1800 and 1900, the world population increased from 600 million to 1.6 billion. The rate of growth then accelerated such that by the year 2000, the global population had increased more than 3.5 times, reaching the 6 billion mark. In fact, it took only 12 years for the population to increase from 5 to 6 billion [UNFPA 2008]. This jump took place despite major factors: 1) a decline in the global rate of population growth as fertility rates in 88 countries fell below the level required for long term replacement of their populations; and 2) the AIDS pandemic that spread across Eastern and Southern Africa, resulting in a significant cut in population growth rates in several countries in these regions [UNFPA 2008]. The global population growth rate has averaged about 1% per year since 1990, about half the rate experienced during the 1950s and 1960s [UNFPA 2008]. Despite this slowdown in population growth, total global population increases more than 320 million people—larger than the population of the United States in 2008—every 4 years [FAO 2009]. There was a noticeable difference in the rate of population growth among country groups during the twentieth century. Developing countries, which had been characterized by high birth and death rates, transitioned to low death rates, thereby fueling global population growth. In developed countries, birth rates and death rates generally declined at the same rate; in some countries, mainly European countries and Japan, however, birth rates declined faster than death rates, leading to nearly stagnant population growth rates. This difference in growth paths led to a lasting change in the dynamic of populations during the past century: 80% of the population growth since 1900 has taken place in developing countries, particularly in the world’s poorer countries [UNFPA 2008]. Among developing countries, the population growth paths are clearly associated with income levels. In Asia, the highest income growth region, the decline in the population growth rate was much faster than in other regions. This is followed by Latin America and then Africa. At the subregion level, the smallest change has occurred in sub-Saharan Africa (SSA), where historical annual population growth of 3% is projected to decline to 2.2% per year through the next decade, which can to a large degree be attributed to the AIDS pandemic. Despite the HIV/AIDS pandemic, SSA’s population nearly tripled, to an estimated 650 million between 1960 and 2000; in 2008, the region’s population was roughly 760 million or about 11% of the global population [UNFPA 2008].
2.3 INCOME GROWTH AND DISPARITY The most significant factor behind global food consumption patterns is income growth. From 1970 to 2005, global per capita income grew by about 70% on average,
Global Food Situation and Unresolved and Emerging Issues
13
increasing from about $3700 to $8000 (in 2000 U.S. dollars) [IMF 2007]. If this trend continues, it will result in a doubling of global per capita income by 2040. As expected, this record of overall growth in average world income, however, was not equally distributed among regions. In East Asia and the Pacific, per capita income rose more than 7-fold; in South Asia, income grew more than 2.5-fold, with a 2-fold increase in developed countries with high incomes and Euro areas. Per capita income in the Middle East and North Africa increased by 64% and by 56% in Latin America. SSA experienced the slowest growth at only 3% in 35 years. To put this into a global context, the region has about 11% of the global population, but less than 1.5% of its income. The disparity in growth at the country level was much higher than at the regional level. Per capita income (measured as per capita gross domestic product, GDP, in 2000 U.S. dollars) in 1970 ranged from $122 (in China and Malawi) to $25,000 in Switzerland. In 2005, it ranged from $89 in DR Congo to $40,000 in Norway. In addition to this wide gap between low- and high-income countries, there were also substantial differences among the low-income countries. For example, between 1970 and 2005, China and Malawi took very different growth paths. Malawi’s per capita income increased to $138, while China’s per capita income jumped more than tenfold to $1451. These global income disparities were also witnessed within countries. In the United States, for example, the top 5% of household income was 2.6 times higher than the 50th percentile in 1967; in 2005, it was 3.6 times higher. This scenario is replicated in both developed and developing countries.* In sum, at the start of the twenty-first century, the world was faced with enormous income disparities both across and within countries. The question at hand is, “If recent trends in income disparity continue, what will be the impact on poverty and food security?” During the past century, growth in global food production surpassed population growth, leading to an overall improvement in per capita food consumption at the aggregate level. The main food production surge occurred during the 1950s and 1960s as new, high-yielding crop varieties were adopted [FAO statistics 2009]. Productivity growth allowed for the release of resources—labor in particular—to the rest of the economy and was responsible for a decline in the prices of major agricultural commodities. It is not certain that future productivity increases will be able to outweigh population growth and this could have disastrous effects on food security.
2.4 PRODUCTION GROWTH AND DISPARITY While growth in global food production far outpaced the growth in the world’s population from 1970 to 2007, there was a significant growth disparity among regions. For example, per capita food production growth in South America was more than 60% [Shapouri et al. 2009]. And while growth in North America, Central America, North Africa, and South Asia fell short of that rate, it was still quite strong. SSA had by far the weakest performance. As can be seen in Figure 2.1, West Africa was the only subregion with positive growth, and that was only 1% [Shapouri et al. 2009]. * http://www.frbsf.org/publications/economics/letter/2007/el2007-28.pdf.
14
World Soil Resources and Food Security Percent change 70 60 50 40 30 20 10 0 –10 –20 –30 –40
E N S W Cent N Cent S S Africa Africa Africa Africa Africa Amer Amer Amer Asia
FIGURE 2.1 Per capita food production from 1970 to 2007.
In addition to changes in food production, returns on agricultural labor (value added per worker), an indicator of prosperity in rural areas, showed an increase of about 40% at the global level from 1980 to 2005 [FAO 2009]. In high-income countries, this return more than tripled from $9000 (in 2000 U.S. dollars) in 1980 to $29,000 by 2005. The second highest return per agricultural worker in 2005 was about $3000 in Latin America and the Caribbean, followed by South Asia at $2300. The lowest return was $348 in SSA. These differences in returns per agricultural worker are mainly due to disparities in agricultural investment and the use of new technology. For example, agricultural machinery use, as measured by tractor use, was by far the lowest in SSA. In high-income countries, 435 tractors were used per 100 square km of arable land in 2003; this was followed by 143 in South Asia and 122 in Latin America and the Caribbean [FAO 2009). In SSA, tractor use was only 12 per 100 square km of arable land [FAO 2009]. Grain yields are also a good proxy for the technological gap among regions. Grains comprise the largest share of the diet of the developing world, and therefore their yield performance directly affects food availability and food security of these countries. Grain yields have outpaced population growth since 1970, and yield growth was comparable for all regions except SSA. In high-income countries as well as Latin America and South Asia, grain yields more than doubled in 35 years [Rosen et al. 2008]. As can be seen in Figure 2.2, even in low-income countries, yields nearly doubled. In SSA, however, yields increased only 30% during the same time period. As a result, in 2005, SSA grain yields were only 20% of those in the highincome countries and about 50% of those in low-income countries in other regions [Rosen et al. 2008]. One major factor affecting yields is technology adoption. Farmers choose between technologies based on land characteristics such as soil quality and access to water, as well as personal characteristics like land tenure, income/wealth, and access to credit and information. The farmer’s choice of practices, such as fertilizer application and residue management, also depends on his/her time horizon. For example, practices
15
Global Food Situation and Unresolved and Emerging Issues 6000 5000 kg/ha
4000 3000 2000 1000 0
Sub-Saharan Africa
High-income Latin America & South Asia countries Caribbean 1970
Low-income countries
2005
FIGURE 2.2 Grain yields.
generating high net returns today may not do so indefinitely if they contribute to land degradation. However, practices that reduce land degradation and offer higher net returns in the long run may require initial investments that can preclude adoption. In recent decades, it is estimated that about half of all gains in crop yields have been attributed to genetic improvements in seeds, while the remainder of gains are due to the increased use of conventional inputs, especially fertilizer and irrigation water [Frisvold et al. 1999]. The World Development Report of 2008 notes that an increase in fertilizer use is responsible for 20% of the growth in developing country agriculture for the past 30 years [World Bank 2007]. Growth in fertilizer consumption per hectare of cropland has been slowing, however, from a global average annual increase of 9% in the 1960s to an average annual decline of 0.1% in the 1990s [FAO statistics 2009]. Developed countries are considered a mature fertilizer market, and support measures in the United States and European Union have been reduced. Also, the increasing awareness of potential environmental harm from overuse is hindering growth in fertilizer use. Growth of fertilizer use in developing countries is expected to exceed that of developed countries in the coming years. The fertilizer consumption of developing countries has increased rapidly since 1970. In fact, these countries’ share of global fertilizer use increased from 10% in the 1960s to more than 60% in 2006 [World Bank 2007]. Among developing regions, fertilizer consumption per hectare increased most rapidly in land-scarce Asia and most slowly in Africa. SSA accounts for only 1% of global fertilizer consumption. In many low-income countries, particularly in SSA and Latin America, almost all fertilizer is imported and the insufficient foreign exchange constrains availability. Fertilizer use is most productive in irrigated areas or areas with sufficient moisture. Therefore, in regions suffering from or vulnerable to dry periods (Latin America or SSA), fertilizer use would not experience the results of areas without similar adversities. Investment in irrigation has boosted food production in many developing countries. These higher yields generate higher revenues. For example, gross crop revenue from irrigated plots has shown to be 79% higher than that of nonirrigated plots in
16
World Soil Resources and Food Security
China, due to higher yields, increased cropping intensity, and shifts to higher-valued crops, all made possible by irrigation [Huang et al. 2006]. On average, the global irrigated area increased nearly 1.4% per year between 1980 and 2002, although the growth rate has declined over time [FAO 2009]. Growth in developing countries, in particular, exceeded this rate and currently more than a quarter of arable land area in developing countries is irrigated. The highest growth in irrigated area in the developing world has occurred in Asia, particularly Bangladesh, Nepal, and Vietnam. There are projections that by 2030, irrigated agriculture in many developing countries will be responsible for a large percentage of the increase in grain production [UNESCO 2006]. Currently in South Asia, more than 39% of the cultivated land is irrigated, and more than 29% is irrigated in East Asia, while only 4% is irrigated in SSA [World Bank 2007]. In addition, expansion of irrigated lands in the region is negligible—0.5% per year since 1990 [FAO 2009]. This rate marks a significant slowdown from growth in the prior decade. Irrigation requires access to water as well as investment in equipment and maintenance—elusive factors in most of SSA. Renewable water resources in Africa are less than 9% of the global renewable resources [FAO 2010]. Population growth and the increasing cost of developing new sources of water will place increasing pressure on world water supplies in the coming decades. Even as demand for irrigation water increases, farmers face growing competition for water from urban and industrial users, as well as pressure to support water’s ecological functions. Genetic improvements to seeds that enhance input responsiveness, resistance to pests and diseases, and tolerance to other stresses have driven much of the gains in yields in recent years. Figure 2.3 shows the extent of growth in scientifically bred varieties by the 1990s. 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
Rice
Corn Developing
Wheat Developed
FIGURE 2.3 Percentage of land used for growth of scientifically bred varieties in the 1990s.
17
Global Food Situation and Unresolved and Emerging Issues
Gains from genetic improvements will continue, but likely at slower rates and increasing costs, particularly because gains in input responsiveness have been almost fully exploited. Moreover, while the use of hybrid seeds has raised yields considerably in some countries, their proliferation may not be possible in many developing countries where conditions are not amenable due to lack of adequate extension services or funding for research.
2.5 FOOD CONSUMPTION AND COMPOSITION The combination of income growth and increasing agricultural production not only resulted in increased food availability, but also changed the composition of diets. Globally, daily per capita calorie availability increased by 17% between 1970 and 2005—from 2134 to 2722 [Rosen et al. 2008]. In 1970, grains accounted for more than half of the calories consumed, as shown in Figure 2.4. Overall, globalization opened markets for products, and many farmers were able to capitalize on these changes by supplying a wide variety of products to growing and evolving markets. Per capita consumption in developing countries exceeded 2722 calories per day in 2005, up from 2134 calories in 1970. The Food and Agriculture Organization (FAO) of the United Nations recommends a daily per capita intake of roughly 2100 calories. Grains account for more than half of the diets in developing countries, but the almost 11% increase in per capita grain consumption between 1970 and 2005 was much lower than the overall increase in calorie consumption [Rosen et al. 2008]. While per capita consumption of some higher-value food items soared, meat still accounted for only about 9% of the diets in developing countries in 2003, compared with 14% in developed countries [Rosen et al. 2008]. Diet composition changes are illustrated in Figures 2.5 through 2.8. The two most populous countries—China and India—with about one-third of the global population, significantly influence the direction of changes in food consumption in developing countries. In China, the roughly 50% increase in per capita food consumption (measured in calories) between 1970 and 2005 included a tenfold 60 50 40 30 1970
20
2005
10 0
Cereals
Sugar and Veg oils sweeteners
FIGURE 2.4 Global diet composition.
Meat
Milk
18
World Soil Resources and Food Security 3% 6%
4% Cereals
8%
Sugar and sweeteners Veg oils Meat Milk 79%
FIGURE 2.5 Diet composition in developing countries in 1970.
increase in the consumption of fruits and an eightfold increase in the consumption of meat and milk. Other significant increases were in vegetables and vegetables oils. Reflecting the diversification of the diet and increased wealth of the population, the largest decline was seen in cereals, whose share of diets declined from 68% in 1970, to 47% in 2005. In India, the increase in per capita food consumption was relatively modest, 20 percent, during 1970 to 2005. Among the main food categories, per capita consumption of vegetable oils, milk, and eggs more than doubled. Like China, cereals’ share of the diet declined, from 66% in 1970 to 59% in 2005. While per capita calorie consumption rose in the least developed countries (those with per capita incomes below $500 per year), the gain was much smaller than in developing countries as a whole. Their consumption increased from 2000 calories per person in 1970 to 2200 in 2000 [Rosen et al. 2008]. However, consumption of nutritionally beneficial foods, such as pulses, vegetables, and fruits, has declined in the least developed countries. The decline was sharpest for vegetables (32%), followed by fruits (9%) and pulses (5%) [Rosen et al. 2008]. Even with the modest
9%
4% Cereals
12%
Sugar and sweeteners Veg oils
9%
Meat 66%
Milk
FIGURE 2.6 Diet composition in developing countries in 2003.
Global Food Situation and Unresolved and Emerging Issues
19
12% Cereals
13%
45%
Sugar and sweeteners Veg oils Meat
11%
Milk 19%
FIGURE 2.7 Diet composition in developed countries in 1970.
increases in overall calorie consumptions in these countries, there is a shift in diets toward fats and sugar and away from traditional diets of vegetables and pulses. A major factor influencing the composition of diets is urbanization. In 1975, unlike developed countries, only 25% of developing countries’ populations resided in urban areas. By 2015, that number is expected to reach 50% [UNDP 2002]. In most developing countries, the rate of urbanization was two to three times higher than the population growth rate during the prior three decades [World Bank 2007a]. Since 1970, in 68 out of 114 countries, population in the largest cities more than doubled and most of the increase was in developing countries [World Bank 2007a]. In 18 countries, 16 in SSA, increases in the populations of the largest city were more than fivefold [World Bank 2007a]. Unlike rural agricultural households, urban residents do not rely solely on homeor locally-grown foods and, therefore, have access to a wider selection of foods. Although detailed data for each country are not available, examining diet compositions across countries shows that in countries with the same income level, those with a higher share of urban population tended to have diets with more fat, both
11% Cereals
14%
41%
Sugar and sweeteners Veg oils Meat
17%
Milk 17%
FIGURE 2.8 Diet composition in developed countries in 2003.
20
World Soil Resources and Food Security
vegetable and animal. For example, the urbanization rate is 67% in Mexico vs. 92% in Uruguay, and daily per capita consumption of fat in Mexico was half that of Uruguay, despite similar per capita income levels ($6172 and $6248, respectively, in 2005) [Rosen et al. 2008]. Similarly, fat consumption in Jordan was more than four times that of Namibia. Although their per capita income was almost the same ($2086 in Jordan and $2083 in Namibia in 2005), the 82% urbanization rate in Jordan was much higher than that of Namibia’s at 35% [Rosen et al. 2008]. However, other factors such as cultural and dietary habits might also contribute to differences. All urban environments are not the same; the openness of an economy, access to mass media, particularly television, and marketing systems can also significantly influence consumers’ choices. Regardless of consumer food choices, however, an urban lifestyle usually means a decline in physical activity and the higher participation of women in the workforce. The latter often translates into less time for food preparation, which often leads to increased consumption of processed foods. In addition to income growth and urbanization, the expansion of international trade through world economic integration has influenced global diets. Trade agreements over the past 3 decades, in addition to expanding global trade, have been catalysts for increased investment in transportation and communication systems. The average ocean freight and port charges for U.S. import and export cargo decreased 60% between 1970 and 1990 [Rosen et al. 2008]. New technologies such as refrigeration also facilitated trade in perishable products. The decline in global trade barriers was followed by liberalization in global financing, altering the food systems of most countries. The evolution of global food systems and increases in the number of supermarkets have promoted convenience shopping and wider food varieties in developing countries. Due to efficiencies of scale, supermarkets are often able to offer lower prices than traditional retail stores. Lower prices have boosted the market shares and profits of supermarkets, which have fueled their expansion. The high growth in market share of supermarkets in Latin America highlights the extent of the change: from a 10%–20% market share in the 1980s to 50%–60% in the 1990s, and now rapidly approaching the U.S. share of about 70%–80% [Reardon et al. 2004]. The experiences of East and South Asia also show a similar pattern. In SSA, with the exception of South Africa, the supermarket share in the retail food market is much smaller, but expansion is under way due to growing investment by South African companies. The growing role of supermarkets has both positive and negative implications for consumers. On the positive side, supermarkets are introducing better quality, greater variety, higher standards, and lower prices to the food systems of developing countries. On the negative side, they increase access to low-cost, high-calorie convenience foods for urban consumers with limited physical activity, thereby fueling obesity problems. Furthermore, food imports have become an important component of food supplies in both developed and developing countries because national food self-sufficiency has declined in many countries during the past few decades. Trade in grains, vegetable oils, and meat increased three to five times during the past 3 decades. Developing countries also became more dependent on imports of staple commodities, such as grains and vegetable oils. Rising consumption of wheat, in the processed form of
Global Food Situation and Unresolved and Emerging Issues
21
bread and pasta, has replaced traditional grains such as millet and sorghum, as well as root crops. Import growth was not limited to staple foods; imports of a variety of commodities, including semiprocessed and processed foods, have also grown. Between 1970 and 2005, global trade volume of highly processed foods (the FAO definition includes food items such as canned meat, breakfast cereals, pastries, and wine) increased more than four times. Import growth for highly processed foods was highest in developing countries—growing more than fivefold between 1970 and 2005 [Rosen et al. 2008].
2.6 UNRESOLVED AND EMERGING ISSUES 2.6.1 Persistent Food Insecurity Despite the growth in food availability, reports (USDA-ERS and FAO) indicate that food insecurity among the poorest countries is intensifying. The global economic crisis, coupled with the financial pressures created by the soaring food and fuel prices during 2006–2008, put food security on the front line of global concerns. The difficult global financial environment may influence food security by limiting commercial imports, which had been growing rapidly in many of these countries, as seen in Figure 2.9. This growing reliance on food imports was spurred by income growth, trade liberalization policies, and improvements in the global transportation system. The global downturn, however, threatened the import capacities of many low-income countries because it led to reduced export earnings as a result of reduced global 120
Million tons
100 80 60 40 20 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 Grain food aid
Grain commercial imports
*in 70 lower income developing countries included in the USDA-ERS Food Security Assessment
FIGURE 2.9 Rising commercial imports.
22
World Soil Resources and Food Security
demand, foreign capital inflows, and remittances from relatives working abroad. Regionally, import dependency for grain, the main staple food consumed by the poor, is lowest in Asia, but increases for SSA, Latin America and the Caribbean, and North Africa. Much of Latin America, the Caribbean, and North Africa import nearly half of their grain supplies. Some countries can forgo imports of other commodities and allocate a much larger share of their import budget to food during a crisis period. However, for those that were highly food insecure at the outset, like many in SSA, the decline in economic growth and import capacity can have adverse food security implications. The near- and medium-term food security of developing countries depends on the depth and the length of the current economic downturn. Tighter credit and weaker global growth are likely to cut into government revenues and investment in areas such as human capital and infrastructure that are essential for sustained growth. Even under the assumption of global economic recovery, food insecurity in many lower income countries is expected to remain precarious into the long term. According to the USDA’s Economic Research Service (USDA-ERS), the number of food-insecure people is projected to remain relatively flat through the next decade, reaching 834 million by 2018 [Shapouri et al. 2009]. However, these impacts will not be distributed evenly; while an improvement is projected in Asia, deteriorating food security is projected for SSA. SSA will remain the most vulnerable region in 2018. The region accounts for 25% of the population of the 70 countries included in the USDA-ERS study, but accounts for 57% of the food-insecure people [Shapouri et al. 2009]. Several countries, such as Somalia and the Democratic Republic of Congo, are likely to remain politically unstable as they continue to be plagued by armed conflicts that have caused significant breakdowns of law and order. The resulting social dislocation, food insecurity, and famine diminish optimism for the future. The slow pace or lack of progress in improving food insecurity in low-income countries was evident even before the current global economic downturn. While the full consequences of this crisis are not known, for low-income countries, food security problems are expected to worsen. The short-term concern is that these countries have few or no domestic safety net programs in place. The international safety nets that do exist are inadequate for stabilizing food supplies for the more vulnerable countries. Food aid, the primary safety net, is not sufficient to meet estimated needs around the world. Integrating international and national resources to redesign these safety net programs could help mitigate the effects of shocks and serve as adjuncts to longer term food security strategies. The challenge, however, is to design efficient programs that minimize costs, while working for longer term solutions.
2.7 GROWING DEMAND FOR BIOFUELS Recent oil price hikes have raised serious concerns in low-income countries, both because of the financial burden of the higher energy import bill and because of potential constraints on imports of necessities like food and raw materials. Adding to these worries, the higher oil prices have sparked global energy security concerns that have spurred the demand for biofuel production. The use of food crops for producing
Global Food Situation and Unresolved and Emerging Issues
23
biofuels, coupled with the increase in food demand spurred by income growth in the most populous countries such as China and India, has reversed the path of declining price trends for several commodities. While rising energy prices have negatively impacted the budget of importing countries, they have also created an opportunity for advances in biofuel technology, which could help fill the growing energy needs of developing countries. Investments in biofuel production by low-income countries could promote rural development, since large shares of their populations depend on agriculture for employment and income. Countries such as Colombia and India have adopted production targets for increasing the share of biofuels in their transportation fuel supplies. Other countries are examining alternative biofuel sources appropriate for their particular environments and resource availabilities. In addition, researchers in several countries of Asia, Latin America, and Africa have pointed toward the potential of several indigenous plants such as Jatropha, which grows wild in these countries, requires little water and few nutrients, and has a relatively high oil yield. The payoff for lowincome countries from investing in research in the area of biofuels can be significant and could help reverse the trend of low national and international investments in agricultural research in low-income countries. Currently, traditional biofuels such as wood fuel account for about one-third of all energy consumed in developing countries. However, these fuel sources provide low yields in terms of heat. For example, a kilogram of wood generates only about onetenth of the heat of a kilogram of liquid petroleum gas. The new sources of biofuels could improve energy efficiency, as well as increase the supply of energy and boost farm incomes and rural employment in poor areas. Success, however, depends on increasing investment in the development of new technology that is consistent with the structure of the agricultural sectors of lowincome countries. Most low-income countries have poor market infrastructures and weak financial systems, which lead to raised costs of production, particularly for newly introduced biofuel commodities that require specific production and distribution facilities. In addition, the financial capacity for investment in low-income countries is limited, and there is concern that increasing investments in biofuel production could compete with food production, thereby intensifying food insecurity.
2.8 NATURAL RESOURCES AND CLIMATE CHANGE Agricultural resource scarcity is an underlying issue that has greatly impacted and will continue to impact any efforts to increase global food supply. Limited water availability, soil degradation (through erosion and nutrient depletion), and deforestation and desertification can have detrimental effects on agricultural production. Water resources, in particular, are considered an integral component of development, especially in rural areas that depend on agriculture. An average of 70% of the water withdrawn from surface or groundwater supplies globally is used for agriculture [FAO 2003]. Of course, there are significant regional variations on how this water is used and the efficiency of those uses. The work Water for Food, Water for Life [Molden 2007] notes that 55% of the value of the world’s food is produced under rainfed conditions. However, given the variability and location of water supplies,
24
World Soil Resources and Food Security
irrigated agriculture plays an increasingly important role, especially for grain production. It is estimated that irrigated agriculture is responsible for 60% of the grain production in the developing world [Huang 2006]. Water resources are subject to multiple demands in addition to agriculture, including energy, industry, domestic, and environmental flows. Presently, around 1.6 billion people live in river basins where water use exceeds recharge rates, and one in three people are already facing water shortages [Molden 2007]. Climate change is projected to result in increased precipitation variability and affect food production through both biophysical and socioeconomic pathways. These impact projections vary greatly among different regions of the world and the magnitude of the changes is unknown. However, temperature increases, changing patterns of precipitation, and more extreme weather will have the most extensive effects on agriculture in the subtropics, where a majority of the least developed countries are located [Molden 2007]. It is estimated that one in six countries will face food shortages each year due to droughts [UNFPA 2009]. A few estimates have been completed regarding the impacts of increased variability and extreme events on agricultural production in these areas. Overall, developing countries are expected to have a 9%–21% reduction in agricultural productivity as a result of global warming by 2080 [Cline 2007]. While the impact projections vary with projections of CO2 levels and assumptions regarding adaptation, the models are robust in the projection of increased aridity in both Mediterranean environments and in the marginal arid and semiarid tropics, especially in SSA [Bates et al. 2008]. Developing economies, in general, are expected to be more affected by climate change due to greater exposure to climate shocks and less adaptive capacity [World Bank 2009]. Despite the unknown climate change impacts on food production in the developing world, there are opportunities to reduce some of the consequences. Productivity gains in low productivity areas (particularly rainfed areas) and more integrated management of water resources could increase food security [Molden 2007]. However, as climate change increases water stress and variability, it is important that water management approaches be integrated into large poverty reduction strategies to help reduce vulnerability to shocks [Molden 2007].
2.9 A CCESS TO NEW TECHNOLOGY TO ENHANCE PRODUCTION GROWTH Global agricultural production more than doubled during the second half of the twentieth century, but the rate of growth among countries differed significantly. With demand growth expected to nearly double by the middle of this century as both population and incomes rise in many of the world’s poorest countries, a substantial increase in scientific and technical efforts will be required to achieve a sustainable system of agricultural production. Genetically modified (GM) seeds have been successful in increasing the productivity of several food crops. GM seeds were originally developed to improve crop protection, improve resistance to insect damage or viruses, and increase tolerance to herbicides, but their impact on yield growth has been considerable. The adoption of GM seeds is more widespread in developed countries, but developing countries have
Global Food Situation and Unresolved and Emerging Issues
25
begun to adopt them as well. In fact, 90% of the farmers growing GM crops in 2007 were smallholders in developing countries, mostly in China, India, South Africa, and the Philippines [PG Economics 2009]. Corn, cotton, canola, and soybeans represent most of the GM crops grown commercially in developing countries and are either herbicide-tolerant, pest-resistant, or both [FAO 2002]. However, countries are performing trials on a much broader range of crops including bananas, cassava, cowpeas, plantains, rice, and sorghum [FAO 2002]. Also, the traits of the crops are expanding to include drought-resistance and quality improvement. Global farm income gains attributed to growing GM crops were estimated at more than $44 billion between 1996 and 2007 [PG Economics 2009]. In 2007 alone, the benefit was valued at more than $10 billion, nearly 60% of which was derived by producers in developing countries [PG Economics 2009]. Included among the many factors that have contributed to these gains are reduced costs from decreased insecticide applications, higher yields from improved weed control and reduced pest damage, and increased production levels resulting from the planting of second-season crops due to the shortening of the production cycle. An additional benefit—for developing countries in particular—is the higher nutritional value of some of the varieties such as golden rice, for example, which can reduce vitamin A deficiency. For all the benefits of GM crops, there are also notable concerns associated with them. Unfortunately, because these are new technologies, the potential effects are simply not known. There is concern—particularly in the developing world—regarding regulating and ensuring the safe use of GM crops. In 2000, for example, a corn variety intended for animal consumption was found in food products [FAO 2002]. Concerns about GM crops, related to human health, revolve around provoking allergies, gene transfer, and outcrossing (moving genes from GM plants to conventional crops in the wild or the mixing of conventional and GM crops) can result in food safety concerns [WHO 2010]. Developing countries will need to weigh the potential benefits of adopting GM crops against the risk factors to determine the appropriate path for them to pursue. For some countries, many in SSA, the technology gap in food production relative to developed countries is huge and any step toward reducing the gap could have significant implications on their farm income and food security.
2.10 RISING OBESITY RATES In contrast to the problem of inadequate food supplies for the poor, the global increase in calorie consumption has led to excess food consumption by some segments of the population in many countries. In developing countries, consumption of high-calorie foods, such as fats and sugar, has risen, and the income elasticity (percentage change in consumption for each 1% change in income) for these products remains positive. While an estimated 800 million to 1 billion people are food insecure, the World Health Organization (WHO) estimated that there are about 1 billion overweight and obese people worldwide. While this problem is more prevalent in Western countries, it is increasing rapidly in developing countries as well, due to dietary shifts [WHO 2009]. An emerging concern in developing countries is obesity-related diseases such as diabetes and hypertension. For example, the WHO reports that in China hypertension
26
World Soil Resources and Food Security
increased 12% (or the equivalent of 160 million people) between 1991 and 2002 [FAO 2006]. Similarly, an estimated 25%–50% of the population in countries such as Mexico, Thailand, and Tunisia suffer from diabetes [FAO 2006]. The WHO assessment indicates that overweight and obesity represent a rapidly growing threat to health in an increasing number of developed and developing countries. It also indicates that, in some countries, overweight and obesity are now replacing the more traditional public health concerns of undernutrition and infectious diseases. The direct cost of obesity is the increased risk of chronic diseases such as diabetes, cardiovascular disease, gallbladder disease, and cancer. If current trends continue, health costs for the developing economies could be substantial. In most developing countries, people are a major resource and public health is a key to economic progress. Research shows that obesity reduces productivity [Thompson et al. 1994]. Moreover, health costs associated with the growing rate of obesity and its related diseases could overwhelm the healthcare systems of developing countries already overburdened with the costs of combating communicable diseases and the effects of malnutrition among lower-income populations. According to the latest World Bank data, average health expenditures, per capita, in developing countries are less than 10% of the health expenditures in developed countries and less than 1% in the least developed countries. Nutritional education is crucial in terms of reaching out to consumers. Because dietary habits are formed at young ages, the nutritional education of children can play a vital role in influencing dietary habits. Other policy interventions can also promote healthy eating. The Scandinavian countries reduced coronary heart disease between 1976 and the 1980s by providing subsidies for the purchase of healthy food items, such as fish [Rosen et al. 2008].
2.11 LOOKING AHEAD An FAO study, How to Feed the World 2050, argues that future demand growth will be influenced by lower population growth rates, higher urbanization rates, and higher incomes that, in turn, will influence the composition of diets. There is consensus that it is possible to produce adequate food in 2050, but achieving that goal requires certain conditions and policy decisions [FAO 2009]. The essential conditions according to the report are investments in productivity and market infrastructures to improve the functioning of markets and implementation of policies to enhance food access of the poor. The question is, “Will the growth in supply be adequate to improve global food security?” An important point to note is that in the past few decades, several large developing countries such as China, India, Brazil, and Mexico have shown significant improvements in their food security situations. On the other hand, food security may remain precarious for low-income countries such as those in SSA. For many low-income countries, growth in agricultural productivity is critical to improving food security for two reasons. First, growth in agricultural productivity translates into increased food supplies and lower food prices for all consumers. This is particularly true in many of the poorest countries that depend heavily on domestic food production for large parts of their food supplies, due in part to financial conditions that constrain imports and the fact that rural economies are not fully integrated
Global Food Situation and Unresolved and Emerging Issues
27
into formal economies. Second, growth in agricultural productivity means higher incomes and, thus, improved abilities to purchase food for the large share of these populations that depend on agriculture for their livelihoods. The World Development Report 2008 emphasized the need to increase investment in agriculture, especially the importance of increasing agricultural productivity in Africa through gains in soil fertility [World Bank 2007]. This would be achieved by increasing fertilizer use in the region, which is currently less than 10% of the level used in other developing regions [World Bank 2007]. Most of the least developed countries are in the early stages of adopting new agricultural technologies, and the potential to increase productivity is enormous, but sustained agricultural growth requires substantial investments in irrigation, rural infrastructure, human capital, and institutions, not just basic inputs. In addition to increasing the productivity of the agricultural sector, support for rural development can provide nonfarm employment and an opportunity for rural communities to diversify their sources of incomes, leading to higher incomes and greater stability. The World Bank has recently devoted much attention to the issue of rural development, especially in the context of increased environmental and economic variability, as seen in the upcoming World Development Report 2010 [World Bank 2009]. In the midst of this variability, developing rural markets can help create lower-risk environments that are essential for sustainable economic growth and improved food security. Unlike the productivity potential that exists in some developing countries in SSA and Latin America, many developed and some developing countries are close to their maximum scientific and technical potential for growing crops. Therefore, maintaining current growth rates may be unlikely in these areas with today’s technologies and practices. These countries must promote investment in agricultural research, technology education, and rural infrastructure. Economic development in developing countries must build upon the lessons learned from previous experience in the region. In some cases, a stronger focus on market integration is necessary, while in other cases, more sustainable use of natural resources is necessary to help compensate for variability and multiple demands.
REFERENCES Bates, B.C., Z.W. Kundzewicz, S. Wu, and J.P. Palutikof, eds. 2008. Climate change and water. Technical paper of the Intergovernmental Panel on Climate Change. Geneva: IPCC. Cline, W. 2007. Global warming and agriculture: Impact assessments by country. Washington, DC: Peterson Institute for International Economics. Food and Agriculture Organization of the United Nations (FAO). 2009. How to feed the world in 2050. The prospects for agriculture. FAO Expert Meeting. October, 2009. Rome: FAO. FAO. 2010. AQUASTAT database. http://www.fao.org/nr/water/aquastat/regions/africa/index .stm (accessed January 13, 2010). FAO. 2006. Fighting hunger and obesity. http://www.fao.org/ag/magazine/pdf/0602-1.pdf. Rome: FAO. FAO. FAOSTAT database. Available at http://faostat.fao.org/default.aspx. FAO. 2003. Agriculture, food, and water. Rome: FAO.
28
World Soil Resources and Food Security
FAO. 2002. Biotechnology and food security. Available at http://www.fao.org/worldfoodsummit/ english/fsheets/biotech.pdf (accessed January 11, 2010). Frisvold, G., J. Sullivan, and A. Raneses. 1999. Who gains from genetic improvements in US crops? AgBioForum 2(3&4):237–246 (available at http://www.agbioforum.org). Huang, Q., S. Rozelle, B. Lohmar, J. Huang, and J. Wang. 2006. Irrigation, agricultural performance and poverty reduction in China. Food Policy 31:130–152. International Monetary Fund (IMF). 2007. Finance and development report. International Monetary Fund (IMF) database. Available at http://www.imf.org/external/data.htm. Molden, D., ed. 2007. Water for food, water for life. London: IWMI Earthscan. PG Economics. 2009. Focus on income, well-being, and food security: Biotech crops: Evidence, outcomes, and impacts 1996–2007. Available at http://www.pgeconomics .co.uk/pdf/focusonincomeeffects2009.pdf (accessed January 11, 2010). Reardon, T., P. Timmer, and J. Berdegue. 2004. The rapid rise of supermarkets in developing countries: Induced organizational, institutional, and technological change in agrifood systems. The Electronic Journal of Agricultural and Development Economics 1(2): 168–183 (available at http://www.fao.org/docrep/Article/ejade/ae226e/ae226e00.htm). Rosen, S., S. Shapouri, K. Quanbeck, and B. Meade. 2008. Food security assessment, 2007. Washington, DC: USDA-ERS. Shapouri, S., S. Rosen, B. Meade, and F. Gale. 2009. Food security assessment, 2008–09. Washington, DC: USDA-ERS. Thompson, D., J. Edelsberg, G.A. Colditz, A.P. Bird, and G. Oster. 1994. Lifetime health and economic consequences of obesity. Arch. Intern. Med. 159(18):2177–2183. United Nations Development Program (UNDP). 2002. Human development report 2002: Deepening democracy in a fragmented world. New York: UNDP. United Nations Educational, Scientific and Cultural Organization (UNESCO). 2006. Water: A shared responsibility. The United Nations World Water Development Report 2. Paris: UNESCO. United Nations Population Fund (UNFPA). 2009. State of world population, facing a changing world: Women, population, and climate. New York: UNFPA. UNFPA. 2008. State of world population, reaching common ground: Culture, gender and human rights. New York: UNFPA. United States Department of Agriculture Economic Research Service (USDA-ERS). 2009. Outlook Report No. (GFA-20). Available at http://www.ers.usda.gov/Publications/ GFA20/. World Bank. 2009. World development report 2010: Development and climate change. Washington, DC: World Bank. World Bank. 2007a. World development report 2008: Agriculture for development. Washington, DC: World Bank. World Bank. 2007b. World development indicators 2007. Available at http://go.worldbank .org/3JU2HA60D0. World Health Organization (WHO). 2010. 20 questions on genetically modified (GM) foods. Available at http://www.who.int/foodsafety/publications/biotech/20questions/en/ (accessed January 11, 2010). WHO. Obesity and overweight facts. Available at http://www.who.int/dietphysicalactivity/ publications/facts/obesity/en. World Policy Institute. 2009. Water wars? A talk with Ismail Serageldin. World Policy Journal 26(4):33–40.
Soil Resources 3 World Opportunities and Challenges H. Eswaran, P. F. Reich, and E. Padmanabhan CONTENTS 3.1 Introduction..................................................................................................... 29 3.2 Global Distribution of Soils............................................................................. 32 3.2.1 Alfisols................................................................................................. 32 3.2.2 Andisols............................................................................................... 32 3.2.3 Aridisols.............................................................................................. 32 3.2.4 Entisols................................................................................................ 37 3.2.5 Gelisols................................................................................................ 37 3.2.6 Histosols.............................................................................................. 37 3.2.7 Inceptisols............................................................................................ 37 3.2.8 Mollisols.............................................................................................. 43 3.2.9 Oxisols................................................................................................. 43 3.2.10 Spodosols............................................................................................. 43 3.2.11 Ultisols................................................................................................. 43 3.2.12 Vertisols............................................................................................... 47 3.3 Historical Trends............................................................................................. 47 3.4 Current Issues.................................................................................................. 49 3.4.1 Opportunities....................................................................................... 49 3.4.2 Challenges........................................................................................... 49 3.5 Future Perspectives.......................................................................................... 50 3.6 Conclusions...................................................................................................... 51 References................................................................................................................. 51
3.1 INTRODUCTION Soil is a product of the interaction of climate with bedrock; the properties of the soil produced are determined by ancillary factors such as the position of the site on the landscape and the vegetation present. All this, in combination in various proportions, influence the kind of biota that dwells on or in the soil, which also contribute to the soil’s development. The intensity of each of these factors and their interactions 29
30
World Soil Resources and Food Security
determine the nature of the product. Many of the factors are rhythmic with intensities varying with time. The changes may be diurnal and/or seasonal and knowing the climatic endowments of the area may be predictable. The human factor in the development of the soil is intriguing and has been the subject of innumerable studies. Most such studies describe the dramatic or obvious changes or changes that are measurable with available devices. However, land use also induces changes that are generally imperceptible. This includes such features as thickness of surface organic rich horizons. Changes in intensity and direction of processes are seldom evaluated unless they result in recognizable features. Soil management requires fertilization and in some instances copious amounts of fertilizers are added to the soil. This alters the microenvironment of the topsoil having consequences on the biota and processes that operate in the soil. This aspect of soil development has not received the study it deserves; most such studies focus on the fate of the organisms or the response of the soil to management and less on the microenvironmental changes within the soil. In the absence of a better understanding of the changes, soil management will remain ad hoc and contribute less to the sustainability of the system. For most practical purposes, the soil at a point in the landscape is a definable entity. It has measurable properties and studies have established that these properties are the result of factors and processes of soil formation at that point in the landscape. By comparison of soils at different points on the landscape and for different kinds of landscapes, the role of the factors and processes can be elucidated. These observations and, in some instances, assumptions are relevant to understanding the global distribution of soils. Climate is generally the most important variable. Climate can determine the kind of soil or preclude the occurrence of some kinds of soils. The organic and base-rich Mollisols of the temperate regions are rare in the tropics and the Oxisols of the tropics are absent in the higher latitudes. Rock weathering and mineral alteration are also controlled by climate, with intensities and directions being determined by ancillary factors, including the nature of the rock. An appropriate climate map of the world provides a first assessment of different soil regions. In the early classification systems, tropical soils were a unique group of soils and climatic parameters were used to delimit them. During these early days, tropical soils were studied and mapped by persons from European institutions that had a separate section or division dedicated to such soils. However, the study and classification of these soils adhered to the principles used for temperate soils. The contributions of the Belgians [Tavernier and Sys 1965] working in Zaire, the French [Aubert 1958] working in the francophone countries, and the Portuguese [Botelho da Costa 1959] working in Angola provided a different dimension to the knowledge of these soils. This knowledge was synthesized with the advent of soil taxonomy. All classification systems then recognized the unique place of the Oxisols and also that there were equivalents of temperate soils in the tropics. In soil taxonomy [Soil Survey Staff 1999], the distinction between the two groups of soils was considered at the family level. The major concepts and principles of soil classification found a general agreement by the 1970s and subsequent to that, much of the effort was devoted to refining the system. An active soil survey program in the United States enabled a continuous
World Soil Resources: Opportunities and Challenges
31
testing of the system and the system was formally published in 1975. Although the U.S. Soil Survey was confident of the system for U.S. purposes, its utility around the world was yet to be tested. A project funded by the U.S. Agency for International Development, the Soil Management Support Services (SMSS) enabled this [SMSS Staff 1991]. SMSS not only collected soil samples from around the world, but also trained soil scientists in these countries in the use and application of the system. Knowledge of the soil resources of the world was facilitated by the Food and Agriculture Organization (FAO) of the United Nations in its Soil Map of the World Project [FAO-UNESCO 1976]. This monumental project provided the first understanding of global soil resources. Using an internationally agreed upon terminology and legend, it remains today the best assessment of global soil resources. This effort will not be replicated in the near future and will remain as the basic study of global soil resources for some time to come. Under the auspices of the International Soil Science Society, the World Reference Base for Soil Classification [FAO, ISRIC, and ISSS 1998] was initiated in the 1980s. Although some progress was made, due to lack of funds and general enthusiasm, it remains on the back burner. The unspoken reason may be that it was difficult to improve on the FAO legend or the U.S. soil taxonomy. Consequently, a steady state was reached in the development of classification systems and with only minor changes in the past 2 decades, and with a better appreciation of the systems by the proponents, we now have a reasonably good assessment of global soil resources. Unfortunately, this status was reached when funds for soil resource inventories were dwindling and in the past 2 decades, particularly in developing countries, there is no effort for updating their soil resource inventories. Soil survey organizations are nonexistent or nonfunctional in many developing countries, particularly in Africa. As a result soil information in most of these countries is at least 40 to 50 years old. The purpose of this discourse is to assess the current status of our understanding and knowledge of global soil resources and to evaluate areas of knowledge gaps and consider research priorities. In the past few decades, land use and the demands on land have changed, even in developing countries. However, with modern technology, dependence on forces of nature for production is considerably less, having been circumvented by appropriate management techniques. Technology has aided the economies of water management, pest control, and sustainability in general. Modern agriculture is akin to an automated factory churning out products tailor-made to specifications dictated by consumers. Streamlining and enabling this process to proceed in an unabated manner is the challenge of modern agriculture. Although this is the trend, the whole spectrum from primitive slash-and burn systems to modern plantations occurs in the world. The continuing challenge is to sustainably use the land resources to feed and clothe the current population, without jeopardizing the opportunities for future generations. The challenge is how to maintain the quality of the resource base and even enhance the quality for future generations. Sustainable systems are based on sedentary agriculture and so the challenge is one of making optimal use of the land with a built-in flexibility to make changes as necessity dictates.
32
World Soil Resources and Food Security
3.2 GLOBAL DISTRIBUTION OF SOILS There are tens of thousands of different kinds of soils throughout the world. Soils are classified into groups according to physical, morphological, and chemical properties. In soil taxonomy, soils are grouped into several categories, the highest being soil orders [Soil Survey Staff 1999]. The Global Soil Regions map (Figure 3.1) is a reclassification of the Food and Agriculture Organization of the United Nations Educational, Scientific and Cultural Organization (FAO/UNESCO) Soil Map of the World to soil taxonomy suborders and is aggregated to soil orders. As soil moisture regimes are used to define suborders, a global soil climate map was used with the FAO soil units to determine the best soil taxonomy equivalent. Differences between the FAO system and soil taxonomy are such that direct correlations were not always possible.
3.2.1 Alfisols Most Alfisols were or are forested, with moderate to high base saturation. Typically they have a light-colored surface layer over a horizon of silicate clay accumulation (argillic). The cooler Alfisols tend to form a belt between the grassland Mollisols and the Spodosols of the more humid climates. Where temperatures are warmer, they form a belt between the Aridisols and the older Ultisols and Oxisols. Along with Mollisols, Alfisols account for a major portion of soils that are used to grow crops in the world (Figure 3.2).
3.2.2 Andisols Soils formed on volcanic ash and cinders and having andic properties are distributed along the circum-Pacific belts and occur sporadically elsewhere. The Andisols have mineralogical compositions ranging from volcanic glass, short-range order minerals such as allophane and immogolite, and variable amounts of halloysite. This mineralogical association gives unique properties to such soils including a high phosphate fixing capacity, low cation retention, and a high water holding capacity. Many of these soils are found on volcanic slopes or are developed through the weathering of plateau basalts. These soils support a high human population density due to their general ease of cultivation and also because of the cool environment of the volcanic mountains, which is generally free of pests and diseases (Figure 3.3).
3.2.3 Aridisols Aridisols, as their name implies, are soils that do not have water available to mesophytic plants for long periods. During most of the time when the soil is warm enough for plants to grow, soil water is held at potentials less than the permanent wilting point, or it is salty, or both. There is no period of 90 consecutive days when moisture is continuously available for plant growth. The concept of Aridisols is based on the low availability of soil moisture for sustained plant performance. In areas bordering deserts, the absolute precipitation may be high but, due to runoff, a very low storage capacity, or both, the actual soil moisture regime is aridic (Figure 3.4).
World Soil Resources: Opportunities and Challenges
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
33
FIGURE 3.1 Global distribution of soil orders.
Suborder Aqualfs Cryalfs Ustalfs Xeralfs Udalfs TOTAL
Global Sq km Percent 1,029,202 0.79 2,531,270 1.94 6,024,432 4.61 892,711 0.68 2,678,021 2.05 13,155,636 10.07
Africa Sq km Percent 392,900 1.32 – – 2,647,065 8.89 83,898 0.28 165,063 0.55 3,288,927 11.04
FIGURE 3.2 Global distribution of Alfisol suborders.
Asia Sq km Percent 34,703 0.08 729,218 1.66 1,102,064 2.51 213,922 0.49 271,186 0.62 2,351,093 5.35
Australia/Oceania Sq km Percent 126,393 1.58 302 0.00 422,037 5.28 267,277 3.34 227,790 2.85 1,043,800 13.05
Europe Sq km Percent 3.90 373,737 12.11 1,161,705 4.07 389,963 1.35 129,190 5.23 501,670 26.66 2,556,265
South America Sq km Percent 56,410 0.32 1,507 0.01 1,028,205 5.81 28,166 0.16 675,208 3.82 1,789,497 10.12
North America Sq km Percent 45,058 0.21 638,538 2.95 435,098 2.01 170,257 0.79 837,103 3.86 2,126,054 9.81
World Soil Resources and Food Security
Map legend
34
Alfisol suborders
Map legend
Suborder Cryands Torrands Xerands Vitrands Ustands Udands Gelands TOTAL
Global Sq km Percent 0.19 250,515 0.00 1,494 0.02 31,763 0.21 280,698 0.05 61,848 0.21 274,747 0.05 61,924 0.74 962,989
Africa Sq km Percent – – 0.00 1,131 – – 0.00 1,087 0.04 11,503 0.11 32,606 – – 0.16 46,326
Asia Sq km Percent 0.13 56,743 0.00 164 – – 0.04 19,552 0.04 18,506 0.25 108,612 0.13 58,959 0.60 262,536
Australia/Oceania Sq km Percent 0.00 71 – – – – 0.37 29,429 0.01 1,072 0.40 31,787 – – 0.78 62,360
World Soil Resources: Opportunities and Challenges
Andisol suborders
Europe Sq km Percent 0.27 25,437 – – 0.13 12,668 0.00 368 0.02 1,897 0.01 618 0.00 10 0.43 40,999
South America Sq km Percent 0.41 71,834 0.00 77 0.01 2,043 0.28 49,183 0.06 11,182 0.46 81,854 0.01 2,216 1.23 218,390
North America Sq km Percent 0.44 96,429 0.00 122 0.08 17,052 0.84 181,079 0.08 17,688 0.09 19,269 0.00 739 1.53 332,378
35
FIGURE 3.3 Global distribution of Andisol suborders.
Suborder Cryids Salids Gypsids Argids Calcids Cambids TOTAL
Global Sq km Percent
1,035,988 1,286,956 679,776 4,677,772 4,886,894 2,919,308 15,486,694
0.79 0.98 0.52 3.58 3.74 2.23 11.85
Africa Sq km Percent 602 162,418 359,577 348,719 1,805,822 847,773 3,524,911
FIGURE 3.4 Global distribution of Aridisol suborders.
0.00 0.55 1.21 1.17 6.06 2.85 11.83
Asia Sq km Percent 399,899 995,337 318,366 1,107,287 2,184,015 1,106,447 6,111,353
0.91 2.27 0.73 2.52 4.97 2.52 13.92
Australia/Oceania Sq km Percent – 78,381 – 1,590,658 598,849 336,519 2,604,407
– 0.98 – 19.88 7.49 4.21 32.56
Europe Sq km Percent 105 7,851 1,833 64,119 1,110 29,038 104,055
0.00 0.08 0.02 0.67 0.01 0.30 1.09
South America Sq km Percent 159,648 32,595 – 501,895 102,844 430,966 1,227,947
0.90 0.18 – 2.84 0.58 2.44 6.94
North America Sq km Percent 475,735 10,373 – 1,065,094 194,254 168,565 1,914,021
2.19 0.05 – 4.91 0.90 0.78 8.83
World Soil Resources and Food Security
Map legend
36
Aridisol suborders
World Soil Resources: Opportunities and Challenges
37
3.2.4 Entisols The Entisols show little or no evidence of soil formation. They are most extensive on subrecent alluvial plains and valleys or on steep slopes where erosion is rapid. The rate of soil formation is reduced for several reasons. Generally, time has not elapsed since deposition of the material for soil-forming processes to act. In some of these soils, peraquic conditions prevail, where the soil is saturated with water during the whole year. The soil is permanently reduced, preventing cambic horizon formation. On steep slopes, rapid erosion results in shallow soils where weathered parent materials rest on hard rock (Figure 3.5).
3.2.5 Gelisols In areas where the mean annual soil temperature is less than 0°C, the soils are frozen for long periods of the year and thaw out during the short warmer spells. The freezing and thawing processes promote physical changes in the soil. If there is sufficient water and the warm period is long enough, vegetation establishes and organic matter accumulates on the soil. Organic-rich soils or peat develop. Due to low temperatures, these Arctic soils have unique features such as ice-lenses or a layer of ice underlying the soil (Figure 3.6).
3.2.6 Histosols Most soil classifications, including soil taxonomy, separate mineral soils from organic soils. Histosols are soils that consist dominantly of organic soil materials. They develop where the rates of organic matter accumulation exceed decomposition and removal. Most of these soils formed under saturated conditions where the soil was saturated or nearly saturated with water most of the year. These soils have been referred to as bogs, moors, peat, or mucks. To be farmed, most Histosols must be drained. Management of the water table depth is critical to their use. When drained, Histosols oxidize and subside, requiring further drainage (Figure 3.7).
3.2.7 Inceptisols The Latin word “inceptum” means beginning, and the central concept of Inceptisols is that of soils in the early stages of soil formation. The initial stages of soil formation are exemplified by several attributes that are the result of the presence or absence of certain processes. Soil formation on rocks consists of the weathering of the rock, which is essentially a geochemical process accompanied by soil-forming processes acting on the weathered products. In cool humid climates, the soil-forming process may be the accumulation of organic matter to give rise to a mollic or umbric epipedon. In warmer climates, cambic horizon formation takes place, which is expressed by clay formation or the release of iron to form a color B horizon (Figure 3.8).
Suborder Aquents Psamments Fluvents Orthents TOTAL
Sq km
Global Percent
108,730 4,447,280 3,055,908 15,834,436 23,446,354
0.08 3.40 2.34 12.12 17.94
Africa Sq km Percent
42,596 2,271,000 666,033 7,896,258 10,875,887
FIGURE 3.5 Global distribution of Entisol suborders.
0.14 7.62 2.24 26.50 36.51
Asia Sq km Percent 65,095 121,746 1,284,418 4,634,084 6,105,343
0.15 0.28 2.93 10.55 13.90
Australia/Oceania Sq km Percent – 1,251,498 8,945 761,496 2,021,938
– 15.64 0.11 9.52 25.28
Europe Sq km Percent – 21,864 342,825 299,360 664,049
– 0.23 3.57 3.12 6.92
South America Sq km Percent
1,039 781,172 631,074 1,122,408 2,535,694
0.01 4.42 3.57 6.35 14.33
North America Sq km Percent – – 122,613 1,120,830 1,243,443
– – 0.57 5.17 5.74
World Soil Resources and Food Security
Map legend
38
Entisol suborders
Map legend
Suborder Histels Turbels Orthels TOTAL
Global Sq km Percent
1,006,321 5,065,473 5,692,048 11,763,841
0.77 3.88 4.36 9.00
Africa Sq km Percent – – – –
– – – –
Sq km
Asia Percent
478,105 1,973,049 4,372,198 6,823,352
1.09 4.49 9.96 15.54
World Soil Resources: Opportunities and Challenges
Gelisol suborders
Australia/Oceania Sq km Percent – – – –
– – – –
Europe Sq km Percent 76,946 53,966 208,194 339,106
0.80 0.56 2.17 3.54
South America Sq km Percent – 23,278 55,256 78,534
– 0.13 0.31 0.44
North America Sq km Percent
451,269 3,015,180 1,056,400 4,522,850
2.08 13.91 4.87 20.87
39
FIGURE 3.6 Global distribution of Gelisol suborders.
Suborder Fibrists Hemists Saprists TOTAL
Global Sq km Percent
195,680 980,686 330,983 1,507,349
0.15 0.75 0.25 1.15
Africa Sq km Percent 17,078 17,078
FIGURE 3.7 Global distribution of Histosol suborders.
0.06 0.06
Asia Sq km Percent 136,211 274,558 249,310 660,079
0.31 0.63 0.57 1.50
Australia/Oceania Sq km Percent 1,125 1,125
0.01 0.01
Europe Sq. km Percent 205,451 53 205,504
2.14 0.00 2.14
South America Sq km Percent 10,353 25,204 35,557
0.06 0.14 0.20
North America Sq km Percent 59,468 489,200 39,338 588,006
0.27 2.26 0.18 2.71
World Soil Resources and Food Security
Map legend
40
Histosol suborders
Map legend
Suborder Aquepts Anthrepts Cryepts Ustepts Xerepts Udepts Gelepts TOTAL
Global Sq km Percent
3,656,927 450,275 2,598,182 2,230,450 682,562 4,102,039 6,043,091 19,763,526
2.80 0.34 1.99 1.71 0.52 3.14 4.63 15.13
Africa Sq km Percent 369,402 183,883 371 763,561 167,253 218,611 – 1,703,081
1.24 0.62 0.00 2.56 0.56 0.73 – 5.72
Sq km
Asia Percent
1,704,743 29,308 1,495,427 589,305 176,546 1,347,136 4,751,603 10,094,067
3.88 0.07 3.41 1.34 0.40 3.07 10.82 22.99
World Soil Resources: Opportunities and Challenges
Inceptisol suborders
Australia/Oceania Sq km Percent 2,764 – 8,672 14,793 261 252,312 – 278,801
0.03 – 0.11 0.18 0.00 3.15 – 3.49
Europe Sq km Percent 194,747 – 350,803 369,529 320,890 767,872 98,592 2,102,432
2.03 – 3.66 3.85 3.35 8.01 1.03 21.92
South America Sq km Percent 413,983 226,968 260,776 180,570 8,286 933,556 1,901 2,026,040
2.34 1.28 1.47 1.02 0.05 5.28 0.01 11.45
North America Sq km Percent 971,289 10,116 482,133 312,693 9,326 582,553 1,190,995 3,559,105
4.48 0.05 2.22 1.44 0.04 2.69 5.49 16.42
41
FIGURE 3.8 Global distribution of Inceptisol suborders.
Suborder Albolls Aquolls Rendolls Xerolls Cryolls Ustolls Udolls Gelolls TOTAL
Sq km
Global Percent
6,651 117,867 261,035 923,418 2,464,068 3,937,001 1,262,703 155,544 9,128,288
0.01 0.09 0.20 0.71 1.89 3.01 0.97 0.12 6.99
Africa Sq km Percent – – 396 74,795 – 4,423 – – 79,613
FIGURE 3.9 Global distribution of Mollisol suborders.
– – 0.00 0.25 – 0.01 – – 0.27
Asia Sq km Percent 6,651 34,861 62,635 256,378 1,593,058 948,629 131,230 150,766 3,184,208
0.02 0.08 0.14 0.58 3.63 2.16 0.30 0.34 7.25
Australia/Oceania Sq km Percent – – 11,831 58,158 – 21,622 25,465 – 117.076
– – 0.15 0.73 – 0.27 0.32 – 1.46
Europe Sq km Percent – 14,930 127,100 371,806 152,896 898,549 78,211 30 1,643,522
– 0.16 1.33 3.88 1.59 9.37 0.82 0.00 17.14
South America Sq km Percent – – – 719 51,060 408,051 527,354 1,470 988,654
– – – 0.00 0.29 2.31 2.98 0.01 5.59
North America Sq km Percent – 68,076 59,074 161,562 667,055 1,655,728 500,444 3,278 3,115,215
– 0.31 0.27 0.75 3.08 7.64 2.31 0.02 14.37
World Soil Resources and Food Security
Map legend
42
Mo l l i s o l s u b o r d e r s
World Soil Resources: Opportunities and Challenges
43
3.2.8 Mollisols To a large extent Mollisols are the breadbasket of the world—the prairies of the United States, the steppes of Russia, and the pampas of Argentina. Most Mollisols are cultivated; in fact, there are only limited areas in the world where they have not been cultivated. Mollisols may initially be farmed with no addition of fertilizers. However, to sustain the high yields of corn, soybeans, sorghum, and small grains of today, fertilizers must be used. Soil temperature and moisture are principally used to separate all but two (Albolls and Rendolls) of the seven suborders of Mollisols (Figure 3.9).
3.2.9 Oxisols Oxisols are reddish, yellowish, or grayish soils. They are most common on the gentle undulating surfaces of geologically old surfaces in tropical and subtropical regions. The most extensive areas of Oxisols are on the interior plateaus of South America, the lower portions of the Amazon Basin, significant portions of the Central African Basin, and important areas in Asia, northern Australia, and several tropical islands of the Pacific. Their profiles are distinctive because of the lack of obvious horizons. Their surface horizons are usually somewhat darker in color than the subsoil, but the transition of subsoil features is gradual (Figure 3.10).
3.2.10 Spodosols A black or reddish brown to strong brown subsoil (spodic) horizon is the primary identifying characteristic of a Spodosol. It is often overlain by a gray to light gray eluvial horizon. These distinctive and contrasting colors make Spodosols easy to identify, although there are always exceptions. The simple explanation for this horizon sequence holds that under cool, humid, or perhumid climates, organic acids from a litter layer leach amorphous mixtures of organic matter and aluminum with or without iron from the eluvial horizon and deposit them in the illuvial spodic horizon. Most Spodosols have formed under such conditions and thus are common in the northern latitudes where most of these soils are to be found (Figure 3.11).
3.2.11 Ultisols Ultisols are similar to Alfisols in that they have a subhorizon of clay accumulation but have few bases, especially at depth. Most Ultisols are acid, although some may have a high pH in the surface horizons due to vegetation recycling or aerosolic additions. From a process point of view, the ideal Ultisol has a subsurface horizon of clay enrichment due to clay translocation from the surface horizons. If the surface horizons are more than 40% clay, these soils—with textural changes with depth— are considered as Ultisols for practical purposes. If there is less than 40% clay in the surface horizons, then they are classified as Oxisols (Figure 3.12).
Suborder Aquox Torrox Ustox Perox Udox TOTAL
Sq km
Global Percent
321,973 31,179 3,114,685 1,166,678 5,233,036 9,867,552
0.25 0.02 2.38 0.89 4.01 7.55
Africa Sq km Percent 321,128 9,140 1,711,581 89,224 1,743,805 3,874,878
FIGURE 3.10 Global distribution of Oxisol suborders.
1.08 0.03 5.75 0.30 5.85 13.01
Asia Sq km Percent
– – 19,391 54,539 77,193 151,124
– – 0.04 0.12 0.18 0.34
Australia/Oceania Sq km Percent 845 4,134 53,604 9,129 24,668 92,379
0.01 0.05 0.67 0.11 0.31 1
Europe Sq km Percent – – – – – –
– – – – – –
South America Sq km Percent – 16,366 1,328,943 1,013,198 3,382,965 5,741,472
– 0.09 7.51 5.73 19.12 32.46
North America Sq km Percent – 1,539 1,166 589 4,405 7,699
– 0.01 0.01 0.00 0.02 0.04
World Soil Resources and Food Security
Map legend
44
Oxisol suborders
Map legend
Suborder Aquods Cryods Humods Orthods Gelods TOTAL
Sq km
Global Percent
167,311 2,574,197 57,145 648,889 1,110,554 4,558,096
0.13 1.97 0.04 0.50 0.85 3.49
Africa Sq km Percent – – – – – –
– – – – – –
Asia Sq km Percent 13,250 64,555 19,236 13,343 107,020 217,404
0.03 0.15 0.04 0.03 0.24 0.50
World Soil Resources: Opportunities and Challenges
Spodosol suborders
Australia/Oceania Sq km Percent – 14 32,708 28,502 – 61,224
– 0.00 0.41 0.36 – 0.77
Europe Sq km Percent
99,330 1,012,855 4,471 265,137 402,126 1,783,920
1.04 10.56 0.05 2.76 4.19 18.60
South America Sq km Percent – 14,648 652 413 – 15,713
– 0.08 0.00 0.00 – 0.09
North America Sq km Percent 54,731 1,482,126 78 341,494 601,408 2,479,836
0.25 6.84 0.00 1.58 2.77 11.44
45
FIGURE 3.11 Global distribution of Spodosol suborders.
Suborder Aquults Humults Udults Ustults Xerults TOTAL
Global Sq km Percent 1,284,545 380,494 5,539,787 3,339,721 18,664 10,563,211
0.98 0.29 4.24 2.56 0.01 8.08
Africa Sq km Percent 55,637 35,699 602,503 1,355,657 900 2,050,396
FIGURE 3.12 Global distribution of Ultisol suborders.
0.19 0.12 2.02 4.55 0.00 6.88
Asia Sq km Percent 256,980 177,677 3,065,910 717,548 2,632 4,220,748
0.59 0.40 6.98 1.63 0.01 9.61
Australia/Oceania Sq km Percent 75,833 25,339 87,712 70,823 – 259,707
0.95 0.32 1.10 0.89 – 3.25
Europe Sq km Percent – – 4,275 – – 4,275
– – 0.04 – – 0.04
South America Sq km Percent 691,509 – 858,411 1,058,014 13 2,607,947
3.91 – 4.85 5.98 0.00 14.74
North America Sq km Percent 204,586 141,780 920,975 137,679 15,120 1,420,139
0.94 0.65 4.25 0.64 0.07 6.55
World Soil Resources and Food Security
Map legend
46
Ultisol suborders
World Soil Resources: Opportunities and Challenges
47
3.2.12 Vertisols Vertisols are clay soils with deep, wide cracks on some occasions during the year and slickensides within 100 cm of the soil surface. They shrink when dry and swell when moistened. Vertisols make up a relatively homogeneous order of soils because of the amount and kind of clay that is common to them. In many countries where Vertisols are common, they are known by their local names; for example, gilgai soils (Australia), adobe (Philippines), sha chiang (China), black cotton soils (India), smolnitza (Bulgaria), tirs (Morocco), makande (Malawi), vleigrond (South Africa), and sonsosuite (Nicaragua) (Figure 3.13).
3.3 HISTORICAL TRENDS Soil science evolved from geology, chemistry, physics, and biology. This field of science has been considered to be an independent area with a strong focus on agriculture for the past century or so [Arnold 1983; Sumner 2000]. Research in agriculture is instrumental in the rapid development of soil science [Sposito and Reginato 1992]. Over the past 2 decades, food insecurity has become a major concern in developing countries. A compounded effect of this insecurity is seen in the decline in soil productivity leading to accelerated land degradation. In contrast, intensive development of the agricultural sector in developed countries has culminated in the advent of several environmental issues. Enhanced techniques in modern agriculture include monocropping, high input and nonlabor intensive agriculture, precision agriculture, and zero-tillage. Concomitant with increasing demands for food and declining soil productivity, the need to implement sustainable agricultural systems worldwide was well recognized about 3 decades ago. The concept of sustainable development was enlarged to include wise use of natural resources in 1992 in Rio de Janeiro [UNCED 1992]. It is interesting to note that many methods that are currently in use (e.g., soil erosion and conservation assessment) were developed after the World War II. Despite some improvements in these methods over the years, there appears to be little progress in terms of their use and adaptability to the current context. In modern times, the soil scientist plays an important ethical role in linking the farmers’ interests with those of the land quality in defining the productivity of the agroecosystem. In the 1980s, the focus of soil science research took a different turn. It was widely recognized at this time that technology used in temperate regions could not be directly transferred to tropical areas. Some of the challenges to soil scientists at that time were summarized by Khanwar [1982]: • Assessing the potential of soil resources for alternative uses • Optimizing agricultural productivity of the land under cultivation • Improving the efficiency of agricultural inputs such as water and fertilizer in combination with high yielding seeds to maximize yield • Curbing the acceleration of soil degradation due to poor management of resources • Rehabilitating degraded lands • Developing a monitoring system for early detection of declines in soil quality
Suborder Aquerts Cryerts Xererts Torrerts Usterts Uderts TOTAL
Sq km
Global Percent
5,441 17,148 98,024 893,583 1,769,861 382,611 3,166,668
0.00 0.01 0.08 0.68 1.35 0.29 2.42
Africa Sq km Percent 688 – 12,317 194,140 715,765 66,074 988,985
FIGURE 3.13 Global distribution of Vertisol suborders.
0.00 – 0.04 0.65 2.40 0.22 3.32
Asia Sq km Percent 3,322 – 46,569 64,698 599,270 110,332 824,190
0.01 – 0.11 0.15 1.36 0.25 1.88
Australia/Oceania Sq km Percent – – 19,905 586,074 242,591 14,360 862,929
– – 0.25 7.33 3.03 0.18 10.79
Europe Sq km Percent – 16,995 17,284 – 28,191 19,768 82,237
– 0.18 0.18 – 0.29 0.21 0.86
South America Sq km Percent 1,430 46 1,026 6,365 38,702 105,618 153,188
0.01 0.00 0.01 0.04 0.22 0.60 0.87
North America Sq km Percent – 107 923 42,306 145,342 66,460 255,138
– 0.00 0.00 0.20 0.67 0.31 1.18
World Soil Resources and Food Security
Map legend
48
Ver t i s o l s u b o r d e r s
World Soil Resources: Opportunities and Challenges
49
Khanwar [1982] pointed out that in industrialized countries, the focus was on maintaining high soil productivity while minimizing degradation. However, in developing countries, the focus was on improving productivity per unit of land per unit of time by enhancing the efficiency of water use and implementing high-input agricultural systems. Interestingly, Khanwar [1982] stressed the lack of interest in the management of forest lands, range lands, and grasslands and that, although priorities may vary, the challenges remain the same and, although the practices may differ, the principles of soil science remain be the same. Since the 1980s, there has been a steady shift in emphasis from soil science to the environment and ecology [Wild 1989; Yaalon 1993; Miller 1993; Gardner 1993; Bouma 1994; Warkentin 1994; Wilding 1994; Mermut and Eswaran 1997, 2001; Sparks 2000]. The Soil Science Society of America joined the American Geological Institute (AGI) in 1993 to further enhance the society’s role in geoscience [Landa 2004]. Understandably, there were some changes in priorities at the different time intervals.
3.4 CURRENT ISSUES 3.4.1 Opportunities Spatial and temporal analysis of soils presents an opportunity to strengthen the discipline of soil science by integrating several branches such as soil genesis, classification, mineralogy, physics, chemistry, biology, biochemistry, and micromorphology. Currently, there is also a need to establish good links between the various disciplines within earth science. This would undoubtedly create an avenue for more research in the areas of ethnopedology [Winkler-Prins and Sandor 2003], hydrogeophysics [Hubbard and Rubin 2002; Muller 2003], hydropedology [Lin 2003], and nanoscience [NSF 2002]. In addition, a holistic approach to soil science would improve the understanding of landscape evolution and increase awareness of the impacts of global climate change or economic globalization on social and ecological environment. Recently, Effland et al. [2009] have indicated that increased demands for effective communication of global soil resource inventories requires continued development of correlation matrices for cross-referencing international and national soil classification systems. A direct impact of this would be enhanced agrotechnology transfer and greater scopes of scientific research. The geoscience division and soil science in the twenty-first century have been noted to share a common agenda. An aspect, among others, of this agenda is to encourage a fundamental knowledge of earth systems that is more systematic, interdisciplinary, dynamic, and process-oriented and to identify early warning systems of natural hazards and resource degradations. Despite identifying this aspect as an opportunity, the very fact that a more holistic approach needs to be undertaken in understanding earth systems renders the process a modern day challenge.
3.4.2 Challenges The widening gap between geologists and soil scientists has been left unmonitored for over 4 decades. Problems associated with this can be seen in a steady decrease of
50
World Soil Resources and Food Security
the number of students enrolling in soil science programs at universities worldwide. This is further evidenced by the closure and renaming of soil science departments, purportedly to stay abreast of the latest developments and requirements of the general public. This lack of focus and leadership in steering the soil science community to greater heights has culminated in the profession earning a poor reputation among the public. Undoubtedly, funds for basic soil science research are shrinking fast and this is directly lined to drastic reduction in the number of scientific papers in highimpact journals. These issues indicate that, in the absence of proper and proactive strategies to revitalize the entire discipline, the long-term sustainability of this program remains very doubtful. On a slightly different note, it is worthwhile to reexamine the criteria for securing research grants, evaluation of research projects, and scientific recognition of the work done. There appears to be a tendency to recognize highly specialized works. This has a negative impact on young researchers who will be quietly discouraged from taking part in interdisciplinary groups. Changes in evaluation criteria will undoubtedly allow soil science to adapt better to current challenges.
3.5 FUTURE PERSPECTIVES Soil science has an important role in enhancing food security and addressing major issues of the twenty-first century such as climate change, water management (Figure 3.14), desertification, and biofuels. Holistic approaches have been advocated on countless occasions to foster stronger ties with all the disciplines. Such paradigm shifts are not restricted to research trends only, but also apply to undergraduate and graduate programs. A direct consequence of this is the strong need to raise the profile of the soil scientist and the profession. The future of soil science looks extremely bright, taking into account the need to increase the profile to meet current and future expectations.
FIGURE 3.14 Improper water management causes flooding, as well as degradation of soil resources.
World Soil Resources: Opportunities and Challenges
51
3.6 CONCLUSIONS This treatise shows very clearly that priorities have changed over the past 4 decades or so and continue to change as our expectations for the land and science change with time. The basic tenets of agrotechnology transfer appear to have been partially successful. The main reason for this partial success is the fact that technology developed in temperate countries was not really adaptable in tropical situations. In many cases, this mismatch in technology and, to a large extent, knowledge, still exists. To encompass a holistic approach to soil science, local conditions will have to be considered. Priorities will tend to vary depending on geographical locations as a result of influence by natural limitations apart from other challenges such as local politics. Administrative restrictions will set the precedence in determining the challenges in soil science for the future. Clearly this is an extremely challenging task, but not an unachievable one. Past successes show that a concerted effort to conduct research in a holistic manner has paid off very well. This, therefore, should be the norm in future research programs.
REFERENCES Arnold, R.W. 1983. Concepts of soils and pedology. In Pedogenesis and soil taxonomy: I. Concepts and interactions, 1st ed., L.P. Wilding, N.E. Smeck, and G.F. Hall, eds., 1–21. Amsterdam: Elsevier. Aubert, G. 1958. Classification des sols. Compte Rendu Reunion Sous-comite. Congo Brazzaville. Botelho da Costa, J.V. 1959. Ferralitic, tropical fersiallitic and tropical semi-arid soils: Definitions adopted in the classification of the soils of Angola. Third Inter-African Soils Conference, Dalaba. I:317–319. Bouma, J. 1994. Proposed activities to seize opportunities for soil science and its applications in the 21st century. Bull. Int. Soc. Soil Sci. 86:13–14. Effland, W.R., Eswaran, H., and Reich, P. 2009. Toward development of a global soil classification correlation matrix (GSCCM). 2009 International Annual Meeting ASA-CSSASSSA. Nov. 1–5, 2009. Pittsburgh, PA. FAO, ISRIC, and ISSS, 1998. World reference base for soil resources. Rome: FAO. FAO-UNESCO. 1971–1976. Soil map of the world. Rome: FAO. Gardner, W.R. 1993. A call to action. Soil Sci. Soc. Am. J. 57:1403–1405. Hubbard, S., and Rubin, Y. 2002. Study institute assesses the state of hydrogeophysics. Eos, Transactions American Geophysical Union 83(51):02–606. Khanwar, J.S. 1982. Managing soil resources to meet the challenges to mankind: Presidential address. 12th International Congress of Soil Science. New Delhi, India. February 8–12, 1982. New Delhi: Rekha Printers Pvt Ltd. Landa, E.R. 2004. Soil science and geology: Connects, disconnects and new opportunities in geoscience education. J. Geosci. Educ. 52:191–196. Lin, H. 2003. Hydropedology—Bridging disciplines, scales and data. Vadose Zone Journal 2:1–11. Mermut, A.R., and Eswaran, H. 1997. Opportunities for soil science in a milieu of reduced funds. Can. J. Soil Sci. 77:1–7. Mermut, A.R., and Eswaran, H. 2001. Some major developments in soil science since the mid1960’s. Geoderma 100:403–426. Miller, F.P. 1993. Soil science: A scope broader than its identity. Soil Sci. Soc. Am. J. 57(299):564.
52
World Soil Resources and Food Security
Muller, M. 2003. Opening doors for geophysics in soil science. American Geophysical Union 84(26):243. National Science Foundation (NSF). 2002. Report of the nanoscience workshop. Berkeley, CA, June 14–16, 2002. Soil Management Support Services (SMSS) Staff. 1991. Soil Management Support Services Final Report 1980–1990. Washington, DC: USDA-SCS. Soil Survey Staff. 1999. Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys. Ag. Handbook No. 436. Washington, DC: USDA-NRCS. Sparks, D.L. 2000. Soils. Geotimes 45:39. Sposito, G., and Reginato, R.J., eds. 1992. Opportunities in basic soil science research. Madison, WI: Soil Science Society of America, Inc. Sumner, M.E., ed. 2000. Handbook of soil science. Boca Raton, FL: CRC Press. Tavernier, R., and Sys, C. 1965. Classification of the soils of Congo. Pedologie (Ghent, Belgium) Special Issue No. 5:91–136. UNCED. 1992. United Nations Conference on Environment and Development. Rio de Janeiro: UN. Warkentin, B.P. 1994. The discipline of soil science: How should it be organized? Soil Sci. Soc. Am. J. 58:267–269. Wild, A. 1989. Soil scientists as members of the scientific community. J. Soil Sci. 40: 209–221. Wilding, L.P. 1994. Surficial studies: Soil science. Geotimes 39:13–14. Winkler-Prins, A.M.G.A., and Sandor, J.A., eds. 2003. Ethnopedology (Special Issue). Geoderma 111(3–4). Yaalon, D. 1993. Soil science in the eyes of the beholder: Better understanding of soil processes and pedology urged. Bull. Int. Soc. Soil Sci. 84:13–14.
Resources and 4 Soil Human Adaptation in Forest and Agricultural Ecosystems in Humid Asia S. Funakawa, T. Watanabe, A. Kadono, A. Nakao, K. Fujii, and T. Kosaki CONTENTS 4.1 Introduction..................................................................................................... 55 4.2 Clay Mineralogy and Its Relationship to Soil Solution Composition in Soils from Different Weathering Environments.............................................. 56 4.2.1 Soil Samples Used for the Analysis..................................................... 57 4.2.2 Mineralogy of the Silt and Clay Fractions........................................... 58 4.2.3 Chemical Composition of Soil Water Extracts....................................60 4.2.4 pH and Activities of Al–OH Species in the Soil Water Extracts........ 63 4.2.5 Transformation of 2:1 Type Clay Minerals..........................................64 4.3 Pedogenetic Alterations of Illitic Minerals Represented by Radiocesium Interception Potential in Soils with Different Soil Moisture Regimes............ 65 4.3.1 Study Soils...........................................................................................66 4.3.2 Quantitative Mineralogical Properties, Cs-Fixed Capacity, and RIP Values...........................................................................................66 4.3.3 Statistical Relationship between RIP and Mineralogical Properties............................................................................................. 67 4.3.4 Nature of Illitic Minerals in the Three Regions in Terms of the Frayed Edge Site.................................................................................. 68 4.3.5 Illite Vermiculitization and Soil Weathering...................................... 70 4.3.6 Weathering Sequence in Upland Soils of Humid Asia........................ 71 4.4 Regional Trends in the Chemical and Mineralogical Properties of Upland Soils in Humid Asia with Special Reference to the WRB Classification Scheme...................................................................................... 72 4.4.1 Soils Studied........................................................................................ 73
53
54
4.5
4.6
4.7
4.8
World Soil Resources and Food Security
4.4.2 Regional Trend in the General Physicochemical and Mineralogical Properties of the Soils.................................................. 73 4.4.3 Relationship between Clay Mineral Composition and Cation Exchange Capacity in the Soils........................................................... 77 4.4.4 Relationship between Clay Mineral Composition and Soil pH........... 77 4.4.5 Classification of the Soils Examined According to WRB and Its Relationship to Mineral Weathering Conditions................................. 79 Factors Controlling Potentially Mineralizable and Recalcitrant Soil Organic Matter in Humid Asia........................................................................ 81 4.5.1 Soils Studied........................................................................................ 82 4.5.2 Soil Properties and a Comparison of Land Use.................................. 83 4.5.3 Amounts of PMC and PMN and the Rate Constants for the Different Land-Use Systems................................................................84 4.5.4 Estimation of PMC and PMN by Stepwise Linear Regression Using the Five Identified Factors by Principal Component Analysis.... 84 4.5.5 Estimation of PMC and PMN by Stepwise Linear Regression Using Soil Properties Representing the Five Identified Factors.......... 86 4.5.6 Factors Controlling ROC and RON..................................................... 87 Comparative Study on Soil Fertility Status under Shifting Cultivation in East Kalimantan, Northern Thailand, and Japan in Relation to Dynamics of Readily Mineralizable SOM and Soil Acidity........................... 89 4.6.1 Description of Study Sites................................................................... 89 4.6.2 General Physicochemical Properties of the Soils Studied..................92 4.6.3 Factors Controlling the Amounts of C0 and N0...................................94 4.6.4 Relationships between Respective Factors and Topography or Land-Use Status in East Kalimantan..................................................97 4.6.5 Dynamics of Soil Fertility Status throughout a Land Rotation System of Shifting Cultivation in Northern Thailand......................... 98 4.6.6 General Consideration on Land Uses in Upland Soils in Respective Regions in Relation to Soil Fertility Status..................... 101 Comparative Study on Proton Budgets in Soils of Cropland and Adjacent Forest in Thailand and Indonesia................................................... 103 4.7.1 Study Plots......................................................................................... 104 4.7.2 Physicochemical Properties of Soils.................................................. 104 4.7.3 Carbon Stock and Flow..................................................................... 104 4.7.4 Soil Solution Composition................................................................. 109 4.7.5 Net Proton Generation and Consumption.......................................... 111 4.7.6 Comparison of Soil Acidification Processes in Forests and Croplands........................................................................................... 115 4.7.7 Evaluation of Cultivation-Induced Soil Acidification in Relation to Organic Matter Dynamics............................................................. 116 Quantitative Analysis of Organic Matter Dynamics under Shifting Cultivation Systems in Northern Thailand with Special Reference to Functions of the Soil Microbial Community................................................. 117 4.8.1 Description of Study Sites................................................................. 117
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
55
4.8.2 Field Measurements of SOM Budgets in Different Stages of Land Use under Shifting Cultivation................................................. 118 4.8.3 In Situ Soil Solution Composition under Shifting Cultivation.......... 127 4.8.4 Fluctuation of Microbial Biomass and Metabolic Quotient, qCO2.................................................................................................. 127 4.8.5 Substrate-Induced Microbial Activities (Respiration and N Assimilation/Nitrification) by Short-Term Laboratory Incubation .. 129 4.8.6 Dynamics of Microbial Activities during Different Stages of Shifting Cultivation and the Function of the Fallow Phase............... 131 4.8.7 Main Functions of the Fallow Phase in Shifting Cultivation by Karen People in Northern Thailand.................................................. 133 4.9 Factors Controlling Soil Organic Matter Decomposition in Small Home Gardens in Different Regions of Indonesia................................................... 134 4.9.1 Description of Study Sites................................................................. 134 4.9.2 Field Measurement of Soil Respiration............................................. 137 4.9.3 Laboratory Incubation to Determine Soil Respiration Rate with Microbial Origin................................................................................ 140 4.9.4 Extensive Survey to Determine Distribution Patterns of SOM- Related Properties of Soils in Java and East Kalimantan........ 143 4.9.5 SOM Dynamics in Soils Situated in Different Climatic and/or Geological Conditions....................................................................... 147 4.9.6 Possible Land Management Systems in Different Regions of Java and East Kalimantan.................................................................. 150 4.10 General Discussion and Conclusion.............................................................. 150 References............................................................................................................... 152 Appendix—Analytical Methods............................................................................. 161 4.A1 Composition of Soil Water Extract for Thermodynamic Analysis............... 161 4.A2 Quantification of the Frayed Edge Site Using Radiocesium Interception Potential (RIP) Methodology......................................................................... 162 4.A3 Theoretical Calculation of Water Fluxes, Soil Acidification Rates, and Net Proton Generation................................................................................... 163 4.A3.1 Water Fluxes................................................................................... 163 4.A3.2 Soil Acidification Rate and Net Proton Generation........................ 164 4.A4 Determination of Potentially Mineralizable Carbon (PMC) and Nitrogen (PMN)............................................................................................. 166
4.1 INTRODUCTION Soils in humid Asia exhibit relatively incipient mineralogical characteristics because of the dominant steep slopes, crust movement, and volcanic activity on young alpine fold belts [FAO 2001] compared with soils developed on stable plains associated with the Precambrian shield in eastern South America or equatorial Africa. Among these continents, geological components are also different; distribution of mafic (or basic) parent materials is limited and sedimentary rocks as well as igneous felsic rocks
56
World Soil Resources and Food Security
are major parent rocks in humid Asia [Geological Survey of Japan 2004], whereas some metamorphic or basic rocks are dominated in the tropics of other continents [FAO/Unesco 1971, 1977]. Because of these essential differences in topographical and geological conditions, dominant upland soils in humid Asia are Ultisols, which is quite contrasting with the humid tropics in other continents where Oxisols are predominant [Soil Survey Staff 1999]. Since much of the information relating to the soils and agriculture in humid tropics has been derived from the achievements in Africa and Latin America [e.g., Mohr et al. 1972; Sanchez 1976], we believe that the actual situation of soil distribution and ecological processes both in natural and agricultural ecosystems in humid Asia are worthwhile to be introduced as integrated information when considering possible solutions as well as countermeasures against land degradation presently observed in the area. The present work comprises two parts. The first part (Sections 4.2 to 4.4) highlights unique clay transforming processes in the upland soil of humid Asia (Indonesia, Thailand, and Japan), which are considered to be affected strongly by both geological and climatic conditions. The pedogenetic alterations of illitic minerals under different soil moisture regimes (ustic or udic) are the main topics discussed in these sections. In the studied regions, sedimentary and felsic igneous rocks are predominantly distributed and fates of dioctahedral mica (illite) are essential to determine soil physicochemical properties such as soil acidity and cation exchange capacity (CEC). In the second part (Sections 4.5 to 4.9), on the basis of information of the previous part, conventional shifting cultivation and upland farming in different regions of Thailand and Indonesia are comparatively analyzed with special reference to the dynamics of soil organic matter (SOM) and soil acidity. There are several differences between shifting cultivation systems on Alisols under rainforest climates and those on Acrisols under monsoon climates, which can closely be linked with the ecosystem and pedogenetic processes and resulting soil properties in the regions. The main conclusion is given that, under low-input managements, subsistence agriculture must be strongly regulated by the respective soil-ecological conditions and, in turn, it could be represented as human adaptation to respective ecological conditions.
4.2 C LAY MINERALOGY AND ITS RELATIONSHIP TO SOIL SOLUTION COMPOSITION IN SOILS FROM DIFFERENT WEATHERING ENVIRONMENTS To assess inherent soil fertility for appropriate large-scale land management, it is important to understand how clay mineralogy relates to geological and weathering conditions and how physicochemical properties are affected by the clay mineralogy of soils. The neoformation of gibbsite, kaolin minerals, and smectite, and the transformation of mica are important processes in the formation of clay minerals. Neoformation is mainly controlled by H 4SiO 04 activity. Gibbsite forms under conditions of strong desilication where H 4SiO 04 activity is low [Huang et al. 2002]. Kaolin minerals form under moderate H 4SiO 04 activity conditions and smectite under high activity [Reid-Soukup and Ulery 2002]. Mica, which is commonly present in felsic
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
57
and sedimentary rocks, weathers to vermiculite and smectite, with a decrease in the layer charge and release of alkaline metals. The increased resistance to weathering of dioctahedral mica means that dioctahedral vermiculite is more common in soils than trioctahedral vermiculite [Malla 2002]. Among the processes, the intensity of vermiculitization of dioctahedral mica (or illitic minerals) is one of the key processes when considering the influence of mineralogy on soil physicochemical properties since the presence of expandable 2:1 minerals strongly affects soil properties such as CEC. In this section, therefore, unique clay transforming processes in upland soils of humid Asia (Indonesia, Thailand, and Japan) are comparatively analyzed relating to climatic conditions and resulting soil moisture regimes. The distribution of clay minerals in the upland soils of humid Asia, namely Japan, Thailand, and Indonesia, has been described previously. In general, hydroxy-Al interlayered vermiculite (HIV), mica, and kaolin minerals dominate in Japan [Matsue and Wada 1989; Araki et al. 1990], while kaolinite, mica, and smectite are dominant in Thailand [Yoshinaga et al. 1995; Yoothong et al. 1997; Kanket et al. 2002], and kaolinite, vermiculite, and smectite are the primary clay minerals found in Indonesia [Goenadi and Tan 1988; Koch et al. 1992; Supriyo et al. 1992a; Prasetyo et al. 2001]. However, each of the above regional reports fails to compare the formation of local clay minerals with formative processes in other regions. In addition, thermodynamic data on soil solution is lacking for humid Asia, and little emphasis has been placed on thermodynamic analysis of soil–water relationships. Soil solution data have been used successfully in the past to quantitatively predict and explain the distribution of clay minerals in soil and explain weathering trends [Kittrick 1973; van Breemen and Brinkman 1976; Karathanasis et al. 1983; Rai and Kittrick 1989; Norfleet et al. 1993; Karathanasis 2002]. Such an approach may help to understand the distribution of different clay minerals throughout humid Asia. The objective of the present section is, therefore, to explain the distribution of clay minerals in humid Asia based on the composition of soil water extracts, as affected by climate, parent material of the soil, and soil age.
4.2.1 Soil Samples Used for the Analysis The regions considered in this study are Japan, Thailand, and the Java, Sumatra, and East Kalimantan regions of Indonesia. Japan is part of a volcanic belt, and contains widely distributed felsic igneous and sedimentary rocks and tephra. In northern and northeast Thailand, sedimentary and felsic igneous rocks are most common. Java and Sumatra are part of a volcanic belt and consist mostly of tephra, andesite, and sedimentary rocks; felsic rocks are only found locally. Most of East Kalimantan is covered with sedimentary rocks where there are no active volcanoes; soils are more weathered on stable rolling landscapes. All the above regions have a humid climate, although Thailand has a distinct dry season with lower annual precipitation. Temperatures in Indonesia and Thailand are higher than those in Japan; this strongly favors weathering. Soil temperature and moisture regime are therefore described as mesic/thermic-udic in Japan, hyperthermic-udic in Indonesia, and hyperthermicustic in Thailand (Figure 4.1). Among these, a moisture deficiency during the dry
58
World Soil Resources and Food Security
Interfrost 40ºN
Ustic Aridic or permafrost
JPN
JPS Udic or perudic
THH Ustic
20ºN
THL
Udic or perudic
IDH
0º
IDL
100ºE
120ºE
FIGURE 4.1 Study area with soil moisture regimes.
season is only expected in Thailand (Figure 4.2). Considering the geology of each region, we collected upland soils from subsurface horizons at 204 sites (37 in Japan, 90 in Northern Thailand, one in Northeastern Thailand, 28 in Java, nine in Sumatra, and 39 in East Kalimantan) for identification of clay minerals. We also selected 53 soils from 38 of the 204 sites with the representative clay mineral composition and various parent materials for each region, for the analysis of the composition of soil water extracts. Soil samples were divided into four groups: Japan (JP1-JP13), Thailand (TH1-TH14), sedimentary rock areas of Indonesia (ID-S1-ID-S10), and volcanic areas of Indonesia (ID-V1-ID-V7). Indonesian samples were subdivided on the basis of parent material, as the volcanic rocks are usually andesitic or mafic and therefore quite different from the typically felsic sedimentary rocks. We use the following abbreviations for samples from Indonesian sites: JV, Java; SM, Sumatra; and EK, East Kalimantan. For JP and TH, parent rocks were mostly felsic igneous and sedimentary rocks.
4.2.2 Mineralogy of the Silt and Clay Fractions X-ray diffractograms of the silt fraction indicate that mica is usually absent in soils with parent materials of andesitic or mafic volcanic ejecta or andesite that are characteristic of ID-V, gabbro or limestone, while mica is present in soils with parent materials of sedimentary or felsic igneous rock (granite and rhyolite), characteristic
Soil Resources and Human Adaptation in Ecosystems in Humid Asia 300
250
250
200
200
150
150
100
100
50
50
0
0
300
Chiang Mai (Thailand)
Ja n Fe b M a Apr Mr ay Ju n J Auul g Se p O c N t ov D ec
Kyoto (Japan)
Ja n Fe b M a Apr Mr ay Ju n J Auul g Se p O c N t o Dv ec
Monthly average (mm)
300
59
Balikpapan (Indonesia)
200 150 100 50 0
Rainfall Potential evapotranspiration Ja n Fe b M a Apr Mr ay Ju n J Auul g Se p O c N t o Dv ec
Monthly average (mm)
250
FIGURE 4.2 Monthly distribution of rainfall and potential evapotranspiration for the three regions: southwestern Japan (Kyoto), northern Thailand (Chiang Mai), and East Kalimantan of Indonesia (Balikpapan). The precipitation data were derived from the Global Historical Climatology Network, version 1 (GHCN 1), and the potential evapotranspiration was estimated with the Thornthwaite method [Thornthwaite and Mather 1957].
of JP, TH, and ID-S (Figure 4.3). In some of soils from Java, smectite is dominant in the silt fraction; this may be inherited from the parent rocks. For the clay fraction from all regions, mica and vermiculite are absent when a mica peak at 1.0 nm was not detected in the silt fraction (ID-V6(SM) in Figure 4.3). In contrast, when mica was present in the silt fraction, mica, vermiculite, and HIV were common in the clay fraction. Mica was most abundant in 2:1 clay minerals in TH, with vermiculite most common in ID-S, and HIV in JP (Figure 4.4). Interlayered materials between the 2:1 layers were generally more common in JP than in ID-S, as the peak tended to remain with the heat treatment. In JP and ID-S, the relative peak intensity of 1.4–1.0 nm was larger in the clay than in the silt fraction, suggesting that 1.4 nm clay minerals
60
World Soil Resources and Food Security
ID-V6(SM) Bw (Andesitic-basaltic volcanic breccia)
TH1 Bt (Granite)
Clay, Mg
Clay, Mg Silt, Mg
Silt, Mg 1.42 1.0 0.72 d-space (nm)
0.485
JP1 BC (Granite)
1.42 1.0 0.72 d-space (nm)
0.485
ID-S7(EK) Bt (Sandstone-mudstone) Clay, K-550ºC
Clay, K-550ºC
Clay, K-25ºC
Clay, K-25ºC
Clay, Mg Clay, Mg Silt, Mg Silt, Mg 1.42 1.0 0.72 d-space (nm)
0.485
1.42 1.0 0.72 d-space (nm)
0.485
FIGURE 4.3 X-ray (Cu–Kα) diffractograms from oriented specimens of silt and clay fractions. Parent materials are shown in parentheses.
formed from mica (Figure 4.3). A broader peak at 0.7 nm among the JP and ID-V samples indicates the lower crystallinity of kaolin minerals compared to ID-S and TH samples. Our clay mineral composition results are generally consistent with previous reports for each region [Goenadi and Tan 1988; Matsue and Wada 1989; Araki et al. 1990; Koch et al. 1992; Supriyo et al. 1992a; Yoshinaga et al. 1995; Yoothong et al. 1997; Prasetyo et al. 2001; Kanket et al. 2002].
4.2.3 Chemical Composition of Soil Water Extracts The pH of soil water extracts from JP and ID-S samples is low, mostly within the range 4.3 to 5.5; however, the pH is relatively high in TH and ID-V soils (pH 5.4– 6.5). The pH of the soil water extracts was generally higher than that determined for soil suspensions, presumably due to the gradual dissolution of minerals that occurs during the 1-week preparation of the soil water extracts. The activities of Al–OH species (Al3+, Al(OH)2+, Al(OH)+2 , Al(OH)30 , and Al(OH) –) are higher in JP than in
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
61
Kaolin minerals 0 10
100 80
30
70
40
(%)
JP TH ID-S ID-V
90
20
60
50
50
60
40
70
30
80
20
90
10
100 Mica
0
(%)
10
20
30
40
50
60
70
80
90 100
(%)
0 1.4 nm minerals
FIGURE 4.4 Clay mineralogical composition of the soils determined based on relative peak areas of kaolin minerals, mica, and 1.4-nm minerals in x-ray diffractograms.
ID-S samples, although the pH of extracts from both regions is low (Figure 4.5). The activities of Al–OH species in TH and ID-V samples are low; in some extracts, the Al concentration is below the detection limit, due to higher pH. The H 4SiO 04 activity of the soil water extracts was highest in JP (approximately 10−3.1−10−3.8 mol L−1) and ID-V (approximately 10−3.2−10−3.9 mol L−1) samples, followed by ID-S (approximately
JP TH ID-S ID-V Gibbsite Amorphous Al(OH)3
log(Al total)
–2
–4
–6 ID-S2(JV)
–8
3
4
5
pH
6
7
ID-S1(JV)
8
FIGURE 4.5 Sum of activities of Al–OH species (Al3+, A1(OH)2+, Al(OH)+2 , Al(OH)30 , Al(OH)−4 ) in soil water extracts with gibbsite and amorphous Al hydroxide solubility lines.
62
World Soil Resources and Food Security JP
ID-S
4
Muscovite
2 0
–4.0
–2.5
–2 –4.5
–4.0
–3.5
–3.0
ID-V 6
pH + log(K+)
Muscovite
–2 –4.5
–4.0
TH Microcline
Gibbsite
Kaolinite
log (H4SiO4)
2 0
Gibbsite
log (H4SiO4)
6 4
–3.0
ID-S1(JV)
2
Kaolinite –3.5
Microcline
Muscovite
4 ID-S2(JV)
0
Gibbsite
–2 –4.5
pH + log(K+)
6
Microcline
pH + log(K+)
pH + log(K+)
6
–3.5
log (H4SiO4)
–3.0
–2.5
Microcline
4
Muscovite
2 0
Kaolinite
–2.5
Gibbsite
–2 –4.5
–4.0
Kaolinite –3.5
log (H4SiO4)
–3.0
–2.5
FIGURE 4.6 Composition of soil water extract from JP, TH, ID-S, and ID-V plotted on a stability diagram.
10 −3.5−10 −4.0 mol L−1) and lowest in TH (approximately 10 −3.6 −10 −4.3 mol L−1) [detailed data were given by Watanabe et al. 2006]. In the stability diagram (Figure 4.6), the compositions of TH extracts plot close to the line where both kaolinite and muscovite are stable, while JP and ID-S samples mostly plot within the field where kaolinite is stable and muscovite unstable; exceptions to this trend are ID-S1(JV), which is derived from limestone, and ID-S2(JV), which contains significant montmorillonite. In the solubility diagram (Figure 4.7), the compositions of JP and TH extracts mostly plot between the solubility lines of amorphous Al hydroxide and gibbsite, which indicates that gibbsite can precipitate. The ID-S samples and a few TH and ID-V samples plot under the kaolinite solubility line, indicating the dissolution of gibbsite and kaolinite.
63
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
Gibbsite
7 6 –4.5
Muscovite
Kaolinite
–3.5
ID-S2(JV) Smectite ID-S1(JV) Gibbsite
Muscovite
7
JP6 –4.0
Amrp. Al(OH)3
9 8
–3.0
Kaolinite
6 –4.5
–2.5
–4.0
log (H4SiO4) Amrp. SiO2 Amrp. Al(OH)3
9
Smectite
8 Muscovite 7 6 –4.5
Gibbsite
–3.5
ID-V
11 10 9 8
Amrp. Al(OH)3
–2.5
–3.0
–2.5
Smectite
Gibbsite Muscovite Kaolinite
7
Kaolinite
–4.0
–3.0
Quartz
TH
log(Al3+) + 3pH
log(Al3+) + 3pH
10
–3.5
log (H4SiO4)
Quartz
11
Amrp. SiO2
10
Amrp. SiO2
8
Smectite
Quartz
Quartz Amrp. Al(OH)3
ID-S
11
log(Al3+) + 3pH
log(Al3+) + 3pH
10 9
Amrp. SiO2
JP
11
6 –4.5
–4.0
log (H4SiO4)
–3.5
–3.0
–2.5
log (H4SiO4)
FIGURE 4.7 Composition of soil water extract from JP, TH, ID-S, and ID-V plotted on a solubility diagram.
4.2.4
pH and Activities of Al–OH Species in the
Soil Water Extracts
The pH and activities of Al–OH species such as Al3+ are very important when considering mineral behavior such as stability or the possibility of neoformation in soils. Extract pH appears to reflect precipitation, evapotranspiration, and geology. The difference between precipitation and potential evapotranspiration in each region was 808 mm in Japan, –390 mm in Thailand, and 457–2709 mm in Indonesia; these trends may explain the lower pH in JP and ID-S samples and the higher pH in TH samples. ID-V samples have a high pH despite intense leaching. This probably occurs because ID-V soils contain easily weathered minerals such as mafic minerals that can neutralize acidity when weathered rapidly; this will be apparent at high H 4SiO 04 activity in high pH extracts (Figures 4.6 and 4.7).
64
World Soil Resources and Food Security
Solubility diagrams (Figures 4.5 and 4.7) for JP and TH samples indicate that the activity of Al–OH species is controlled by the dissolution and precipitation of Al hydroxide, which has a crystallinity intermediate between gibbsite and amorphous Al hydroxide. In ID-S samples, the dissolution of kaolinite controlled lower activities. Al–OH species record a higher activity in JP samples than ID-S. As both regions are characterized by low pH, the difference in activity may reflect the freshness of JP soils.
4.2.5 Transformation of 2:1 Type Clay Minerals In the case of TH samples, both mineralogical and thermodynamic analyses indicate that mica is relatively stable and did not transform easily into 1.4 nm minerals (Figures 4.4 and 4.6). Araki et al. [1998] investigated the weathering of Tanzanian soils, which experience a distinct dry season similar to that in Thailand, and reported the weathering of mica without the formation of 1.4 nm minerals. The authors assumed that it takes a long time, perhaps several million years, for mica to weather to kaolinite or gibbsite. According to the stability diagram, both pH and K+ activity in solution are important for the weathering of mica. Mica is more stable than kaolinite with increasing pH or K+ activity (Figure 4.6). For example, Rausell-Colom et al. [1965] demonstrated that the removal of K from micas is strongly dependent on the K concentration of the solution. In our experiment, however, K+ activity in TH extracts is almost the same as that in JP and ID-S extracts, ranging from 10 –3.8 to 10 –4.8 mol L−1 (Figures 4.6 and 4.7). This indicates that pH plays an important role in the in situ weathering of mica. The apparent limits of pH(H2O) and pH of the soil water extract in terms of mica weathering to 1.4 nm minerals are 5.0–5.5 and 5.5–6.5, respectively, above which mica is stable. In contrast, mineralogical analysis of JP and ID-S samples show the formation of 1.4-nm clay minerals such as HIV and vermiculite from mica (Figure 4.3). The stability and solubility diagrams indicate the dissolution of mica, which is interpreted in this dissolution process as 2:1 opened layers forming 1.4 nm minerals such as HIV and vermiculite as transitional products (Figures 4.6 and 4.7). In JP samples, low crystalline clay minerals and easily weathered primary minerals dissolve and release Al to the soil solution; the released Al may precipitate as Al hydroxide contaminants between 2:1 layers as well as gibbsite. The 2:1 layers of HIV are thought to be stable, probably because the buffering capacity of interlayered materials is such that it can react with acid more easily than the 2:1 layers. Such a preferential dissolution of interlayered materials in 2:1 layers is widely observed in the surface horizons of acidic forest soils [Ross and Mortland 1966; Bouma et al. 1969; Gjems 1970; Hirai et al. 1989; Funakawa et al. 1993a]. In ID-S samples, small to negligible amounts of easily weathered minerals are present in soils from a tropical rainforest climate and relatively stable landscape conditions; this indicates limited potential for acid neutralization via mineral dissolution. Accordingly, interlayered materials and gibbsite may already have dissolved preferentially to 2:1 layers and vermiculite formed. Dissolution of 2:1 layers or a reduction in layer charge may result in the formation of smectite from vermiculite [Borchardt 1989].
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
65
4.3 P EDOGENETIC ALTERATIONS OF ILLITIC MINERALS REPRESENTED BY RADIOCESIUM INTERCEPTION POTENTIAL IN SOILS WITH DIFFERENT SOIL MOISTURE REGIMES X-ray diffraction (XRD) analysis normally is used to identify illitic minerals in a complex soil matrix. In some aspects, the nature of the minerals may differ from the pure illite formed in geological environments, but the XRD patterns at least indicate the presence of a K-bearing 2:1 phyllosilicate layer (i.e., an illitic layer). Computational disintegration analysis of the XRD patterns can yield a quantitative estimation of a soil’s illitic layer [Lanson 1997]; the total K content in soil clay also affords a reasonable estimate of their content, provided that K-feldspars are not present [Thompson and Ukrainczyk 2002]. Thus, we can determine the mass of illitic layers from the XRD pattern and total amount of K in clay. However, these conventional methods barely reveal the details of the nature of the illitic minerals, such as their degree of partial vermiculitization. In this section, we focused on the amount of frayed edge site as an indicator of changes in illitic minerals in soil. The weathering front in the interlayers of illitic minerals expands into a wedge shape, referred to as the frayed edge site [Sawhney 1972] (Figure 4.8). Illitic minerals reportedly enlarge the site by transformation to vermiculite [Le Roux and Rich 1969]; on the other hand, hydroxy-Al interlayering in the vermiculitic layers can reduce it at the site [Maes et al. 1999a]. The frayed edge site is characterized as having very high selectivity of Cs against hydrated cations [Brouwer et al. 1983]. Based on its selective-charge characteristics, Radiocesium interception potential (RIP) methodology was established to quantify the frayed edge site in soil [Cremers et al. 1988; Wauters et al. 1996]. The detailed methodology as well as experimental method for RIP measurement is given in the Appendix and Nakao et al. [2008].
1.0 nm
1.4 nm
K+ Hydrated cation Dehydrated cation Frayed edge site (FES) (RIP/KcFES(Cs-K)) FES + Vermiculitic expanded layer (Cs-fixed capacity)
FIGURE 4.8 Schematic illustration of the layer expansion of micaceous minerals.
66
World Soil Resources and Food Security
4.3.1 Study Soils Soils were collected from southwestern Japan (JP), northern Thailand (TH), and East Kalimantan in Indonesia (ID), representing, respectively, temperate, subtropical, and tropical climates. We used subsurface soils with B-horizon properties. Soils strongly affected by volcanic ashes, basaltic rocks, or ultramafic rocks were excluded because dioctahedral-type micas usually are uncommon in these types of parent materials. Figure 4.2 shows the monthly distributions of both rainfall and potential evapotranspiration. In total, 63 soils were sampled from 35 profiles in the three regions described above (JP, n = 21; TH, n = 19; ID, n = 23) and used for the RIP experiment as well as several soil characteristics including the acid-oxalate-extractable Fe (Feo) [McKeague and Day 1966], dithionite-citrate-bicarbonate (DCB) extracted-Fe (Fed) [Mehra and Jackson 1960], and total Fe (Fet) of the soils (for <2 mm fraction) and hot sodium citrate extracted-Al and Si (Alcit and Sicit, respectively) [Tamura 1958], total K content (TKclay), and Cs-fixed capacity of the clay fraction as indicators of interlayered materials of 2:1 minerals, clay mica content, and the vermiculitic layer [Komarneni and Roy 1980], respectively.
4.3.2 Q uantitative Mineralogical Properties, Cs-Fixed Capacity, and RIP Values Although JP soil clays usually showed a trace level 1.0 nm XRD peak after the Mg-25°C treatment compared to TH soils, as already shown in Figure 4.3, there was no significant difference in mean TKclay values between JP (27 g kg−1) and TH soil clays (33 g kg−1) and between those of JP and ID (20 g kg−1) clays (p < 0.05). This contradictive result suggests that the dominating 1.4 nm peak may overlap the 1.0 nm peak, or that the hydroxy-interlayered minerals in JP soils may occlude the illitic layers [Harris et al. 1992]. The mean kaolinite content of JP soil clays (119 g kg−1) was significantly smaller than that in the TH (409 g kg−1) and ID (482 g kg−1) soil clays, corresponding closely to the differences in the relative abundances of the 0.72 nm XRD peaks. This trend in kaolinite contents in temperate- and tropical-regions confirms common understanding [Velde 2001; Thiry 2000]. The mean Alcit in JP was significantly larger than that in TH and ID, suggesting that the HIV is more common in JP. However, unexpectedly, TH soil clays had a mean Alcit of 5.6 g kg−1, approximately half that of JP soil clays (11 g kg−1), despite the trace level of the 1.4 nm XRD peak. At the same time, the amounts of Sicit for both TH and ID were significantly larger than that for JP (p < 0.05). This might reflect the dissolution of some aluminosilicate during this extraction [Ndayiragije and Delvaux 2003]. For TH and ID, kaolinite is assumed to dissolve in the extraction. To avoid overestimating hydroxy-Al interlayering in the TH and ID samples, in the following statistical analysis we adopted the molar subtraction of Sicit from Alcit (i.e., Alcit – Sicit) in place of Alcit as the variable indicating the amount of hydroxy-Al polymers in the vermiculitic layers. The mean calculated Alcit – Sicit values for JP, TH, and ID, respectively, were 0.27, 0.04, and 0.01 mol kg–1. We note that this subtraction might underestimate the amount of hydroxy-interlayered minerals, as the interlayer spaces also might fix hydroxy-aluminosilicate in addition to hydroxy-aluminum [Sterte and Shabtai 1987].
67
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
The mean Cs-fixed capacity was highest for JP (7.5 cmolc kg–1) followed by ID (3.4 cmolc kg–1) and TH (1.3 cmolc kg–1) (p < 0.05), i.e., consistent with the relative abundances of the 1.4 nm XRD peaks. Therefore, it can be regarded as a representative value indicating the amount of vermiculitic layer charge. Thus, the Cs-fixed capacity and mineralogical properties described above agree with the mineralogical characteristics represented by XRD patterns of soil clays in each sampling region. We compared them with the RIP values in the following discussions. On the other hand, there was no significant difference in the mean values of RIP between JP (12.0 mol kg−1), TH (8.1 mol kg−1), and ID (13.4 mol kg−1). Even the minimum RIP values were larger than that of standard kaolinite (0.01 mol kg−1) and montmorillonite (0.1 mol kg−1) [Nakao et al. 2008], and those of organic soils (<0.02 mol kg−1) [Rigol et al. 2002]. Many soil clays in the present study have larger RIP values than standard illite [Montana illite, 12.6 mol kg−1; Delvaux et al. 2001].
4.3.3 Statistical Relationship between RIP and Mineralogical Properties Table 4.1 represents the Pearson correlation matrices for the mineralogical and Cs-adsorption (Cs-fixed capacity, RIP) indices. Because of the skewed distribution of RIP in each regional group, and TKclay and Cs-fixed capacity in JP, the variables TABLE 4.1 Pearson Correlation Coefficients for the Mineralogical Properties of JP, TH, and ID Soil Clays (JP)
log RIP
Kaolinite
1/TKclaya
Kaolinite 1/TKclaya Alcit − Sicit log(Cs-fixed)a DCB-oxides
−0.41 −0.59** −0.74*** 0.73*** −0.34
0.44* 0.46* −0.23 −0.40
0.55** −0.32 0.36
(TH)
log RIP
Kaolinite
Kaolinite TKclay Alcit − Sicit Cs-fixed DCB-oxides
−0.83*** 0.95*** −0.28 0.17 −0.38
Alcit – Sicit
log(Cs-Fixed)a
−0.85*** 0.50*
–0.49*
TKclay
Alcit − Sicit
Cs-Fixed
−0.77*** −0.06 −0.46* 0.01
−0.48* 0.06 −0.47*
0.42 0.57*
0.20
TKclay
Alcit − Sicit
Cs-Fixed
−0.26 −0.20
−0.02
(ID)
log RIP
Kaolinite
Kaolinite TKclay Alcit − Sicit Cs-fixed DCB-oxides
−0.59** 0.46 0.24 0.34 0.21
−0.87*** −0.21 0.16 −0.49
0.19 –0.19 0.40
Note: Significant at the *p = 0.05, **p = 0.01, and ***p = 0.001 level. a TK clay was reciprocated and Cs-fixed was transformed in the logarithm to normalize the data (JP).
68
World Soil Resources and Food Security
were normalized by either logarithmic- or reciprocating-transformation for the correlation analysis. In JP soil clays, the logarithm of RIP significantly correlated with Alcit − Sicit (r = −0.74), the logarithm of Cs-fixed capacity (r = −0.73), and 1/TKclay (r = −0.59). In TH soil clays, strong correlations existed between the logarithm of RIP and TKclay (r = 0.95) and the amount of kaolinite (r = –0.83). The two independent variables also were negatively correlated (r = −0.77). Similar to TH group, ID soil clays showed correlations between the logarithm of RIP and TKclay (r = 0.46) and the amount of kaolinite (r = −0.59), and negative correlations between the two independent variables (r = −0.87). There were three types of statistical relationships between the logarithm of RIP and the other variables: positive (TKclay and Cs-fixed capacity), negative (amount of kaolinite and Alcit − Sicit), and no correlation (amount of DCB-oxides). The negative correlations of the amount of kaolinite with the logarithm of RIP found for TH and ID reflect the enrichment of kaolinite in tropical–subtropical soils that directly decreases the proportion of illitic minerals in soil clays. The lack of a significant correlation between the logarithm of RIP and the amount of DCB-oxides in all sampling regions agrees with the findings of Dumat et al. [1997] showing that free oxides may not affect 137Cs selective adsorption on soil clays. Therefore, we do not take these two variables into account in the following discussion on alterations in illitic mineral in clays.
4.3.4 Nature of Illitic Minerals in the Three Regions in Terms of the Frayed Edge Site The absence of vermiculitic layers in TH soil clays was confirmed by the trace 1.4 nm XRD peak and the very small Cs-fixed capacity. The ustic soil moisture regime with a distinct dry season causes a higher soil pH and more K+ to be held in the solid phase than does the udic soil moisture regime [Mizota et al. 1988; Malavolta 1985], which may reduce the alteration of illitic minerals into vermiculite, as supposed in the previous section [Watanabe et al. 2006]. The strong positive correlation between the logarithm of RIP and TKclay observed for the TH soil clays indicates that the amount of the frayed edge site is related directly to the amount of illitic minerals. Therefore, it is reasonable to define the correlation line between the logarithm of RIP and TKclay in TH soil clays as being a no vermiculitization line, as depicted in Figure 4.9. The general understanding on Belgian and northwest European soils, belonging to the udic soil moisture regime [Boucneau et al. 1996], is that the amount of the frayed edge site is correlated with that of the vermiculitic layers [Delvaux et al. 2001]. Maes et al. [1999b] suggested the 137Cs-selective sites are not illitic sensu stricto, but vermiculitic layers associated with illitic interlayers in Belgian forest soils. We expect that the amount of the frayed edge site in highly vermiculitized illitic minerals might be related to the amount of the vermiculitic layers, rather than that of the illitic layers. The same may be true for the udic soils in humid Asia (i.e., JP and ID). They showed a weaker correlation between the logarithm of RIP and TKclay and a stronger one between the logarithm of RIP and Cs-fixed capacity than did the TH soil clays (Table 4.1). For ID soil clays, all logarithms of RIP values were
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
69
log (RIP) (mol kg clay–1)
2.0
1.5
1.0
0.5
0.0
JP TH ID 0
10
20
30
40
TKclay (g kg clay–1)
50
60
FIGURE 4.9 Scatter diagram of log RIP and TKclay. The solid line through the data for TH soil clays is a correlation line between log RIP and TKclay (r = 0.95, p < 0.001).
above the no vermiculitization line (Figure 4.9), and the mean Cs-fixed capacity was significantly larger than that for TH soil clays, although the simple correlation between the logarithm of RIP and Cs-fixed capacity was weak. Thus, apparently, in ID soils the formation of the frayed edge site through the vermiculitization of illitic minerals proceeds. As the low Alcit – Sicit and weak correlation between the logarithm of RIP and Alcit – Sicit indicate, the ID soil clays are not affected greatly by hydroxy-Al interlayering. The effect of hydroxy-Al interlayering on both illitic- and vermiculitic-layers is more important in JP than in ID soil clays. In the former, the logarithm of RIP correlated best and negatively with Alcit – Sicit, indicating that the frayed edge site is reduced with an increasing degree of hydroxy-Al interlayering. The hydroxy-Al polymers may well directly reduce the frayed edge site since some JP soil clays with a large Alcit – Sicit value were below the no vermiculitization line shown in Figure 4.9, despite there being a larger mass of vermiculitic layers in JP soil clays, represented by the Cs-fixed capacity, than in TH soil clays. Maes et al. [1999a] described such a process after investigating the decrease in magnitude of 137Cs fixation in vermiculitized biotite after artificially intercalating hydroxy-Al polymers. On the other hand, they reported the in situ transformation of illitic minerals through the release of hydroxy-Al polymers from the vermiculitic layer, as assessed in relation to 137Cs fixation in the weathering sequence from Dystric Cambisol to Haplic Podzol [Maes et al. 1999b]. The release of hydroxy-Al polymers through monomerization and complexation with organic acids can occur in highly acidic soils in southwestern Japan, as one of the representative mineralogical transformation processes occurring with podzolization [Funakawa et al. 1993a; Hirai 1995]. Pedogenetic alteration of illitic
70
World Soil Resources and Food Security
minerals in JP soils, associated with the frayed edge site, might be similar to that in soils from Belgium or northwest Europe.
4.3.5 Illite Vermiculitization and Soil Weathering The weak vermiculitization of illitic minerals in TH soil clays does not indicate that the soils are in a juvenile stage. The degree of their weathering can be assessed from the soil’s Fe status, that is, the activity ratio (Feo/Fed) and the crystallinity ratio (Fed – Feo)/Fet [Nagatsuka 1972; Arduino et al. 1986; Tsai et al. 2007]. Soil with lower Feo/Fed and higher (Fed – Feo)/Fet ratios is assumed to have been weathered more intensively. The TH soils showed relatively small Feo/Fed and large (Fed – Feo)/ Fet ratios mostly within the region of Feo/Fed < 0.4 and (Fed – Feo)/Fet > 0.5, namely the red soil region in Japan [Nagatsuka 1972]. In contrast, most JP and approximately one half the ID soils (with an elevation above 600 m) are within the region of Feo/Fed > 0.4 and (Fed – Feo)/Fet < 0.5 (the brown forest soil region), indicating that the TH soils are highly weathered relative to the latter soils. Therefore, the weak vermiculitization in TH soil clays is considered as the accumulative effect of soil development. Belgian forest soils, under well-drained conditions with an udic soil moisture regime, show prominent 1.4 nm XRD peaks [Brahy et al. 2000] that could be ascribed to the partial vermiculitization of illitic minerals. On the other hand, they have large Feo/Fed and small (Fed – Feo)/Fet ratios compared with the soils that we studied, suggesting, in terms of Fe status, that the Belgian soils are less weathered. ID soils from an elevation below 100 m have Feo/Fed and (Fed – Feo)/Fet ratios similar to those of TH soils, whereas the soils collected from around 600 m are within the region of brown forest soils, indicating that altitude largely affects the degree of soil weathering. However, it seems that the ID soil clays at both higher and lower altitudes exhibited more intense vermiculitization than that of TH soil clays. Thus, the pedogenetic alteration of illitic minerals may not correspond to the weathering stage of soils, but be strongly determined by whether the soil moisture regime is ustic or udic. The lack of relationship between soil weathering and pedogenetic alterations of illitic minerals may rest partly on the difference in the response to wet-dry repetitions. These periodic repetitions in the ustic soils enhanced weathering of primary silicates to secondary minerals (e.g., feldspar–kaolinite) [Inskeep et al. 1993]. On the other hand, wet-dry repetitions may enhance K fixation to the expanded 2:1 layers, rather than its release from illitic minerals, as shown in some laboratory experiments [Šucha and Širáńová 1991; Olk et al. 1995]. Since laboratory experiments cannot properly reproduce the in situ effect of wet-dry cycles on the pedogenetic alteration of illitic minerals, direct evidence is required, such as transformations of test illite inserted in soil in the respective regions, to clarify the effect of climate on illitic mineral alterations. Another factor might cause the RIP values in the ID soil clays and some of JP ones to be larger than expected from the amount of TKclay. Thus, chloriteto-vermiculite transformation may increase the proportion of the latter in the soil clays [Barnhisel and Bertsch 1989] and so contribute to the RIP value. However, there has been no attempt to determine whether chlorite, or vermiculite derived from it, has 137Cs selective sites like a frayed edge site. To attribute the regional
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
71
characteristics of the RIP values to pedogenetic alteration of illitic minerals, further investigations are required.
4.3.6 Weathering Sequence in Upland Soils of Humid Asia Based on the discussions so far, the mineral weathering sequence for each region is shown in Figure 4.10. In Thailand, under higher pH conditions associated with the ustic moisture regime, mica is relatively stable whereas other primary minerals such as feldspars are unstable and dissolve to form kaolinite and gibbsite. Under lower pH conditions of the udic moisture regime in Japan and Indonesia, mica weathers to form 1.4 nm minerals. Japanese soils are young and have much Al–OH species in soil solution, resulting in Al hydroxide between the 2:1 layers and gibbsite, while Al hydroxides and gibbsite have not formed or were already removed from the more highly weathered Indonesian soils. In soils derived from mica-free parent material, which for clay mineral compositions are kaolin minerals and smectite, the clay mineral composition changes to the kaolin apex along the mineral axis of the 1.4 nm minerals–kaolin. Based on the analysis of the frayed edge site as an indicator of the alteration of illitic minerals in the soil using RIP methodology, the details of vermiculitization of illitic minerals were traced for each of the regions studied, as illustrated in Figure 4.11. In TH soils, a small degree of vermiculitization was demonstrated by the direct correlation between the frayed edge site and the amount of illitic minerals, and the very small Cs-fixed capacity. Under an udic soil moisture regime, illitic minerals weather to form vermiculite, while in ID, such vermiculitization is directly reflected in the enlarging of the frayed edge site. In the case of JP, however, hydroxy-Al interlayering reduces the frayed edge site that forms through vermiculitization. The release of hydroxy-Al polymers from vermiculitic layers can induce the progressive reexposition of the frayed edge site under podzolization.
Kaolin minerals
Higher pH (Ustic)
TH
ID-S Mica-free Lower pH (Udic)
Mica
JP
1.4 nm minerals
FIGURE 4.10 Assumed mineral weathering sequence in upland soils in humid Asia.
72
World Soil Resources and Food Security Under ustic soil moisture regime
Thailand
Indonesia
Japan
Few expandable minerals are formed. Mica slowly weathers to kaolin or other secondary minerals. Under udic soil moisture regime Vermiculitization (i.e., formation of expandable 2:1 minerals) occurs due to intensive acid load to interlayered K+ Secondary Al hydrixides fill up the interlayer space along with vermiculitization and then 2:12:1:1 intergrades are formed.
FIGURE 4.11 Vermiculitization of illitic minerals under different moisture conditions.
4.4 R EGIONAL TRENDS IN THE CHEMICAL AND MINERALOGICAL PROPERTIES OF UPLAND SOILS IN HUMID ASIA WITH SPECIAL REFERENCE TO THE WRB CLASSIFICATION SCHEME Recently a new classification scheme of the world’s soils was proposed, the World Reference Base for Soil Resources (WRB) [ISSS-ISRIC-FAO 1998a, 2006]. One of the new ideas in the WRB is the introduction of clay activity (or cation exchange capacity [CEC]/clay) to classify soils in the highest category level (i.e., reference soil groups [RSG]), which is used to discriminate soils that have an argic horizon into Lixisols and Luvisols (high base saturation [>50%] soils) or Acrisols and Alisols (low base saturation [<50%] soils). Upland soils in humid Asia are predominantly acidic because of excessive precipitation and, hence, Acrisols and Alisols are in the major RSGs if argic horizons are recognized [ISSS-ISRIC-FAO 1998b]. The regional distribution patterns of these two soils remain, however, controversial in this region, and there is very little data on which to base any discussion. Although both Acrisols and Alisols are acidic and agricultural production is strongly restricted on these soils [FAO 2001], higher CEC/clay values in Alisols indicate that the levels of exchangeable and soluble Al achieve toxic quantities and cause more serious constraints for agricultural use in Alisols than in Acrisols. It is necessary to understand the regional distribution patterns of the two soils with reference to pedogenetic factors, such as geology and climate. As discussed in the preceding sections, different pathways of weathering of illitic minerals were postulated in our study. Since it is generally known that the presence of expandable 2:1 minerals strongly affects soil properties such as cation exchange capacity (CEC), regional distribution patterns of Acrisols and Alisols are possibly regulated by soil moisture regimes through difference in the mineral weathering sequences.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
73
4.4.1 Soils Studied A total of 186 B-horizon soils from upland soil profiles with minimum disturbance (i.e., under forest or cropland with low-input management), a majority of which is overlapped with the samples used in Section 4.2, were used for the analysis in this section. These soils are grouped into seven categories based on pedogenetic conditions (Table 4.2). According to Figure 4.1, a large part of East Asia is situated under either ustic, udic or perudic soil moisture regimes and, hence, the present study was considered to cover major soils in terms of soil moisture conditions. As soils derived from intermediate to mafic volcanic rocks and limestone are considered to exhibit different properties, judging from the results presented in a previous section, they are separated from the other soils and grouped as mafic. These soils were derived mainly from volcanic rocks in the Java and Sumatra Islands of Indonesia (ID-V soils in Section 4.2), and partly from limestone in northern Thailand and mafic intrusive rocks in Japan. The sample groups of IDL and IDH comprise soils derived from sedimentary rocks (mostly sandstone and mudstone) or felsic materials in low (<600 m) and high (>600 m) elevation areas of Indonesia, respectively, and equivalent to the ID-S soils in Section 4.2. Although both elevation areas would be classified as having a hyperthermic soil temperature regime according to the U.S. Taxonomy [Soil Survey Staff 2006], they are divided into two groups in the present study because the percolation of organic matter into the B-horizon soils is apparently more extensive at elevations above 600 m according to soil survey results, presumably because of the extremely humid conditions (i.e., perudic soil moisture regime) there. The THL and THH soils, equivalent to the TH soils in Section 4.2, were collected from Thailand, mostly from the northern mountainous region, which has a monsoon climate. The border of the two groups is approximately 800 m above sea level, above which there is predominately evergreen forest, presumably because of decreasing water stress during the dry season. The soil temperature regimes of THL and THH are hyperthermic and thermic, respectively. A wide variety of sedimentary rocks and granite make up most of the parent materials of these soils. The JPS and JPN soils, equivalent to the JP soils in Section 4.2, were collected from warm and cool temperate forests in Japan and, therefore, the soil temperature regimes of these soils were thermic and mesic. The parent materials are predominantly sedimentary rocks (predominantly mudstone with minor occurrences of sandstone and shale) and partly felsic igneous rocks. Soils strongly affected by volcanic ejecta were not included in the samples. General physicochemical properties of the soils collected were analyzed. The CEC derived solely from mineral components (CECmin) was calculated by eliminating the contribution of organic C [see details in Funakawa et al. 2008].
4.4.2 Regional Trend in the General Physicochemical and Mineralogical Properties of the Soils The average values of selected chemical and mineralogical properties of the B-horizon soils from different regions, with ANOVA results, are summarized in
74
TABLE 4.2 Average Values of Selected Chemical and Mineralogical Properties of the Soils from Different Regions Clay Mineral Composition pH(H2O) Sample Group
Number of Samples
AVE
Clay
SE
AVE
Total C
CECmin/Clay
a
SE
AVE
(%)
SE
AVE
(g kg−1)
Fed
SE
AVE
(cmolc kg−1)
1.4 nm
a
SE
AVE
(g kg−1)
1.0 nm
SE
AVE
(%)
0.7 nm
SE
AVE
(%)
SE
(%)
Sorted by regions and parent materials 39
4.58
0.07
a
38.6
2.1
ab
5.4
0.5
a
32.4
1.5
bc
20.7
1.5
a
37
3
b
7
1
ab
56
3
cd
IDH
10
4.52
0.06
a
34.0
2.3
ab
6.5
1.0
ab
45.2
2.9
d
22.1
3.1
abc
37
5
b
18
4
abc
45
4
bc
THL
40
5.44
0.07
c
53.0
2.3
c
9.6
0.5
bc
22.2
1.3
a
37.3
2.6
cd
4
1
a
31
3
c
66
3
de
THH
24
5.50
0.09
c
47.2
3.2
bc
14.6
2.0
c
25.7
1.4
ab
40.8
5.0
bcd
12
2
a
17
4
b
71
4
ef
JPS
30
4.71
0.04
ab
35.1
2.4
a
12.9
2.6
bc
40.0
2.1
d
23.6
2.6
ab
58
4
c
11
2
ab
31
3
b
JPN
20
4.59
0.05
a
47.1
1.9
bc
33.4
3.0
d
39.9
2.2
cd
34.0
1.9
bcd
80
3
d
9
2
ab
11
2
a
mafic
23
5.02
0.16
b
66.5
2.3
d
11.2
0.9
bc
31.4
2.9
bc
65.4
6.7
d
14
5
a
4
3
a
81
5
f
Sorted by reference soil groups Andosols
4
4.66
0.06
ab
50.9
1.7
ab
46.0
5.8
c
31.2
4.8
ab
30.4
2.2
a
83
3
c
9
3
ab
8
1
a
Podzols
6
4.38
0.09
a
43.7
2.1
ab
33.3
3.6
c
43.4
2.5
b
42.0
1.6
a
75
5
c
15
5
ab
10
1
a
Acrisols
32
5.28
0.09
b
56.4
2.6
b
9.4
0.7
ab
18.1
0.8
a
39.1
4.1
a
6
1
a
25
3
b
70
4
c
Alisols
64
4.76
0.06
a
45.7
1.8
a
6.8
0.5
a
33.2
1.1
b
31.6
2.5
a
33
3
b
9
1
a
57
3
bc
Luvisols and Lixisols
9
5.81
0.20
c
56.3
7.0
ab
9.9
1.3
ab
30.9
4.0
b
47.8
7.5
a
6
3
ab
15
7
ab
79
7
bc
Cambisols
71
4.95
0.07
ab
41.0
1.9
a
15.7
1.5
b
36.2
1.6
b
32.6
3.0
a
39
4
b
15
2
a
46
4
b
All
186
4.96
0.04
46.3
1.2
12.5
0.8
31.9
0.9
34.4
1.7
32
2
14
1
54
2
Note: The values with the same letters are not significantly different by Tukey test (p < 0.05). AVE, average; SE, standard error. a ANOVA was applied after logarithmic transformation for normalizing dataset.
World Soil Resources and Food Security
IDL
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
75
Table 4.2. Although clay mica was always the dominant component in the 1.0-nm minerals, judging from virtually no expansion of the 1.0-nm diffraction peak after glycerol solvation as well as a clear diffraction peak remaining after 550°C heating in our samples, there is a possibility that halloysite is a minor component. As shown in Table 4.2 and Figure 4.12, the majority of soils were acidic. Base saturation (by CEC at pH 7) is usually below 50%. Among the B-horizon soils, soils from the ustic soil moisture regime in Thailand show significantly higher pH(H2O) values. Some soils from mafic parent materials also exhibit higher base saturation above 50% (Figure 4.12). According to Table 4.2, Alisols and Podzols are strongly acidic, followed by Andosols, Cambisols, and Acrisols. As for clay content, the soils from Indonesia are less clayey than those from Thailand or from mafic parent materials. Total C content clearly increases as soil temperature decreases in the following order: IDL, IDH, THL (hyperthermic) < THH, JPS (thermic) < JPN (mesic). The values of CECmin/clay are plotted together with ECEC/clay (effective CEC [sum of exchangeable bases and Al] divided by clay content) or exchangeable Al/clay in Figure 4.13. The CECmin/clay of the soils derived from sedimentary rocks (excluding limestone) or felsic materials showed a clear regional trend; that is, it was usually higher than 24 cmolc kg−1 (corresponding to Alisols if the argic horizon is recognized) under the udic or perudic soil moisture regimes in Indonesia (IDL and IDH) and Japan (JPS and JPN), whereas it was predominantly lower than 24 cmolc kg−1 (corresponding to Acrisols) under the ustic soil moisture regime in Thailand (THL). The values of THH soils are occasionally higher than 24 cmolc kg−1, presumably because the more percolating soil moisture conditions in high mountains result in similar conditions to the udic soil moisture conditions. Most of the cation exchange sites represented by ECEC seem to be occupied by exchangeable Al (Figure 4.13b) and, therefore, the level of Al toxicity should be higher in the soils from the udic or perudic soil moisture regimes of Indonesia and Japan compared with those from the ustic soil moisture regime of Thailand. Similarly high CEC/clay values were reported for soils with an udic soil moisture regime, for example, 34.0 and 26.3 cmolc kg−1 in
Base saturation (%)
100
r = 0.70 (n = 186, p < 0.01)
80 60 40
IDL IDH THL THH JPS JPN mafic
Base saturation = 50%
20 0
4
5
pH(H2O)
6
7
FIGURE 4.12 Acidity and base saturation of the soils studied.
76
World Soil Resources and Food Security
ECEC/clay (cmolc kg–1)
(a)
CECmin/clay = 24 cmolc kg–1
60
IDL IDH THL THH JPS JPN mafic
40 20 0
0
20
40
CECmin/clay (cmolc (b) Exch. Al/clay (cmolc kg–1)
1:1
60
CECmin/clay = 24 cmolc kg–1
60
80
kg–1) 1:1 For Acrisols and Alisols y = 0.52x – 3.57 r2 = 0.48, n = 96
40 20 0
0
20
40
60
CECmin/clay (cmolc kg–1)
80
IDL IDH THL THH JPS JPN mafic
FIGURE 4.13 Characteristics of the cation exchange capacity (CEC) of the soils. CECmin/ clay, CEC value per unit clay content; ECEC/clay, effective CEC (sum of exchangeable bases and Al), divided by clay content.
subsoil layers of upland soils in northern Sarawak, Malaysia [Tanaka et al. 2005], or average values of 34.7 and 41.2 cmolc kg−1 for 25 and 17 upland soils, respectively, in southern Sarawak, Malaysia [Tanaka et al. 2007]. In contrast, for soils with an ustic soil moisture regime, lower values of CEC/clay were often reported. Toriyama et al. [2007] reported the CEC/clay values for subsoils of well-drained Acrisols in central Cambodia (60–100 m above sea level) to be approximately 10 cmolc kg−1. According to Watanabe et al. [2004], the clay mineralogy and CEC/clay values in upland soils in northern Laos (600–1100 m above sea level) were essentially similar to our results in that illite and kaolinite were dominant in the clay fraction, and the values of CEC/ clay of the B-horizon soils were almost equal to or lower than the threshold value (i.e., 24 cmolc kg−1). All these reports indicate that our results obtained in some limited regions in humid Asia can be applied to different countries with similar geological or climatic conditions. The relationship obtained between CECmin/clay and exchangeable Al/clay for Acrisols and Alisols (i.e., exchangeable Al/clay = −3.57 + 0.52 CECmin/clay) is close to that reported in FAO [2001] for soils in Indonesia, the Caribbean region, Rwanda,
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
77
Cameroon, Peru, and Colombia; that is, Al/clay = −1.28 + 0.64 CEC/clay. Thus, the overall properties of Al retention and perhaps of the proportion of permanent and variable negative charges of the soils in the present study are considered to be similar to those in Acrisols and Alisols from other regions. There is a unique characteristic of the soils derived from mafic parent materials compared to the soils from sedimentary rocks (excluding limestone) or felsic volcanic materials; these soils are more variable in pH (Figure 4.12) or CECmin/clay (Figure 4.13a). According to Figure 4.13b, the majority of the soils do not retain appreciable amounts of exchangeable Al, even when the value of CECmin/clay is high. One reason for this must be the relatively higher pH and base saturation for some soils (Figure 4.12). Another possible explanation is that variable negative charges derived from free oxide surfaces can contribute to the apparent increase of CEC at higher pH ranges (i.e., 7), and the actual CEC at the lower pH range at which Al can dominate is small. Variation in the chemical properties of soils derived from mafic materials (including limestone) in Java Island was also reported by Supriyo et al. [1992b].
4.4.3 Relationship between Clay Mineral Composition and Cation Exchange Capacity in the Soils It is widely known that clay mineralogical properties of soils strongly affect soil physicochemical properties. In fact, the CECmin/clay in the present study is influenced by the relative abundance of 1.4 nm minerals in the clay fraction (Figure 4.14a). The following equation is obtained:
CECmin/clay (cmolc kg−1) = 0.201 × (1.4 nm minerals in %) + 25.5
(r 2 = 0.24, p < 0.01).
(4.1)
This equation suggests that the 1.4 nm minerals contribute to a CEC increase of 20.1 cmolc kg–1. As most of the 2:1 minerals in the Japanese soils are modified by hydroxy-interlayered materials [Hirai 1995; Kitagawa 2005], the CEC values of these soils could be estimated to be lower than the possible negative charge derived from the 2:1 lattice structure [Funakawa et al. 2003]. On the contrary, among soils with the lowest amounts of expandable 2:1 minerals (e.g., Thai soils), the relative contribution of other mineral components increases, and the CECmin attributable to the 1.4 nm minerals should be appreciably lower than the value of the intercept of the equation, that is, 25.5 cmolc kg−1. As a result, the actual contribution of expandable 2:1 minerals to CEC may be considerably higher than 20.1 cmolc kg−1 clay. Thus, the CEC of the upland soils in humid Asia is primarily determined by the relative abundance of expandable 2:1 minerals, in which exchangeable Al is dominant as represented by the equation: Exch. Al/clay = –3.57 + 0.52 CECmin/clay.
4.4.4 Relationship between Clay Mineral Composition and Soil pH Soil pH is an important index that affects a wide range of agronomical and environmental functions of soils [Robson 1989]. In our present study, a negative correlation
78
World Soil Resources and Food Security (a)
80
CECmin/clay (cmolc kg–1)
r = 0.49 (n = 186, p < 0.01) 60 y = 0.201x + 25.5 r2 = 0.24
40
20 CECmin/clay = 24 cmolc kg–1 0
20
40
60
80
100
Relative abundance of 1.4 nm minerals (%)
(b) Relative abundance of 1.4 nm minerals (%)
0
IDL IDH THL THH JPS JPN mafic
100 80
r = –0.46 (n = 186, p < 0.01) IDL IDH THL THH JPS JPN mafic
60 40 20 0
4
5
6
7
pH(H2O)
FIGURE 4.14 Relationship between mineralogical composition and chemical properties of the soils. Relationships between relative abundances of expandable 1.4-nm minerals in clay fraction and (a) CECmin/clay and (b) pH(H2O) of soils.
is observed between the relative abundance of 1.4 nm minerals and soil pH (Figure 4.14b), and the following equation is obtained:
pH(H2O) = 5.06 – 0.00803 × (1.4 nm minerals [%]) + 0.00451 × (Fed [g kg–1])
(r 2 = 0.23, p < 0.01).
(4.2)
As most of the soils studied are acidic and base saturation is lower than 50%, the dominant cation retained on the cation exchange sites of soils is Al3+. The equation above indicates that the presence of expandable 2:1 minerals contributes to pH reduction by 0.008 units per clay percentage. In solution, the Al3+ ion behaves as a weak acid through a hydrolytic reaction: that is, Al3+ + OH− = Al(OH)2+ (pKa = 5.0), and soils rich in exchangeable Al show a distinct buffer zone against OH− addition during
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
79
titration [Funakawa et al. 1993b]. The strong acidity far below pH 5.0 of some soils must be affected by the presence of a stronger acid(s), that is, H+, although quantitative analysis of exchangeable H+ is difficult, and the selectivity coefficient between Al3+ and H+ ions on the permanent negative sites of soils is rarely determined. In contrast, the oxide surfaces represented by the Fed fraction are considered to increase soil pH by 0.00451 units per g Fed kg−1 soil. As is well known, the oxide surfaces act as weak acids that have a zero point of charge (ZPC) of approximately 6–9 [McBride 1989]. In lower pH regions, it might mitigate H+ in solution through the protonation reaction: M–OH + H+ = M − OH +2 , where M represents the metal ions (Fe, Al) at the surface of the soil particles, and thereby contributes to increased soil pH. Soils derived from mafic volcanic materials are often rich in the Fed fraction, which was originally inherent in the parent materials, and soil pH was relatively high (Table 4.2). In humid Asia, the major distribution of the mafic parent materials is generally limited to specific regions, such as the volcanic belts of Java and Sumatra Islands. In the humid tropics of other continents, for example, eastern Latin America or equatorial Africa, however, mafic materials cover wider regions. It can be said that acid mitigation by oxide surfaces is relatively limited in humid Asia, unlike in other continents. The general results obtained in the present study are considered to be specific for some regions of humid Asia, in which sedimentary rocks or felsic volcanic materials are the dominant parent materials of the soils.
4.4.5 Classification of the Soils Examined According to WRB and Its Relationship to Mineral Weathering Conditions Table 4.3 summarizes the classification of the soils examined according to WRB [ISSS-ISRIC-FAO 2006]. The RSGs to which mafic soils are classified as diverse, including Alisols, Acrisols, Luvisols, and Cambisols. More detailed surveys focusing on the relationship between parent materials or primary minerals and properties of the soils that developed there are necessary. Most of the members of IDL and IDH belong to Alisols according to WRB, except for those on relatively steep slopes in high mountains that are classified as Cambisols (dystric) because of a lack of a clear argic horizon. Higher soluble Al3+ in Alisols may be a more serious constraint for agricultural production compared with Acrisols. In contrast, Acrisols are dominant in THL, whereas the proportion of Alisols and Cambisols (dystric) increased in THH. In Japan, the majority of the soils are Cambisols (dystric) and some from JPS are Alisols. In JPN, some soils are classified into Andosols or Podzols because of the relatively high influence of amorphous materials under cool temperate climates and/or the presence of nearby active volcanoes. It should be noted that no extremely weathered soils such as Ferralsols or Plinthosols were found in the present study, although it is necessary to extend future surveys to whole regions of humid Asia, especially to the relatively stable plains of Northeast Thailand, and to central Cambodia. Thus, of the 186 soil samples, only nine and eight soils are classified into Luvisols (or Lixisols) and Cambisols (eutric), respectively, indicating that the upland soils in humid Asia are predominantly acidic. In the preceding sections, dioctahedral mica inherent to sedimentary rocks or felsic igneous rocks was thought to weather to form
80
TABLE 4.3 Summary of the Classification of the Soils According to the World Reference Base Number of Soils Classified into Different RSG Cambisols Sample Group
Number of Samples 39 10 40 24 30 20 23 186
Andosols
Podzols
Alisols
Acrisols
Luvisols (or Lixisols)
(Dystric)
(Eutric)
0 0 0 0 0 4 0 4
0 0 0 0 0 6 0 6
34 5 5 6 7 0 7 64
2 0 20 6 0 0 4 32
0 0 3 (1) 1 0 0 4 9
2 5 9 9 23 10 5 63
1 0 2 2 0 0 3 8
World Soil Resources and Food Security
IDL IDH THL THH JPS JPN mafic Total
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
81
expandable 2:1 minerals, that is, vermiculitization, under the lower pH conditions associated with the udic or perudic soil moisture regime. In contrast, mica is relatively stable under the higher pH conditions associated with the ustic soil moisture regime, whereas other primary minerals, such as feldspars, are unstable and dissolve to form kaolin minerals and gibbsite. These processes should be critically analyzed further considering the possibility of selective dissolution of interlayered K+ ion of mica under acidic conditions from a clay mineralogical viewpoint [see Fanning et al. 1989]. Our finding in the present study suggests that the regional distribution patterns of Acrisols and Alisols in humid Asia are strongly related to pedogenetic conditions through clay mineral formation. In turn, such a close relationship suggests that the soils in humid Asia are predominantly on the course of pedogenesis under the present bio-climatic conditions, probably unlike tropical soils on plains in other continents. Thus, it was observed that chemical and mineralogical properties of soils and associated ecosystem processes in humid Asia were largely different depending on geological and climatic conditions. In the following sections, soil fertility status is first analyzed in terms of soil acidity and SOM-related properties and then conventional cropping systems in different regions (e.g., shifting cultivation and home garden management), are comparatively evaluated from the viewpoint of human adaptation to respective soil environments under low-input conditions.
4.5 F ACTORS CONTROLLING POTENTIALLY MINERALIZABLE AND RECALCITRANT SOIL ORGANIC MATTER IN HUMID ASIA SOM has been traditionally regarded as one of the key components for maintaining the soil fertility of agricultural land. However, in relation to the problem of global warming, considerable attention has recently been focused on SOM dynamics as both a possible source and sink of carbon dioxide. In this context, very many attempts have been made to quantitatively simulate SOM dynamics using models such as CENTURY [Parton et al. 1987] and RothC [Jenkinson 1990], which have proved successful under certain environmental conditions. Generally these models were developed under mild climatic conditions such as a temperate climate with rather dry conditions, where steppe vegetation or rain-fed cereal cropping provides relatively stable C fluxes. It is, therefore, still questionable whether such simulations can directly apply to SOM dynamics under highly fluctuating conditions, which are very common in the humid tropics due to conversion of land use from forest to cropland or vice versa. Recently, models with conceptual SOM fractions have been improved by using functional SOM pools determined by physical and chemical fractionation methods [e.g., Shirato et al. 2004; Zimmermann et al. 2006]. These methods, however, do not yield homogeneous SOM pools with respect to the turnover rates [von Lützow et al. 2007]. One of the methods used to determine the turnover rate of the SOM pool is decomposition. Soils are incubated for several months under constant and favorable conditions for microbes. The released CO2 and/or mineral N are measured and fitted to first-order kinetics to calculate the rate constant and potentially mineralizable carbon (PMC) or nitrogen (PMN) [e.g., Stanford and Smith 1972]. To investigate the relationship between the potentially mineralizable OM and other soil properties,
82
World Soil Resources and Food Security
including SOM fractions and climatic conditions, Kadono et al. [2008] tested 41 surface soils in the Eurasian steppe and concluded that PMC and PMN could be determined by light fraction-organic matter (LF-OM), clay content and soil pH. On the other hand, recalcitrant organic C (ROC) and N (RON), equivalent to the difference between the total organic C (OC) or nitrogen (ON) content and the PMC content (or PMN content), comprise most of the slowly decomposing SOM group. To understand the mechanism of C sequestration in soil, the factors controlling ROC and RON need to be determined. In humid Asia, including Thailand, Indonesia, and Japan, however, there is a wide variety of soils under different climatic conditions and with different parent materials. In general, humid regions are characterized by a mean annual precipitation (MAP) to potential evapotranspiration ratio greater than 0.65 [Middleton and Thomas 1992] and covered by forest ecosystems (i.e., natural vegetation). Although these regions are important in terrestrial C and N cycles, owing to the vast stock in vegetation and rapid C turnover between the vegetation, atmosphere, and SOM, the factors controlling potentially mineralizable and recalcitrant OM in this region have not been studied in detail. In addition, this area is also characterized by the presence of volcanos and is covered by volcanic parent materials. Although the relatively higher soil organic carbon (SOC) stock in the volcanic soils results from a higher amount of active metals (e.g., Al and Fe) in the amorphous clay minerals [Hiradate et al. 2004], the contribution to potentially mineralizable OM and recalcitrant OM has not been clearly determined. In the present study, we tested soil samples from humid regions in Asia to reveal the relationships between potentially mineralizable OM, recalcitrant OM, and soil and meteorological properties.
4.5.1 Soils Studied A total of 89 surface soil samples (0–10 cm) were collected from Japan (n = 26), northern Thailand (n = 9), and Indonesia (n = 54), covering a range of climatic conditions and parent materials. In Japan, forest sites (n = 14) were selected in the west with varying vegetation and management (seminatural evergreen broadleaved forest, and mixed and fallow forests), and cropland sites (n = 12) were selected to cover a range of climates (from cool temperate to subtropical) and agricultural uses (paddy fields and upland cultivation). Parent materials in the forest sites included tephra, rhyolite, and sedimentary rocks, whereas in the cropland sites the parent materials included peat, tephra, sedimentary rocks, and alluvial materials. Detailed information about the Japanese cropland sites was given in Sano et al. [2004]. In northern Thailand, soils were sampled from forest fallow sites under shifting cultivation at various altitudes. The parent material of the lowland (<180 m) sites was sedimentary rock, whereas the parent material of the highland (>1000 m) sites was granite. All nine sampling sites in Thailand were classified as forest sites in the present study. In Indonesia, soil sampling was conducted on Java, Sumatra, and Borneo Islands in areas with varying parent materials (e.g., sedimentary rocks, volcanic rocks, limestone, and tephra) and land uses (e.g., natural tropical rainforest, secondary, or fallow forest and cultivated land). Thirty-seven sites were classified as forest sites and 17 sites were cropland sites.
83
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
The sampling sites were categorized into two land uses, namely cropland (n = 29) and forest (n = 60). The forest sites included tropical rainforests, deciduous forests, mixed forests, secondary forests, and fallow forests under shifting cultivation. Fresh soil samples were used to determine PMC and PMN (see details in Appendix) in addition to conventional determination of general physicochemical properties of the soils using air-dried samples. The contents of the light fraction (LF) and the heavy fraction (HF) were also determined with density fractionation using sodium iodide solution (1.60 g cm−3) and centrifugation at 2600 g [modified from Spycher et al. 1983].
4.5.2 Soil Properties and a Comparison of Land Use The average soil properties for all the land uses are shown in Table 4.4. Higher coefficients of variation (CV) were observed for variables of SOM and were well explained
TABLE 4.4 Averages and Coefficients of Variation for Soil Properties in Different Land-Use Systems Total (n = 89) TC TN TC/TN pH Sand Silt Clay LFW LFC LFN LF C/N HFC HFN HF C/N LFC/TC LFN/TN Sio Feo Alo
(g C kg soil) (g N kg−1 soil) −1
(%) (%) (%) (%) (g C kg−1 soil) (g N kg−1 soil) (g C kg−1 soil) (g N kg−1 soil) (%) (%) (g kg−1) (g kg−1) (g kg−1)
Cropland (n = 29)
Forest (n = 60)
AVE
CV (%)
AVE
AVE
49.5 3.4 14.7 5.2 42 25 33 2.1 5.8 0.26 24.7 43.7 3.1 14 11.9 7.6 1.9 8.1 7.9
69 62 28 17 48 39 49 89 93 99 30 70 62 30 76 89 279 98 141
30.8b 2.2b 14.6a 5.7a 46a 23a 31a 0.9b 1.9b 0.09b 25.5a 28.9b 2.1b 14.2a 7.3b 4.6b 3.4a 8.1a 9.0a
58.5a 3.9a 14.8a 5.0b 41a 25a 34a 2.6a 7.6a 0.34a 24.3a 50.8a 3.6a 14.0a 14.1a 9.1a 1.1a 8.2a 7.4a
Note: For each variable for land use, a different letter in the same row indicates a significant difference (p < 0.05). AVE, average; CV, coefficient of variation; TC, total organic carbon; TN, total nitrogen; LFW, LF content of the soil weight; LFC and LFN, carbon and nitrogen contents in LF, respectively; HFC and HFN, carbon and nitrogen contents in HF, respectively; Sio, Feo, Alo, acid oxalate soluble Si, Fe and Al, respectively.
84
World Soil Resources and Food Security
by differences in the land use. Significantly higher amounts of SOM accumulated in forest sites than in cropland sites. Total C and TN in the forest sites (58.5 g C kg−1 and 3.9 g N kg−1, respectively) were 1.9-fold and 1.8-fold the TC and TN in the cropland sites (30.8 g C kg−1 and 2.2 g N kg−1, respectively), whereas the TC/TN ratio did not differ with land use. Although the total variation in the pH values was relatively low, variation in the pH values for different land uses was observed in the present study. Soil texture did not differ with land use. The LF-OM was only 2.1% of the whole soil by weight, and LFC and LFN accounted for 11.9% of TC and 7.6% of TN, respectively. The LF-OM in the forest sites was approximately 4-fold larger than that in the cropland sites, and LFC/TC in the forest sites (14.1%) was higher than LFC/TC in the cropland sites (7.3%). Light-fraction C/N and HF C/N did not differ with land use. Properties relating to amorphous oxides (Sio, Feo, and Alo) had the highest CV, owing to the presence of samples with extremely high values; the maximum values recorded for Sio, Feo, and Alo content were 31 g kg−1, 53 g kg−1, and 57 g kg−1, respectively. There were 14 samples with Alo + 1/2Feo contents greater than 20 g kg−1, 10 of which were classified as Andisols. The higher pH in the cropland sites (5.7) than in the forest sites (5.0) results from liming in crop production.
4.5.3 Amounts of PMC and PMN and the Rate Constants for the Different Land-Use Systems As eight samples had kC values of <0.003, the total number of samples fitted to the equations was 81. The PMC amounts of 69 samples were fitted to the first-order equation, six samples were fitted to the double first-order equation, and six samples were fitted to the Gompertz equation. The PMN of 37 samples were fitted to the firstorder equation, 24 samples were fitted to the Gompertz equation, and 23 samples were fitted to the logistic equation. Five samples did not converge for any of the three equations. The average amounts of PMC and PMN and the proportions of total organic C and N for all the soils and for each land use are shown in Table 4.5. The PMC and PMN had high CV (92% and 85%, respectively), which is partly explained by land use. In the present study, both PMC and LFC in the forest sites were approximately 4-fold higher than in the cropland sites, whereas the difference in PMN with land use (2.4-fold) was not so conspicuous as LFN (4-fold). As the PMC/PMN in the forest sites (26.0) was similar to the LF C/N (24.7, n = 89), whereas PMC/PMN in the cropland sites (12.6) was much closer to the HF C/N (14.0, n = 89), the major source of potentially mineralizable OM might differ with land use, that is, LF-OM in the forest sites and HF-OM in the cropland sites.
4.5.4 Estimation of PMC and PMN by Stepwise Linear Regression Using the Five Identified Factors by Principal Component Analysis The soil pH, sand, silt and clay content, LFC, LFN, HFC, HFN, C/N ratio of the LF and HF, Sio, Feo, Alo, MAT, and MAP were summarized as five factors by principal component analysis. Based on a correlation between the factors and variables, these factors were named LF, amorphous, clay, humid and warm, and C/N (Table 4.6).
85
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.5 Average Amounts for Potentially Mineralizable Organic Carbon and Potentially Mineralizable Organic Nitrogen, Their Proportions in Total C and Total N, and Their Rate Constants for Different Land-Use Systems Total PMC kC (Fi) PMC/ TC PMN kN (Fi) PMN/TN PMC/PMN
(mg C kg−1) (d−1) (%) (mg N kg−1) (d−1) (%)
Cropland
Forest
n
AVE
CV (%)
n
AVE
n
AVE
81 69 81 84 37 84 77
3878 0.010 8.1 200 0.012 7.1 21.8
92 56 59 85 58 87 89
27 24 27 26 12 26 24
1261b 0.012a 5.1b 100b 0.009a 7.3a 12.6b
54 45 54 58 25 58 53
5186a 0.009a 9.6a 244a 0.013a 7.1a 26.0a
Note: For each variable for land use, a different letter indicates a significant difference (p < 0.05). AVE, average; CV, coefficient of variation; TC, total organic carbon; TN, total nitrogen; PMC and PMN, potentially mineralizable C and N, respectively; kC (Fi) and kN (Fi), rate constants for PMC and PMN that were fitted with the first-order equation, respectively.
TABLE 4.6 Correlation Matrix between the Factors and Soil Properties Variable pH Sand Silt Clay LFC LFN LF C/N HFC HFN HF C/N Sio Feo Alo MAT MAP Proportion (%)
PC1 −0.55** −0.04 0.14 −0.03 0.93** 0.93** −0.26* 0.71** 0.70** 0.15 −0.03 0.09 0.19 −0.27* 0.20 22 LF
PC2
PC3
PC4
PC5
−0.08 0.07 0.13 −0.16 −0.06 0.00 −0.23* 0.43** 0.44** 0.03 0.87** 0.86** 0.94** −0.24* 0.08 20 Amorphous
0.02 −0.98** 0.66** 0.84** −0.01 −0.07 0.28** 0.34** 0.36** −0.14 −0.21* 0.16 −0.14 −0.14 0.07 17 Clay
−0.17 0.01 −0.34** 0.19 −0.01 −0.09 0.34** −0.23* −0.16 −0.19 −0.03 −0.03 −0.09 0.75** 0.84** 11 Humid and warm
−0.20 0.02 0.13 −0.10 0.00 −0.15 0.60** 0.11 −0.12 0.83** −0.16 0.15 −0.07 −0.07 0.01 8 C/N
Note: Significant correlation coefficients at **p < 0.01 and *p < 0.05. LFC and LFN, C and N contents in the light fraction, HFC and HFN, C and N contents in the heavy fraction; Sio, Feo and Alo, acid oxalate soluble Si, Fe and Al, respectively, MAP, mean annual precipitation, MAT, mean annual temperature.
86
World Soil Resources and Food Security
Using linear regression analysis with the stepwise method, PMC and PMN were estimated by the five identified factors as follows:
PMC (mg C kg−1) = 4044 + 2629 × (LF factor) + 1072 × (clay factor) + 751 × (humid and warm factor)
(n = 81, r 2 = 0.63**)
PMN (mg N kg−1) = 201 + 67 × (LF factor) + 53 × (clay factor)
(n = 84, r 2 = 0.26**).
(4.3)
(4.4)
As the factor scores are standardized values, the coefficient for each factor indicates the relative contribution of the factor in each equation. Compared with the clay factor, the LF factor showed a larger contribution to PMC than to PMN; the coefficient of the LF factor was 2.5-fold larger than that of the clay factor in PMC and 1.3-fold larger in PMN. Similar to the Eurasian steppe soils [Kadono et al. 2008], the LF factor had a larger contribution to PMC and PMN than the clay factor in the present study. This results from the labile nature of LF-OM compared with the clay-associating OM. There are several reports stating that the LF is enriched in carbohydrates relative to both the whole soil and the HF [Dalal and Henry 1988; Murayama et al. 1979; Oades 1972; Whitehead et al. 1975]. Von Lützow [2007] reviewed articles on the turnover times of different SOM fractions separated by physical fractionation schemes and showed that the stability of OM increased from the free LF (or free particulate OM [POM]) to the occluded LF (or POM) and further to mineral-associating SOM. The humid and warm factor contributed positively to PMC in the present study. This results from the higher NPP in the higher MAT and MAP sites. A strong correlation between NPP and MAP has been reported for sites with less than approximately 3000 mm year−1 precipitation [Gower 2002].
4.5.5 Estimation of PMC and PMN by Stepwise Linear Regression Using Soil Properties Representing the Five Identified Factors Using linear regression analysis with the stepwise method, PMC was estimated by LFC (mg C kg−1), Alo (g kg−1), clay (%), MAP (mm), and HF C/N as representative variables of the five factors. The following equation was obtained:
PMC (mg C kg−1) = −2942 + 0.46 × LFC (mg C kg−1) + 41.5 × clay (%) + 1.2 × MAP (mm)
(n = 81, r 2 = 0.58**).
(4.5)
Similarly, a regression analysis was conducted for PMN. The equation obtained by this method was as follows:
PMN (mg N kg−1) = 10.0 + 0.24 × LFN (mg N kg−1) + 3.9 × clay (%)
(n = 84, r 2 = 0.25**).
(4.6)
87
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
According to these equations, scatter plots between the measured and estimated PMC and PMN values are shown in Figure 4.15. As shown in the equations, PMC and PMN could be estimated by LFC, LFN, clay content, and MAP. Amorphous oxide (Alo) did not contribute to PMC and PMN. The coefficients of clay content in Equations 4.5 and 4.6 indicated that 1% of clay contributed to 41.5 mg C kg−1 of PMC and 3.9 mg N kg−1 of PMN, respectively. This might suggest that clay-associated OM was partially decomposable and/or clay-occluded LF-OM was mineralizable. As von Lützow et al. [2007] reported that the 14C mean residence time of the clay fraction ranged from 75 to 4409 years, clay-associated OM shows a wide range of decomposability and might be a mixture of labile and recalcitrant OM. As the clay content did not differ with land use, the difference in the amount of PMC and PMN between the forest and cropland sites is attributed to the LF-OM. Therefore, the lower content of LF-OM in the cropland sites results from the relatively higher contribution of clayassociating OM to PMC and PMN.
4.5.6 Factors Controlling ROC and RON Using linear regression analysis with the stepwise method, ROC and RON were estimated by the five identified factors as follows: ROC (g C kg−1) = 45.3 + 24.5 × (LF factor) + 12.6 × (amorphous factor) + 8.8 × (clay factor) − 7.2 × (humid and warm factor) + 3.4 × (C/N factor) (n = 81, r 2 = 0.91**)
Estimated PMC (mg C kg–1)
20000
R2 = 0.58**
1000
1:1
15000 10000 5000 0
0
5000
10000
15000
20000
Estimated PMN (mg N kg–1)
Measured PMC (mg C kg–1)
1:1
R2 = 0.25**
750
500
250
0
–5000
(4.7)
0
500
1000
Measured PMN (mg N kg–1)
1500
FIGURE 4.15 Scatter plots showing the measured potentially mineralizable organic C (PMC) and the estimated PMC (n = 81), and the measured potentially mineralizable organic N (PMN) and estimated PMN (n = 84). Black circles indicate samples from cropland sites and triangles indicate samples from forest sites.
88
World Soil Resources and Food Security
RON (g N kg−1) = 3.11 + 1.52 × (LF factor) + 0.84 × (amorphous factor) + 0.61 × (clay factor) − 0.29 × (humid and warm factor) − 0.26 × (C/N factor)
(n = 84, r 2 = 0.89**).
(4.8)
According to the regression equations obtained, ROC and RON were explained by all five factors; the amorphous factor had a higher contribution than the clay factor, the contribution of the humid and warm factor was negative, and the C/N factor had the opposite affect effect on PMC and PMN. Using original variables closely related to the respective factors, ROC was estimated as follows:
ROC (g C kg−1) = −23.3 + 0.48 × clay (%) + 3.67 × LFC (g C kg−1) + 1.26 × Alo (g kg−1) + 1.51 × HF C/N
(n = 81, r 2 = 0.71**).
(4.9)
Similarly, RON was estimated as follows:
RON (g N kg−1) = −0.42 + 0.034 × clay (%) + 4.65 × LFN (g N kg−1) + 0.076 × Alo (g kg−1) + 0.00022 × MAP (mm)
(n = 84, r 2 = 0.67**).
(4.10)
According to the coefficient of LFC and LFN in Equations 4.9 and 4.10, respectively, the amount of ROC proportional to LFC was 3.67-fold larger than the amount of LFC itself, whereas the RON controlled by LFN was 4.65-fold as much as LFN. As the LF-OM partly represents the amount of NPP in each site, it would suggest that the higher amount of recalcitrant OM was partly attributed to the higher intensity of OM input by the vegetation. The contribution of the clay content is attributed to the SSA as the capacity to sorb dissolved OM [Kaiser et al. 1996; Nambu and Yonebayashi 2000]. In the present study, 0.48 g C kg−1 of ROC and 0.034 g N kg−1 of RON was accumulated at an increment of 1% clay content. In humid Asia, amorphous minerals played a significant role in both ROC and RON, whereas they did not contribute to PMC and PMN. According to the coefficient of Alo in Equations 4.9 and 4.10, 1 g kg−1 of Alo contributed to 1.26 g C kg−1 of ROC and 0.076 g N kg−1 of RON (i.e., mol ratio of C to Alo and N to Alo was 2.8 and 0.15, respectively). In general, soils from volcanic ash are characterized by a high accumulation of OC because of the ligand exchange of carboxyl groups to aluminum surface groups or complexes of reactive Al to humic and fulvic acids [Boudot 1992; Parfitt et al. 1977]. Kaiser and Guggenberger [2003] found that the specific surface area (SSA) of amorphous Al hydroxide (285 m2 g−1) was 3.7-fold and 18-fold that of illite (77 m2 g−1) and kaolinite (16 m2 g−1), respectively. Both the higher affinity of amorphous mineral surfaces to OM and the larger SSA contribute to the higher ROC and RON in the volcanic soils.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
89
4.6 C OMPARATIVE STUDY ON SOIL FERTILITY STATUS UNDER SHIFTING CULTIVATION IN EAST KALIMANTAN, NORTHERN THAILAND, AND JAPAN IN RELATION TO DYNAMICS OF READILY MINERALIZABLE SOM AND SOIL ACIDITY Shifting cultivation, which has often been referred to as slash-and-burn or swidden agriculture, is an extensive farming system that incorporates both cropping and fallow periods in its land rotation [Kunstadter and Chapman 1978; Nye and Greenland 1960]. Recently population increases in the mountainous areas of Southeast Asia has lead to an expansion of cropland, widespread forest degradation, and a shortening of the fallow period in traditional farming systems. This has resulted in an irreversible degradation of soil fertility. To alleviate forest degradation while maintaining or improving the living conditions of the hill people, it is essential to introduce improved and/or alternative land-use systems that allow continuous cultivation. It is widely believed that tropical soils are generally infertile and shifting cultivation is an adaptation to such poor soil conditions. However, as discussed so far, ecological conditions in the tropics are surprisingly diverse, and systems of shifting cultivation are consequently also quite variable. The main objective of this study is to compare soil fertility status under shifting cultivation systems in different bioclimatic conditions with special reference to the dynamics of labile organic matter in soils. This fraction of organic matter is expected to be a good indicator for evaluating the overall sustainability of the fallow-cropping system, and may give an insight into how to achieve more efficient management of organic matter resources in forest-soil ecosystems.
4.6.1 Description of Study Sites This study was carried out in three areas under different bioclimatic conditions: East Kalimantan, Indonesia (EK), northern Thailand (NT), and Japan (JP) (Figure 4.16). In EK, two broadly separated regions were selected: the lower part of the R. Mahakam watershed with an elevation of less than 100 m (EKL), and the mountainous area known as Apokayan on the Malaysian Sarawak border with an elevation of between 600 and 800 m (EKH). Both areas belong to Köppen’s Af climate type and do not have a distinct dry season [Kottek et al. 2006], hence the soils have an udic soil moisture regime. The mean annual temperature is above 25°C and annual precipitation exceeds 2000 mm in EKL. The EKH region is very humid and the annual precipitation exceeds 6000 mm. Although both regions have an equally hyperthermic soil temperature regime (>22°C), accumulation of organic matter in soil profiles is apparently higher in EKH soils according to field observation, presumably due to the extremely humid conditions there. Parent materials are mainly tertiary sandstone and/or mudstone, except for a limited number of plots in EKH, in which andesitic volcanic rocks are distributed. Among the EKL soils, however, limestone sometimes influences the soil, and in this case the soils were expected to be less acidic than other soils in the region. Most soils in the two regions were classified as Alisols [World Reference Base for Soil Resources, ISSS-ISRIC-FAO 2006], corresponding
90
World Soil Resources and Food Security
DP
20°N
RP(RPc) NT
EKH SM EKL BS(BSc)
0°
BG LB 100°E
PC 120°E
FIGURE 4.16 Locations of the experimental plots in Thailand and Indonesia.
to Udults in U.S. soil taxonomy [Soil Survey Staff 2006] (Figure 4.17a). In EK, farmers usually plant upland rice for 1 year only after clearing and burning the forest cover, and then leave the land for more than 20 years as fallow (Figure 4.17b). On relatively fertile soils near the river however, cultivation of crops with a shorter fallow of around eight years is practiced. In NT, activities of shifting farmers were observed mainly in a mountainous area with an elevation of between 600 and 1300 m. The climate type falls between Köppen’s Aw and Cw classifications, with a distinct dry season. Mean annual temperature is in the range of 20°C to 25°C and annual precipitation is between 1200 and 2000 mm. Parent materials are mainly granite, shale, mudstone, and limestone. The soils are mostly classified as Acrisols in WRB and Typic Haplustults or Ustic Dystropepts in U.S. Soil Taxonomy (Figure 4.17c). Although the traditional land use by the Karen people is 1 year of cropping and more than 10 years of fallow, the fallow period has recently decreased to as little as 3 years as a result of increasing population pressure (Figure 4.17d). Our survey was mainly carried out at a particular Karen village (DP village), at which the fallow period is still eight years. Upland rice is the main crop. In JP, traditional shifting cultivation is difficult to find. Our study site, located in Ishikawa Prefecture, may be one of the last examples of shifting cultivation in JP. This area belongs to Köppen’s Cf climate type, with an annual precipitation of 2600 mm and a mean annual temperature of 12°C. Parent material is mudstone with some influence from volcanic ejecta. Soils are classified as Cambisols (WRB) or Andic Dystrudepts (U.S. Soil Taxonomy) (Figure 4.17e). Traditionally, farmers plant cereals for about five years, followed by a long period of fallow (Figure 4.17f).
(c)
(b)
(e)
(d)
(f )
91
FIGURE 4.17 Representative soil profiles and landscape at the experimental plots in Indonesia, Thailand, and Japan with East Kalimantan, Indonesia (a) and (b); northern Thailand (c) and (d); and Japan (e) and (f).
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
(a)
92
World Soil Resources and Food Security
At the end of each dry season from 1997 to 2001, a total of 115 composite samples were collected from the surface 5-cm layer of soil in fields under different land-use stages of shifting cultivation. Then mineralization patterns of readily mineralizable organic C and N were analyzed using the same methodology as in the previous section (Section 4.5).
4.6.2 General Physicochemical Properties of the Soils Studied Table 4.7 shows the general physicochemical properties of the soils studied. The soils from EK were generally more acidic than those from NT, with higher Al, lower bases, and lower pH values. Among the EKH soils, however, those from riverside were higher in exchangeable Mg and Ca as well as in base saturation than those from upper slopes, suggesting relatively high fertility in the former. The soils from JP were somewhat intermediate between those of EK and NT in terms of exchangeable cation composition, since they had higher amounts of bases than EK or NT, but at the same time had high Al. This trend is similar to the results obtained for B-horizon soils in these regions (Section 4.4). Under the udic soil moisture regime in EK or JP, both subsoils and surface soils are subjected to more intensive acidification. The main difference between EKL and EKH was observed for the properties relating to SOM; that is, total C and N were significantly higher in EKH than in EKL. In addition, the EKL soils had a relatively coarse texture. This was presumably because of progressive clay translocation in the soil profiles due to a relatively gentle topography in the coastal or Mahakam lowland areas, compared with the mountainous region at the inland site of the island. The C mineralization pattern of the soils was fitted either by the Fi or Fi + Fi model and the value of C0 was determined in most cases. Only three EKL values and five EKH values were eliminated from further discussion because of low reliability with extremely low k values (i.e., <0.003). The average values of C0 are also given in Table 4.7. In EKH, the C0 values and the percentage of C0/TC were significantly higher, and fewer C0 values were able to be calculated than in the others. Thus it is assumed that the amount of readily mineralizable C is especially high in EKH, presumably because of the extremely high precipitation in that region and the resulting high net primary production and the slightly lower temperatures that may decrease the rate of SOM decomposition. As regards N mineralization, it was remarkable that NH +4 often accumulated during the incubation experiment, especially in the EK soils. The NH +4 accumulation over NO3− formation was commonly observed in the soils from the upper slopes of EKL or EKH (Figure 4.18a), indicating a delay in nitrification relative to N mineralization (or ammonification). Accumulation of NH +4 may lead to enhanced immobilization, and this may be why N0 is quite low relative to C0, compared with the other soils (Figure 4.18b). However, this figure for N mineralization was often missing among the soils from lower slopes or riverside areas of EKH. In these cases, little NH +4 remained at the end of the incubation experiment, and the pattern of nitrification appeared to be smooth, as was the case in the NT soils. Enhanced nitrification in a downslope was also detected using aerobic incubation [Kutsuna et al. 1988] or the 15N dilution technique [Tokuchi et al. 2000] in a Japanese conifer plantation forest in which the soils were similarly acidic.
Exchangeable K Area
n
pH(H2O)
EKL
36
4.7
EKH
34
NT
27
JP
5
Mg
pH(KCl)
a
3.9
4.9
a
5.6
b
4.8
a
NH4
Sum of Exch. Bases
Al
Base Saturation
(cmolc kg–1)
a
0.45
a
1.22
3.9
a
0.51
ab
4.8
b
0.95
b
4.3
a
1.00
ab C0
Area
Ca
(%)
a
1.80
a
0.41
1.66
a
2.38
ab
2.86
b
4.36
b
2.60
ab
9.10
c
N0
Total C
Total N
(% as TC)
(% as TN)
(g kg )
(g kg )
(g kg )
(g kg )
b
3.13
b
3.90
a
43.5
0.72
c
4.35
b
5.28
ab
52.5
a
0.25
ab
0.92
a
8.21
bc
85.4
b
0.04
a
3.19
ab
12.89
c
76.0
ab
Available Phosphate
a
Particle Size Distribution Sand
Silt
Clay
CN Ratio
(mg kg )
EKL
30.5
a
2.2
a
3.98 (13)a
a(b)
0.15 (7.0)
a(b)
14.2
a
39.4
a
56.9
b
20.7
a
22.4
a
EKH
63.4
b
4.4
bc
12.03 (18)b
b(c)
0.28 (6.2)
b(b)
14.3
a
57.8
a
36.8
a
26.2
a
37.0
c
NT
54.0
b
3.7
b
5.21 (9.3)
a(ab)
0.22 (5.9)
ab(b)
14.5
a
45.0
a
43.8
a
26.8
a
32.5
bc
JP
68.3
b
5.2
c
1.87 (2.9)
a(a)
0.09 (1.7)
a(a)
13.1
a
321.9
b
53.3
ab
23.6
a
23.1
ab
–1
–1
–1
–1
(%)
–1
Note: Soils were collected from cropland or fallow forest under shifting cultivation in Ishikawa Prefecture, Japan. The values with the same letters are not significantly different (p < 0.05). a n = 33. b n = 29.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.7 General Physicochemical Properties of the Soils Studied
93
94
World Soil Resources and Food Security
NH4-N accumulation at 133 d (g kg–1)
(a) 200
100
0
–100
N0 (g kg–1)
(b)
1:1
0
100
200
300
400
500
–100 Net NO3-N formation at 133 d (mg kg–1) EKL/upper slope EKH/upper slope EKH/riverside NT JP
0.8 0.6 0.4 0.2 0.0
0
10
C0 (g kg–1)
20
30
FIGURE 4.18 Scattergrams among the mineralization indices of soil organic matter determined by incubation experiments. C0 and N0 are the amounts of readily mineralizable organic C and N.
4.6.3 Factors Controlling the Amounts of C0 and N0 To analyze the effects of physicochemical properties on the amounts of C0 and N0, or N mineralization properties, principal component analysis followed by stepwise multiple linear regression was conducted separately for the soils from EK and NT, since physicochemical properties of soils were largely different between the two regions. Variables employed included pH(H2O), pH(KCl), exchangeable K, Mg, Ca, and Al, available phosphate, total C and N, CN ratio, and contents of sand, silt, and clay fractions. Tables 4.8 and 4.9 show the factor patterns for the first four principal components determined for the EK and NT soils, which accounted for 86% and 82% of their total variances, respectively. The factor patterns in both soils were rather similar, with factors relating to the quantity of organic matter, soil acidity, and NC ratio commonly determined. In the next step, stepwise multiple regression analysis was conducted to examine the contribution of each factor (determined above) to C0, N0, and NH +4 accumulation by the 133rd day (NH4 (133d)), and net nitrification by the 133rd day (NO3 (133d)),
95
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.8 Factor Pattern for the First Four Principal Components Determined for the Surface Soil Samples from East Kalimantan (n = 70) Variable pH(H2O) pH(KCl) Exch. K Exch. Mg Exch. Ca Exch. Al Avail. P Total C Total N CN ratio Sand Silt Clay Proportion (%)
PC1
PC2
PC3
PC4
0.13 0.13 0.09 0.16 0.20 0.50** 0.29 0.93** 0.94** 0.01 –0.84** 0.58** 0.91** 31 SOM and texture
–0.89** –0.93** –0.18 –0.83** –0.91** 0.80** 0.04 –0.03 –0.11 0.25 0.18 –0.25 –0.09 31 Acidity
0.05 0.01 0.84** 0.16 –0.01 –0.12 0.82** 0.18 0.18 –0.02 –0.18 0.26 0.09 12 P and K
–0.16 –0.18 –0.07 –0.19 –0.12 –0.10 –0.02 0.26 0.11 0.87** 0.44 –0.54** –0.28 12 CN ratio
Note: Significant correlation coefficients at **p < 0.01 and *p < 0.05.
TABLE 4.9 Factor Pattern for the First Four Principal Components Determined for the Surface Soil Samples from Northern Thailand (n = 27) Variable pH(H2O) pH(KCl) Exch. K Exch. Mg Exch. Ca Exch. Al Avail. P Total C Total N CN ratio Sand Silt Clay Proportion (%)
PC1
PC2
PC3
PC4
−0.78** −0.90** −0.68** −0.91** −0.94** 0.83** −0.18 0.14 0.04 0.32 0.26 −0.16 0.18 35 Acidity
−0.48 −0.27 0.25 0.22 0.07 0.38 −0.05 0.94** 0.98** −0.41 −0.37 −0.13 0.01 21 SOM
0.17 0.20 0.38 0.03 0.02 0.12 0.80** −0.13 0.07 −0.70* 0.70* 0.13 −0.19 15 P and NC ratio
−0.08 0.15 −0.07 −0.14 0.10 0.04 −0.01 −0.10 −0.09 −0.02 −0.08 0.82** 0.82** 11 Texture
Note: Significant correlation coefficients at **p < 0.01 and *p < 0.05.
96
World Soil Resources and Food Security
using the dataset of 62 samples from EK, for which all the variables were determined. The latter two variables are used because it is difficult to determine convergent values for mineralized NH4 and NO3. The following equations were obtained for the EK samples:
C0 (mg kg−1) = 7630 + 4925 × (SOM and texture factor) + 1139 × (acidity factor) + 1384 × (CN ratio factor)
(r 2 = 0.59**)
(4.11)
N0 (mg kg−1) = 212.5 + 76.2 × (SOM and texture factor) − 38.1 × (acidity factor) + 24.8 × (P and K factor) + 23.3 × (CN ratio factor)
(r 2 = 0.40**)
(4.12)
NH4 (133d) (mg kg−1) = 37.9 + 24.7 × (SOM and texture factor) + 48.8 × (acidity factor) − 8.41 × (P and K factor) + 10.4 × (CN ratio factor)
(r 2 = 0.68**)
(4.13)
NO3 (133d) (mg kg−1) = 103.2 + 19.0 × (SOM and texture factor) − 64.7 × (acidity factor) + 27.3 × (P and K factor)
(r 2 = 0.53**).
(4.14)
Since all the scores were already standardized, each coefficient directly gave the relative contribution of the factors. If we introduce standardized C0 value (C0-std) into the regression to take account of the effect of mineralizable C on N mineralization, Equations 4.12 and 4.14 can be improved with relatively high coefficients for C0-std as follows: N0 (mg kg−1) = 211.4 + 128 × (SOM and texture factor) − 26.0 × (acidity factor) + 20.3 × (P and K factor) + 38.0 × (CN ratio factor) − 69.8 × (C0-std)
(r 2 = 0.49**)
(4.15)
NO3 (133d) (mg kg−1) = 102.6 + 55.0 × (SOM and texture factor) − 56.3 × (acidity factor) + 24.2 × (P and K factor) + 14.3 × (CN ratio factor) − 48.3 × (C0-std)
(r 2 = 0.61**).
(4.16)
On the other hand, for NT soils, the following equations were obtained:
C0 (mg kg−1) = 5215 + 2249 × (SOM factor) − 851 × (texture factor)
(r 2 = 0.44**)
(4.17)
97
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
N0 (mg kg−1) = 220.3 − 30.7 × (acidity factor) + 60.3 × (SOM factor)
(r 2 = 0.74**)
(4.18)
Since NH +4 accumulation was scarcely observed for the NT soils, NH4 (133d) could not be explained by the regression analysis. NO3 (133d) showed a similar trend to N 0. These equations indicate that:
1. Becaues the OM factor positively contributes to C0, N0, NH4 (133d), and NO3 (133d), the quantity of readily mineralizable organic matter reflects the overall organic matter accumulation in the soils. 2. Equations 4.12, 4.13, 4.14, and 4.18 indicate that the factors relating to the acidic and/or oligotrophic nature of the soils, such as high acidity and P depletion, suppress nitrification and hence accelerate NH +4 accumulation, resulting in a reduction in net N mineralization. 3. C0 contributes negatively to nitrification and net N mineralization, implying that accumulation of readily mineralizable C accelerates microbial activity and N immobilization during the incubation experiment (Equations 4.15 and 4.16).
It is noteworthy that suppression of nitrification, which was often observed for the EK soils, during the long-term incubation experiment was usually not observed in the NT soils. The NT soils showed lower soil acidity than the EK soils, except for some cases where soil was collected under forests with long-term fallow phases. Thus, the more acidic and/or oligotrophic conditions of EK soils may suppress nitrification. In turn, this retards further loss of bases and slows soil acidification. Furthermore, the accumulation of C0 under the highly productive rainforest in EK may enhance the immobilization of NH +4 through the activity of soil microbes. Hence the ecology of EK is not so beneficial for shifting farmers in terms of N supply for agricultural production. It is thus indispensable to retain a long fallow phase, so that N can be released through the intensive soil burning effect that only occurs when a large amount of plant biomass is burned [Tanaka et al. 2001].
4.6.4 Relationships between Respective Factors and Topography or Land-Use Status in East K alimantan Based on Table 4.8 and Figure 4.19, the EKH soils showed high scores for the first component (i.e., the SOM and texture factor) compared to the EKL soils, indicating a relatively high accumulation of organic matter-related resources in high elevation regions, as was observed for the C0 values in Table 4.7. Regarding the quality of the organic matter, the scores of the CN ratio factor in the EKL soils were more variable than in the EKH soils (Figure 4.19b), suggesting that extremely low N status relative to C can often affect agricultural production on some EKL soils, in spite of similar values of average CN ratios in both soils (Tables 4.7 and 4.10). Among the EKH soils, those from riverside areas showed lower acidity (Table 4.10). At the same time, the
98
World Soil Resources and Food Security
Acidity factor
(a)
2 1 –2
0
–1
0
1
2
3
–1 –2 –3
SOM and texture factor (b)
4
CN ratio factor
2
–2
–1
0
–2
0
P and K factor
1
2
6
7
EKL/upper slope EKH/upper slope EKH/riverside
FIGURE 4.19 Distribution of factor scores determined by the principal component analysis in relation to elevation and topography in soils from East Kalimantan.
riverside soils of EKH often exhibited higher scores for P and K factors (Table 4.10 and Figure 4.19b), suggesting that the soil nutritional conditions relating to bases and P are all more favorable in soils from riverside areas. In EKH soils, where the influence of limestone is negligible, the scores of the acidity factor showed relatively clear trends in terms of the length of the fallow period (Table 4.10). That is, they tended to be low in the early stage of fallow (from the first to the third year), presumably due to the effect of ash input on slash-and-burn of forest before cultivation, and then later increased along with the growth of fallow forest. This high acidity in the middle to later stages of fallow may lead to a poor N nutrient status through suppressing nitrification and enhancing immobilization, as shown in the regression analysis discussed earlier.
4.6.5 Dynamics of Soil Fertility Status throughout a Land Rotation System of Shifting Cultivation in Northern Thailand The dynamics of soil fertility status were more clearly observed in the shifting cultivation system by Karen people in northern Thailand. Figure 4.20 shows the
99
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.10 Average Factor Scores in Different Areas of East Kalimantan or Under Different Land-Use Stages of Shifting Cultivation in EKH Area For all EK soils EKL/slope EKH/slope EKH/riverside For EKH soils Early stage of fallow (1st to 3rd year) Middle stage of fallow (4th to 8th year) Later stage of fallow (after 9th year)
n
SOM and Texture Factor
Acidity Factor
P and K Factor
CN Ratio Factor
36 21 13
−0.660 0.860 0.439
a b b
0.071 0.331 −0.732
b b a
−0.018 −0.288 0.514
ab a b
0.135 −0.124 −0.174
a a a
9
0.970
a
−0.849
a
0.352
a
−0.174
a
18
0.518
a
0.128
b
−0.082
a
−0.075
a
7
0.816
a
0.395
b
−0.152
a
−0.278
a
Note: The values with the same letters are not significantly different (p < 0.05).
dynamics of the first, second, and third factors (Table 4.9) throughout different landuse stages of shifting cultivation. In the initial stage of fallow, the scores of the acidity factor were kept low, presumably reflecting the ash input at the beginning of cultivation (Figure 4.20a). It then sharply increased in the middle stage of fallow (3–6 years), and then decreased again in the late stage of fallow. As a similar but more prolonged acidification was reported for subsoils under fallow in the area [Tanaka et al. 1997], soil acidification may commonly occur as a result of cation uptake by fallow vegetation. The values of SOM factor were low throughout the cropping year and the initial stage of fallow (0–6 years), and then increased along with the establishment of secondary forest and an increasing supply of forest litter (7–8 years) (Figure 4.20b). Similarly, the values of C0 and N0 were also low through the cropping year and the initial stage of fallow (0 to 4 years), and then increased along with the establishment of secondary forest and an increasing supply of forest litter (5 to 8 years) (Figure 4.21). As the acidity factor showed a decreasing trend at the same time in the late stage (Figure 4.20a), this litter may be supplying bases (obtained by tree roots from further down the soil profile) to the surface soil. This simultaneous change in acidity and SOM factors in the surface soil through forest-litter deposition in a late stage of fallow has an increasing effect on nutritional elements, and can be considered essential to the maintenance of this forest-fallow system. Agricultural production can thus be maintained with a relatively short fallow period of around ten years. Conversely, fallow periods of less than five years can be considered to be less sustainable. However, since in the long fallow the acidity factor scores were often high, cumulative cation uptake by forest vegetation may cause soil acidification. It is necessary to investigate in more detail whether prolonged fallow can accumulate bases
100
World Soil Resources and Food Security (a)
Factor 1
Scores of acidity factor
2
Long fallow (>15 y)
1 0
0
2
4
6
8
10
–1 –2
Fallow period (y)
–3
(b)
Factor 2
Scores of SOM factor
3
Long fallow (>15 y)
2 1 0
0
2
4
6
8
10
–1 –2
Scores of P and NC ratio factor
(c)
Fallow period (y) Factor 3
3
Long fallow (>15 y)
2 1 0
–1
0
2
4
6
8
10
–2 –3
Fallow period (y)
FIGURE 4.20 Distribution of factor scores in relation to land-use stages of shifting cultivation in northern Thailand.
in surface soil layers. In the long fallow, the values of “SOM” factor scores were usually high due to cumulative litter input. The P and NC ratio factor was occasionally high, but only at the cropping phase and the initial stage of fallow, presumably after higher ash input at the burning; it was then low throughout the fallow stages, including long fallow. The fallow practice does not seem to improve the soil properties relating to this factor. The slashand-burn practice was, therefore, considered to be indispensable for improving the soil properties relating to this factor for the next cropping. This is in contrast to the
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
C0 (mg kg–1)
(a)
N0 (mg kg–1)
Long fallow (>15 y)
10000
5000
0
(b)
C0
15000
101
0
2
4
6
8
10
Fallow period (y)
N0
800
Long fallow (>15 y)
600 400 200 0
0
2
4
6
8
10
Fallow period (y)
FIGURE 4.21 Distribution of C0 and N0 in relation to land-use stages of shifting cultivation in northern Thailand.
acidity and/or SOM-related properties of the soils, which already showed improvements in the late stage of fallow. Figure 4.22 shows the fluctuations of C0 and N0 in the shifting cultivation system in JP. Unlike EK or NT, neither C0 nor N0 fluctuate appreciably during the cropping and fallow phases, presumably because the northern temperate climate produces lower amounts of forest litter during the fallow period. Such a mild climate also prevents SOM from actively decomposing, allowing continuous cultivation for several years, as introduced earlier.
4.6.6 General Consideration on Land Uses in Upland Soils in Respective Regions in Relation to Soil Fertility Status As shown in Table 4.10 for EK, the prolonged fallow does not improve the low fertility status of the soils relating to the P and K factor, indicating that a high ash input, probably after long period of fallow, is indispensable for achieving sufficient yields
102
World Soil Resources and Food Security (a)
Long fallow (>15 y)
C0
4000
C0 (mg kg–1)
3000 2000 1000 0
0
2
4
6
8
10
12
Fallow period (y) (b)
Long fallow (>15 y)
N0
250
N0 (mg kg–1)
200 150 100 50 0
0
2
4
6
8
Fallow period (y)
10
FIGURE 4.22 Distribution of C0 and N0 in relation to land-use stages of shifting cultivation in Japan.
in the next cropping phase in the traditional shifting cultivation system. As far as P supply, the situation of NT was not very much different, i.e., the factor scores in the P and NC ratio factor were not improved even in the late stage of the fallow phase in NT (Figure 4.20c). In EK, the high ash input after a long period of fallow may also contribute to amelioration of soil acidity problems, resulting in an increased N utilization by crops. This is not the case for the relatively fertile soils in terms of N supply under shifting cultivation in NT. This may be a key factor in explaining why it has been difficult to establish cropping systems without long fallow periods in the traditional shifting cultivation in highly acidic upland soils in EK. Smithson and Giller [2002] also concluded that additions of P and N fertilizer should be seen as necessary in low-nutrient soils of the tropics. From these results, we can conclude that it will be difficult to establish intensive cropping systems in most of the upland areas unless very high amounts of fertilizer as well as liming are applied in EK. Topographic factors should be taken into consideration for developing alternative, more intensive agriculture; that is, the riverside soils should be the first priority for more intensive
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
103
uses. As discussed earlier, the riverside soils showed higher net nitrification relative to NH4 formation, and higher N0/C0 ratios than the other EK soils from upper slopes. Soil fertility status relating to acidity as well as P and K factors was also apparently favorable in the riverside soils. Thus the riverside soils were exceptionally fertile in terms of soil acidity, P and K status, and N supply potential, promising a better agricultural response at this location. On the other hand, in the shifting cultivation system by Karen people in NT, some soil properties relating to soil acidity improve at the same time as the SOM-related properties increase the late stage of fallow. The litter input may be supplying bases (obtained by tree roots from further down the soil profile) to the surface soil. This simultaneous increase in SOM and bases in the surface soil, through forest-litter deposition in the late stage of fallow, has an increasing effect on nutritional elements. These functions of the fallow phase can be considered essential to the maintenance of this forest-fallow system. Agricultural production can therefore be maintained with a fallow period of around 10 years in NT, which is somewhat shorter than widely believed. Traditional shifting cultivation in EK, NT, and JP can be seen to be well adapted to their respective soil-ecological conditions. Socioeconomic conditions are, however, drastically changing, making it difficult to sustain these systems. Even so, these systems provide an insight into basic soil processes—an understanding of which is essential to the better management of organic matter in all agroecosystems.
4.7 C OMPARATIVE STUDY ON PROTON BUDGETS IN SOILS OF CROPLAND AND ADJACENT FOREST IN THAILAND AND INDONESIA Soil acidification is a natural process, accelerated by agriculture, in humid regions under a climate where precipitation exceeds evapotranspiration [Helyar and Porter 1989; Juo et al. 1996]. In the croplands, proton generation associated with nitrification was reported to accelerate soil acidification owing to the enhanced mineralization of soil organic nitrogen [Tanaka et al. 1997; see Section 4.8 in this volume], limited vegetation uptake at the beginning of the cropping season [Poss et al. 1995], and nitrogen fertilization [Bouman et al. 1995]. However, soil acidification is also contributed to by proton sources other than nitrification, i.e., acidic deposition, dissociation of organic anions and carbonic acid and excess uptake of cations over anions by vegetation. Since most of proton-generating processes are associated with the organic matter cycles, they are influenced by the cultivation-induced changes in organic matter cycles, typically the loss of SOM owing to the increased organic matter decomposition and product removal. However, the effects of cultivation on individual processes are various and thus, their influence on soil acidification is still unclear. As discussed in the previous section (Section 4.6), soil acidity fluctuated considerably in different stages of land uses in shifting cultivation systems that incorporated both cropping and fallow phases in their land rotations. The objective of the present section was to evaluate the influence of cultivation on soil acidification by
104
World Soil Resources and Food Security
quantifying proton budget in a soil-vegetation system including solute leaching, vegetation uptake, and organic matter decomposition in croplands and adjacent forests in northern Thailand and East Kalimantan, Indonesia.
4.7.1 Study Plots Experimental plots consisted of one forest and one cropland plot in both Thailand and Indonesia. The forest and cropland plots in Thailand (RP and RPc, respectively) were located in Ban Rakpaendin, Chiang Rai Province (Figure 4.16). The corn (Zea mays L.) had been cultivated during the wet season without fertilization in RPc for 3 years since the conversion of forest to cropland. Soils were derived from sedimentary rocks (RP) and sedimentary rocks associated with granite intrusion (RPc) and classified as Typic Haplustults [Soil Survey Staff 2006]. On the other hand, the forest and cropland plots in Indonesia (BS and BSc, respectively) were located in the experimental forest of the Tropical Rainforest Research Center, Mulawarman University, Bukit Soeharto, East Kalimantan Province (Figure 4.16). The chili (Capsicum sp.) had been cultivated for 2 years after deforestation in BSc. Soils were derived from sedimentary rocks and classified as Typic Paleudults [Soil Survey Staff 2006]. 0.71 Mg DW ha−1 yr−1 of poultry manure was applied at the beginning of cropping season (October 2004–October 2005). Soil solution and precipitation (for the cropland plots) or throughfall (for the forested plots) were collected and the proton budget was calculated (see Appendix for detailed methodology).
4.7.2 Physicochemical Properties of Soils The physicochemical properties of soils are presented in Table 4.11. Soil pH was low throughout the soil profiles in BS (3.8–4.3) and BSc (4.2–4.3) than in RP (4.6–5.0) and RPc (5.4–5.5). Clay contents in soils were higher in RP (70%–75%) and RPc (40%–51%) than in BS (23%–31%) and BSc (19%–33%), although the differences in clay contents between RP and RPc might be due to the influence of granite in RPc. Total carbon contents in the A horizons in the cropland plots were lower (27 and 14 g kg−1 in RPc and BSc, respectively) than in the adjacent forest plots (63 and 23 g kg−1 in RP and BS, respectively).
4.7.3 Carbon Stock and Flow Carbon stock is presented in Table 4.12. In the forest plots, organic carbon was stored as the aboveground biomass (169.1 and 292.6 Mg C ha−1 in RP and BS, respectively) as well as SOM (65.2 and 26.6 Mg C ha−1 in RP and BS, respectively) in the upper 30-cm layers of soil. The higher aboveground biomass and the lower stock of organic matter in the mineral soil in the present study, as compared to temperate forests, were consistent with a previous report [Nakane 1980]. In the cropland plots, C stock in the mineral soil was lower than in the adjacent forest plots. C stock in the mineral soil in RPc and BSc (55.8 and 24.6 Mg C ha−1, respectively) was 9.4 and 2.0 Mg C ha−1 less than in RP and BS, respectively (Table 4.12).
Depth
pH
Exchangeable Total C
Horizon
(cm)
(H2O)
(KCl)
Total N
Bases
(g kg−1)
Al
Particle Size Distribution CEC
Sand
(cmolc kg−1)
Silt
Clay
(%)
Thailand RP
RPc
A
0–7
5.0
4.1
62.6
3.8
5.8
1.9
27.6
5
25
70
BA
7–20
4.9
3.9
19.8
1.5
2.0
3.5
19.9
4
23
73
Bt
20–45+
4.6
4.0
8.9
1.0
1.4
3.2
20.1
6
19
75
Ap
0–7
5.4
4.7
26.8
2.0
8.0
0.2
11.5
42
18
40
BA
7−20
5.5
4.8
15.0
1.3
5.9
0.2
14.4
32
13
55
Bt
20−45+
5.5
4.8
10.8
1.0
4.3
0.4
14.4
39
10
51
Indonesia BS
BSc
0−5
4.0
3.9
22.9
1.6
2.2
3.0
8.5
52
25
23
BA
A
5−25
3.8
3.8
4.2
0.5
0.8
3.9
6.2
49
27
24
B1
25−40
4.0
3.8
3.5
0.5
0.8
4.8
5.0
43
30
27
Bt
40+
4.3
3.8
2.5
0.4
1.0
7.0
5.0
34
35
31
Ap1
0−5
4.3
3.9
14.3
1.6
2.1
1.9
6.8
64
16
19
Ap2
5−20
4.2
3.8
4.8
0.5
1.8
3.3
8.7
54
19
28
B1
20−40
4.2
3.8
3.7
0.5
1.2
4.6
10.3
46
21
33
Bt
40−60+
4.3
3.8
3.2
0.4
0.9
5.9
11.3
54
19
26
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.11 Physicochemical Properties of Soils in Thai and Indonesian Plots
105
106
TABLE 4.12 Stock and Annual Flow of Carbon in the Thai and Indonesian Plots RP 169.1
RPc
BS
(−)
5.8
(0.8)
292.6
2.4 2.2 1.7
(0.3) (0.2) (0.1)
0.1 0.1 –
(0.0) (0.0)
2.6 65.2
(0.1) (5.0)
– 55.8
(a)
5.5
(0.3)
(b) (c) (d)
5.2 4.0 1.2 – 5.8 −0.2
(0.2) (0.2) (0.1)
(e)
(−) (0.4)
BSc (−)
1.2
(0.2)
1.0 1.2 2.1
(0.4) (0.3) (0.4)
0.2 0.1
(0.1) (0.0)
(2.3)
3.5 26.6
(0.3) (0.8)
24.6
(0.3)
8.2
(0.3)
5.4
(0.5)
3.6
(0.3)
4.3 4.1 0.2 1.7 – −3.9
(0.7) (0.7) (0.1) (0.3)
5.0 4.1 0.9 – 10.6 −0.4
(0.5) (0.5) (0.1)
1.4 1.1 0.3 0.2 – −2.2
(0.1) (0.1) (0.1) (0.0)
(0.8)
(–) (0.8)
Note: The figures in parentheses represent standard errors. a Organic carbon in soil at 0–30-cm depths was counted. b The C input was calculated as the sum of litterfall and root litter (b = c + d). c The annual rates of root litter production were assumed to be 20% of the fine root biomass in the forest [Nakane 1980]. d The C budget in soil was calculated as the difference between organic matter decomposition and C input (e = b − a).
(0.3)
World Soil Resources and Food Security
C stock (Mg C ha−1) Aboveground biomass Fine root biomass O, A horizon BA horizon B1 horizon Soil organic matter O horizon Mineral soil horizonsa C flow (Mg C ha−1 yr−1) Organic matter decomposition Organic matter production C input to soilb Litterfall Root litterc Product removal Wood increment C budget in soild
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
107
Seasonal fluctuations of soil temperature and volumetric water content in the soil and the rates of organic matter decomposition are shown in Figures 4.23 and 4.24, respectively. In RP and RPc, the rates of organic matter decomposition were positively correlated with volumetric water content in the soils (RP: r = 0.91, n = 9, p < 0.01; RPc: r = 0.63, n = 9, p < 0.10), while, in BS and BSc, they were independent of volumetric water content in the soils. Therefore, the annual rates of organic matter decomposition in RP and RPc were calculated as totals of the simulated emission rates using the regression equations and monitored volumetric water content in the soil, whereas, in BS and BSc, the rates were calculated using the average rates of CO2 emission measured. Volumetric water content in soil (L L–1)
Soil temperature (ºC) 35
Thailand RP
30 25
0.5 0.4
20
0.3
15
0.2
10
0.1
5 0 Apr-04
0.6
Jun-04
Aug-04
Oct-04
Dec-04
Feb-05
0.0
Date Volumetric water content in soil (L L–1)
Soil temperature (ºC) 35
Indonesia BS
30 25
0.3
15
5 0 Sep-04
0.5 0.4
20
10
0.6
0.2 Soil temperature at 5-cm depth (Cropland) Soil temperature at 5-cm depth (Forest) 0.1 Volumetric water content in soil at 5-cm depth (Cropland) Volumetric water content in soil at 5-cm depth (Forest) 0.0 Nov-04 Jan-05 Mar-05 May-05 Jul-05 Sep-05
Date
FIGURE 4.23 The seasonal fluctuations of soil temperature and volumetric soil water content.
108
World Soil Resources and Food Security C flux (µ gC m2 s–1) 60
Thailand RP
40
20
0 Feb-04 Apr-04 Jun-04 Aug-04 Oct-04 Dec-04 Feb-05 Apr-05
Date
C flux (µ gC m2 s–1) 60
Indonesia BS Forest Cropland
40
20
0 Nov-04
Jan-05
Mar-05
May-05
Jul-05
Sep-05
Nov-05
Date
FIGURE 4.24 Seasonal fluctuations of the rates of organic matter decomposition in the forest and cropland plots. Bars indicate standard errors (n = 5).
The annual flow of carbon is presented in Table 4.12. In the forest plots, the annual rates of organic matter decomposition thus determined were 5.5 and 5.4 Mg C ha−1 yr−1 in RP and BS, respectively. These values were consistent with those reported in tropical seasonal forests (4.1 Mg C ha−1 yr−1) (Section 4.8) and tropical rain forests (4.8–8.9 Mg C ha−1 yr−1) [Bond-Lamberty et al. 2004], respectively. Assuming that annual root litter rates were 20% of the fine root biomass [Nakane 1980], the annual rates of organic matter decomposition were almost balanced with C inputs as the sum of litterfall and root litter (5.2 and 5.0 Mg C ha−1 yr−1 in RP and BS, respectively) in the forest plots. Organic matter was accumulated in the ecosystems as wood increment in the growth stage of forests in RP and BS.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
109
In the cropland plots, the annual rates of organic matter decomposition were higher in RPc and BSc (8.2 and 3.6 Mg C ha−1 yr−1, respectively) than C inputs (4.3 and 1.4 Mg ha−1 yr−1, respectively). These values suggest a loss of SOM (3.9 and 2.2 Mg C ha−1 yr−1 in RPc and BSc, respectively) in the cropland plots. The SOM loss caused by cultivation can account for the lower SOM stock in the cropland plots, although the lower clay content in the soil of RPc might also contribute to the lower SOM stock than in RP.
4.7.4 Soil Solution Composition The annual volume-weighted mean concentrations of ions in precipitation, throughfall, and soil solution are presented in Table 4.13. The soil solution pH was relatively high in RP and RPc (6.1–6.2 and 5.7–5.8, respectively). The soil solution pH was moderately low (5.2–5.6) in BSc, while the soil solution pH was very low (4.2–4.4) in BS. The concentrations of bicarbonate in soil solution were significant (0.03–0.05 mmolc L−1) in all plots except for BS, where they were negligible owing to the low pH of soil solution. The concentrations of organic anions in soil solution were higher in the O and A horizons in BS (0.20–0.26 mmolc L−1) and in the Ap1 and Ap2 horizons in BSc (0.09–0.13 mmolc L−1), as compared to RP and RPc (0.01–0.02 mmolc L−1) (Table 4.13). The concentrations of organic anions in soil solution were correlated with those of DOC in BS and BSc, and these relationships were expressed by the following regression:
(Orgn–) = 0.087 × (DOC) + 0.042 (r = 0.77***, n = 111, p < 0.01)
(4.19)
(Orgn–) = 0.100 × (DOC) + 0.008 (r = 0.66***, n = 77, p < 0.01)
(4.20)
where (Orgn–) represents the concentration of organic anions in soil solution (mmolc L−1) and DOC represents the concentration of DOC in soil solution (mmol L−1). The higher concentrations of DOC in soil solution in the surface soil horizons in BS and BSc (17.2–34.7 and 9.3–9.7 mg C L−1, respectively), as compared to RP and RPc (3.1–3.9 and 2.2–4.5 mg C L−1, respectively), contributed to proton generation associated with dissociation of one acidic functional group for 11.5 and 10.0 C atoms of DOC, respectively. In the cropland plots, the concentrations of nitrate in soil solution were higher both in RPc (0.34–0.38 mmolc L−1) and BSc (0.06–0.14 mmolc L−1), as compared to the respective adjacent forest plots. The concentrations of nitrate in the Ap horizons were highest especially at the beginning of the cropping season in RPc (April 2004– June 2004) and BSc (October 2004–January 2005) owing to the enhanced decomposition of SOM and the small amount of biomass (Figure 4.25). Although nitrification is generally suppressed in the acidic soils [Kemmitt et al. 2006], the concentrations of nitrate in soil solution were still higher in BSc than in BS. This was consistent with the report by Killham [1990], in which nitrate could be produced by autotrophic bacteria even in the highly acidic soils in croplands. The concentrations of nitrate in soil solution were correlated with the concentrations of Ca (r = 0.99, n = 83, p < 0.01)
110
TABLE 4.13 Water Flux and Annual Volume-Weighed Mean Concentrations of Ions in Throughfall and Soil Solution
Thailand RP
RPc
BSc
a b
Horizon
pH
DOC (mg C L−1)
TF a A BA Bt
2083 1602 1162 825
6.09 6.21 6.13 6.05
2.7 3.9 3.1 3.1
0.029 0.034 0.029 0.029
0.015 0.020 0.012 0.012
0.017 0.021 0.025 0.018
0.058 0.072 0.060 0.069
0.001 0.001 0.001 0.001
0.012 0.007 0.007 0.007
0.104 0.123 0.103 0.108
0.000 0.007 0.006 0.008
0.001 0.009 0.008 0.004
0.003 0.061 0.054 0.060
P a Ap BA B1
2223 1414 1064 1064
6.22 5.80 5.66 5.83
2.9 4.5 2.2 1.9
0.030 0.034 0.048 0.047
0.008 0.017 0.009 0.012
0.014 0.371 0.380 0.338
0.052 0.117 0.076 0.095
0.001 0.002 0.002 0.001
0.009 0.007 0.008 0.008
0.088 0.508 0.486 0.439
0.000 0.004 0.001 0.002
0.001 0.004 0.005 0.006
0.003 0.149 0.120 0.101
TF a O A B1
2031 1619 1196 545
5.23 4.44 4.22 4.39
9.0 34.7 17.2 9.9
0.009 0.001 0.000 0.003
0.101 0.257 0.196 0.119
0.036 0.038 0.044 0.032
0.104 0.136 0.100 0.116
0.006 0.037 0.061 0.041
0.038 0.088 0.035 0.018
0.197 0.256 0.197 0.176
0.003 0.019 0.009 0.005
0.007 0.040 0.038 0.027
0.017 0.070 0.068 0.051
P a Ap1 Ap2 B1
2187 1144 824 824
6.09 5.62 5.57 5.21
4.2 9.7 9.3 5.9
0.024 0.042 0.051 0.038
0.016 0.126 0.087 0.038
0.009 0.143 0.137 0.061
0.022 0.084 0.097 0.083
0.001 0.002 0.003 0.006
0.014 0.020 0.018 0.023
0.045 0.311 0.298 0.187
0.000 0.010 0.008 0.006
0.001 0.046 0.034 0.019
0.001 0.061 0.083 0.067
HCO3–
Orgn− b
NO3–
CI– + SO24–
H+
NH+4
(mmolc L–1)
P and TF represent precipitation and throughfall, respectively. Orgn− represents anion deficit, the negative charge of organic acids.
Fe2+
Aln+ b
(mmolc L–1)
Si (mmol L−1)
World Soil Resources and Food Security
Indonesia BS
Na+ + K+ + Mg2+ + Ca2+
Water Flux (mm)
Soil Resources and Human Adaptation in Ecosystems in Humid Asia Nitrate concentration (mmolc L–1) RP
Nitrate concentration (mmolc L–1) BS
0.5
0.5
A horizon
0.4
B1 horizon
0.2 0.1
0.1 0 Apr-04
B horizon
0.3
Bt horizon
0.2
O horizon
0.4
BA horizon
0.3
Jun-04
Aug-04
Date
Oct-04
Dec-04
Nitrate concentration (mmolc L–1)
0.0 Nov-04
Feb-05
May-05 Aug-05 Nov-05
Date
Nitrate concentration (mmolc L–1)
RPc
3.0
BSc
3.0
2.5
Ap horizon
2.5
Ap1 horizon
2.0
BA horizon
2.0
Ap2 horizon
1.5
B horizon
1.5
B1 horizon
1.0
1.0
0.5
0.5
0.0 Apr-04
111
Jun-04
Aug-04
Date
Oct-04
Dec-04
0.0 Nov-04
Feb-05
May-05 Aug-05 Nov-05
Date
FIGURE 4.25 Seasonal fluctuations of concentrations of nitrate in soil solution. Bars indicate standard errors (n = 5).
and Mg (r = 0.99, n = 83, p < 0.01) in soil solution in RPc. The basic cations, mainly Ca and Mg, were leached out with nitrate (Figure 4.26).
4.7.5 Net Proton Generation and Consumption NPGNtr, NPGCar, and NPGOrg (where NPG stands for net proton generation) were calculated from the fluxes of solutes (Figure 4.26). Since cation contents exceed anion contents in plant materials in all plots (Table 4.14), excess cation charge is compensated for by net proton release from root to soil as NPGBio. Within NPGBio, since litterfall was the circulating fraction, NPGBio attributable to litter can be neutralized by cations released from fallen litter. NPGBio (wood increment or product removal) in each of the soil horizons was calculated by distributing it based on the distribution of the fine root biomass in the soil profiles (Table 4.12), according to Shibata et al. [1998]. Using NPGNtr, NPGCar, NPGOrg, and NPGBio, proton budget was calculated in each of the soil horizons (Figure 4.27). In RP, the concentrations of organic anions and nitrate in soil solution were low (0.01–0.03 mmolc L−1) owing to the rapid mineralization of organic anions and nitrate uptake by vegetation, respectively, and thus, NPGOrg and NPGNtr in each of the soil horizons were negligible (Figure 4.27). Although bicarbonate was a dominant anion in soil solution in RP, contribution of NPGCar to soil acidification was minor. In RP,
112
World Soil Resources and Food Security
Horizon RP
TF
Horizon SO42– NO
HCO3–
3
+
NH4 Na+ NH2+ Ca2+ Fe
A
Orgn–
Cl–
TF
3+
HCO3– NO3–
Ca2+
Na+
Fe3+
Aln+
O
Aln+
K+
BS
Orgn–
SO42–
Cl– H+
–2
0
NH4+
K+
Mg2–
A
BA Bt
B1
H+
–8
–6
–4
–2
0
2 4 6 8 (kmolc ha–1 yr–1)
–8
Fluxes of cations (+) and anions (–) Horizon
Orgn–
2 4 6 8 (kmolc ha–1 yr–1)
Fluxes of cations (+) and anions (–)
BSc
HCO3– Na+ NH + 4
SO42–
Ap
–4
Horizon
RPc P
–6
P
Fe3+ NO3–
Cl–
Mg2+
BA
Cl–
SO42–
Ap1
Ca2+
K+
HCO3–
Orgn–
NH4+ Na+ Fe3+ K+ Mg2+Ca2+
NO3–
Aln+
n+
Al
Ap2
Bt
B1
H+
–8
–6
–4
–2
0
2 4 6 8 (kmolc ha–1 yr–1)
Fluxes of cations (+) and anions (–)
–8
–6
–4
–2
0
H+
2 4 6 8 (kmolc ha–1 yr–1)
Fluxes of cations (+) and anions (–)
FIGURE 4.26 Fluxes of solutes at each horizon. P and TF represent precipitation and throughfall, respectively. O, A, Ap1, Ap2, BA, B, B1, and Bt represent soil horizons.
NPGBio (wood increment) distributed in the mineral soil horizons (0.6–0.8 kmolc ha−1 yr−1) and it contributed to soil acidification in each soil horizon (Figure 4.27). Since protons were consumed in the same horizon, proton leaching was negligible in RP (Figure 4.27). In BS, although protons were produced by the dissociation of organic anions in the O horizon (NPGOrg: 2.10 kmolc ha−1 yr−1), they were consumed owing to the mineralization and adsorption of organic anions in the A and B horizons (NPGOrg: −1.8 and −1.7 kmolc ha−1 yr−1, respectively) (Figure 4.27). An increase in the flux of NH +4 leaching from the O horizon (0.6 kmolc ha−1 yr−1) (Figure 4.26) suggested that protons were mainly consumed in the mineralization of organic nitrogen to NH +4 in the O horizon (NPGNtr: −0.78 kmolc ha−1 yr−1) (R–NH2 + H2O + H+ = NH +4 + R–OH) (Figure 4.27). A decrease in the fluxes of NH +4 leaching from the A horizon (1.0 kmolc ha−1 yr−1) (Figure 4.26) suggested that protons were released associated with the excess uptake of NH +4 over NO3− by biomass or adsorption of NH +4 on clays in the A horizon (NPGNtr: 0.9 kmolc ha−1 yr−1) (Figure 4.27). NPGBio (wood increment) distributed in the A and B horizons (0.7 and 2.1 kmolc ha−1 yr−1, respectively), while it also distributed in the O horizon (1.0 kmolc ha−1 yr−1) (Figure 4.27). NPGOrg
OMa production
Na
K
Ca
(Mg ha−1 yr−1) Thailand RP Wood increment Litterfall RPc
Product removal Litterfall Indonesia BS Wood increment Litterfall BSc
a b
Product removal Litterfall Fertilizerb
Mg
Fe
Al
Cl
S
P
(Cation)bio
(kg ha−1 yr−1)
(Anion)bio
NPGBio
(kmolc ha−1 yr−1)
NPGBio/OM Production (molc mol−1 C)
5.8
9.2
20.9
8.7
2.7
4.4
4.0
0.8
0.4
0.7
2.13
0.07
2.06
0.004
3.4
2.8
23.2
40.7
18.8
0.9
1.3
0.6
2.3
2.4
3.90
0.41
3.49
0.012
1.7
2.5
12.6
0.4
2.2
0.2
0.2
0.3
0.3
1.7
0.66
0.08
0.58
0.004
4.1
4.1
42.6
5.9
8.8
0.9
2.0
0.1
3.5
2.4
2.47
0.30
2.17
0.006
10.6
14.7
33.0
29.3
15.8
10.0
0.5
1.0
8.6
2.2
4.39
0.64
3.75
0.004
4.1
3.6
45.2
35.7
18.5
1.7
3.3
0.7
5.0
2.5
4.57
0.41
4.15
0.012
0.2
0.5
11.4
0.0
0.0
0.0
0.0
0.7
0.6
0.0
0.31
0.05
0.26
0.020
1.0 –
1.7 0.4
19.6 2.2
12.9 2.7
3.5 0.8
0.6 0.1
1.1 0.1
4.4 0.1
3.3 0.2
1.2 1.0
1.48 0.25
0.37 0.05
1.11 0.20
0.013 –
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.14 Uptake of Cations and Anions by Vegetation
OM represents organic matter. Poultry manure was applied at the rate of 0.71 Mg DW ha−1 at the beginning of rainy season.
113
114
World Soil Resources and Food Security Horizon
Horizon
RP
A BA
A
Bt
B1
Total
Total
–10
–5
0
5
10
(kmolc ha–1 yr–1)
Net proton generation (+) or consumption (–)
–10
–5
0
5
10
(kmolc ha–1 yr– 1)
Net proton generation (+) or consumption (–)
Horizon
Horizon
Ap BA
BS
O
Ap1
RPc
Ap2
Bt
B1
Total
Total
(H+)in – (H+)out
BSc
ΔANC NPGNtr
NPGOrg NPGBio
–10
–5
0
5
10
(kmolc ha–1 yr–1)
Net proton generation (+) or consumption (–)
–10
NPGCar
–5
0
5
10
(kmolc ha–1 yr–1)
Net proton generation (+) or consumption (–)
FIGURE 4.27 Net proton generation and consumption in the soil profiles. TF represents throughfall. O, A, Ap1, Ap2, BA, B, B1, and Bt represent soil horizons.
and NPGBio (wood increment) contributed to acidification of the O horizon (1.8 kmolc ha−1 yr−1) (Figure 4.27). Although most of the protons produced in the O horizon were consumed in the same horizon, 0.6 and 0.7 kmolc ha−1 yr−1 of protons leached from the O and A horizons, respectively (Figure 4.26). In RPc and BSc, protons were produced by nitrification in the Ap horizons (Ap1 horizon in BSc) (NPGNtr: 5.0 and 1.5 kmolc ha−1 yr−1 in RPc and BSc, respectively), most of which were neutralized by basic cations in the Ap horizons and this resulted in leaching of basic cations (Figure 4.26). Owing to the fine root biomass concentrated in the Ap horizons (Table 4.12), NPGBio (product removal) was concentrated in the Ap horizons (0.4 and 0.2 kmolc ha−1 yr−1 in RPc and BSc, respectively) in the cropland plots (Figure 4.27). Although NPGOrg in RPc and BSc was lower than the respective forest plots, dissociation of organic anions also contributed to proton generation in the Ap horizon of BSc (1.1 kmolc ha−1 yr−1). On the other hand, loss of SOM contributes to acid neutralization owing to cation release from the decomposed SOM. The cation release from the decomposed SOM can be estimated using the rates of SOM loss (3.93 and 2.24 Mg C ha−1 yr−1 in RPc
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
115
and BSc, respectively) (Table 4.12), the organic matter to carbon ratio of 1.8, and the cation exchange capacity of SOM (CECSOM) [Poss et al. 1995]. According to the soil pH-CECSOM equation proposed by Helyar and Porter [1989], the average CECSOM can be calculated as 124 and 90 cmolc kg−1 SOM in RPc and BSc, respectively, which corresponds to 0.027 and 0.019 molc for 1 mol of soil organic C, respectively. The basic cation release associated with loss of SOM was estimated as 8.8 and 3.6 kmolc ha−1 yr−1 in RPc and BSc, respectively, which can account for most of the depletion of the acid neutralizing capacity in the Ap horizons (ΔANC: −6.6 and −4.1 kmolc ha−1 yr−1 in RPc and BSc, respectively) (Figure 4.27). Therefore, it was considered that protons were mainly consumed by cations released from the decomposed organic matter in the present study. The contribution of cation release from the applied poultry manure to acid neutralization is low (0.2 kmolc ha−1 yr−1) in BSc (Table 4.14).
4.7.6 Comparison of Soil Acidification Processes in Forests and Croplands Contribution of NPGOrg to soil acidification was different between the topsoil horizons of RP and BS (Figure 4.26). NPGOrg in BS is as high as those in Spodosols under temperate forests (1.2–3.7 kmolc ha−1 yr−1) [Guggenberger and Kaiser 1998]. In BS, the higher fluxes of DOC leaching from the O horizon (360 kg C ha−1 yr−1), as compared to RP (62 kg C ha−1 yr−1), owing to the limited mineralization of DOC in the highly acidic soils are considered to result in higher NPGOrg. In the forest plots, NPGNtr in each soil horizon was different depending on the differences in behaviors (mineralization, uptake, and leaching) of NH +4 and NO3−. Ammonium leaching from the O horizon and its uptake in the A horizon contributes to the spatial heterogeneity of NPGNtr in the soil profile of BS, while mineralization and uptake by vegetation in the same horizon results in negligible NPGNtr throughout the soil profile in RP. Since NPGOrg and NPGNtr could be consumed owing to the mineralization and adsorption of organic anions on clays (e.g., ligand exchange and cation bridge) and vegetation uptake, respectively, their contributions to soil acidification in the complete soil profiles are minor in the forest plots (Figure 4.27). NPGBio is a dominant proton source in the forest plots. Soil acidification in the entire soil profiles is mainly caused by excess cation accumulation in wood in the growth stage of forests (Figure 4.27; Table 4.10). In contrast to the forest plots, the increased decomposition of organic matter is considered to contribute to nitrification predominating over plant uptake and result in higher NPGNtr in the Ap horizons in the cropland plots (5.0 and 1.5 kmolc ha−1 yr−1 in RPc and BSc, respectively) (Figure 4.27). Despite no fertilization, NPGNtr in RPc is comparable to the higher reported values in the fertilized croplands (1.4–11.5 kmolc ha−1 yr−1) [Ridley et al. 1990; Bouman et al. 1995; Poss et al. 1995; Lesturgez et al. 2006; Noble et al. 2008]. Since, in the present study, NPGNtr is caused by the net mineralization and nitrification of soil organic nitrogen, the higher rate of SOM loss in RPc is considered to be responsible for the higher NPGNtr in the Ap horizon in RPc than in BSc. The lower NPGNtr in BSc is considered to be partly attributed to the lower soil pH [Kemmitt et al. 2006]. Although NPGNtr is enhanced by cultivation, its extent is various depending on soil pH, as well as the rates of SOM loss.
116
World Soil Resources and Food Security
Although cultivation results in a decrease of DOC fluxes in the topsoil owing to a loss of the O horizon and adsorption of organic anions in the Ap horizon of BSc, the DOC fluxes in the Ap horizon of BSc (111 kg C ha−1 yr−1) were still higher than in RPc (62 kg C ha−1 yr−1) (Figure 4.26). The higher fluxes of DOC leaching results in the higher NPGOrg in BSc (1.1 kmolc ha−1 yr−1), as compared to RPc (<0.1 kmolc ha−1 yr−1) (Figure 4.27). On the other hand, as in RP, NPGOrg was negligible in RPc owing to the mineralization of organic matter to CO2 and adsorption of DOC on clays. In the cropland plots, product removal, as well as nitrification, contributed to proton generation (Figure 4.27; Table 4.10). This is consistent with the pattern of soil acidification in pastures [Bolan et al. 1991]. However, although variability of the amounts of cations released associated with SOM loss was high (SE: 0.8–1.8 kmolc ha−1 yr−1), judging from the fact that the amounts of cations released associated with SOM loss predominates over NPG owing to nitrification and product removal, cultivation results in an increase of both proton generation associated with nitrification and acid neutralization associated with SOM loss (Figure 4.27).
4.7.7 Evaluation of Cultivation-Induced Soil Acidification in Relation to Organic Matter Dynamics In BS (Indonesia), cultivation results in a significant decrease in NPGOrg in the topsoil horizon (Figure 4.27). In contrast, in RP (Thailand), irrespective of land use, contribution of NPGOrg to soil acidification was minor owing to mineralization and adsorption in the moderately acidic and clayey soils in RP and RPc (Figure 4.27). In the forest plots, contribution of NPGNtr to soil acidification in the entire soil profile was negligible. Judging from the almost even balance between organic matter decomposition and C inputs in forest plots (Table 4.12), the proton budget is considered to be balanced in a complete N cycle between mineralization and vegetation uptake (Figure 4.27), as mentioned by Binkley and Richter [1987]. On the other hand, in the cropland plots, where organic matter decomposition being greater than C inputs resulted in a loss of SOM, net mineralization and nitrification of soil organic nitrogen, and limited vegetation uptake due to the small amount of biomass at the beginning of the cropping season contributed to high NPGNtr. The loss of SOM contributed to NPGNtr at the rate of 0.015 and 0.008 molc for the loss of 1 mol SOC in RPc and BSc, respectively. The effects of cultivation on NPGOrg and NPGNtr were dependent on the soil properties (soil pH and texture), as well as the budget of organic matter. Soil acidification and alkalinization are dependent on the rates of both production and decomposition of organic matter. Cation excess accumulation in wood and product is a major proton-generating process in whole soil profiles in the forest and cropland plots, respectively. In the present study, production of organic matter contributed to proton generation at the rate of 0.004–0.020 molc for production of 1 mol organic C. On the other hand, in the cropland plots, the loss of SOM contributes to acid neutralization at rates of 0.026 and 0.019 molc for the loss of 1 mol SOC in RPc and BSc, respectively. The amounts of cations released associated with SOM loss can contribute to complete acid neutralization in cropland plots, at least during the initial stage of cultivation in tropical regions.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
117
Both proton generation owing to nitrification and acid neutralization by cation release from SOM on a basis of the loss of 1 mol SOC were higher in RPc than in BSc owing to the higher nitrification activity and the higher CECSOM of the moderately acidic soils of RPc. The influence of cultivation on the proton budget is different between Thailand and Indonesia, depending on the budget of organic matter and soil types (soil pH and texture).
4.8 Q UANTITATIVE ANALYSIS OF ORGANIC MATTER DYNAMICS UNDER SHIFTING CULTIVATION SYSTEMS IN NORTHERN THAILAND WITH SPECIAL REFERENCE TO FUNCTIONS OF THE SOIL MICROBIAL COMMUNITY Traditional shifting cultivation systems consist of two distinct phases: a cropping phase that occurs after the slash-and-burn of the forest and the fallow phase. The main effects of burning on crop production are considered to be the supply of bases and phosphorus via ash input [Alegre et al. 1988; Ewel et al. 1981; Nye and Greenland 1960; Raison et al. 1985; Tanaka et al. 1997]. In addition, N supply will be increased by enhanced mineralization after burning [Knoepp and Swank 1993; Marion et al. 1991; Nye and Greenland 1960; Sakamoto et al. 1991; Tanaka et al. 2001]. On the other hand, the function of the fallow phase is said to guarantee enough accumulation of nutritional elements into the forest vegetation [Richter et al. 2000] and SOM via litter input [Ruark 1993] for the next cropping phase to be successful. The effects of burning and subsequent ash input during shifting cultivation systems have been intensively studied [e.g., DeBano et al. 1998]. Although quantitative analysis of the fallow phase is still limited, this information is thought to be imperative for a comprehensive understanding of shifting cultivation in relation to its historical sustainability. When considering possible functions of the fallow phase, we could ask the question: How many years of fallow period are needed to maintain the productivity of the shifting cultivation in northern Thailand? In order to answer this question, several field experiments were conducted. In the present section, in situ SOM budget under shifting cultivation in northern Thailand is analyzed with special reference to functions of the soil microbial community.
4.8.1 Description of Study Sites The study was carried out in a farmer’s field in the village of Ban Du La Poe, located in the Mae La Noi District, Mae Hong Son Province, northern Thailand (DP village; see Figure 4.16). The village is inhabited by Karen people. The cropping fields of the villagers are mostly on steep slopes (usually exceeding 20°) of a hilly landscape, with an elevation of 1,100–1,300 m above sea level. There is a distinct dry season in the area, which is under the influence of a monsoon climate, and precipitation is concentrated in the rainy season from April to October. In the year in which the field experiments were carried out (2001), the annual precipitation and mean annual air temperature at our monitoring site were recorded as 1222 mm and 20.2°C,
118
World Soil Resources and Food Security
respectively. Parent materials of the soils are derived from partly metamorphosed fine-textured sedimentary rocks and granite. Most of the soils studied were classified into Ustic Haplohumults in the USDA classification system [Soil Survey Stuff 2006]. The shifting cultivation system of the village was similar to a traditional land rotation system, in which cropping for upland rice is limited to only 1 year, followed by at least 7 years of fallow. According to our field observation, this fallow period seems to be the longest one in this area. In fact, at many of the nearby villages, the fallow period has recently been shortened to around 4 years. Only a limited area is used for the cultivation of cash crops, such as cabbage.
4.8.2 Field Measurements of SOM Budgets in Different Stages of Land Use under Shifting Cultivation The field experiments were conducted from April 2001 to April 2002. To compare SOM dynamics under different stages of land use (cropped field after burning, fallow forest, and natural forest), six plots were set up as follows: cropland in 2001 (CR01); first, second, fourth, and sixth years of fallow forest (F1, F2, F4, and F6, respectively); and unburned seminatural forest stand aged about 30 years (NF) for comparison (Figure 4.28). To estimate the amount of C input into the soils, we measured litter input in CR01, F1, F2, F4, F6, and NF from 19 April 2001 to 19 April 2002, as follows. First, litter input in cropland is assumed to be equal to the amount of rice residues after harvest, which were collected for 1-m2 plots in triplicate just after harvest. Second, in the early stage of fallow (first to third year), fallowed fields are occupied by perennial herbaceous species, among which Eupatorium odoratum is dominant. Tree species gradually increase, but their biomass and litter supply are still limited at this stage. As it was difficult to collect litter fall in these fields, we postulate that in the first and second year the amount of litter input is equivalent to the amount of leaves at the end of the rainy season, and in the third year all the herbaceous biomass and half of the litter input of the establishing tree vegetation is returned to the soils. The biomass of herbaceous species was measured in F1 and F2 at the end of the rainy season (in 1-m2 plots in triplicate). Third, in the later stage of fallow (fourth to seventh year), secondary forest is successively established and litter fall from tree vegetation comprises a major part of litter input. The amount of litter fall was collected every month in F4, F6, and NF (in approximately 0.4-m2 plots in five replications). In addition, the amounts of litter stock were measured for 1-m2 plots in triplicate in F4, F6, and NF on 16 May 2001 and 19 April 2002. The results of litter input are given in Table 4.15. Annual litter fall was calculated to be 6.87, 5.37, and 9.00 Mg ha−1 yr−1 in F4, F6, and NF, respectively. These values can be converted to carbon: 3.19, 2.61, and 4.48 Mg C ha−1 yr−1, respectively. For these forest stands, as it is difficult to measure the amount of annual supply of dead root litter, we used the ratio of root litter to litter fall of 0.32, according to the simulation of Nakane [1980], which was proposed for matured forests under different climates. In F1 and F2, the total biomass of herbaceous species was 12.0 and 14.3 Mg ha−1 yr−1, respectively, of which approximately 47% consisted of leaves. In CR01, only 1.31 Mg C ha−1 yr−1 of rice residues remained
Soil Resources and Human Adaptation in Ecosystems in Humid Asia (a)
(b)
(c)
(d)
(e)
(f )
119
(g)
FIGURE 4.28 Landscapes of the study plots: (a) CR01 (April); (b) CR01 (August); (c) F1 (April); (d) F2 (April); (e) F4 (April); (f) F6 (April); and (g) NF (April).
120
TABLE 4.15 Litter Layer and Annual Litter Input in the Cropland and Fallow Fields under Shifting Cultivation in Northern Thailand Little Layer in May 2001
Amount
C
N
P
Content
Content
Content
(Mg C
(kg N
(kg P
ha−1)
ha−1)
ha−1)
Amount
(Mg ha−1)
Annual Litter Input Herbacious
Annual
Biomass at
Litter-Fall
Total Litter Input Estimateda
C
N
P
the End of
Input (C
C
N
P
Content
Content
Content
Rainy Season
Content)
Amount
Content
Content
Content
(mg
(mg C
(kg N
(kg P
(Mg C
(kg N
(kg P
(Mg ha−1
(mg ha−1
ha–1
ha−1
ha−1
ha−1
ha−1)
ha−1)
ha−1)
yr−1)
yr−1)
yr–1)
yr−1)
yr−1)
yr−1)
3.07
3.07
1.31
17.8
0.355
F1
11.95
5.56
2.59
21.4
0.428
F2
9.18
4.27
1.99
22.3
0.446
CR01
F4
5.29
2.66
70.4
1.41
4.69
2.35
54.9
1.10
6.87 (3.19)
9.04
4.19
99.2
3.48
F6
3.67
1.50
29.4
0.588
5.93
2.88
39.3
0.532
5.37 (2.61)
7.06
3.44
49.3
1.51
NF
7.52
3.63
93.0
1.86
6.81
3.43
97.8
1.96
9.00 (4.08)
11.84
5.89
143.1
4.21
a
For the herbacious vegetation in F1 and F2, only leaves are assumed to be incorporated into soils. The coefficient is 0.47. On the other hand, the amount of root-litter under the forest stand (F4, F6, and NF) is estimated to be litter-fall × 0.12/0.38 according to simulation of Nakane [1980].
World Soil Resources and Food Security
(Mg ha−1)
Little Layer in April 2002
121
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
on the soil after harvest. As a result, total litter input is calculated to be 4.3–9.0 Mg ha−1 yr−1 in the fallowed plots, which equals 2.0–4.2 Mg C ha−1 yr−1. It is noteworthy that especially high concentrations of N or P could be returned to the soils via the forest litter compared with via the herbaceous litter (Table 4.15). On the other hand, the soil C output was estimated by measuring the in situ field soil respiration rate using a closed chamber method for several times during the experimental period. At the same time, soil temperature at a depth of 5 cm and volumetric moisture content of the soil at depths of 0–15 cm were measured at all sites and also monitored continuously using data-loggers at two sites (CR01 and NF). Figure 4.29 shows the seasonal fluctuations of soil temperature (5 cm), volumetric water content of the soils (0–15 cm), and soil respiration rate (mol C ha−1 h−1) with or without root respiration (Cem+R and Cem–R) measured at the plots. Soil temperature at CR01 was the highest, within the range of 23°C–32°C, whereas that at NF was the lowest, within the range of 18°C –21°C. In the fallowed plots, soil temperature was between CR01 and NF. Volumetric water content of the CR01 soil was lowest among (b)
0.4
2/1/02
4/1/02 4/1/02
Date
(d) 200
200
150
150
100
Date
10/1/01
0
8/1/01
50
6/1/01
4/1/02
2/1/02
12/1/01
10/1/01
8/1/01
4/1/01
6/1/01
50
4/1/01
100
0
Soil respiration including plant roots (mol C ha–1 h–1)
(c) Soil respiration excluding plant roots (mol C ha–1 h–1)
2/1/02
Date
12/1/01
0.0
12/1/01
0.2 10/1/01
4/1/02
2/1/02
12/1/01
10/1/01
4/1/01
15
8/1/01
20
0.6
8/1/01
25
0.8
6/1/01
30
CR01 F1 F2 F4 F6 NF
1.0
4/1/01
35
6/1/01
Soil temperature (ºC)
40
Volumetric soil water content (L L–1)
(a)
Date
FIGURE 4.29 Fluctuation of soil temperature, soil moisture, and soil respiration rates measured from April 2001 to April 2002.
122
World Soil Resources and Food Security
all the plots throughout the year. The temperature and moisture data monitored by the data-loggers also exhibited the same trend. The values of Cem–R were usually lower than Cem+R, indicating that whole-soil respiration is also contributed by plant root respiration. Both the Cem+R and Cem–R at all the plots exhibited the highest values in the late rainy season (September to October). The Cem+R and Cem–R values were low throughout the dry season (December to March). In order to estimate total soil respiration throughout the year, we first established an equation that represented a relationship of the in situ hourly soil respiration rate and environmental factors (such as soil temperature and moisture) by multiple regression analysis. We then calculated hourly soil respiration rate by substituting each parameter of the equation using monitored data, and summed up hourly soil respiration rates for a given period. In the first step, we assume an Arrhenius-type relationship between soil temperature and respiration rate, as follows:
Cem = a θb exp(–E/RT) exp(cD)
(4.21)
where Cem is an hourly soil respiration rate with or without root respiration (mol C ha−1 yr−1), θ is a volumetric soil moisture content (L L−1), E is the activation energy (J mol−1), R is the gas constant (8.31 J mol−1 K−1), T is an absolute soil temperature (K), D is days after the rainy season started on 14 May (until 13 November, on which θ was falling below 0.3), and a, b, and c are coefficients. As soil respiration rate and microbial biomass N showed an increasing trend in some of the plots during the rainy season, a parameter D was introduced. For the samples in the dry season (after 13 November, θ < 0.3), the value of D was set to 0. The equation was then rewritten in the logarithm form:
lnCem = lna + blnθ – E/RT + cD.
(4.22)
Following this, a series of coefficients (a, b, c, and E) are calculated by stepwise multiple regression analysis (p < 0.15) using the measured data, Cem, θ, T, and D [SPSS Inc. 1998]. The results of regression analysis are summarized in Table 4.16. The coefficient relating to soil temperature (E) was often rejected (p < 0.15). For example, this occurred in F1 and F4 for Cem–R and in CR01 for Cem+R, indicating that under the subtropical conditions in this study, the effects of seasonal fluctuations of temperature on the soil respiration rate were limited. In contrast, coefficient b usually caused a fluctuation in the soil respiration rate due to the presence of a distinct dry season. In CR01, the influence of both soil temperature and moisture was uncertain, and only the date factor, c, was strongly related to fluctuations of soil respiration. A drastic change in the microbial community after slash-and-burn may cause such a unique behavior of soil respiration, which continuously increased throughout the rainy season. Using these regression equations, and monitoring soil temperature and moisture data in CR01 and NF, cumulative soil respiration during the year was calculated. As the monitored data was available for only two of the six plots, fluctuations in soil temperature and moisture for the remaining four plots were assumed based on the relationship between the actual data and either monitored data of CR01 or of NF with
Cem = aθbexp(–E/RT) exp(cD)
Coefficients
Site Excluding roots (Cem − R) CR01 F1 F2 F4 F6 NF Whole respiration (Cem + R) CR01 F1 F2 F4 F6 NF
c
r2
n
(at T = 293 K, θ = 0.4 L L−1, D = 0) (mol C ha−1 h−1)
Annual Soil Respiration (Mg C ha−1 yr−1)
lna
b
E (kJ mol−1)
−9.39* 3.73*** 167*** 4.50*** 59.0*** 58.6*
– 0.394* 0.769*** 0.993*** 0.612*** 0.484*
−32.3* – 398*** – 135** 134*
5.96E–03*** 3.41E−03* 3.01E−03* – – 6.24E−03**
0.81*** 0.52*** 0.75*** 0.55*** 0.45** 0.77***
15 15 15 15 15 15
49.1 29.1 18.1 36.3 19.2 28.2
6.34 3.02 3.35 2.99 2.15 4.14
4.67*** 45.5*** 53.3*** 62.5*** 49.2*** 85.3***
0.360*** 0.931*** 0.732*** 0.393*** 0.708*** 0.621***
– 99.3*** 118** 141*** 109*** 196***
4.27E−03*** – – – 2.69E−03* –
0.81*** 0.81*** 0.92*** 0.73*** 0.74*** 0.74***
15 15 15 15 15 15
76.4 47.2 52.2 74.1 55.6 74.2
8.23 5.76 4.72 6.96 7.03 6.64
123
Note: Significant at *15%, **5%, and ***1% levels, respectively. Cem = aθbexp(−E/RT) exp(cD); Cem, rate of CO2 emission (mol C ha−1 h−1); T, temperature (K); θ, volumetric moisture content of soil (L L−1); D, days after rainy season started on May 14 (D = 0 for the samples in dry season, after November 13, θ < 0.3); R = 8.315 (J K−1 mol−1); E, activation energy (J mol−1); a, b, and c, coefficients. This equation is converted to; lnCem = lna + blnθ – E/RT + cD.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.16 Parameters Determined by Stepwise Multiple Regression Analysis
124
World Soil Resources and Food Security
a higher correlation. We used the monitoring data of CR01 for simulation in CR01, F1, and F2, and that of NF for simulation in NF, F4, and F6. The results of these calculations are also given in Table 4.16. Annual soil respiration without root respiration was highest in CR01, followed by NF, and then F2, F1, F4, and F6, ranging from 2.15 to 6.34 Mg C ha−1 yr−1. The contribution of root respiration on whole-soil respiration was calculated to be 23%, 48%, 29%, 57%, 69%, and 38% in CR01, F1, F2, F4, F6, and NF, respectively (an average of 55% in forest plots F4, F6, and NF), which was similar to the estimation (50%) by Nakane [1980]. Table 4.17 summarizes the amounts of SOM stock (up to a 15-cm depth) and C budget in the study plots. Total SOM stock, including litter layer, in the soil profiles ranges from 37.3 to 66.7 Mg C ha−1. Compared with the total stock, annual output of soil C under the fallow forest is small (2.15–3.35 Mg C ha−1 yr−1), as is that of cropland (6.34 Mg C ha−1 yr−1); this is equivalent to only 4.5%–10.3% of total SOM. This may be one of the reasons why soil degradation did not become pronounced after 1 year of cropping in this area. Considering the fact that, even in NF, the SOM budget is still positive, the SOM level after the forest ecosystem reaches its equilibrium would possibly be higher. In turn, the SOM level under the current shifting cultivation practice can be said to be a secondary one, in which SOM stock had already been decreasing during repeated cultivation, compared with natural forest. Figure 4.30 shows annual input and output of organic matter in the soils with a cumulative budget based on the data in Table 4.17. For the fifth year, the average value of F4 and F6 was used for the calculation and, for the third year, the average value of F1 and F2 and half of the litter input of F4 were assumed to consist of litter input. During the cropping phase and the initial stage of fallow, the annual C budget was negative. It became positive when the initial herbaceous vegetation was succeeded by tree vegetation (after the third or fourth year of fallow) and a cumulative budget became positive after the sixth year of fallow. As the litter fall supplied may require 6 months to 1 year to be completely incorporated into the soils after initial decomposition (based on the C stock in litter layer and the annual C input from litter fall (see Table 4.17), approximately 6–7 years of fallow are needed to balance the SOM budget under the current fallow/cropping practice. A possible error can derive, in this study, from uncertainty regarding the estimation of dead plant roots in the burning stage or during the later stage of fallow. Even so, the minimum fallow period required would not fall below 4 years, judging from the fact that the SOM budget in the system strongly depends on incorporation of initial herbaceous biomass into the soil system after establishment of tree vegetation (in approximately the fourth year). It was possible to establish a multiple regression equation between the annual soil respiration and soil parameters for the five noncultivated plots (excluding CR01), as follows:
Annual soil respiration (Mg C ha−1 yr−1) = 1.291 + 0.483 × C0 (g kg−1) + 0.643 × pH(H2O) − 0.116 × clay content (%); (r 2 = 0.93, p < 0.25, n = 5). (4.23)
These parameters were chosen from variables that correlated with the different principal components determined in Section 4.6. Although the reason for the negative contribution of soil acidity is uncertain, it seemed worthwhile to investigate further.
Site CR01 F1 F2 F4 F6 NF
Total Litter Input Estimated
Annual C Output by Decomposition
Annual Budget of SOM
C Stock in Litter Layer
C Stock in 0–15-cm Depth
(from Table 4.15) (a)
(from Table 4.16) (b)
(a – b)
(Mg C ha )
(Mg C ha )
(Mg C ha yr )
(Mg C ha yr )
(Mg C ha−1 yr−1)
2.7 1.5 3.6
61.9 64.4 61.8 64.0 35.8 50.8
1.31 2.59 1.99 4.19 3.44 5.89
6.34 3.02 3.35 2.99 2.15 4.14
−5.03 −0.43 −1.36 1.20 1.29 1.75
−1
−1
−1
−1
−1
−1
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.17 Stock and Budget of Soil Organic Carbon from 19 April 2001 to 19 April 2002
125
126
10
5
0
NF
F6
F4
F2
–10
F1
–5
CR
Input/output of SOM (Mg C ha–1) and its cumulative budget (Mg C ha–1 yr–1)
World Soil Resources and Food Security
Litter input Litter input, estimated Soil respiration
FIGURE 4.30 Annual input and output of organic matter of the soils, with a cumulative budget.
Annual soil respiration estimated (Mg C ha–1 yr–1)
Using this equation, the annual soil respiration was estimated for all the soils analyzed in Section 4.6 (excluding those in the cropping year) and plotted in Figure 4.31. In most cases, the annual soil respiration in the fallowed plots ranged from 2 to 4 Mg C ha−1 yr−1 and was almost consistent with the measured values. Therefore, it is possible to conclude that the annual soil respiration or the annual decay of SOM in the studied village falls in this range, and that amounts of organic material should be incorporated into soils annually to keep the present level of SOM.
10 Long fallow (>20 y) 8 6 4 2 0
0
2
4
6
8
10
Fallow period (y)
FIGURE 4.31 Annual soil respiration estimates based on the regression equation obtained.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
127
4.8.3 In Situ Soil Solution Composition under Shifting Cultivation As decomposed products of SOM, such as NO3−, were expected to be present in the soil solution, we collected the soil solution by a porous-cup method [Funakawa et al. 1992] to analyze the process of N translocation in soil profiles. Figure 4.32 shows that the NO3− concentration in the soil solution was high in May and sharply decreased in early June at all the plots, except for CR01, in which the NO3− concentration remained as high as 0.5 mg N L−1 at a depth of 45 cm, even on 22 June. These data were consistent with the results given in Section 4.7 for RPc in northern Thailand (Figure 4.25) and suggested that a higher amount of NO3− could have been leached out from the top 45 cm of the soil layer in CR01 along with rainfall events. In contrast, a low concentration of NO3− was detected in the soil solution at a depth of 45 cm in NF throughout the rainy season, indicating that almost no N leaching from the forested ecosystem occurred. Figures for fallow forest were between that of CR01 and NF, in that the high concentrations of NO3− were observed only at the initial stage of the rainy season (May), and they soon dropped to a low level.
4.8.4 Fluctuation of Microbial Biomass and Metabolic Quotient, qCO2 To estimate the dynamics of a microbial biomass and its activity in relation to the dynamics of C and N, the amounts of microbial biomass C and N were determined several times during the experimental period by the chloroform fumigationextraction method [Brookes et al. 1985; Vance et al. 1987]. The metabolic quotient, qCO2, was calculated from the soil respiration rate divided by the microbial biomass C. Figure 4.33a shows that microbial biomass C in NF was the highest, ranging from 530 to 1849 mg kg−1, whereas that in CR01 was usually the lowest throughout the year (85–614 mg kg−1). Salt extractable C in CR01 increased remarkably, up to 600 mg kg−1 in the dry season (Figure 4.33b), in which microbial biomass C dropped. Such
2 1 0
FIGURE 4.32 Concentration of NO3− in a soil solution.
NF F6 F4 F2 CR01
NF F6 F4 F2 CR01
4-Sep
9-Jul
22-Jun
2-Jun
0
18-May
1
F6 NF
3
4-Sep
2
4
9-Jul
3
CR01 F2 F4
45 cm
2-Jun
F6 NF
5
22-Jun
4
(b)
18-May
CR01 F2 F4
15 cm
NO3– concentration (mg N L–1)
5
NO3– concentration (mg N L–1)
(a)
128
World Soil Resources and Food Security
400
Date (d)
Date
0
4/1/02
2
2/1/02
4/1/02
2/1/02
12/1/01
10/1/01
8/1/01
4/1/01
6/1/01
50
12/1/01
100
4
10/1/01
150
6
8/1/01
200
6/1/01
250
CR01 F1 F2 F4 F6 NF
8
4/1/01
Metabolic quotient, qCO2 (mg C g Bc–1 h–1)
300
0
4/1/02
Date
(c) Microbial biomass N (mg N kg–1)
2/1/02
0
12/1/01
200
4/1/02
2/1/02
12/1/01
10/1/01
8/1/01
0
6/1/01
500
600
10/1/01
1000
800
8/1/01
1500
1000
6/1/01
2000
4/1/01
Salt-extractable C (mg C kg–1)
(b)
4/1/01
Microbial biomass C (mg C kg–1)
(a)
Date
FIGURE 4.33 Seasonal fluctuation of microbial biomass C (a), salt-extractable C (b), microbial biomass N (c), and metabolic quotient (d) during the experiments.
an inverse fluctuation of biomass C and salt extractable C indicates an increase of microbial debris during the dry season due to exposure to a strongly fluctuating environment (such as the soil temperature), with a limited vegetation cover. According to Figure 4.33c, microbial biomass N was the lowest in CR01, whereas the highest was seen in NF. Microbial biomass N exhibited an increasing trend throughout the rainy season, probably due to continuous assimilation of N. Similarly, the value of qCO2 was the highest in the late rainy season (Figure 4.33d) and then sharply dropped at the start of the dry season. Soil respiration also decreased, whereas the microbial biomass C level remained high, except for CR01 (Figure 4.33a). It is considered that a large proportion of the soil microbes could have survived in the forested soils without a drastic decrease of biomass during the dry season, presumably by suppressing their activity.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
129
4.8.5 S ubstrate-Induced Microbial Activities (Respiration and N Assimilation/Nitrification) by Short-Term Laboratory Incubation To analyze the behavior of the microbial community as a controlling factor of N dynamics in shifting cultivation in more detail, short-term responses of soil microbes after the addition of C and N substrates were comparatively investigated for forest and cropland soils from the study village using fresh soil samples collected in March 2003 (late dry season), early June and July 2003 (rainy season), from a field that was used for rice cropping in 2003 (CR03), a field in its second year of fallow (CR0103), a field in its sixth year of fallow (F4 03), and a seminatural forest stand (NF03). The latter three are the same plots as those used for the field experiments in 2001. Then substrate-induced microbial activities were traced using a fresh soil after addition of glucose (equivalent to 4000 mg C kg−1 soil) with or without 0.0168 g of NH4NO3 (equivalent to approximately 400 mg N kg−1 soil). In treatment with NH +4 , enough N was added to eliminate the microbial activity limitations that occur as a result of N shortage. Then the mineralized C was measured continuously up to 94 h at 25°C. Similarly short-term N transformation after addition of NH +4 –N as an N source was traced up to 168 h at 25°C under aerobic conditions. Figure 4.34a compares cumulative CO2 emissions after the adjustment of glucose levels with or without N in the cropland (CR03) and forest (NF03) soils. Irrespective of N addition, glucose-induced respiration in NF03 was significantly delayed compared with that in CR03. In the case of CR03, N addition resulted in further acceleration of respiration, suggesting that the soils of CR03 originally did not contain enough N for the microbial community to demonstrate its maximum potential or the microbial community could utilize additional N efficiently for multiplication of the community. However, such a trend was scarcely observed for the forest soil, NF03. A higher and more efficient utilization of additional C and N resources is obvious in CR03 compared with NF03. Such a response of the microbial community was observed only in CR03 during the rainy season (after burning and moistening) (Figure 4.34b). The soils from the second and sixth years of fallow forests (CR0103 and F4 03) showed a similar trend as NF03. Even in CR03, the soil collected before the burning event (but after forest clearcutting) did not exhibit active respiration after glucose addition with N. Therefore, the unique property of the microbial community in CR03 was probably introduced after slash-and-burn of fallow forest. At the end of the incubation period (42 h), 2170 mg kg−1 of soil C was mineralized in CR03 (in the +C+N treatment), which amounted to 54% of added C. Coody et al. [1986] reported that, using a 14C-labelling technique, 27%–44% of added glucose was mineralized and most of the remaining glucose was already assimilated by soil microbes within 96 h at 25°C. It is highly probable in this experiment that the added glucose was almost completely consumed, either through respiration or via assimilation by soil microbes—even considering that our results may involve some degree of priming effect and, therefore, an overestimation of substrate decomposition. On the contrary, the soils from fallowed (CR0103 and F4 03) or matured (NF03) forests exhibited slow decomposition rates compared with CR03, and the amounts of mineralized C at 42 h were only 8%, 21%, and 13% of added glucose, respectively. Since
130
World Soil Resources and Food Security
Cumulative CO2 emission (mg C kg–1)
(a) 2500 2000 1500 1000
100
CR03, June, +C CR03, June, +C+N NF03, June, +C NF03, June, +C+N
50
CR03, June, +C+N CR0103, June, +C+N F403, June, +C+N NF03, June, +C+N Cr03, March, +C+N NF03, March, +C+N
500 0
0
20
40
60
Incubation period (h)
80
Cumulative CO2 emission (mg C kg–1)
(b) 2500 2000 1500 1000 500 0
0
10
20
30
Incubation period (h)
40
FIGURE 4.34 Short-term CO2 emission (cumulative) after glucose addition, with or without NH +4 to the fresh soils; (a) effect of NH +4 addition to cropped (CR03) and forest (NF03) soils, and (b) comparison of the soils in different land-use stages.
the possibility of a deficiency in N was practically eliminated in this experiment, such a slow utilization of the added C could be attributed to a specific property of the microbial community in these forest plots. Figure 4.35 demonstrates the fluctuation of levels of inorganic nitrogen (NH +4 and NO3−) after the addition of NH +4 in the incubation experiment. The nitrifying activity was detected only in the CR03 samples collected in the rainy season (June and July). Minimal nitrifying activity, if any, was traced in the soils from fallowed (CR0103 and F4 03) or matured forest (NF03). As the concentration of NH +4 increased within the period of incubation (168 h), gross mineralization of organic N was superior to gross immobilization of NH +4 . So, in the forested plots including in the initial stage of fallow (CR0103 in the 2nd year), both the nitrifying activity and the NH +4 assimilation are noticeably low. In contrast, judging from the fact that the sum of the remaining NH +4 –N and NO3−–N was decreasing during the experiment, N immobilization also occurred at the same time as active nitrification in cropland (CR03). These results
131
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
200
CNTL 0h 72 h 168 h
0
CNTL 0h 72 h 168 h
100
March
200 100 0
June
July
300 200
CNTL 0h 72 h 168 h
0
CNTL 0h 72 h 168 h
100
Time after addition of NH4
NO3–N NH4–N
NF03
(d) 400
March
June
July
300 200 100 0
CNTL 0h 72 h 168 h
March
Time after addition of NH4
NH4–H and NO3–N (mg kg–1)
400
CNTL 0h 72 h 168 h
NH4–H and NO3–N (mg kg–1)
NO3–N NH4–N
F403
July
300
Time after addition of NH4
(c)
June
CNTL 0h 72 h 168 h
300
400
CNTL 0h 72 h 168 h
July
CNTL 0h 72 h 168 h
June
CNTL 0h 72 h 168 h
March
NO3–N NH4–N
CR0103
CNTL 0h 72 h 168 h
400
(b) NH4–H and NO3–N (mg kg–1)
NO3–N NH4–N
CR03
CNTL 0h 72 h 168 h
NH4–H and NO3–N (mg kg–1)
(a)
Time after addition of NH4
FIGURE 4.35 Short-term transformation of NH4 –N added to the fresh soils.
were generally consistent with the fact that, in most cropped ecosystems, NH +4 is mineralized and then readily nitrified to form NO3− if it is not immobilized again rapidly. In contrast, in forest soils, such an active nitrification is sometimes retarded due to low activity of nitrifying bacteria [Robertson 1982].
4.8.6 Dynamics of Microbial Activities during Different Stages of Shifting Cultivation and the Function of the Fallow Phase Figure 4.33d shows that the values of qCO2 in CR01 were much higher than those in the other plots, indicating a higher activity of soil microbes per unit volume of biomass despite apparent lower biomass there (Figure 4.33a). This is partly because the method used in this study (the chloroform fumigation extraction method) could evaluate the whole biomass including any inactive biomass that may be dominant in the forest soils. Two explanations are possible for this apparent high microbial
132
World Soil Resources and Food Security
activity in CR01. First, this may be an effect of disturbance, as well as slash-andburn, in CR01. Mamilov and Dilly [2002] reported that stress, such as drying and rewetting, causes high microbial activity, which is expected to be more pronounced in the field just after forest reclamation (as in CR01). According to Wardle and Ghani [1995], who described how qCO2 could be used as a bioindicator of disturbance, the cropland ecosystem in the present study (CR01) is considered to be a more disturbed one than the forested ecosystems. Second, the soil microbial community in CR01 drastically changes after slash-and-burn of the forest due to a sharp increase in soil pH along with ash addition. As was discussed previously, a burning event generally produces higher pH values in soils a few years later than is found in fallow or matured forest in the same area (Figure 4.20a). In any case, the distinct difference in the values of qCO2 between the cropland (CR01) and the others suggests some essential difference in composition and/or function of soil microbial communities. On the other hand, soil solution composition revealed that, in the cropland (CR01) soil, a fairly high concentration of NO3− was released into the soil solution, unlike in the forest soil (NF) (Figure 4.32). The values for fallow forests were between these. A small stand of upland rice with possibly poor nutrient-uptake ability in CR01 during the early rainy season may further increase the chance of NO3− leaching. Ellingson et al. [2000] and Neill et al. [1999] also reported that deforestation and burning increased the concentration of NO3− in the soils in Mexican tropical dry forests and Amazonian forests, respectively. There are two explanations in which a higher amount of NO3− could be released into the solution in CR01 in the present study: there was higher nitrification activity of soil microbes in CR01, and/or higher N immobilization (assimilation) activity of soil microbes in NF. As shown in Figure 4.34, the laboratory incubation experiment for analyzing the substrate-induced respiration clearly showed that the soil microbes in the cropland (CR03) soil responded very quickly to addition of glucose, unlike fallowed or matured forest soils (CR0103, F4 03, and NF03). In the same way, soil microbes rapidly nitrified NH +4 added to NO3− in CR03, while partially assimilating the NH +4 at the same time (Figure 4.35). In the remaining plots, neither accelerated glucose consumption nor active NH +4 utilization were observed. Therefore, all the microbial activities that were tested in CR03 were induced quite rapidly after the addition of the substrates. These properties of the microbial community in CR03 were probably introduced after slash-and-burn of fallow forest. This, together with the fact that a higher qCO2 was observed despite an apparent lower microbial biomass in CR01 (Figure 4.33), indicates that the microbial community in the cropped soils consists of a smaller, but more active, number of microbes compared with that in the fallowed or matured forests, in which a larger number of soil microbes coexisted with a low activity. The main reason for suppressing NO3− release into the soil solution is therefore a lower nitrifying activity of the microbial community in the forest soils than in the cropped soils. Such a high nitrifying activity in cropped soils under shifting cultivation is also reported by Tulaphitak et al. [1985a] and Tanaka et al. [2001]. On the other hand, the contribution of reimmobilization of NH +4 once mineralized is secondary to suppressing NO3− release in the forest soil, as in NF03 and the fallowed
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
133
plots (CR0103 and F4 03), where the microbial response for immobilization of added NH +4 was slow in the incubation experiment. Conversely, the seasonal fluctuation of microbial biomass C and salt extractable C suggests that the microbial community in CR01 was easily destroyed after sharp drought conditions during the dry season. With a succession of secondary vegetation, such a community seemed to be replaced by a more stable one, which did not show a clear decrease of microbial biomass C, even in the dry season. At the same time, it does not increase, or multiply, rapidly even under favorable conditions such as the addition of substrates, as demonstrated in the short-term incubation experiment. The increasing trend of microbial biomass N that is observed during the rainy season (Figure 4.33c) is considered to be a result of slow utilization of N by the microbial community and may contribute to accumulation of N into the SOM pool.
4.8.7 Main Functions of the Fallow Phase in Shifting Cultivation by K aren People in Northern Thailand Based on the results obtained in this study, the functions of the fallow phase in shifting cultivation that have ensured the long-term sustainability of the system can be summarized as follows: First, as discussed in Section 4.6, some soil properties relating to soil acidity improve simultaneiously as the SOM-related properties increase in the late stage of fallow. The litter input may be supplying bases (obtained by tree roots from further down the soil profile) to the surface soil. This simultaneous increase in SOM and bases in the surface soil, through forest-litter deposition in the late stage of fallow, has an increasing effect on nutritional elements. Second, the decline in soil organic C during the cropping phase could be compensated by litter input during 6–7 years of fallow. With regards to the overall budget, the organic matter input through incorporation of initial herbaceous biomass into the soil system after establishment of tree vegetation (approximately in the fourth year) was indispensable for maintaining the SOM level. Third, the succession of the soil microbial community from rapid consumers of resources to stable and slow utilizers, along with establishment of secondary forest, retards further N loss through leaching and enhances N accumulation into the forest-like ecosystems. It is noteworthy that, during the fallow period, nitrifying activity of soil microbes, which was once activated in the cropping phase, is apparently suppressed. As a result, NO3− effluent from soil layers was remarkably low, even in the initial stage of fallow. The functions of the fallow phase listed above can be considered essential to the maintenance of this forest-fallow system. Agricultural production can therefore be maintained with a relatively short fallow period of around 10 years. Traditional shifting cultivation in the study village can be seen to be well adapted to the respective soil-ecological conditions. Socioeconomic conditions are, however, drastically changing, making it difficult to sustain this system. Under such conditions, we should search for alternative technical tools that could maintain the SOM level, suppress the nitrifying activity of soil microbes and avoid depletion of bases while mitigating soil acidification. This is imperative if the subsistence agriculture seen in this village is intended to continue in the near future.
134
World Soil Resources and Food Security
4.9 F ACTORS CONTROLLING SOIL ORGANIC MATTER DECOMPOSITION IN SMALL HOME GARDENS IN DIFFERENT REGIONS OF INDONESIA In the Republic of Indonesia, an intensive land management system of small home gardens, or pekarangan, has been developed mainly on Java Island. This land management system is a kind of agroforestry system that allocates tree vegetations and annual crops in a small space and it has both economic and environmental significance for small-scale farmers [Wiersum 1982; Jensen 1993; Hayashi and Ochiai 2004]. On Java Island, land resources are very intensively managed under high population pressure, whereas on surrounding islands such as Kalimantan, degradation of forest and land resources has been accelerated and has become a serious environmental problem. Such degradation is the result of extensive timber cutting and/or government policies for transmigration that aim to distribute the population throughout the entire nation [Sunderlin and Resosudarmo 1996]. Decreases in SOM content are often observed in the degradation processes that occur with conversion of land use from forest to agricultural land (Section 4.5). We consider that SOM dynam ics established in pekarangan, which has a long history on the native islands of many transmigrants, would provide important insights into SOM management strategy in different environments. Even though the practice of pekarangan has been naturally limited to a small farm scale, such information could be useful for tackling the problem of widespread land degradation. To obtain basic information for land management in terms of SOM dynamics under the pekarangan system, the main objectives of the present study were to 1) make a quantitative analysis of field soil respiration, which is derived from SOM decomposition and plant-root respiration; 2) analyze factors that control soil respiration under different climatic conditions in the tropics, i.e., rainforest and savanna climates; and 3) compare different land-use patterns that include different densities of tree species under the pekarangan system.
4.9.1 Description of Study Sites For the intensive study of soil respiration, we installed nine experimental plots at farmers’ home gardens in four regions of Indonesia that differed in terms of climate and geology. The regions were named as follows: BGF and BGC in Bogor, West Java; LBF and LBC in Lembang, in the highlands of West Java; PCF and PCU in Pacet, East Java; and SMF, SMFC, and SMC in Samarinda, East Kalimantan (Figure 4.16). General climatic and geological conditions of the study plots are given in Table 4.18. In each region, two farmers’ home gardens that contained predominately tree species or annual crops were selected for the present study, and are hereafter termed as –F and –C, respectively, except for the plots at Samarinda. For Samarinda, we selected three plots (SMF, SMFC, and SMC), with a decreasing domination of tree species in this order. The home gardens were generally as small as around several hundred square meters. In the F plots, predominated trees reached >10 m in height and the soil surface was usually covered by the tree canopy and occasionally received substantial amounts of litter-fall, whereas, in the C plots, repeated disturbance and minimal residue input with weeding may accelerate SOM depletion. Soil properties
Region Bogor, West Java
Köeppen’s Classification Af
Lembang, West Java
Am/Csb
Pacet, East Java
Aw
Samarinda, East Kalimantan
Af
a b c
Location
Elevation (m)
Mean Annual Temperature (°C)a
S06°36′, E106°52′ S06°48′, E107°36′
250
25.3
3671
Andesite
1250
19.3
1394
350
24.8
2139
Volcanic ejecta, intermediate to mafic Andesite
<50
26.1
2468
S07°37′, E112°32′ S00°23″, E117°07′
Monitored at our experimental plots. Soil Survey Staff [2006]. World Reference Base for Soil Resources [FAO 2006].
Annual Precipitation (mm)a
Parent Materials
Tertially sedimentary rock
Soil Classification (US Taxonomyb/ WRBc)
Name of Plots
Typic Haplohumults/ Haplic Acrisols Typic Melanudands/ Melanic Andosols
BGF, BGC
Lithic Haplustalfs/ Leptic Lixisols Typic Paleudults/ Haplic Alisols
PCF, PCC
LBF, LBC
SMF, SMFC, SMC
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.18 Geological and Climatic Conditions of the Intensive Study Plots in Indonesia
135
136
TABLE 4.19 General Properties of Surface Soils (0–5 cm) at the Monitoring Plots Bulk Density Plot
a b c d
pH
(H2O)
(KCl)
(g cm−3) 1.09 0.88 0.73 1.00 1.29 1.19 1.06 1.26 1.22
Particle Size Distribution
Exchangeable Basesa
Al
Sand
(cmolc kg−1) 4.91 5.09 5.56 6.57 6.29 7.83 6.17 5.30 5.40
4.47 4.53 5.17 6.03 5.27 7.25 5.72 4.45 4.40
10.44 6.39 14.88 19.51 10.15 20.07 14.66 9.88 12.59
0.82 0.82 0.00 0.00 0.00 0.00 0.00 0.26 0.56
Silt
Available Pb
Clay
(%) 15 48 60 36 73 57 69 32 31
33 21 33 29 15 23 16 32 34
51 31 6 36 12 21 15 36 35
(as P2O5)
Total C
Total N
Microbial Biomass Cc
C0d
C0/TC
(mg kg−1)
(g kg−1)
(g kg−1)
(mg kg−1)
(mg kg−1)
(%)
31.4 21.9 68.1 29.0 14.6 12.7 42.1 20.0 10.0
2.98 2.11 5.62 2.86 1.18 1.10 2.88 1.71 1.20
526 211 145 119 121 65 209 225 68
2355 948 1108 1368 1283 1712 2938 2460 798
7.5 4.3 1.6 4.7 8.8 13.5 7.0 12.3 8.0
61.0 118.6 182.8 519.2 224.9 1303.9 63.4 18.6 241.2
Sum of Na, K, Mg, Ca, and NH4. Determined by Bray-II method. Determined by chloroform fumigation-extraction method [Vance et al. 1987]. Determined by aerobic incubation method followed by nonlinear fitting supposing the first-order kinetic model (see Appendix 4.A4).
World Soil Resources and Food Security
BGF BGC LBF LBC PCF PCC SMF SMFC SMC
pH
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
137
are given in Table 4.19. Total C contents were usually higher in the F plots than in the comparable C plots. Both the soil pH and the content of exchangeable bases in SM soils were high compared to forest soils under shifting cultivation in comparable regions in East Kalimantan (Section 4.6). The available P contents were especially high in annual crop-based gardens. These findings suggest that the home gardens studied had received considerably high amounts of manure, which is in fact a common observance in pekarangan.
4.9.2 Field Measurement of Soil Respiration The field soil respiration rate was determined at each of the plots using the closed chamber method, essentially following the guidelines of Anderson [1982], on a total of 15 to 18 occasions between February 2000 and August 2001. At the same time, soil temperature and volumetric water content were measured. Soil temperature (at a depth of 5 cm) and moisture (at depths of 0–15 cm) as well as meteorological data such as rainfall and air temperature were also monitored using data-loggers at BGF, LBF, PCF, SMFC, and SMC. To estimate total soil respiration throughout the year, the procedure discussed in the previous section was used; we first established an equation relating the in situ hourly soil respiration rate to environmental factors such as soil temperature and moisture by multiple regression analysis. We then calculated hourly soil respiration rates using the continuously monitored environmental data, summing these hourly results over a given period. Figure 4.36 gives the soil respiration rate (in kg C ha−1 d−1) measured at each field during the study period. Soil respiration rates at some plots (PCF and PCC) clearly dropped during the dry season, indicating a relatively large annual fluctuation of soil respiration at these two plots. In LBF and LBC situated in the highlands, soil respiration rates were markedly low throughout a year compared to the other plots located in low elevation regions. In some cases, there is a daily fluctuation of the soil respiration rate as well as soil temperature with a weak correlation between them (data not shown). Using the respiration data and soil temperature and moisture measured at the same time, parameters relating to the temperature and/or moisture dependency of soil respiration are calculated as shown in Table 4.20. In the case of BGF, no significant parameters were obtained. Similarly, the parameters obtained for SMF poorly explained the dataset (r 2 = 0.08). These two plots are situated under forest-like vegetation in a rainforest climate, indicating that the fluctuation of climatic conditions (e.g., soil temperature and moisture) is somewhat too limited to yield a clear seasonal trend of the soil respiration rate. However, in the adjacent cropland-like plots (BGC, SMFC, and SMC), the contribution of the moisture parameter, b, is more significant, indicating that periodic desiccation due to occasional tillage may cause the fluctuation of the soil respiration rate. The moisture parameter contributes to the soil respiration rate more determinatively for PCF and PCC (r 2 = 0.75 and 0.86, respectively), which are situated under a tropical savanna climate (Table 4.18). On the other hand, the coefficient relating to soil temperature (E) was often rejected (p < 0.15), although not for data obtained for the Lembang and Pacet plots. The high contribution of the temperature factor on soil respiration rates in temperate zones is often reported
100
Mar-02
Dec-01
SMF, measured SMF, calculated SMFC, measured SMFC, calculated SMC, measured SMC, calculated
SM
200
Sep-01
Jun-01
Mar-01
Dec-00
Sep-00
Date
Jun-00
0
Mar-00
50
150 100
FIGURE 4.36 Comparison of measured and simulated soil respiration rates at the experimental plots.
Mar-02
Sep-01
Jun-01
Dec-01
Date
Mar-01
0
Dec-00
50
Sep-00
Mar-02
Dec-01
Sep-01
Jun-01
Dec-00
Mar-01
Date
Sep-00
0
Jun-00
50
LBF, measured LBF, calculated LBC, measured LBC, calculated
150
Jun-00
100
LB
Mar-00
PCF, measured PCF, calculated PCC, measured PCC, calculated
150
200
World Soil Resources and Food Security
PC
200
Soil respiration (kg C ha–1 d–1)
Mar-02
Dec-01
Sep-01
Jun-01
Mar-01
Dec-00
Sep-00
Jun-00
0
Mar-00
50
Soil respiration (kg C ha–1 d–1)
100
Mar-00
Soil respiration (kg C ha–1 d–1)
BGF, measured BGF, calculated BGC, measured BGC, calculated
150
Date Soil respiration (kg C ha–1 d–1)
BG
138
200
Parameters for Regression Equation Determining Cem Plot
n
BGF BGC LBF LBC PCF PCC SMF SMFC SMC
17 18 17 18 15 15 17 17 17
lna 4.46*** 36.96*** 48.22*** 47.86** 41.04*** 3.62*** 4.48*** 27.70***
b
r2
Annual Soil Respiration (Mg C ha−1)
Annual Average Soil Temperature (°C)
Annual Average Soil Moisture (L L−1)
0.45*** 0.33*** 0.47*** 0.75*** 0.86*** 0.08* 0.29** 0.66***
20.45a 20.06 12.72 9.92 26.81 26.52 28.69 25.51 23.96
25.1 25.7 19.9 20.9 24.9 27.8 26.1 27.0 28.3
0.497 0.342 0.280 0.327 0.270 0.247 0.415 0.355 0.358
E (kJ mol–1)
No parameters were determined. 1.41*** −b b − 83.9*** 0.51* 110.9*** 1.38*** 106.1* 1.88*** 88.4*** 0.35* −b 1.24** −b 1.61*** 57.5**
Note: Significant at *15%, **5%, and ***1% levels, respectively. a The value was determined based on the average of measured soil respiration. b The parameter was rejected in the stepwise regression (p = 0.15).
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
TABLE 4.20 Parameters Determined by Stepwise Multiple Regression Analysis Using Field Data and Applying the Equation: Cem (in mmol C m−2 h−1) = aθb exp (−E/RT) and Its Legal Form: in Cem = lna + blnθ − E/RT
139
140
World Soil Resources and Food Security
[Shinjo et al. 2006], and partially explains the cases at Lembang in the highlands. Under the tropical climate in the present study, it therefore appears that the effects of seasonal temperature fluctuations on soil respiration rates were rather limited. Using the regression equations and data from the data-loggers, fluctuations of soil respiration rates during one consecutive year were simulated (Figure 4.36; solid lines). Cumulative soil respiration throughout the period was calculated to range from 9.9 (LBC) to 28.7 (SMF) Mg C ha−1 (Table 4.20). As the monitored data were available for five of the nine plots, fluctuations in soil temperature and moisture for the remaining four plots (BGC, LBC, PCC, and SMF) were assumed based on the relationship between the actual data and monitored data of adjacent plots (i.e., BGF, LBF, PCF, and SMC). In the case of BGF, for which no significant parameters were obtained, the average value of the obtained data was used for calculation. The values obtained here appear to be somewhat high compared to those reported previously for other tropical regions in Asia, for example, 8.5 Mg C ha–1 yr–1 in the rainforests of peninsular Malaysia [Kira 1976], 7.4 Mg C ha−1 yr−1 in savanna forests of northeastern Thailand [Tulaphitak et al. 1985b], and 7.7 to 8.6 Mg C ha−1 yr−1 in secondary forests of northeastern Thailand [Funakawa et al. 2006b], all of which are estimations excluding plant root respiration. If estimations include plant root respiration, the annual soil respiration in the latter report increases to 14.8 to 15.0 Mg C ha−1 yr−1, with significant suppression during the dry season. According to the review of Raich and Schlesinger [1992], the annual soil respiration ranges from 8.9 to 15.2 Mg C ha−1 yr−1 (n = 10) in tropical and subtropical moist forests and the values in the present study are highest. When considering the Cem values at the fixed condition (T = 298 K and θ = 0.3 L L−1), however, the values obtained in the present study, i.e., 37.8 to 82.5 kg C ha−1 d−1, are not necessarily higher than those calculated for secondary forests in northern Thailand in our previous study (27.6 to 68.9 kg C ha−1 d−1) [Funakawa et al. 2006a], suggesting that prolonged favorable conditions of soil temperature/moisture may be primarily responsible for the high annual soil respiration in the present study. Another possible explanation for such high soil respiration in the present study is that a high soil fertility of home gardens may increase the SOM decomposition rates and, possibly, primary production when compared with natural ecosystems, as introduced earlier. Cleveland and Townsend [2006] observed that P or N fertilization of tropical forests increased soil respiration, though a controversial result was also obtained in another study [Olsson et al. 2005]. It is still necessary to specify the reasons for, as well as the processes involved in, the possible rapid C cycling in home gardens.
4.9.3 Laboratory Incubation to Determine Soil Respiration Rate with Microbial Origin To compare the soil respiration rate originating solely from microbial respiration between each of the study plots, laboratory incubation was carried out using undisturbed core soil samples. In total, five or six core samples were collected from surface 5-cm layers of soils at each of the nine plots. After the volumetric moisture contents of the undisturbed soil samples were adjusted to different levels within the
141
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
available range of the actual field conditions by slow drying in a refrigerator, the amounts of mineralized C were determined several times up to 40 days under aerobic incubation at a constant temperature of 25°C (or 20°C for LBF and LBC). After determination of the C0 and k values obtained by the simulation using the first-order kinetic model [Ct = C0{1 − exp(−kt)}], the initial CO2 emission rates (Cem) under different moisture conditions are calculated to be kC0. Based on the linear relationship observed between the volumetric soil moisture content of the core samples and Cem, moisture parameters for determining Cem were obtained (Table 4.21). In this case, the values of Cem calculated for the fixed condition (T = 298 K and θ = 0.3 L L−1) varied widely, with apparently higher values for the forest-like plots in each of the regions, except for the Pacet plots. Figure 4.37a shows the relationship between Cem values under the fixed condition determined by field and laboratory incubation data. Only the values from PCF and PCC disturb the linear relationship, suggesting that the contribution of roots to the whole respiration is noticeably high under tropical savanna climates. According to the review by Jackson et al. [1996], a higher root/ shoot ratio is observed for tropical deciduous forests (0.34) compared to evergreen forests (0.19), though there is no clear difference in the root biomass itself (4.9 and 4.1 kg m−2, respectively). Although our F plots, including PCF, are covered mostly by evergreen species, such a difference in root distribution as cited above may affect the relative contribution of plant roots to soil respiration under different climates. Using these parameters, hourly soil respiration originating solely from microbial respiration is simulated after temperature correction using the Q10 relationship,
TABLE 4.21 Parameters for Determining Cem (kg C ha−1 d−1) under Different Moisture Contents by Incubation of 100-mL Undisturbed Core Samples from Surface 5-cm Layers Parameters for Determining Cem Cem0 = aθ + b Plot BGF BGC LBF LBC PCF PCC SMF SMFC SMC a b
a
b
r2
n
Cem at T = 298K and θ = 0.3 L L−1 (kg C ha−1 d−1)
86.49 34.59 32.43a 15.49a 19.40 23.99 68.94 61.97 19.41
−13.15 −3.91 0.47a 0.04a −0.97 1.38 −1.32 −6.15 −1.75
0.99 0.90 0.64a 0.64a 0.89 0.73 0.67 0.96 0.87
6 6 5 6 6 5 5 6 5
12.80 6.47 14.42b 6.63b 4.85 8.58 19.36 12.44 4.07
The parameters were determined by the incubation at 20°C. The value of Cem was corrected under the supposition of Q10 = 2.
Annual Soil Respiration from Surface 5-cm Layers of Soil (Mg C ha−1) 10.51 2.94 3.09 1.77 1.54 3.24 10.71 6.59 2.37
142
World Soil Resources and Food Security
Cem at T = 298K and θ = 0.3 L L–1 determined by incubation experiment (kg C ha–1 d–1)
(a)
30 25 20 15 10
PCC PCF
5 0
0
20
40
60
80
100
Cem at T = 298K and f Æ = 0.3 L L–1 determined by field measurement (kg C ha–1 d–1)
Annual soil respiration estimated by incubation experiment (Mg C ha–1)
(b)
40
1:1
30 20 10 0
0
10
20
30
40
Annual soil respiration determined by field measurement (Mg C ha–1)
FIGURE 4.37 Comparison of soil respiration rates at fixed conditions (298 K, 0.3 L L−1) (a) and annual soil respiration (b) determined by field measurement and laboratory incubation using 100-mL undisturbed soil core samples. ▲, 0–5 cm; ●, estimation for 0–15 cm; ▼, 0–30 cm.
then summed for one consecutive year to calculate annual soil respiration (Table 4.21). Since the core samples were collected from surface 5-cm layers of soils, the annual soil respiration would be underestimated due to the contribution of deeper soil layers. In Figure 4.37b, therefore, possible CO2 emission from surface 15- and 30-cm layers of soils is estimated based on the SOM distribution pattern in the soil profiles and compared with the annual soil respiration determined by field data. Assuming that the soil respiration originated only from surface 5-cm layers of soil, the microbial respiration estimated based on the incubation experiment would correspond to only 22% of field soil respiration on average (Table 4.20). If we extend the depth up to 15- and 30-cm layers of soils based on the distribution of total C stock in the soils, the ratio increases by 55% and 98%, respectively, of the field soil
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
143
respiration (Figure 4.37b). According to Nakane [1980], approximately 50% of whole soil respiration would be derived from plant root respiration. In previous papers, we obtained similar results for fallow forest of shifting cultivation in northern Thailand (31%–71%) and for tree plantations on sandy soils in Northeast Thailand (52%–57%) [Funakawa et al. 2006a, 2006b]. Raich and Schlesinger [1992] estimated that, at a minimum, 24% of the soil respiration might be derived from plant root respiration based on the difference of net primary production and soil respiration, whereas Hanson et al. [2000] gave a higher value of percent root contribution to total soil respiration throughout an entire year or growing season as 45.8% and 60.4% for forest and nonforest vegetation, respectively. Although comparative measurements by labeling techniques for determination of the relative contribution of roots to whole soil respiration are scarce in tropical forests, it is reported to be 20% for 2-year-old poplar trees [Horwath et al. 1994] and 55% in temperate forests in the United States [Andrews et al. 1999]. Given these previously reported values, it would be reasonable to assume that the surface 15-cm layers of soils, which accounted for 55% of the whole soil respiration measured in the field, is primarily responsible for the CO2 emission derived from microbial decomposition of SOM.
4.9.4 Extensive Survey to Determine Distribution Patterns of SOM- Related Properties of Soils in Java and East K alimantan To estimate the general distribution patterns of soil respiration in wide areas of Java and East Kalimantan based on soil properties and climatic/geological parameters such as SOM content, particle size distribution, mean annual temperature (MAT), and mean annual precipitation (MAP), extensive soil sampling was carried out in the relatively dry season of August 2000. A total of 53 soil samples were collected from surface 5-cm layers of soils as composite samples from fields in landscapes that differed in terms of relative domination of tree species, i.e., forest, pekarangan, and annual-crops field. Samples were then subjected to chemical analysis. The analytical data obtained for the soils from the extensive survey are summarized for each region in Table 4.22, together with parameterized climatic conditions. The data includes nine monitoring plots. The survey covered a wide range of elevations in Java (i.e., 0 to 1800 m above sea level), as well as different geology and land uses. The values of MAT and MAP for each plot are cited from the nearest meteorological station and MAT is corrected with the assumption of the decreasing rate of temperature with increasing elevation as 0.6°C/100 m. According to Table 4.22, MAT is highest in the plots of East Kalimantan, whereas the values of MAP are higher in West Java than in Central or East Java. Soil pH is lowest in East Kalimantan, followed by West, Central, and East Java. Higher precipitation, as well as parent materials (mostly Tertiary sedimentary rocks), affect lower pH values of the soils in East Kalimantan. On the contrary, the higher pH of the soils in Central and East Java might derive from parent materials (volcanic origin) and drier climatic conditions. It is notable that SOM-related properties (i.e., total C, total N, and C0) do not exhibit a clear regional trend, presumably because different land uses may increase the variation of these properties within respective regions.
144
TABLE 4.22 Average Values of Meteorological Indices and Soil-Analytical Data of the Study Plots in Extensive Survey MAT
Regions
AVE
MAP
SE
AVE
(°C)
pH(H2O)
SE
AVE
Clay Content
SE
AVE
(mm)
22 14 8 18
23.4 23.7 24.4 26.8
0.5 1.0 1.0 0.1
62
24.6
0.4
a a ab b
SE
AVE
(%)
2824 2331 2207 2717
161 94 47 27
2602
68
a b b ab
5.62 6.13 6.34 5.04
0.17 0.22 0.42 0.13
5.66
0.12
ab a a b
Total C SE
AVE
(g kg−1)
45.5 36.8 33.8 27.2
4.5 6.3 4.6 3.0
36.7
2.5
a ab ab b
35.8 41.5 34.3 19.5
5.3 8.9 7.2 2.3
32.2
3.1
Note: The values with the same letters are not significantly different (p < 0.05). AVE, average; SE, standard error.
C0
Total N SE
AVE
(g kg−1) a a a a
2.97 3.38 2.36 1.58
0.45 0.71 0.56 0.17
2.58
0.25
SE
(mg kg−1) a a a a
1576 2110 2511 1287
208 580 591 191
1675
177
a a a a
World Soil Resources and Food Security
West Java Central Java East Java East Kalimantan All
Number of Samples
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
145
To analyze factors that can control the soil respiration rate and annual soil respiration, principal component analysis to summarize various environmental and soil factors and then stepwise multiple linear regression were conducted using the dataset presented in Table 4.22. Variables employed included land-use class, elevation, MAT, MAP, potential evapotranspiration (PEP), wetness (MAP/PEP), coefficient of variation (CV) for monthly precipitation, clay content, pH (H2O), total C content, and C0. The latter four variables were determined for surface 5-cm layers of soils. Table 4.23 shows the factor patterns for the first four principal components, which accounted for 85% of their total variances. The first component showed high coefficients, positive or negative, for elevation, MAT, PEP, wetness, and total C. These variables corresponded to the properties derived from the temperature gradient and resulting total C content of soils (i.e., decreasing elevation resulted in increased MAT and PEP and decreased wetness, as well as decreased total C pool). Hence, the first component was referred to as the temperature and decreasing SOM factor. As shown in Figure 4.38a and 38b, soils from the Java Island highlands—usually of tephra origin—seemed to show low scores for the first component, while those from East Kalimantan showed the highest scores. Figure 4.38b shows that soils derived from sedimentary rock or alluvial deposits were generally low in this component.
TABLE 4.23 Factor Pattern for the First Four Principal Components Relating to Soil and Environmental Variables Variable Land-use pattern Elevation Mean annual temperature Mean annual precipitation (MAP) Potential evapotranspiration (PEP) Wetness (MAP/PEP) CV value for monthly precipitation Clay content pH Total C content C0 Eigenvalue
PC1
PC2
PC3
PC4
0.00 −0.96 0.98 0.12 0.97 −0.79 −0.39 0.19 0.02 −0.73 −0.16 4.23 Temperature and decreasing SOM factor
0.85 0.06 −0.06 −0.12 −0.04 −0.05 0.05 0.22 −0.07 0.49 0.85 1.78 Vegetation and mineralizable C factor
0.10 −0.18 0.10 0.92 0.04 0.58 −0.46 0.41 −0.25 −0.21 −0.21 1.75 Precipitation factor
0.02 −0.04 0.01 −0.16 −0.12 −0.07 0.66 0.70 0.79 −0.02 0.06 1.59 Desiccation and high pH factor
146
World Soil Resources and Food Security (a)
3
West Java Central Java East Java East Kalimantan Java, > 1000 m
BGF
BGC
Scores of factor 3 (Precipitation factor)
2 1 0
–3
–2
–1
LBF
SMF
1 SMFC
0
SMC
–1
LBC
PCC
–2
2
PCF
Scores of factor 1 (Temperature and decreasing SOM factor) Sedimentary rocks and alluvial soils Limestone Volcanic rocks and sediments Tephra
Scores of factor 4 (Desiccation and high pH factor)
(b)
2
–3
PCC
1
LBC
–2
0
–1 LBF
–1
PCF
0 BGF
1SMC
BGC
SMF
2
SMFC
–2
Scores of factor 1 (Temperature and decreasing SOM factor)
(c)
Scores of factor 3 (Desiccation and high pH factor)
3
West Java Central Java East Java East Kalimantan Java, > 1000 m
3 2 PCC LBC PCF
–2
1 0
BGC SMC
0
LBF
–1
BGF
SMFC
2
4
SMF
–2
Scores of factor 2 (Vegetation and mineralizable factor)
FIGURE 4.38 Scattergrams of the factor scores determined for each study plot with different regions or parent materials.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
147
The second component showed positive coefficients for land-use patterns and C0, suggesting that under forest-like land use, the amount of mineralizable C increases due to the high addition of root and/or leaf litter to soils, as discussed in Section 4.5. Hence, this component can be referred to as the vegetation and mineralizable C factor. Since our extensive survey samples include a variety of plots in terms of land use, no regional trend is found for the distribution of this factor (Figure 4.38c). The third component showed a high coefficient for AP and was considered to be a precipitation factor. As shown in Figure 4.38a, the plots in West Java exhibit higher scores for this factor, followed by those in East Kalimantan and East/Central Java. The fourth component had a close relationship with the CV value for monthly precipitation, clay content, and pH of soils. A higher value of CV for precipitation indicates the presence of a distinct dry season, which often causes higher pH values in soils under monsoon climates compared to those under rainforest climates (Sections 4.4 and 4.6). Thus, the fourth component can be referred to as the desiccation and high pH factor. The scores are generally high among the plots derived from volcanic rock and sediment (mostly andesite) and from limestone, indicating that this factor involves the influence of parent materials of soils. Most of the variables were closely related to only one component with high coefficients above 0.6, with the exception of wetness, which was affected by both factors relating to temperature and precipitation. In the next step, stepwise multiple regression analysis (p < 0.25) was conducted to examine the contribution of each factor of annual soil respiration or Cem under the fixed condition (T = 298 K and θ = 0.3 L L−1), which was determined by field measurement or laboratory incubation data. The parameters obtained are given in Table 4.24 and are summarized as follows. Annual soil respiration based on field measurement is primarily determined by the temperature and decreasing SOM factor, indicating that—along with temperature increases—C turnover through both plant and soil microbial activities seems to be accelerated. Although the negative contribution of the precipitation factor is small, its ecological reason is not clear. In contrast, annual soil respiration determined by the laboratory incubation data positively influenced the vegetation and mineralizable C and desiccation and high pH factors, indicating that soil respiration with a microbial origin is mainly determined by the pool size of mineralizable C with positive influences from favorable pH soil conditions. On the other hand, Cem (at T = 298 K and θ = 0.3 L L−1) determined by the field conditions is poorly explained by the factors obtained (r 2 = 0.24, p < 0.25), while that determined from laboratory incubation data is mainly related to the vegetation and mineralizable C and temperature and decreasing SOM (negative) factors. Thus, as the SOM pool size increases, Cem at the fixed condition also increases.
4.9.5 S OM Dynamics in Soils Situated in Different Climatic and/or Geological Conditions Using the equations given in Table 4.24 and factor scores of each soil sample from the extensive survey, annual soil respiration determined by field measurements and laboratory incubation data are plotted against the total SOM stock (Mg C ha−1) in the
148
TABLE 4.24 Factors Affecting Soil Respiration Parameters Coefficients and Probability for Each Factor
Parameters
Constant
Factor 1
Factor 2
Factor 3
Factor 4
Temperature and Decreasing SOM Factor
Vegetation and Mineralizable C Factor
Precipitation Factor
Desiccation and High pH Factor
r2
For field measurement 21.482
***
6.265
***
–
–
−0.728
*
–
–
0.86
***
59.248
***
–
–
9.087
*
−5.560
*
–
–
0.24
*
Annual soil respiration from surface 5-cm layers of soil (Mg C ha−1) Cem at T = 298 K and θ = 0.3 L L−1 (kg C ha−1 d−1)
5.778
***
–
–
5.872
***
–
–
1.263
*
0.81
***
10.943
***
−1.872
*
7.163
***
–
–
–
–
0.62
**
Note: Significant at *25%, **5%, and ***1% levels, respectively.
World Soil Resources and Food Security
Annual soil respiration (Mg C ha−1) Cem at T = 298 K and θ = 0.3 L L−1 (kg C ha−1 d−1) For incubation experiment
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
149
surface 5-cm soil layers (Figure 4.39). The amount of SOM stock is calculated based on the total C content (g kg−1) and bulk density (g cm−3) estimated using the following equation, which was obtained for the surface soils in the monitoring plots:
Bulk density = 1.82 – 0.23 ln (total C content)
(r 2 = 0.59, p < 0.05).
(4.24)
Within each of the regions, despite differences in vegetation and SOM stocks in surface soil layers, annual soil respiration determined by field measurement is essentially identical (Figure 4.39a). Whole soil respiration is strongly controlled by climatic factors. As a result, annual soil respiration estimated for the plots shows a clear regional trend; that is, it is highest in East Kalimantan, followed by the low-elevation areas of Java, and then the high elevation areas of Java.
Annual soil respiration rate determined by field experiment (Mg C ha–1)
(a)
30
West Java PCC
25
SMF SMC
20
Central Java
SMF
PCF
East Java East Kalimantan Java, > 1000 m
BGF
BGC
15
LBF
10
LBC
5 0
0
50
100
150
Total C stock in surface 5-cm layers of soil (Mg C ha–1)
Annual soil respiration estimated by incubation experiment (Mg C ha–1)
(b)
Upland crops Upland crops with trees Trees with upland crops Forest 'T' indicates tephra origin
30 25
T
20
y = 0.72x – 3.73 r2 = 0.43 (Excluding 'T'plots)
15 10
SMFC
5
PC SM
0 –5
0
BGF BGC LBF, T PCP
10
T
20
T
SM T T LBF, T
T T
30
40
50
Total C stock in surface 5-cm layers of soil (Mg C ha–1)
FIGURE 4.39 Scattergrams between total C stock in surface 5-cm layers of soils and the amount of annual soil respiration estimated based on field (a) or laboratory incubation (b) experiments.
150
World Soil Resources and Food Security
On the other hand, annual soil respiration with microbial origin is higher under more forest-like land use (Figure 4.39b). However, plots under such forest-like use are also expected to receive more litter input and actually tend to accumulate higher amounts of SOM in the upper soil layers. It is notable that the soils derived from tephra accumulate exceptionally high amounts of SOM while suppressing SOM decomposition, indicating that these soils are very advantageous for SOM management by farmers.
4.9.6 Possible Land Management Systems in Different Regions of Java and East K alimantan As described above, annual soil respiration, including both plant root and microbial respiration, is highest in the East Kalimantan plots (Figure 4.39a), indicating that primary production is rapidly consumed through high biological activity and would not be readily accumulated as SOM in tropical rainforest climates. In Figure 4.39b, a relationship can be seen between SOM stock (SOM0–5) and annual soil respiration with microbial origin (ASR MB0–5) in surface 5-cm layers of soils, i.e., ASR MB0–5 = 0.72 SOM0–5 − 3.73, when excluding soils derived from volcanic tephra. This finding indicates that a high proportion of SOM stock, equivalent to 70%, can be decomposed within one year. Even though an additional litter supply may compensate for the loss of SOM, SOM stock can be easily exhausted after the conversion of land use from a more forest-like system to upland-cropping systems. A continuous input of organic materials into soils is therefore indispensable for maintaining the SOM level over a long period of time. Hayashi [2002] observed a gradual increase of soil fertility, including SOM levels, in home gardens of transmigrants, in which tree vegetation predominated, since the Imperata field had been cleared in South Kalimantan. High input of organic or inorganic resources may, however, have been easier through the use of external resources, such as domestic waste or crop residues from surrounding cropping fields, under traditional management of pekarangan in Java or Bali Island with high population density. Additionally, it is also notable that tephraderived soils, which are considered to be more resistant to rapid decreases in SOM because of their high potential for retaining SOM, are commonly found on these islands. Jensen [1993] counted the existence of a “medium fertile soil with large nutrient reserve” as one of the important requirements for sustainability of the home gardens in Java. Considering the lack of these advantages found in Java and Bali, additional efforts may be required to maintain the SOM level in home gardens in Kalimantan. Thus, land management—including a more forest-like system, which usually supplies higher amounts of litter-fall—is a more viable option in Kalimantan under present circumstances. It may also be applicable to land-use strategy on an entire regional scale when considering the high decomposition rate relative to the stock of SOM under nonvolcanic soils in East Kalimantan.
4.10 GENERAL DISCUSSION AND CONCLUSION As discussed so far, intensive land management, historically, has been available only in limited areas of upland soils in humid Asia, especially in tropical regions.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
151
There were several constraints that limited the development of intensive agriculture. From the aspects of soil management, rapid decomposition of SOM, high acidity, and probably high erodability should be counted as main difficulties. In the present work, the former two points are discussed in relation to ecosystem- and pedogeneticprocesses. Soils developed from volcanic tephra, mafic volcanic materials, or their secondary sediments are apparently advantageous to both SOM and soil acidity management. Although the volcanic materials mixed with 2:1 minerals often exhibited strong acidic natures, as typically observed in “nonallophonic” Andisols in Japan [Shoji et al. 1985], the distribution of such soils were virtually limited in the regions studied (i.e., Java and Sumatra Islands). As clearly described in Sections 4.5 and 4.9, SOM-related soil fertility is higher in the soils from volcanic materials than in others. Intensive agricultural management is, therefore, typically observed in this volcanic area. Besides Andisols and Cambisols, soils distributed in the tropics could generally be categorized into four groups with regard to land management and pedogenesis; Alisols with extremely high acidity (i.e., high exchangeable Al with high CEC), Acrisols with relatively lower contents of exchangeable Al, Arenosols (sandy soils), and Ferralsol-related soils (Ferralsols, Plinthosols, and probably Nitisols). Since the distribution of the Ferralsols group is limited in humid Asia, the remaining three compose the majority of soils in this region. The sandy soils may exhibit unique properties in terms of water and nutrient retention in soils, as well as drastic fluctuations of soil environments such as temperature and moisture, and hence, must be investigated separately in more detail. The main discussion in the present work is focused on the differences between Acrisols and Alisols. As discussed in Sections 4.2 through 4.4, from a pedological viewpoint, distribution patterns of Acrisols and Alisols in humid Asia can be related to the difference in the vermiculitization process of dioctahedral mica, which is explained mainly by climatic factors. That is, dioctahedral mica inherent to sedimentary rocks or felsic igneous rocks is thought to weather to form expandable 2:1 minerals (vermiculitization) under the lower pH conditions associated with the udic or perudic soil moisture regime, whereas, in contrast, mica is relatively stable under the higher pH conditions associated with the ustic soil moisture regime and other primary minerals such as feldspars dissolve to form kaolin minerals and gibbsite. The distribution pattern of clay mineralogy and pedogenetic-processes summarized above are considered to have restricted possible options in agricultural activities by human beings under low-input conditions. There are several differences between shifting cultivation systems on Alisols under rainforest climates and those on Acrisols under monsoon climates. In East Kalimantan of Indonesia, farmers traditionally planted upland rice for one year only after clearing and burning the forest cover, and then left the land for more than 20 years as fallow. Only on relatively fertile soils near rivers was cultivation of crops with a shorter fallow of around eight years practiced. Extremely high acidity and resulting N mineralization characteristics, as well as low P supplying potential and high decomposition rates of SOM (Sections 4.6 and 4.9), had forced farmers to keep relatively long periods of fallow with heavy dependence on forest ecosystems for fertility recovery of soils. Under this condition, even transformation of land use to a pekarangan system did not seem
152
World Soil Resources and Food Security
to be easy. Plantation enterprises for tree crops are progressively widespread today in this region, at the expense of forest reclamation. In contrast, in the shifting cultivation system on Acrisols under the monsoon climate in northern Thailand, the positive function of the fallow phase was more clearly observed, even after relatively short periods of fallow. Karen people in our study village usually planted upland rice only for 1 year, followed by approximately 7 years of fallow. In the late stage of fallow, some soil properties relating to soil acidity improved at the same time as the SOM-related properties increased. This simulta neous increase in SOM and bases in the surface soil, through forest-litter deposition in the late stage of fallow, had an increasing effect on nutritional elements. The decline in soil organic C during the cropping phase could be compensated by litter input during the 6–7 years of fallow. Moreover, the succession of the soil microbial community from rapid consumers of resources to stable and slow utilizers, along with establishment of secondary forest, retards further N loss through leaching and enhances N accumulation into the forest-like ecosystems. These functions of the fallow phase can be considered essential to the maintenance of the forest-fallow system under monsoon climates. Agricultural production can therefore be maintained with a relatively short fallow period of around 10 years. The shifting cultivation system under monsoon climates was demonstrated to be dependent upon relatively fertile soils with consciousness of possible risk for soil erosion on the generally steep slopes in the region. The development of infrastructure presently stimulates the agriculture activities in this region to move in more commercially oriented directions. Although it is difficult to find shifting cultivation activities in temperate zones, a set of data obtained in Japan indicated that neither C0 nor N0 fluctuate appreciably during the cropping and fallow phases, presumably because the northern temperate climate produces lower amounts of forest litter during the fallow period. Such a mild climate also prevented SOM from actively decomposing, allowing continuous cultivation for several years, which were recorded to be rather usual in Japan. Commercial cropping with intensive land managements is dominant today in upland areas of Japan, which is largely affected by volcanic ejecta. Traditional shifting cultivation systems in each of the regions can be seen to be well adapted to the respective soil-ecological conditions. Socioeconomic conditions are, however, drastically changing, making it difficult to sustain this system. Under such conditions, we should search for alternative technical tools that could maintain the SOM level, suppress the nitrifying activity of soil microbes and avoid depletion of bases whilst mitigating soil acidification. This is imperative if the subsistence agriculture seen in these regions is intended to continue in the near future.
REFERENCES Adams, F. 1971. Ionic concentrations and activities in soil solutions. Soil Sci. Soc. Am. Proc. 35:420–426. Alegre, J.C., D.K. Cassel, and D.E. Bandy. 1988. Effect of land clearing method on chemical properties of an Ultisol in the Amazon. Soil Sci. Soc. Am. J. 52:1283–1288. Alexiades, C.A., and M.L. Jackson. 1965. Quantitative determination of vermiculite in soils. Soil Sci. Soc. Am. J. 29:522–527.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
153
Anderson, J.P.E. 1982. Soil respiration. In Methods of soil analysis, A.L. Page, R.H. Miller, and D.R. Keeney, eds., 831–871. Madison: Soil Science Society of America. Andrews, J.A., K.G. Harrison, R. Matamala, and W.H. Schlesinger. 1999. Separation of root respiration from total soil respiration using carbon-13 labeling during free-air carbon dioxide enrichment (FACE). Soil Sci. Soc. Am. J. 63:1429–1435. Araki, S., K. Kyuma, and S. Funakawa. 1990. Comparative study of red yellow colored soils of Thailand, China, Korea and Japan. Transactions of 14th International Congress of Soil Science, August 12–18, 1990, Kyoto, Japan, V:328–329. Araki, S., B.M. Msanya, J.P. Magogo, D.N. Kimaro, and Y. Kitagawa. 1998. Characterization of soils on various plantation surfaces in Tanzania. Proc. of the 16th World Congress of Soil Science, August 20–26, 1998, Montpellier, France. Arduino, E., E. Barberis, F.A. Marsan, E. Zanini, and M. Franchini. 1986. Iron oxides and clay minerals within profiles as indicators of soil age in northern Italy. Geoderma 37:45–55. Barnhisel, R.I., and P.M. Bertsch. 1989. Chlorites and hydroxy-interlayered vermiculite and smectite. In Minerals in soil environments, J. B. Dixon and S. B. Weed, eds., 729–788. Madison: Soil Science Society of America. Binkley, D., and D. Richter. 1987. Nutrient cycles and H+ budgets of forest ecosystems. Advances in Ecological Research 16:2–51. Bolan, N.S., M.J. Hedley, and R.E. White. 1991. Processes of soil acidification during nitrogen cycling with emphasis on legume based pastures. Plant Soil 134:53–63. Bond-Lamberty, B., C. Wang, and S.T. Gower. 2004. A global relationship between the heterotrophic and autotrophic components of soil respiration? Global Change Biology 10:1756–1766. Borchardt, G. 1989. Smectites. In Minerals in soil environments, J. B. Dixon and S. B. Weed, eds., 675–727. Madison: Soil Science Society of America. Boucneau, G., M. van Meirvenne, J. Desmet, and G. Hofman. 1996. A methodology to evaluate the reliability of the Belgian soil map for predicting the actual water table characteristics. Geoderma 69:193–207. Boudot, J.P. 1992. Relative efficiency of complexed aluminum, noncrystalline Al hydroxide, allophane and imogolite in retarding the biodegradation of citric acid. Geoderma 52:29–39. Bouma, J., J. Hoeks, L. van der Plas, and B. van Scherrenburg. 1969. Genesis and morphology of some Alpine podzol profiles. J. Soil Sci. 20:384–398. Bouman, O.T., D. Curtin, C.A. Campbell, V.O. Biederbeck, and H. Ukrainetz. 1995. Soil acidification from long-term use of anhydrous ammonia and urea. Soil Sci. Soc. Am. J. 59:1488–1494. Brahy, V., H. Titeux, A. Iserentant, and B. Delvaux. 2000. Surface podzolization in Cambisols under deciduous forest in the Belgian loess belt. Eur. J. Soil Sci. 51:15–26. Bredemeier, M., E. Matzner, and B. Ulrich. 1990. Internal and external proton load to forest soils in Northern Germany. J. Environ. Quality 19:469–477. Brookes, P.C., A. Landman, G. Pruden, and D.S. Jenkinson. 1985. Chloroform fumigation and the release of soil nitrogen: A rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biol. Biochem. 17:837–842. Brouwer, E., B. Baeyens, A. Maes, and A. Cremers. 1983. Cesium and rubidium ion equilibria in illite clay. J. Phys. Chem. 87:1213–1219. Cleveland, C.C., and A.R. Townsend. 2006. Nutrient additions to a tropical rain forest drive substantial soil carbon dioxide losses to the atmosphere. P. Natl. Acad. Sci. USA 103:10316–10321. Coffman, C.B., and D.S. Fanning. 1974. “Vermiculite” determination on whole soils by cation exchange capacity methods. Clays Clay Miner. 22:271–283. Coody, P.N., L.E. Sommers, and D.W. Nelson. 1986. Kinetics of glucose uptake by soil microorganisms. Soil Biol. Biochem. 18:283–289.
154
World Soil Resources and Food Security
Cremers, A., A. Elsen, P. De Preter, and A. Maes. 1988. Quantitative analysis of rediocaesium retention in soils. Nature 335:247–249. Dalal, R.C., and R.J. Henry. 1988. Cultivation effects on carbohydrate contents of soil and soil fractions. Soil Sci. Soc. Am. J. 52:1361–1365. DeBano, L.F., D.G. Neary, and P.F. Ffolliott. 1998. Fire’s effects on ecosystems. New York: John Wiley & Sons. De Koning, A., A.V. Konoplev, and R.N.J. Comans. 2007. Measuring the specific caesium sorption capacity of soils, sediments and clay minerals. Appl. Geochem. 22:219–229. Delvaux, B., N. Kruyts, and A. Cremers. 2000. Rhizosperic mobilization of rediocaesium in soils. Environ. Sci. Technol. 34:1489–1493. Delvaux, B., N. Kruyts, E. Maes, and E. Smolders. 2001. Fate of radiocesium in soil and rhizosphere. In Trace elements in the rhizosphere, G.R. Gobran, W.W. Wenzel, and E. Lombi, eds., 61–91. London: CRC Press. Dumat, C., M.V. Cheshire, A.R. Fraser, C.A. Shand, and S. Stauton. 1997. The effect of removal of soil organic matter and iron on the adsorption of radiocaesium. Eur. J. Soil Sci. 48:675–683. Ellingson, L.J., J.B. Kauffman, D.L. Cummings, R.L. Sanford, and V.J. Jaramillo. 2000. Soil N dynamics associated with deforestation, biomass burning, and pasture conversion in a Mexican tropical dry forest. Forest Ecol. Manage. 137:41–51. Ewel, J., C. Berish, B. Brown, N. Price, and J. Raich. 1981. Slash and burn impacts on a Costa Rican wet forest site. Ecology 62:816–829. Fanning, D.S., V.Z. Keramidas, and M.A. El-Desoky. 1989. Micas. In Minerals in soil environments, J.B. Dixon and S.B. Weed, eds., 551–634. Madison: Soil Science Society of America. FAO. 2001. Lecture notes on the major soils of the world. In World soil resources reports 94, P.M. Driessen, J.A. Deckers, O.C. Spaargaren, and F.O. Nachtergaele, eds., 334. Rome: Food and Agriculture Organization of the United Nations. FAO/Unesco. 1971. Soil map of the world. Volume IV, South America. Paris: FAO/Unesco. FAO/Unesco. 1977. Soil map of the world. Volume IX, Africa. Paris: FAO/Unesco. Funakawa, S., M. Ashida, and K. Yonebayashi. 2003. Charge characteristics of forest soils derived from sedimentary rocks in Kinki District, Japan, in relation to pedogenetic acidification process. Soil Sci. Plant Nutr. 49:387–396. Funakawa, S., Y. Hayashi, I. Tazaki, K. Sawada, and T. Kosaki. 2006a. The main functions of the fallow phase in shifting cultivation by the Karen people in northern Thailand: A quantitative analysis of soil organic matter dynamics. Tropics 15:1–27. Funakawa, S., H. Hirai, and K. Kyuma. 1992. Soil-forming processes under natural forest north of Kyoto in relation to soil solution composition. Soil Sci. and Plant Nutr. 38:101–112. Funakawa, S., H. Hirai, and K. Kyuma. 1993a. Speciation of Al in soil solution from forest soils in northern Kyoto with special reference to their pedogenetic process. Soil Sci. Plant Nutr. 39:281–290. Funakawa, S., K. Nambu, H. Hirai, and K. Kyuma. 1993b. Pedogenetic acidification of forest soils in northern Kyoto. Soil Sci. Plant Nutr. 39:677–690. Funakawa, S., T. Watanabe, and T. Kosaki. 2008. Regional trends in the chemical and mineralogical properties of upland soils in humid Asia: With special reference to the WRB classification scheme. Soil Sci. Plant Nutr. 54:751–760. Funakawa, S., J. Yanai, Y. Hayashi, T. Hayashi, T. Watanabe, C. Noichana, T. Panitkasate, R. Katawatin, T. Kosaki, and E. Nawata. 2006b. Soil organic matter dynamics in a sloped sandy cropland of Northeast Thailand with special reference to the spatial distribution of soil properties. Jpn. J. Trop. Agric. 50:199–207. Geological Survey of Japan. 2004. Digital geological map of East and Southeast Asia. 1:2,000,000, 2nd ed. Japan: AIST.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
155
Gjems, O. 1970. Mineralogical composition and pedogenic weathering of the clay fraction in Podzol soil profiles in Zalesine, Yugoslavia. Soil Sci. 110:237–243. Goenadi, D.H., and K.H. Tan. 1988. Differences in clay mineralogy and oxidic ratio of selected LAC soils in temperate and tropical regions. Soil Sci. 146:151–159. Gower, S.T. 2002. Productivity of terrestrial ecosystems. In Encyclopedia of global change, H.A. Mooney and J. Canadell, eds., 516–521. Oxford: Blackwell Scientific. Guggenberger, G., and K. Kaiser. 1998. Significance of DOM in the translocation of cations and acidity in acid forest soil. Z. Pflanzenernähr Bodenkd. 161:95–99. Hanson, P.J., N.T. Edwards, C.T. Garten, and J.A. Andrews. 2000. Separating root and soil microbial contributions to soil respiration: A review of methods and observations. Biogeochem. 48:115–146. Harris, W.G., A.A. Morrone, and S.E. Coleman. 1992. Occluded mica in hydroxy-interlayered vermiculite grains from a highly-weathered soil. Clays Clay Miner. 40:32–39. Hayashi, Y. 2002. Establishment of homegardens (pekarangan) by transmigrants in South Kalimantan, Indonesia. J. Intl. Soc. Southeast Asian Agric. Sci. 8:47–62. Hayashi, Y., and A. Ochiai. 2004. Changes in farming systems and pekarangan homegarden in the Citarum River Basin, West Java. J. Intl. Soc. Southeast Asian Agric. Sci. 10:58–68. Helyar, K.R., and W.M. Porter. 1989. Soil acidification, its measurement and the processes involved. In Soil acidity and plant growth, A.D. Robson, ed., 61–101. Sydney: Academic Press Australia. Hiradate, S., T. Nakadai, H. Shindo, and T. Yoneyama. 2004. Carbon source of humic substances in some Japanese volcanic ash soils determined by carbon stable isotope ratio, Δ13C. Geoderma 119:133–141. Hirai, H. 1995. Studies on the genesis of Brown Forest soils and their related soils in Japan. Doctor thesis. Kyoto University. Hirai, H., S. Araki, and K. Kyuma. 1989. Clay mineralogical properties of brown forest soils in northern Kyoto with special reference to their pedogenetic process. Soil Sci. Plant Nutr. 35:585–596. Hodges, S.C. 1987. Aluminum speciation, a comparison of five methods. Soil Sci. Soc. Am. J. 51:57–64. Horwath, W.R., K.S. Pregitzer, and E.A. Paul. 1994. 14C Allocation in tree-soil systems. Tree Physiol. 14:1163–1176. Huang, P.M., M.K. Wang, N. Kampf, and D.G. Schulze. 2002. Aluminum hydroxides. In Soil mineralogy with environmental applications, J.B. Dixon and D.G. Schulze, eds., 261–289. Madison: Soil Science Society of America. Inskeep, W.P., J.L. Clayton, and D.W. Mogk. 1993. Naturally weathered Plagioclase grains from the Idaho Batholith: Observation using scanning electron microscopy. Soil Sci. Soc. Am. J. 57:851–860. ISSS-ISRIC-FAO. 1998a. World reference base for soil resources. World soil resources reports 84. Rome: Food and Agriculture Organization of the United Nations. ISSS-ISRIC-FAO. 1998b. World reference base for soil resources: Atlas, E.M. Bridges, N.H. Batjes, and F.O. Nachtergaele, eds. Rome: Food and Agriculture Organization of the United Nations. ISSS-ISRIC-FAO. 2006. World reference base for soil resources 2006—A framework for international classification, correlation and communication. World soil resources reports 103. Rome: Food and Agriculture Organization of the United Nations. Jackson, R.B., J. Canadell, J.R. Ehleringer, H.A. Mooney, O.E. Sala, and E.D. Schulze. 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108:389–411. Jenkinson, D.S. 1990. The turnover of organic carbon and nitrogen in soil. Philosophical Transactions of the Royal Society B. 329:361–368.
156
World Soil Resources and Food Security
Jensen, M. 1993. Soil conditions, vegetation structure and biomass of a Javanese homegarden. Agroforestry Systems 24:181–186. Juo, A.S.R., K. Franzluebbers, A. Dabiri, and B. Ikhile. 1996. Soil properties and crop performance on a kaolinitic Alfisol after 15 years of fallow and continuous cultivation. Plant Soil 180:209–217. Kadono, A., S. Funakawa, and T. Kosaki. 2008. Factors controlling mineralization of soil organic matter in the Eurasian steppe. Soil Biol. Biochem. 40:947–955. Kaiser, K., and G. Guggenberger. 2003. Mineral surfaces and soil organic matter. Eur. J. Soil Sci. 54:219–236. Kaiser, K., G. Guggenberger, and W. Zech. 1996. Sorption of DOM and DOM fractions to forest soils. Geoderma 74:281–303. Kanket, W., A. Suddhiprakarn, I. Kheoruenromne, and R.J. Gilkes. 2002. Clay mineralogy of some Thai Ultisols. Paper presented at 17th World Congress of Soil Science, August 14–21, 2002, Bangkok. Karathanasis, A.D. 2002. Mineral equilibria in environmental soil systems. In Soil mineralogy with environmental applications, J.B. Dixon and D.G. Schulze, eds., 109–151. Madison: Soil Science Society of America. Karathanasis, A.D., F. Adams, and B.J. Hajek. 1983. Stability relationships in kaolinite, gibbsite, and Al-interlayered vermiculite soil systems. Soil Sci. Soc. Am. J. 47:1247–1251. Kemmitt, S.J., D. Wright, K.W.T. Goulding, and D.L. Jones. 2006. pH regulation of carbon and nitrogen dynamics in two agricultural soils. Soil Biol. Biochem. 38:898–911. Killham, K. 1990. Nitrification in coniferous forest soils. Plant Soil 128:31–44. Kira, T. 1976. Terrestrial ecosystem, 146–147. Tokyo: Kyoritsu. (in Japanese) Kitagawa, Y. 2005. Characteristics of clay minerals in podzols and podzolic soils. Soil Sci. Plant Nutr. 51:151–158. Kittrick, J.A. 1973. Mica-derived vermiculites as unstable intermediates. Clays Clay Miner. 21:479–488. Klute, A. 1986. Water retention: Laboratory methods. In Methods of soil analysis. Part 1. Physical and mineralogical methods, 635–662. Madison: Soil Science Society of America. Klute, A., and C. Dirksen. 1986. Hydraulic conductivity and diffusivity: Laboratory methods. In Methods of soil analysis. Part 1. Physical and mineralogical methods, A. Klute, ed., 687–734. Madison: Soil Science Society of America. Knoepp, J.D., and W.T. Swank. 1993. Site preparation burning to improve southern Appalachian pine-hardwood stands: Nitrogen responses in soil, water, and streams. Can. J. Forest Res. 23:2263–2270. Koch, C.B., M.D. Bentzon, E.W. Larsen, and O.K. Borggaard. 1992. Clay mineralogy of two Ultisols from Central Kalimantan, Indonesia. Soil Sci. 154:158–167. Komarneni, S., and D.M. Roy. 1980. Hydrothermal effects on cesium sorption and fixation by clay minerals and shales. Clays Clay Miner. 28:142–148. Kottek, M., J. Grieser, C. Beck, B. Rudolf, and F. Rubel. 2006. World map of the KöppenGeiger climatic classification updated. Meteorol Z. 15:259–263. Kunstadter, P., and E.C. Chapman. 1978. Problem of shifting cultivation and economic development in northern Thailand. In Farmers in the forest, P. Kunstadter, E.C. Chapman, and S. Sabhasri, eds., 3–23. Honolulu: The University Press of Hawaii. Kutsuna, S., M. Suzuki, and I. Nioh. 1988. Comparison of nitrogen mineralization and nitrifying activities of different forest soil types on the same slope with Japanese cedar. J. Jpn. For. Soc. 70:127–130. (in Japanese) Lanson, B. 1997. Decomposition of experimental x-ray diffraction patterns (profile fitting): A convenient way to study clays. Clays Clay Miner. 45:132–146. Le Roux, J., and C.I. Rich. 1969. Ion selectivity of micas as influenced by degree of potassium depletion Soil Sci. Soc. Am. Proc. 33:684–690.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
157
Lesturgez, G., R. Poss, A. Noble, O. Grunberger, W. Chintachao, and D. Tessier. 2006. Soil acidification without pH drop under intensive cropping systems in Northeast Thailand. Agr. Ecosys. Environ. 114:239–248. Lindsay, W.L. 1979. Aluminum, silica, aluminosilicate. In Chemical equilibria in soils, 34–77. New York: Wiley. Maes, E., A. Iserentant, J. Herbauts, and B. Delvaux. 1999a. Fixation of radiocaesium traces in a weathering sequence mica–vermiculite–hydroxy interlayered vermiculite. Eur. J. Soil Sci. 50:117–125. Maes, E., L. Vielvoye, W. Stone, and B. Delvaux. 1999b. Influence of the nature of clay minerals on the fixation of radiocaesium traces in an acid brown earth-podzol weathering sequence. Eur. J. Soil Sci. 50:107–115. Malavolta, E. 1985. Potassium status of tropical and subtropical region soils. In Potassium in agriculture, R.D. Munson, ed., 163–200. Madison: American Society of Agronomy. Malla, P.B. 2002. Vermiculites. In Soil mineralogy with environmental applications, J.B. Dixon and D.G. Schulze, eds., 501–529. Madison: Soil Science Society of America. Mamilov, A-Sh., and O.M. Dilly. 2002. Soil microbial eco-physiology as affected by shortterm variations in environmental conditions. Soil Biol. Biochem. 34:1283–1290. Matsue, N., and K. Wada. 1989. Source minerals and formation of partially interlayered vermiculites in dystrochrepts derived from tertiary sediments. J. Soil Sci. 40:1–7. Marion, G.M., J.M. Moreno, and W.C. Oechel. 1991. Fire severity, ash deposition, and clipping effects on soil nutrients in chaparral. Soil Sci. Soc. Am. J. 55:235–240. McBride, M.B. 1989. Surface chemistry of soil minerals. In Minerals in soil environments, J.B. Dixon and S.B. Weed, eds., 35–88. Madison: Soil Science Society of America. McKeague, J.A., and J.H. Day. 1966. Dithionite- and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. Can. J. Soil Sci. 46:13–22. Mehra, O.P., and M.L. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite– citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7:317–327. Middleton, N.J., and D.S. Thomas. 1992. World atlas of desertification. London: Arnold. Mizota, C., I. Kawasaki, and T. Wakatsuki. 1988. Clay mineralogy and chemistry of seven pedons formed in volcanic ash, Tanzania. Geoderma 43:131–141. Mohr, E.C.J., F.A. van Baren, and J. van Schuylenborgh. 1972. Tropical soils. The Hague: Mouton. Mualem, Y., and G. Dagan. 1978. Hydraulic conductivity of soils: Unified approach to the statistical models. Soil Sci. Soc. Am. J. 42:392–395. Murayama, S., M.V. Cheshire, C.M. Mundie, G.P. Sparling, and H. Shepherd. 1979. Comparison of contribution to soil organic matter fractions, particularly carbohydrates, made by plant residues and microbial products. J. Sci. Food Agr. 30:1025–1034. Nagatsuka, S. 1972. Studies on genesis and classification of soils in warm-temperature region of Southwest of Japan. Part 3. Some features in distribution and mode of existence of free iron and aluminum oxides in the soil profile. Soil Sci. Plant Nutr. 18:147–154. Nakane, K. 1980. Comparative studies of cycling of soil organic carbon in three primeval moist forests. Jpn. J. Ecol. 3:155–172. Nakao, A., Y. Thiry, S. Funakawa, and T. Kosaki. 2008. Characterization of the frayed edge site of micaceous minerals in soil clays influenced by different pedogenetic conditions in Japan and northern Thailand. Soil Sci. Plant Nutr. 54:479–489. Nambu K., and K. Yonebayashi. 2000. Quantitative relationship between soil properties and adsorption of dissolved organic matter onto volcanic ash and non-volcanic ash soils. Soil Sci. Plant Nutr. 46:559–569. Ndayiragije, S., and B. Delvaux. 2003. Coexistence of allophane, gibbsite, kaolinite and hydroxy-Al-interlayered 2:1 clay minerals in a perudic Andosol. Geoderma 117:203–214.
158
World Soil Resources and Food Security
Neill, C., C.M. Piccolo, M.J. Mellio, A.P. Steudler, and C.C. Cerri. 1999. Nitrogen dynamics in Amazon forest and pasture soils measured by 15N pool dilution. Soil Biol. Biochem. 31:567–572. Noble, A.D., S. Suzuki, W. Soda, S. Ruaysoongnern, and S. Berthelsen. 2008. Soil acidification and carbon storage in fertilized pastures of Northeast Thailand. Geoderma 144:248–255. Norfleet, M.L., A.D. Karathanasis, and B.R. Smith. 1993. Soil solution composition relative to mineral distribution in Blue Ridge Mountain Soils. Soil Sci. Soc. Am. J. 57:1375–1380. Nye, P.H., and D.J. Greenland. 1960. The soil under shifting cultivation. Harpenden: Commonwealth Bureau of Soils. Oades, J.M. 1972. Studies on soil polysaccharides: III, Composition of polysaccharides in some Australian soils. Aust. J. Soil Res. 10:113–126. Olk, D.C., K.G. Cassman, and R.M. Carlson. 1995. Kinetics of potassium fixation in vermiculitic soils under different moisture regimes. Soil Sci. Soc. Am. J. 59:423–429. Olsson, P., S. Linder, R. Giesler, and P. Högberg. 2005. Fertilization of boreal forest reduces both autotrophic and heterotrophic soil respiration. Global Change Biology 11:1745–1753. Parfitt, R.L., A.R. Fraser, and V.C. Farmer. 1977. Adsorption on hydrous oxides. III. Fulvic acid and humic acid on goethite, gibbsite, and imogolite. Eur. J. Soil Sci. 28:289–296. Parton, W.J., D.S. Schimel, C.V. Cole, and D.S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Sci. Soc. Am. J. 51:1173–1179. Popov, V.E. 2006. Effect of 137Cs accumulation by organomineral coarse particles in sandy soils contaminated during the Chernobyl NPP accident. Eur. J. Soil Sci. 39:307–313. Poss, R, C.J. Smith, F.X. Dunin, and J.F. Angus. 1995. Rate of soil acidification under wheat in a semi-arid environment. Plant Soil 177:85–100. Prasetyo, B.H., N. Suharta, H. Subagyo, and H.S. Hikmatullah. 2001. Chemical and mineralogical properties of Ultisols of Sasamba area, East Kalimantan. Indonesian J. Agric. Sci. 2:37–47. Rai, D., and J.A. Kittrick. 1989. Mineral equilibria and the soil system. In Minerals in soil environments, J. B. Dixon and S. B. Weed, eds., 161–198. Madison: Soil Science Society of America. Raich, J.W., and W.H. Schlesinger. 1992. The global carbon dioxide flux in soil respiration and its relationship to vegetation and climate. Tellus 44B:81–99. Raison, R.J., P.K. Khanna, and P.V. Woods. 1985. Mechanisms of element transfer to the atmosphere during vegetation fires. Can. J. Forest Res. 15:132–140. Rausell-Colom, J.A., T.R. Sweatman, C.B. Wells, and K. Norrish. 1965. Studies in the artificial weathering of mica. In Experimental pedology, E. G. Hallsworth and D. V. Crawford, eds., 40–72. London: Butterworths. Reid-Soukup, D.A, and A.L. Ulery. 2002. Smectites. In Soil mineralogy with environmental applications, J.B. Dixon and D.G. Schulze, eds., 467–499. Madison: Soil Science Society of America. Richter, D.D., D. Markewitz, P.R. Heine, V. Jin, J. Raikes, K. Tian, and C.G. Wells. 2000. Legacies of agriculture and forest regrowth in the nitrogen of old-field soils. Forest Ecol. Manage. 138:233–248. Ridley, A.M., W.J. Slattery, K.R. Helyar, and A. Cowling. 1990. The importance of the carbon cycle to acidification of a grazed annual pasture. Austr. J. Exper. Agric. 30:529–537. Rigol, A., M. Vidal, and G. Rauret. 2002. An overview of the effect of organic matter on soil– radiocaesium interaction: Implications in root uptake. J. Environ. Radioact. 58:191–216. Robertson, G.P. 1982. Nitrification in forested ecosystems. Philos. Trans. R. Soc. Lond. B 296:445–457. Robson, A.D. 1989. Soil acidity and plant growth. Sydney: Academic Press Australia. Ross, R.J., and M.M. Mortland. 1966. A soil beidellite. Soil Sci. Soc. Am. Proc. 30:337–343.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
159
Ross, G..J., P.A. Schuppli, and C. Wang. 1989. Quantitative determination of vermiculite by a rubidium fixation method. Soil Sci. Soc. Am. J. 53:1588–1589. Ruark, G.A. 1993. Modeling soil temperature effects on in situ decomposition rates for fine roots of loblolly pine. Forest Sci. 39:118–129. Sakamoto, K., T. Yoshida, and M. Satoh. 1991. Comparison of carbon and nitrogen mineralization between fumigation and heating treatments. Soil Sci. Plant Nutr. 38:133–140. Sanchez, P.A. 1976. Properties and management of soils in the tropics. New York: Wiley. Sano, S., J. Yanai, and T. Kosaki. 2004. Evaluation of soil nitrogen status in Japanese agricultural lands with reference to land use and soil types. Soil Sci. Plant Nutr. 50:501–510. Sawhney, B.L. 1972. Selective sorption and fixation of cations by clay minerals: A review Clays Clay Miner. 20:93–100. Schimmack, W., and K. Auerswald. 2004. The radiocaesium interception potential (RIP) at an agricultural site in Germany. J. Environ. Radioact. 77:143–157. Shibata, H., M. Kirikae, Y. Tanaka, T. Sakuma, and R. Hatano. 1998. Proton budgets of forest ecosystem on canogenous Regosols in Hokkaido. Water, Air, and Soil Pollution 105:63–72. Shinjo, H., A. Kato, K. Fujii, K. Mori, S. Funakawa, and T. Kosaki. 2006. Carbon dioxide emission derived from soil organic matter decomposition and root respiration in Japanese forests under different ecological conditions. Soil Sci. Plant Nutr. 52:233–242. Shirato, Y., T. Hakamata, and I. Taniyama. 2004. Modified Rothamsted carbon model for Andosols and its validation: Changing humus decomposition rate constant with pyrophosphate-extractable Al. Soil Sci. Plant Nutr. 50:149–158. Shoji, S., T. Ito, M. Saigusa, and I. Yamada. 1985. Properties of nonallophanic Andosols from Japan. Soil Sci. 140:264–277. Smithson, P.C., and K.E. Giller. 2002. Appropriate farm management practices for alleviating N and P deficiencies in low-nutrient soils of the tropics. Plant Soil 245:169–180. Smolders, E., K.V.D. Brande, and R. Merckx. 1997. Concentration of 137Cs and K in soils solution predict the plant availability of 137Cs in soils. Environ. Sci. Technol. 31:3432–3438. Soil Survey Staff. 1999. Soil taxonomy. A basic system of soil classification for making and interpreting soil surveys, 2nd ed. Washington, DC: US Government Printing Office. Soil Survey Staff. 2006. Keys to soil taxonomy, 10th ed. Washington, DC: US Government Printing Office. SPSS Inc. 1998. SYSTAT 8.0. Statistics. Chicago: SPSS Inc. SPSS Inc. 2002. SigmaPlot 8.0. user’s guide. Chicago: SPSS Inc. Spycher, G., P. Sollins, and S. Rose. 1983. Carbon and nitrogen in the light fraction of a forest soil: Vertical distribution and seasonal patterns. Soil Sci. 135:79–87. Stanford, G., and S.J. Smith.1972. Nitrogen mineralization potentials of soils. Soil Sci. Soc. Am. Proc. 36:465–472. Sterte, J., and J. Shabtai. 1987. Cross-linked smectites. V. Synthesis and properties of hydroxy-silicoamuminum montmorillonites and fluorhectorites. Clays Clay Miner. 35:429–439. Šucha, V., and V. Širáńová. 1991. Ammonium and potassium fixation in smectite by wetting and drying. Clays Clay Miner. 39:556–559. Sunderlin, W.D., and I.A.P. Resosudarmo. 1996. Rates and causes of deforestation in Indonesia: Towards a resolution of the ambiguities. CIFOR, Occasional paper, No. 9. Bogor: CIFOR. Supriyo, H., N. Matsue, and N. Yoshinaga. 1992a. Chemical and mineralogical properties of volcanic ash soils from Java. Soil Sci. Plant Nutr. 38:443–457. Supriyo, H., N. Matsue, and N. Yoshinaga. 1992b. Chemistry and mineralogy of some soils from Indonesia. Soil Sci. Plant Nutr. 38:217–225. Tamura, T. 1958. Identification of clay minerals from acid soils. J. Soil Sci. 9:141–147.
160
World Soil Resources and Food Security
Tanaka, S., T. Ando, S. Funakawa, C. Sukhrun, T. Kaewkhongka, and K. Sakurai. 2001. Effect of burning on soil organic matter content and N mineralization under shifting cultivation system of Karen people in northern Thailand. Soil Sci. Plant Nutr. 47:547–558. Tanaka, S., S. Funakawa, T. Kaewkhongkha, T. Hattori, and K. Yonebayashi. 1997. Soil ecological study on dynamics of K, Mg, and Ca, and soil acidity in shifting cultivation in northern Thailand. Soil Sci. Plant Nutr. 43:695–708. Tanaka, S., J.J. Kendawang, N. Yoshida, K. Shibata, A. Jee, K. Tanaka, I. Ninomiya, and K. Sakurai. 2005. Effects of shifting cultivation on soil ecosystems in Sarawak, Malaysia. IV. Chemical properties of the soils and runoff water at Niah and Bakam experimental sites. Soil Sci. Plant Nutr. 51:525–533. Tanaka, S., M.E.B. Wasli, T. Kotegawa, L. Seman, J. Sabang, J.J. Kendawang, K. Sakurai, and Y. Morooka. 2007. Soil properties of secondary forests under shifting cultivation by the Iban of Sarawak, Malaysia in relation to vegetation condition. Tropics 16:385–398. Thiry, Y., N. Kruyts, and B. Delvaux. 2000. Respective horizon contribution to cesium-137 soil-to-plant transfer: A pot experiment approach. J. Environ. Qual. 29:1194–1199. Thompson, M.L., and L. Ukrainczyk. 2002. Micas. In Soil mineralogy with environmental applications, J.B. Dixon and D.G. Schulze, eds., 431–466. Madison: Soil Science Society of America. Thornthwaite, C.W., and J.R. Mather. 1957. Instructions and tables for computing potential evapotranspiration and the water balance. Pubis Clim. 10:185–311. Tokuchi, N., M. Hirobe, and K. Koba. 2000. Topographical differences in soil N transformation using 15N dilution method along a slope in a conifer plantation forest in Japan. J. For. Res. 5:13–19. Toriyama, J., S. Ohta, M. Araki, E. Ito, M. Kanzaki, S. Khorn, P. Pith, S. Lim, and S. Pol. 2007. Acrisols and adjacent soils under four different forest types in central Cambodia. Pedologist 51:35–49. Tsai, H., Z.Y. Hseu, W.S. Huang, and Z.S. Chen. 2007. Pedogenic approach to resolving the geomorphic evolution of the Pakua river terraces in central Taiwan. Geomorphology 83:14–28. Tulaphitak, T., C. Pairintra, and K. Kyuma. 1985a. Changes in soil fertility and tilth under shifting cultivation. II. Changes in soil nutrient status. Soil Sci. Plant Nutr. 31:239–249. Tulaphitak, T., C. Pairintra, and K. Kyuma. 1985b. Changes in soil fertility and tilth under shifting cultivation. III. Soil respiration and soil tilth. Soil Sci. Plant Nutr. 31:251–261. Van Breemen, N., and R. Brinkman. 1976. Chemical equilibria and soil formation. In Soil chemistry: Basic elements, G.H. Bolt and M.G.M. Bruggenwert, eds., 141–170. Amsterdam: Elsevier. Van Breemen, N., C.T. Driscoll, and J. Mulder. 1984. Acidic deposition and internal proton in acidification of soils and waters. Nature 307:599–604. Van Genuchten, M.Th. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44:892–898. Vance, E.D., P.C. Brookes, and D.S. Jenkinson. 1987. An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem. 19:703–707. Velde, B. 2001. Clay minerals in the agricultural surface soils in the Central United States. Clay Minerals 36:277–294. Von Lützow, M., I. Kögel-Knabner, K. Ekschmitt, H. Flessa, G. Guggenberger, E. Matzner, and B. Marschner. 2007. SOM fractionation methods: Relevance to functional pools and to stabilization mechanisms. Soil Biol. Biochem. 39:2183–2207. Wageneers, N., E. Smolders, and R. Merckx. 1999. A statistical approach for estimating the radiocesium interception potential of soils. J. Environ. Qual. 28:1005–1011. Wardle, D.A., and A. Ghani. 1995. A critique of the microbial metabolic quotient (qCO2) as a bioindicator of disturbance and ecosystem development. Soil Biol. Biochem. 27:1601–1610.
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
161
Watanabe, T., S. Funakawa, and T. Kosaki. 2006. Clay mineralogy and its relationship to soil solution composition in soils from different weathering environment of humid Asia: Japan, Thailand and Indonesia. Geoderma 136:51–63. Watanabe, E., K. Sakurai, Y. Okabayashi, N. Lasay, and C. Alounsawat. 2004. Soil fertility and farming systems in a slash and burn cultivation area of northern Laos. Southeast Asian Studies 41:519–537. Wauters, J., A. Elsen, A.V. Konoplev, A.A. Bulgakov, and R.N.J. Comans. 1996. Prediction of solid/liquid distribution coefficients of radiocaesium in soils and sediments. Part 1. Applied Geochem. 11:589–594. Whitehead, D.C., H. Buchan, and R.D. Hartley. 1975. Components of soil organic matter under grass and arable cropping. Soil Biol. Biochem. 7:65–71. Wiersum, K.F. 1982. Tree gardening and taungya on Java: Examples of agroforestry techniques in the humid tropics. Agroforestry Systems 1:53–70. Yoothong, K., L. Moncharoen, P. Vijarnson, and H. Eswaran. 1997. Clay mineralogy of Thai soils. Applied Clay Sci. 11:357–371. Yoshinaga, N., N. Matsue, P. Virakornphanich, B. Tantiseri, and B. Supa-Udomleark. 1995. Clay mineralogy of soils from upper North Thailand. Soil Sci. Plant Nutr. 41:337–343. Zimmermann, M., J. Leifeld, N.W.I. Schmidt, P. Smith, and J. Fuhrer. 2006. Measured soil organic matter fractions can be related to pools in the RothC model. Eur. J. Soil Sci. 58:658–667.
APPENDIX—ANALYTICAL METHODS 4.A1 C OMPOSITION OF SOIL WATER EXTRACT FOR THERMODYNAMIC ANALYSIS Soil water extracts were collected by continuously shaking the soils for 1 week at 25°C and at 1 atm with a soil (40 g) to water ratio of 1:2, followed by filtering through a 0.025-μm pore membrane filter (Millipore). For these samples, pH and F concentration were determined with glass electrodes. Na+, K+, Cl−, NO3− , and SO 2− 4 concentrations were determined via high-performance liquid chromatography (Shimadzu, Ion chromatograph HIC-6A equipped with a conductivity detector CDD-6A and shim-pack IC-C3 for cations and shim-pack IC-A1 for anions). The concentrations of Si, Al, Ca, Mg, Fe, and Mn were determined using ICP–AES. To eliminate Al complexed with organic matter, the extracts were passed through a column filled with a partially neutralized (pH 4.2) cation exchange resin [Amberlite IR-120B(H); Hodges 1987]. The amount of Al not adsorbed by the resin was determined by ICP–AES and assigned to Al complexed with organic matter, as opposed to the fraction retained that was assigned to inorganic monomeric Al. Ionic activities were calculated with the extended Debye–Hükel equation, using a successive approximation procedure [Adams 1971]. Inorganic monomeric Al was distributed among Al3+, Al(OH)2+, Al(OH)+2 , Al(OH)30 , Al(OH)−4 , AlF2+ and Al(SO4)+ according to the equilibrium constants of Lindsay [1979]. Clay minerals were assumed to be at or very near to equilibrium with the soil water extracts. The stability of minerals was evaluated using stability diagrams and solubility diagrams; these diagrams represent the relative stability and solubility of minerals, as outlined by van Breemen and Brinkman [1976] and Karathanasis [2002], respectively. Thermodynamic mineral data used in these diagrams are from
162
World Soil Resources and Food Security
Karathanasis [2002] for gibbsite, kaolinite, quartz, and amorphous silica, and from Lindsay [1979] for amorphous Al hydroxide, muscovite, and microcline.
4.A2 Q UANTIFICATION OF THE FRAYED EDGE SITE USING RADIOCESIUM INTERCEPTION POTENTIAL (RIP) METHODOLOGY Quantification of the frayed edge site using a simple chemical adsorption-desorption experiment, such as determining the CEC, has proven difficult because the frayed edge site is not accessible to hydrated cations, but can irreversibly adsorb monovalent cations with low hydration energy (e.g., Cs+, Rb+, K+, and NH +4 ) with the formation of an inner-sphere complex. Although the irreversible adsorption sites can be quantified using a K fixation method [Alexiades and Jackson 1965; Coffman and Fanning 1974] or a Rb fixation method [Ross et al. 1989], these methods clearly overestimate the amount of the frayed edge site because K and Rb saturation and drying in these methods induces the irreversible collapse of the expanded layers and, therefore, fixation mostly occurs in expanded layers with a vermiculitic nature [Komarneni and Roy 1980]. However, Cremers et al. [1988] obtained a quantitative index of the frayed edge site from the solid/liquid distribution coefficient of Cs and the concentration of K in solution, which was described as:
FES −1 K Cs D ·mK = K c(Cs− K) ·[FES] = RIP (mol kg )
(4.A1)
−1 where K Cs D is the solid/liquid distribution coefficient of Cs (L kg ), mK is the K conFES −1 centration in solution (mol L ), K c(Cs−K) is the selectivity coefficient of Cs against K in the frayed edge site, and [FES] is the amount of the frayed edge site in soil (mol kg−1). Cs As K FES c(Cs− K) is a constant, then K D ·mK is regarded as proportional to the amount of the frayed edge site. Two experimental conditions are necessary to make Equation 4.A1 valid. First, exchangeable sites must be masked from Cs and K with silver thiourea so that Cs and K adsorption occurs only on the frayed edge site. Based on this assumption, the Cs–K exchange reaction on the frayed edge site can be expressed as:
[K FES] + mCs = [CsFES] + mK
(4.A2)
where [K FES] (or [CsFES]) is the amount of K (or Cs) adsorbed on the frayed edge site (mol kg−1), and mK (or mCs) is the concentration of K (or Cs) in solution (mol L –1). As this reaction is homovalent exchange, the selectivity coefficient (i.e., K FES c(Cs− K)) can be expressed as:
K FES c(Cs− K) = {[Cs FES ]·mK}/{[K FES ]·mCs}.
(4.A3)
Second, the amount of Cs adsorbed on the frayed edge site (i.e., [CsFES]) must be minimized using carrier-free 137Cs so that it can be assumed that the amount of K adsorbed on the frayed edge site (i.e., [K FES]) is identical to the amount of the frayed
163
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
edge site (i.e., [FES] in Equation 4.A1). Based on this assumption, Equation 4.A3 becomes Cs K FES c(Cs− K) (Cs → 0) = K D ·mK/[FES].
(4.A4)
Thus, Equation 4.A1 is accomplished by transposing the denominator on the righthand side to the left-hand side in Equation 4.A4. The method of Cremers et al. [1988] was followed by Wauters et al. [1996], who found that silver thiourea, which was used to mask the exchangeable sites, could be successfully replaced with a specific Ca/K ratio in solution (i.e., 0.1 mol L−1 CaCl2 + 0.5 mmol L−1 KCl) to make the method more available. The K Cs D ·mK value from the method of Wauters et al. [1996], namely the radiocesium interception potential (RIP), is now widely accepted by many researchers in Europe as a quantitative index of the frayed edge site to fix 137Cs in soils [e.g., Delvaux et al. 2000; Popov 2006; Schimmack and Auerswald 2004; Smolders et al. 1997; Wageneers et al. 1999]. As the RIP does not result in an overestimation of the frayed edge site because of the collapse of the expanded layers [de Koning et al. 2007], it can be used in the prediction of 137Cs dynamics in soils, such as 137Cs transfer from soil to plant at a realistic contamination level [Delvaux et al. 2000]. For the RIP determination, 0.1 g of soil clay was weighed into triplicate dialysis bags (cellulose dialysis membrane, Wako Chemicals USA, Inc.) containing 5 mL of 0.1 mol L−1 CaCl2 and 0.5 mmol L−1 KCl solution. The bags were transferred to 250 mL plastic bottles holding 200 mL of 0.1 mol L−1 CaCl2 and 0.5 m mol L –1 KCl solution, and then shaken for 2 hours twice a day for 5 days. We replenished the outer solution each time before shaking to maintain equilibration. Afterward, each dialysis bag was transferred to a 50 mL plastic bottle with 25 mL of Ca–K–137Cs solution (i.e., 0.1 mol L−1 CaCl2 and 0.5 m mol L−1 KCl solution spiked with 10 kBq of carrier-free 137Cs). After continuously shaking the bottles for 5 days, 137Cs activity (Bq mL−1) in the outer solution was measured with a liquid scintillation counter (Aloka, LSC6100, Radioisotope Research Center of Kyoto University), and the distribution coef137Cs in the solution. The ficient of Cs (K Cs D ) calculated from the depletion in the concentration of K in the solution (mK) was assumed to be 0.5 mmol L−1.
4.A3 T HEORETICAL CALCULATION OF WATER FLUXES, SOIL ACIDIFICATION RATES, AND NET PROTON GENERATION 4.A3.1 Water Fluxes Water fluxes of soil solution percolated from the bottom of the O and B horizons were estimated by applying Darcy’s law to the unsaturated hydraulic conductivity and the gradient of the hydraulic heads in surface soil (0–5-cm depth) and subsoil (40–45-cm depth). The one dimensional, vertical flow equation (Richard’s equation) in the unsaturated soil zone was written as follows: C ( h)
δh δ δh = K (h) +1 δt δ z δz
− S ( h)
(4.A5)
164
World Soil Resources and Food Security
where C (m−1) is the differential water capacity, t (day) is the time, Z (m) is the height, h (m) is the soil water pressure head, K (m day−1) is the unsaturated hydraulic conductivity and S (day−1) is the sink term accounting for water uptake by vegetation and lateral water flow. The unsaturated hydraulic conductivity characteristic described by Mualem-van Genuchten functions [Mualem and Dagan 1978; van Genuchten 1980] was written as follows:
K = K S × Se0.5 × 1 − 1 − S
1 m e
2
m
(4.A6)
Se =
(θ − θ ) = (θ − θ ) r
s
1 + (−α h)n
−m
(4.A7)
r
where Se is the effective saturation. Ks (m day−1) is the saturated hydraulic conductivity. θs (L L−1) and θr (L L−1) are the saturated and residual water contents, respectively. n is a fitting parameter and m = 1 − 1/n. Ks and water retention curves were experimentally obtained for undisturbed soils sampled by 0.1 L cores according to the methods of Klute [1986] and Klute and Dirksen [1986], respectively. Based on the water retention curve, pressure heads were calculated from volumetric soil water contents monitored every 30 minutes with TDR sensors (CS615, Campbell Scientific) and data loggers (CR 10X, Campbell Scientific). Parameters were optimized based on the water retention curve (θ − h) by Sigmaplot 8.0 [SPSS Inc. 2002]. Parameters were adjusted and validated to balance the water budget during the winter season. Daily water fluxes were calculated using the parameters of Mualemvan Genuchten functions to describe the physical properties of the soils and volumetric soil water contents monitored during the experimental periods. Fluxes of water percolated from the bottom of the A horizon were estimated by the following equation:
(
)
qA = qO − qO − qB ×
rA rA + rB
(4.A8)
where qO, qA, and qB (mm d−1) are the daily fluxes of water percolated from the bottom of the O, A, and B horizons, respectively. rA and r B (Mg ha−1) are fine root biomass in the A and B horizons, respectively.
4.A3.2 Soil Acidification Rate and Net Proton Generation The soil profile was divided into three compartments corresponding to the natural soil stratification. Soil acidification rate and net proton generation (NPG) resulting
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
165
from cation excess uptake by vegetation, nitrification, dissociation of organic acids, dissociation of carbonic acid, and net proton influx from the overlying horizon were evaluated in each soil horizon compartment. Soil acidification was defined as the decrease of acid neutralizing capacity (ANC) of the solid phase of soil [van Breemen et al. 1984]. ANC(ref pH = 3) (kmolc ha−1) was defined as the sum of basic cation equivalence minus the sum of strongly acidic anion equivalence at a reference soil pH of 3, written as follows:
ANC(ref pH=3) = 2(Na2O) + 2(K2O) + 2(CaO) + 2(MgO) + 2(FeO) + 6(Al2O3) – 2(SO3) – 2(P2O5) – (HCl)
(4.A9)
where parentheses denote molar concentration. The soil acidification rate was calculated as the change in ANC over a given period, i.e., ΔANC (kmolc ha−1 yr−1), which was induced by net proton generation. NPG can be divided into several sources. NPGBio was calculated from cation excess uptake by vegetation, assuming that vegetation uptake was equal to the sum of wood increment and litterfall. NPG resulting from cation excess uptake by vegetation in each soil horizon compartment was allocated to the three soil horizon compartments based on the distribution of fine root biomass in each soil horizon [Shibata et al. 1998]. Soil acidification rate and NPG resulting from nitrification, dissociation of organic acids, dissociation of carbonic acid, and net proton influx from the overlying horizon were calculated from the input–output budget of ions, i.e., the difference between ion fluxes percolating from the overlying horizon and ion fluxes leaching out of the soil horizon compartment [Bredemeier et al. 1990]. Ion fluxes were calculated from ionic concentrations multiplied by the water fluxes during each period. Using ion fluxes, soil acidification rate (ΔANC) was calculated to be:
ΔANC = {(Anion)out – (Anion)in} – {(Cation)out – (Cation)in} – {(Cation) bio – (Anion) bio}
(4.A10)
where (Cation) and (Anion) represent the equivalent sum of cations and anions, respectively, which are responsible for ANC. The ionic species counted in the present + + 2+ 2+ 3+ n+ study were Cl–, H 2PO −4 , and SO 2− 4 for anions and Na , K , Mg , Ca , Fe , and Al for cations. The notations in parentheses represent the following: the suffix “in” represents ion fluxes percolating into the soil compartment by throughfall and soil solution from the overlying horizon (or throughfall for the O horizon); the suffix “out” represents ion fluxes leaching out of the soil compartment by soil solution; and the suffix “bio” represents ion fluxes caused by vegetation uptake (kmolc ha−1 yr−1). Net proton generation by cation excess uptake by vegetation (NPGBio) is
NPGBio = (Cation)bio – (Anion)bio.
(4.A11)
Net proton generation by transformation of nitrogen (NPGNtr) is
{
} {
}
NPG Ntr = ( NH +4 )in − ( NH +4 )out + (NO3− )out − ( NO3− )in .
(4.A12)
166
World Soil Resources and Food Security
Net proton generation by the dissociation of organic acids (NPGOrg) is
NPGOrg = (Orgn–)out – (Orgn–)in
(4.A13)
where Orgn– represents the negative charge of organic acids. Net proton generation by carbonic acid dissociation (NPGCar) is
NPG Car = (HCO3− )out − (HCO3− )in .
(4.A14)
Net proton influx from the overlying horizon is
{(H+)in − (H+)out}.
(4.A15)
The sum of the acid load is neutralized by the decrease of ANC stoichiometrically. Theoretically, ΔANC, NPGBio, NPGNtr, NPGOrg, NPGCar, and net proton influx from the overlying horizon have the following relationship:
ΔANC = NPGBio + NPGNtr + NPGOrg + NPGCar + {(H+)in − (H+)out}. (4.A16)
4.A4 D ETERMINATION OF POTENTIALLY MINERALIZABLE CARBON (PMC) AND NITROGEN (PMN) Fresh soil samples were passed through a 2-mm mesh sieve and their moisture contents adjusted to 60% of field moisture capacity. Mineralized C and N were determined in duplicate more than six times over a period of 133 days, using aerobic incubation at a constant temperature of 30°C. The C and N mineralization patterns were fitted to equations by the least squares method using SigmaPlot 8.0 [SPSS Inc. 2002]. As model functions, first-order kinetic models with single (Fi model) or double (Fi + Fi model) sets of parameters were first prepared for mineralized C:
Ct = C0{1 – exp(–kCt)}
(4.A17)
Ct = αC0{1 – exp(–kC1t)}+ (1 – α)C0{1 – exp(–kC2t)}
(4.A18)
where Ct (mg kg−1) is the amount of mineralized C at time t (d), C0 (mg kg–1) is the pool of readily mineralizable C (mg kg−1) (i.e., PMC), and kC, kC1, and kC2 are rate constants (d−1; kC1 > kC2). In some cases, the time course of C mineralization was simulated well by the Gompertz equation as follows:
Ct = C0exp[−exp{−kC(t − t0)}]
(4.A19)
where t0 (d) is the time when C equals 1/e of C0. If kC for the Gompertz or first-order model or kC2 for the double first-order model was less than 0.003 d−1, the calculated PMC was not used for the following analysis because k < 0.003 means that less than one-third of the PMC is mineralized during the 133-day incubation experiment and
Soil Resources and Human Adaptation in Ecosystems in Humid Asia
167
the value of PMC is calculated with an extreme extrapolation, indicating less reliability of the fitting calculation. For simulating N mineralization patterns, in addition to Fi and Gompertz models (Equations 4.A20 and 4.21), a logistic model (Lo model; Equation 4.A22) was also prepared:
Nt = N0(1 − exp(−k Nt))
(4.A20)
Nt = N0exp[–exp{–k N(t – t0)}]
(4.A21)
Nt = Nmax/{1 + (Nmax/Nint – 1)exp(–k Nt)}
(4.A22)
where Nt (mg N kg−1) is the cumulative N released in time t (d), Nt (mg N kg−1) (i.e., PMN) is potentially mineralizable N, Nmax is the calculated maximum amount of mineral N, Nint is the calculated initial amount of mineral N, k N (d–1) is the rate constant, and t0 (d) is the calculated time when N equals 1/e of N0. The third equation and PMN in this equation is calculated as the difference between Nmax and Nint.
Acidification 5 Pedogenetic in Upland Soils under Different Bioclimatic Conditions in Humid Asia S. Funakawa, T. Watanabe, A. Nakao, K. Fujii, and T. Kosaki CONTENTS 5.1 Introduction................................................................................................... 171 5.2 Pedogenetic Soil Acidification by Different Proton Sources in Forested Ecosystems in Japan...................................................................................... 172 5.2.1 Experimental Plots............................................................................ 172 5.2.2 Soil Solution Composition................................................................. 176 5.2.3 NPGOrg, NPGNtr, NPGBio, and ΔANC................................................ 178 5.2.4 Pedogenetic Implication of the Proton Budget.................................. 183 5.3 Fates of Soil Acidity and Dissolved Organic Matter during Pedogenetic Soil Acidification of Forest Soils in Japan..................................................... 185 5.3.1 Study Soils......................................................................................... 186 5.3.2 Physicochemical Properties and Titratable Alkalinity and Acidity of the Soils............................................................................ 186 5.3.3 Soil Solution Composition................................................................. 196 5.3.4 Eluvi-Illuviation of Inorganic Al and Subsequent Adsorption of DOC................................................................................................... 196 5.3.5 Comigration of Al and DOC in the Soil Profiles..............................200 5.3.6 Dynamics of Titratable Alkalinity and Acidity in the Pedogenetic Processes....................................................................... 201 5.4 Changes in the Blockage Effect of Hydroxy-Al Polymers on the Frayed Edge Site of Illitic Minerals during the Process of Pedogenetic Acidification in Japan.................................................................................... 203 5.4.1 Soils Studied...................................................................................... 203 5.4.2 Soil Types and Clay Mineralogy....................................................... 203 5.4.3 RIP Variation and the Sequential Transformation of HIV to Vermiculite........................................................................................205 5.4.4 Blockage Effect of Hydroxy-Al Polymers on the Frayed Edge Site.....................................................................................................207 169
170
World Soil Resources and Food Security
5.5 Charge Characteristics of Forest Soils Derived from Sedimentary Rocks in Japan in Relation to Pedogenetic Acidification Processes.............209 5.5.1 Soil Samples......................................................................................209 5.5.2 Analytical Methods........................................................................... 212 5.5.3 General Charge Characteristics of the Soils...................................... 213 5.5.4 Contribution of Each Component to the Charge Characteristics of the Soils—Statistical Analysis...................................................... 215 5.5.5 Contribution of Each Component to the Charge Characteristics of the Soils—Analysis by Successive Removal of Soil Components.......219 5.5.6 Changes in the Charge Characteristics of the Soils through Pedogenetic Acidification.................................................................. 220 5.5.7 Conclusions of Sections 5.2 to 5.5..................................................... 221 5.6 Fluxes of DOC under Tropical Forests under Different Geological Conditions in East Kalimantan, Indonesia.................................................... 223 5.6.1 Experimental Plots and Experimental Design.................................. 223 5.6.2 Physicochemical Properties of Soils and C Stock in Soils and Ecosystems........................................................................................224 5.6.3 Organic Matter Decomposition......................................................... 232 5.6.4 Concentrations and Fluxes of DOC in Throughfall and Soil Solution.............................................................................................. 232 5.6.5 Influence of Parent Rocks on DOC Dynamics.................................. 236 5.7 Contribution of Different Proton Sources to Soil Acidification under Tropical Forests under Different Geological Conditions in East Kalimantan, Indonesia................................................................................... 236 5.7.1 Chemical Properties of the Soils Studied.......................................... 237 5.7.2 Soil Solution Composition................................................................. 237 5.7.3 Fluxes of Ions in Solute Leaching and Vegetation Uptake and Proton Budgets in Soils..................................................................... 238 5.7.4 Dominant Soil Acidification Processes in Tropical Forests.............. 242 5.7.5 Proton Generation and Consumption in Soil Profiles....................... 242 5.7.6 Acid Neutralization in Soils.............................................................. 243 5.7.7 Implication of Proton Budgets for Pedogenetic Soil Acidification...................................................................................... 243 5.8 Relationship between Chemical and Mineralogical Properties and the Rapid Response to Acid Loads of Soils in Humid Asia................................ 245 5.8.1 Soils Studied......................................................................................246 5.8.2 Acid Titration Experiment and Source of Soil Alkalinity................248 5.8.3 Column Experiment for Eight Selected Soils.................................... 251 5.8.4 Interpretation of Acid-Neutralizing Reactions under Laboratory and Field Conditions.......................................................................... 255 5.9 General Discussion on the Pedogenetic Acidification Process...................... 256 5.9.1 Organic Matter Dynamics................................................................. 257 5.9.2 Soil Acidification Rate in the Entire Soil Profile.............................. 257 5.9.3 Factors Controlling Proton Generation and Consumption in Relation to Organic Matter Cycles....................................................260 5.9.4 Pedogenetic Implication for Proton Budget in Soil........................... 261
Pedogenetic Acidification in Humid Asia
171
5.9.5 Fates of Soil Acidity and Dissolved Organic Matter during Pedogenetic Soil Acidification of Forest Soils.................................. 262 5.9.6 Response of Soil Minerals against Further Acid Load..................... 262 References............................................................................................................... 263
5.1 INTRODUCTION Soils in humid Asia exhibit relatively incipient mineralogical characteristics because of the dominant steep slopes, crust movement, and volcanic activity on young alpine fold belts [FAO 2001] compared with soils developed on stable plains associated with the Precambrian shield in eastern South America or equatorial Africa. Among these continents, geological components are also different; distribution of mafic (or basic) parent materials is limited and sedimentary rocks as well as igneous felsic rocks are major parent rocks in humid Asia [Geological Survey of Japan 2004], whereas some metamorphic or basic rocks are dominated in tropics of other continents [FAO/Unesco 1971, 1977]. Because of these essential differences in topographical and geological conditions, dominant upland soils in humid Asia are Ultisols, in contrast to the humid tropics in other continents where Oxisols are predominant [Soil Survey Staff 1999]. In Chapter 4, the distribution pattern of upland soils in humid Asia (i.e., Indonesia, Thailand, and Japan) was presented and the ecological processes under conventional upland farming and shifting cultivation in these regions were comparatively analyzed. In the present work, pedogenetic soil acidification processes under relatively natural undisturbed conditions in these regions are analyzed to clarify the ecosystem processes that would bring the difference in soil properties and human activities presented in Chapter 4. Soil acidification is one of the major pedogenetic processes in soils under climates where precipitation exceeds evapotranspiration. Both external and internal acid loads such as proton extrusion concomitant with excess uptake of cation over anion by vegetation, nitrification, dissociation of organic acids, and dissociation of carbonic acid are responsible for soil acidification under leaching conditions [van Breemen et al. 1983; Binkley and Richter 1987; Ulrich 1989]. Although many studies on soil acidification have tended to focus on the effects of acidic deposition or external acid loads, the importance of understanding pedogenetic soil acidification induced by internal acid loads is recognized [Krug and Frink 1983; Hallbäcken and Tamm 1986]. It is one of the major driving forces of mineral weathering and could also be regarded as an ecosystem process that favors acquiring mineral resources from soil and/or the lithosphere. The contribution of internal acid loads to pedogenetic soil acidification was reported to vary greatly from soil to soil. Acidity originating from the dissociation of organic acids and nitrification in surface soils was reported to be responsible for podzolization [Funakawa et al. 1992; Brahy et al. 2000a]. Organic acids, particularly low molecular weight organic acids, were found to contribute to congruent mineral dissolution and Al eluviation caused by complexation in podzolic soils [Lundström 1993; van Hees et al. 2000]. In contrast, acidity originating from the dissociation of carbonic acid was reported to be responsible for andosolization and brunification.
172
World Soil Resources and Food Security
Carbonic acid in ando soils contributed to incongruent dissolution of minerals derived from volcanic materials and formation of amorphous Al silicates [Ugolini et al. 1988; Ugolini and Sletten 1991], while in brown forest soil it contributed to incongruent mineral dissolution and resulted in accumulation of oxides [Ugolini et al. 1990]. In the following sections, pedogenetic acidification processes under different geological and/or climatic conditions are comparatively analyzed using the proton budget method. The fate of organic matter, which is one of the key processes to regulate soil acidification and mineral weathering, is also traced in detail in relation to the properties of soils as cumulative results of these processes. Finally these processes are interpreted from the viewpoint of an ecosystem process that enhances to acquire mineral resources from soil and/or the lithosphere into biosphere.
5.2 P EDOGENETIC SOIL ACIDIFICATION BY DIFFERENT PROTON SOURCES IN FORESTED ECOSYSTEMS IN JAPAN Since pedogenetic soil acidification is controlled by the balance of proton generation and consumption (e.g., dissociation and decomposition of organic acids, nitrification, and nitrate uptake by vegetation) in the biogeochemical cycle of forested ecosystems, a quantitative analysis of both the proton-generating and -consuming processes is required to understand the respective pedogenetic processes in relation to soil acidification. Calculation of proton budget allows us to evaluate the contribution of different proton sources to soil acidification [van Breemen et al. 1984]. In particular, the contribution of external acid loads to soil acidification was successfully quantified in the whole soil compartment by application of the theory of proton budget [van Breemen et al. 1983; Bredemeier et al. 1990]. However, predominant protongenerating and -consuming processes that might be responsible for pedogenetic soil acidification must surely be different in each soil horizon; for example, nitric acid and organic acids produced in surface soil horizons may work as the temporary acids and be consumed in deeper soil horizons by nitrate uptake of vegetation and decomposition of organic acids. Therefore, it is necessary to evaluate the protongenerating and -consuming processes in each soil horizon compartment in order to analyze the pedogenetic processes in a whole soil profile based on the proton budget. In this section, we attempt to apply the theories of proton budget and acidneutralizing capacity to the analysis of the acidifying processes in each soil horizon compartment, and then accordingly to analyze the dominant soil forming processes in three representative forest soils in Japan, i.e., ando soil, podzolic soil, and brown forest soil.
5.2.1 Experimental Plots Three plots were selected to include ando soil under a cold temperate forest, podzolic soil under a cold temperate forest, and brown forest soil under a warm temperate forest (Figure 5.1); these sites are representative of 91.7% of total forest area in Japan (i.e., 11.5% for ando soils, 3.6% for podzolic soils, and 76.5% for brown forest soils),
173
Pedogenetic Acidification in Humid Asia Kyoto soils, JP1–4
TG
130º E
140º E
NG
40º N
KT
JP5–10
30º N
Mie soils
FIGURE 5.1 Location of the experimental plots used in the analysis in Sections 5.2 to 5.5.
according to Classification of Forest Soil in Japan [Forest Soil Division 1976]. Site descriptions are given in Table 5.1. Detailed information was given in Shinjo et al. [2006]. At the Nagano plot (NG), the soil formed by the process of andosolization was derived from secondary sediments of volcanic products and was classified as Acrudoxic Melanudands [Soil Survey Staff 2006]. Compared to the other two plots,
TABLE 5.1 Site Description of Japanese Plots Coordinates Mean annual air temperature (°C) Mean annual precipitation (mm) Elevation (m) Soil typea Parent materials Vegetation Slope Position a b
NG
TG
KT
N35°57′, E138°28′ 6.9b
N35°37′, E135°10′ 10.7
N35°01′, E135°47′ 15.9b
1422b
1782b
1490b
1440 Acrudoxic Melanudands Volcanic ash
675 Andic Haplohumods Sedimentary rocks, granite Fagus crenata
90 Typic Dystrudepts Sedimentary rocks
Quercus mongolica var. crispula S70°E, sloping (5%) Lower slope
S76°W, sloping (19%) Upper slope
Quercus serata Shiia cuspidata N85°W, steep (47%) Upper slope
Soils were classified according to Soil Taxonomy [Soil Survey Staff 2006]. Average of 1984−2004 taken from nearest meteorological stations, i.e., Nobeyama, Miyazu, and Kyoto.
174
World Soil Resources and Food Security
soil pH was relatively high, ranging from 4.6 (A horizon) to 5.0 (Bw horizon) (Table 5.2). At the Tango plot (TG), located in Tango Peninsula, Kyoto Prefecture, the soil was affected by podzolization and derived from sedimentary rocks and granite. It was classified as Andic Haplohumods. Soil pH was low, ranging from 3.8 (AE horizon) to 4.5 (Bs horizon). At the Kyoto plot (KT), located on Mt. Yoshida, Kyoto Prefecture, the soil was affected by brunification and derived from sedimentary rocks. It was classified as Typic Dystrudepts. Soil pH was 4.2 throughout the profile (Table 5.3). At these three plots, throughfall and soil solution was collected by precipitation collectors and tension-free lysimeters, draining a surface area of 200 cm2, beneath the O, A, and B horizons (or the A2 horizon at NG) in each 5 replications, respectively, during the period from June 2003 to February 2005. After filtration through 0.45-μm cellulose acetate membrane filters, the chemical composition of the solution samples was determined, followed by analysis using proton budget method (see the Appendix of Chapter 4 (4.A3) for detailed methodology). Water fluxes of soil solution percolated from the bottom of the O and B horizons were estimated by applying Darcy’s law to the unsaturated hydraulic conductivity and the gradient of the hydraulic heads in surface soil (0- to 5-cm depth) and subsoil (40- to 45-cm depth) (also described in Appendix of Chapter 4). The annual water fluxes, as the
TABLE 5.2 Stock and Annual Flow of Carbon in the Japanese Plots
C Stock (Mg C ha−1) Aboveground biomass Fine root biomass O horizon A horizona B horizona Soil organic matter O horizon Mineral soil horizonsb C Flow (Mg C ha−1 yr −1) Soil respiration Root respiration Decomposition of soil organic matter Decomposition of O horizon Decomposition of mineral soil Litterfall Wood increment a b
NG
TG
KT
78 2.0 0.4 0.6 1.0
83 2.9 2.1 0.6 0.2
115 4.2 1.4 1.5 1.3
3.6 187
30.6 121
3.4 66
5.0 1.5 3.4 1.6 1.9 1.7 1.5
8.2 2.8 5.5 3.4 2.0 2.1 2.5
9.3 4.2 5.1 4.1 1.0 2.9 10.1
The A and B horizons corresponded to the A1 and A2 horizons, respectively, at NG. Organic carbon in soil at 0−30 cm depths was counted.
pH Depth Site
Particle Size Distribution
Exchangeable Total C
Bases
Al
CEC
Sand
(cmolc kg−1)
Silt
Clay
Sio
(%)
Feo
Alo
(g kg−1)
Fed
Ald
Horizon
(cm)
(H2O)
(KCl)
(g kg−1)
(g kg−1)
NG (Andisols)
A1 A2 A3 BA Bw
0–5 5–15 15–35 35–50 50–76+
4.6 4.6 4.8 4.9 5.0
4.3 4.2 4.4 4.6 5.2
212 167 134 98 23
6.6 1.7 0.5 0.3 0.7
2.4 3.2 3.0 1.7 0.2
57.3 53.6 42.1 37.2 17.3
25 26 24 27 37
31 31 25 24 20
44 43 51 50 43
3.5 6.1 6.8 13.2 16.5
16.3 14.4 20.2 20.1 18.7
24.7 25.1 32.8 36.7 36.3
26.7 25.3 30.1 34.7 35.4
24.3 20.3 26.3 20.8 13.3
TG (Spodosols)
AE Bhs Bs BC C
0–9 9–35 35–72 72–86 86–95+
3.8 4.4 4.5 4.6 4.7
3.5 4.4 4.5 4.2 4.3
125 66 18 5 3
0.8 0.5 0.6 0.8 0.9
10.1 5.6 4.5 6.1 5.5
49.5 36.3 19.9 15.9 12.4
12 14 27 56 71
40 42 32 18 14
49 45 41 27 15
0.1 2.2 2.2 0.3 0.3
11.6 14.9 7.0 2.8 1.3
4.9 17.5 11.7 3.1 2.3
27.8 29.6 23.3 18.4 12.9
6.5 18.3 11.9 5.3 3.2
KT (Inceptisols)
A Bw1 Bw2 Bw3
0–7 7–15 15–45 45–65+
4.2 4.2 4.2 4.2
3.7 3.9 4.1 4.1
45 13 8 5
0.4 0.2 0.1 0.2
6.8 4.5 3.7 3.5
19.5 13.9 10.2 10.0
47 43 48 50
6 14 4 1
47 43 48 50
0.1 0.1 0.2 0.2
2.6 1.9 1.6 1.7
4.4 2.1 2.2 1.8
12.4 17.4 17.6 14.6
6.3 4.4 3.7 3.2
Pedogenetic Acidification in Humid Asia
TABLE 5.3 Physicochemical Properties of Soils in Japanese Plots
175
176
World Soil Resources and Food Security
sum of daily water fluxes of throughfall and soil solution, are summarized in Table 5.4. Water fluxes of throughfall were 1155, 1676, and 1657 mm yr−1 at NG, TG, and KT, respectively. Water fluxes leached from the bottom of the B horizon were 616, 873, and 660 mm yr−1 at NG, TG, and KT, respectively. The water losses, i.e., the differences between water fluxes of throughfall and fluxes of water percolated from the bottom of the B horizon, were 539, 803, and 997 mm yr−1 at NG, TG, and KT, respectively (Table 5.4). These figures were comparable to potential evapotranspiration estimated by Thornthwaite [1948], i.e., 561, 598, and 822 mm yr−1 at NG, TG, and KT, respectively.
5.2.2 Soil Solution Composition Solution pH at NG (Andisol site) was relatively high (6.0–6.1) throughout the soil profile, but at KT (Inceptisol site) it was moderately low (4.3–4.5) (Table 5.4). At TG (Spodosol site), it was low in the O horizon (3.9), followed by an increase in pH with depth (4.1 for the A horizon and 4.6 for the B horizon, respectively). The concentrations of dissolved organic carbon (DOC) in soil solution in the O horizon were highest at TG (28.9 mg C L−1), followed by KT (15.4 mg C L−1), and NG (6.8 mg C L−1) (Table 5.4). The DOC concentrations in soil solution were highest in the O horizon, followed by a decrease with depth due to adsorption, and microbial decomposition at TG and KT (Table 5.4). The DOC concentrations of soil solution in the O horizon (11.4–70.6 mg C L−1) were correlated with soil temperature (1.7°C –20.7°C) at TG (r = 0.50**, n = 18, p < 0.05), indicating that DOC was produced primarily by the microbial decomposition of litter in the O horizon. At NG, DOC concentrations in soil solution were low throughout the soil profile (1.4–6.8 mg C L−1) due to DOC decomposition in the O horizon and the high adsorption capacity of amorphous materials in mineral soil. Organic acids were dominant among the biologically produced anions (organic acids, nitrate, and bicarbonate) at all plots (Table 5.4). The concentrations of organic acids were positively correlated with DOC concentrations at NG (r = 0.71, n = 171, p < 0.01), TG (r = 0.73, n = 210, p < 0.01), and KT (r = 0.77, n = 277, p < 0.01). According to linear regression analysis, the negative charge per 1 mol DOC was calculated to be 0.30 at NG, which corresponded to one dissociated acidic functional group for 3.3 C atoms. A low carbon to charge ratio in soil solution of Andisol was also reported by Ugolini et al. [1988], suggesting the presence of low molecular weight organic acids. The negative charge per 1 mol DOC at TG and KT was 0.08 molc and 0.13 molc, respectively. These figures correspond to one dissociated acidic functional group for 12.2 C and 7.4 C atoms at TG and KT, respectively. The high carbon to charge ratios in soil solution at TG and KT suggest the presence of high molecular weight fulvic acids, which contain 7 C atoms for each acidic functional group [Thurman 1985]. The NO3− concentrations in soil solution were highest at TG (0.07–0.30 mmolc −1 L ), followed by NG (0.08–0.12 mmolc L−1), and KT (0.00–0.08 mmolc L−1) (Table 5.4). In all plots, the NO3− concentrations in soil solution were highest in the O horizon (O and A horizons at KT) due to mineralization and nitrification, followed by a decrease with depth due to vegetation uptake (Table 5.4). It was considered that NO3−
Water Flux
DOC
HCO
− 3
Org
n− a
NO
− 3
−
CL + SO
2− 4
H
+
+ 4
NH
Fe2+
Aln+
(mm)
pH
(mg C L )
NG
b
TF O A1 A2
1155 779 718 616
6.0 6.0 6.1 6.0
5.3 6.8 1.5 1.4
0.039 0.022 0.023 0.018
0.059 0.222 0.070 0.074
0.021 0.116 0.107 0.075
0.067 0.091 0.109 0.079
0.001 0.001 0.001 0.001
0.014 0.019 0.001 0.009
0.159 0.406 0.302 0.238
0.000 0.001 0.000 0.000
0.004 0.012 0.003 0.002
0.049 0.526 0.426 0.183
TG
TFb O A B
1676 1189 945 873
5.4 3.9 4.1 4.6
4.7 28.9 13.0 2.3
0.025 0.001 0.000 0.001
0.105 0.333 0.255 0.138
0.024 0.298 0.215 0.074
0.248 0.455 0.393 0.376
0.004 0.123 0.080 0.026
0.009 0.186 0.087 0.006
0.386 0.703 0.548 0.487
0.000 0.009 0.006 0.000
0.064 0.140 0.066 0.003
0.012 0.137 0.105 0.037
KT
TFb O A B
1657 973 805 660
5.2 4.3 4.4 4.5
5.6 15.4 14.0 4.3
0.048 0.022 0.016 0.013
0.031 0.121 0.134 0.076
0.043 0.071 0.081 0.004
0.125 0.224 0.216 0.258
0.003 0.062 0.034 0.027
0.025 0.039 0.045 0.006
0.216 0.287 0.288 0.271
0.000 0.005 0.003 0.000
0.003 0.056 0.078 0.041
0.003 0.062 0.107 0.211
b
−1
(mmolc L )
Si
Horizon
a
(mmolc L )
Na+ + K+ + Mg2+ + Ca2+
−1
(mmol L−1)
−1
Pedogenetic Acidification in Humid Asia
TABLE 5.4 Water Flux and Annual Volume-Weighed Mean Concentrations of Ions in Throughfall and Soil Solution
Orgn− represents anion deficit, the negative charge of organic acids. TF represents throughfall.
177
178
World Soil Resources and Food Security
was produced from organic nitrogen by heterotrophic fungi in the acidic forest soils, since autotrophic nitrification is limited in acidic soils [Killham 1990]. The concentrations of bicarbonate while low were significant throughout the soil profile at NG (0.02 mmolc L−1) and KT (0.01–0.02 mmolc L−1), but were negligible in soil solution at TG because solution pH was low (Table 5.4). Judging from the fact that the concentrations of bicarbonate in soil solution were much lower than those of organic acids and NO3− , the contribution of dissociation of carbonic acid to soil acidification was negligible at all plots in the present study. The concentrations of the essential nutrient elements including K, Mg, and Ca showed maximum levels in the O horizon, followed by a decrease with depth as K, Mg, and Ca were taken up by vegetation at all plots (Table 5.4). The sum of K, Mg, and Ca concentrations were 0.19–0.39 mmolc L−1, 0.17–0.40 mmolc L−1, and 0.17–0.22 mmolc L−1 at NG, TG, and KT, respectively. One of the reasons for the highest concentrations of DOC and basic cations at TG was considered to be due to the presence of ectomycorrhiza in the Fagus crenata. In this case, the ectomycorrhiza could enhance cation availability by enhancing DOC production by fungi [Griffiths et al. 1994]. The Al concentrations of soil solution at NG were negligible throughout the soil profile presumably due to the formation of insoluble complexes with humic acids (Table 5.4) [Ugolini and Sletten 1991]. At TG, Al concentrations of soil solution in the A horizon (0.14 mmolc L−1) were higher than those in the O horizon (0.06 mmolc L−1), followed by a decrease in the B horizon (0.07 mmolc L−1) due to removal of counter anions, especially organic acids and NO3− , in soil solution (Table 5.4). The same pattern was observed at KT (0.06, 0.08, and 0.04 mmolc L−1 in the O, A, and B horizons, respectively). Given that Al content in litterfall was negligible at TG, as discussed later in detail, high Al concentrations of soil solution in the O horizon were probably due to Al release by weathering of aluminosilicate minerals admixed to the O horizon and organic matter decomposition of dead roots at TG [Rustad and Cronan 1995]. In contrast, at KT, the Al released by litter decomposition also contributed to the high Al concentrations of soil solutions in the O horizon, as Al input by litterfall was high at about 9.0 kg ha−1 yr−1 (data not shown). Aluminum mobilization could be enhanced in the O and A horizons due to complexation with organic acids. On the other hand, mobility of organic acids was decreased by saturation with Al. The ratio of charge of Al to the negative charge of organic acids increased at TG as soil solution percolated in the A (0.55) and B horizons (0.48) compared to the O horizon (0.19), while it was similarly high at KT throughout the O (0.46), A (0.58), and B horizons (0.54) (Table 5.4).
5.2.3 NPGOrg, NPGNtr, NPGBio, and ΔANC Based on fluxes of ions at each horizon (Figure 5.2), and excess cation uptake by vegetation, NPGOrg, NPGNtr, and NPGBio (where NPG stands for net proton generation), and ΔANC were calculated according to the method presented in Appendix of Chapter 4 (4.A3) (Figure 5.3). NPGOrg in the O horizon was 1.0, 2.2, and 0.7 kmolc ha−1 yr−1 at NG, TG, and KT, respectively (Figure 5.3). These values are generally consistent with the report by Guggenberger and Kaiser [1998] in which the fluxes of DOCassociated protons leached from the O horizon ranged from 0.4 to 3.7 kmolc ha−1 yr−1
179
Pedogenetic Acidification in Humid Asia (b)
(a) HCO–3 H+ + Na + Fe3+ SO42– NH4
TF O
TF
K+ Mg2+ Ca2+
Orgn– NO–3
–2
Orgn–
NO–3
Cl–
H+ Na+ NH4+K+Mg2+
Aln+
B
NG
–4
Fe3+
A
A2 –6
HCO–3
Ca2+
O
Aln+
Cl–
A1
SO42–
0
2
(c)
4
–15
–10
–5
0
(kmolc ha–1 yr–1)
5
10
15
(kmolc ha–1 yr–1)
HCO–3
TF O
TG
6
Mg2+
NH+4
Orgn– SO42– NO3– Cl– H+ Na+
K+
Ca2+
Fe3+
Aln+
A B KT
–15
–10
–5
0
5
10
15
(kmolc ha–1 yr–1)
FIGURE 5.2 Fluxes of solutes at each horizon. TF represents throughfall. O, A, A1, A2, and B represent soil horizons.
in the acidic forest soils. On the other hand, protons were consumed by decomposition and adsorption of organic acids in the NG, TG, and KT subsoils (Figure 5.3). At TG, the DOC fluxes of throughfall (TF) and the O, A, and B horizons were 78, 344, 123, and 20 kg C ha−1 yr−1, respectively, and hence the O horizon was the main source of DOC. At KT, the DOC fluxes of throughfall and the O, A, and B horizons were 93, 149, 113, and 28 kg C ha−1 yr−1, respectively; here the DOC fluxes of throughfall were appreciably high and, therefore, both the canopy and the O horizon were the main sources of DOC at KT. Leaching of secondary metabolites as tannin and microbial metabolites on leaves could be responsible for DOC in throughfall. In contrast, at NG, the DOC fluxes of throughfall and the O, A, and B horizons were 61, 53, 11, and 9 kg C ha−1 yr−1, respectively, and there was no clear increase in DOC flux in the O horizon. Based on the aforementioned decomposition rates of organic matter in the O horizon (Table 5.3) and the DOC fluxes leached from the O horizon, the latter were high compared to mineralized carbon (the sum of the effluxes of CO2 released by organic matter decomposition and DOC fluxes leached from the O horizon) in the O horizon in the highly acidic (podzolic) soil of TG (10.1%), in comparison with NG (3.3%) and KT (3.6%). DOC leached from the O horizon was reported to be the
180
World Soil Resources and Food Security
(a)
(b) NG
O A1 A2 Total –6
O
(H+)in–(H+)out ΔANC NPGNtr
B
NPGOrg NPGBio
Total
NPGCar
–4
TG
A
–2
0
2
4
6
–6
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
–4
–2
0
2
4
6
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
(c) KT
O A B Total –6
–4
–2
0
2
4
6
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
FIGURE 5.3 Vertical variation of NPGOrg, NPGNtr, NPGBio, and ΔANC. TF represents throughfall. O, A, A1, A2, and B represent soil horizons.
water soluble lignocellulose degradation product in the course of lignin decomposition by litter-decomposing fungi and white-rot fungi [Guggenberger and Zech 1994]. Predominant fungal activity and inhibited bacterial activity might therefore result in high DOC production in highly acidic soils. According to Figure 5.3, in the O horizon, protons were produced by nitrification that was predominant over nitrate uptake by vegetation and microorganisms, since NPGNtr was positive at all plots except for KT, where NPGNtr was small. On the other hand, protons were consumed by predominant nitrate uptake by vegetation and microorganisms, since NPGNtr was negative in the A and B horizons at all plots. Protons were produced by mineralization and nitrification in the O horizon, 0.7 and 1.1 kmolc ha−1 yr−1 at NG and TG, respectively, since NO3− leaching from the O horizon, 0.9 and 3.5 kmolc ha−1 yr−1 at NG and TG, respectively, was higher than NO3− flux of throughfall (Figure 5.2). High NO3− flux leached from the O horizon at TG was presumably due to low NO3− uptake by ectomycorrhizal roots [Harley and Smith 1983]. Judging from the fact that NO3− leaching from the O horizon, 0.7 kmolc ha−1 yr−1, was almost equivalent to NO3− flux of throughfall at KT, the canopy of evergreen forests was the main source of NO3−. Protons were produced in the canopy by nitrification on leaves and NO3− exudation from leaves at KT. At all
Pedogenetic Acidification in Humid Asia
181
plots, NPGOrg and NPGNtr in the O horizon were almost completely consumed in the A and B horizons, and hence the contribution of NPGOrg and NPGNtr to soil acidification in the whole soil compartment was low. Cation contents exceeded anion contents in litter and wood materials at all plots (Table 5.5). Excess cation charge was compensated for by the net proton load to soils as NPGBio. NPGBio in the whole soil compartment was 3.4, 3.3, and 7.2 kmolc ha−1 yr−1 at NG, TG, and KT, respectively. NPGBio was highest among proton sources in the whole soil compartment at all plots, although NPGBio for litter production would be almost neutralized by litterfall returned to soil. NPGBio was then distributed to each soil horizon based on the pattern of fine root biomass in the soil profile. At TG, NPGBio was concentrated in the O horizon, 2.4 kmolc ha−1 yr−1, while it was also distributed in the A and B horizons at NG, 0.7–1.7 kmolc ha−1 yr−1 and at KT, 2.2–2.6 kmolc ha−1 yr−1 (Figure 5.3). Concentrated fine root biomass in the O horizon at TG was responsible for the high NPG due to cation excess uptake by vegetation (Figure 5.3; Table 5.3). Regarding one of the controlling factors of fine root distribution, which was responsible for NPGBio, Aber et al. [1985] reported that a fine root and ectomycorrhiza system developed in the O horizon enhanced NH +4 mobility in acidic soils. This finding is consistent with the higher NH +4 concentrations of soil solution found in the O horizon at TG, 0.19 mmolc L−1, compared to NG, 0.02 mmolc L−1, and KT, 0.04 mmolc L−1, in the present study (Table 5.4). The major proton consuming processes were cation release by organic matter decomposition, exchange reactions, and mineral weathering. Protons were neutralized by basic cations at NG, while protons were neutralized partly by Al mobilization in acidic soils at TG and KT. The fluxes of Al leached from the O and A horizons were 0.77 and 1.32 kmolc ha−1 yr−1 at TG, and 0.55 and 0.63 kmolc ha−1 yr−1 at KT, respectively (Figure 5.2). According to Figure 5.3, at NG, the soil acidification rate (−ΔANC) was low throughout the soil profile, −2.0, 0.4, and −1.0 kmolc ha−1 yr−1 in the O, A, and B horizons, respectively. Protons produced in each horizon were completely consumed in the same horizon at NG, since {(H+)in − (H+)out} was negligible in the A and B horizons. Also seen from Figure 5.3, at TG, the soil acidification rate was high in the O horizon, with ΔANC at −3.9 kmolc ha−1 yr−1. Since the protons produced (the sum of NPGOrg, NPGNtr, NPGCar, and NPGBio) were not completely compensated by cation release in the O horizon, 1.4 kmolc ha–1 yr–1 of protons {(H+) in – (H+)out} were leached from the O horizon and consumed in the A and B horizons (Figures 5.2 and 5.3). At KT, the soil acidification rate was moderately high throughout the soil profile, with ΔANC ranging from –2.6 to –1.3 kmolc ha–1 yr–1. Although most of the protons produced in the O horizon were consumed in the same horizon, 0.5 kmolc ha–1 yr–1 of protons {(H+)in – (H+)out} were leached from the O horizon and consumed mainly in the A horizon (Figures 5.2 and 5.3). According to van Breemen et al. [1984], soil acidification rates in the whole soil compartment range from 7.5 to 16 kmolc ha−1 yr−1 in soils at neutral pH due to carbonic acid dissociation, from 2.5 to 7.5 kmolc ha−1 yr−1 in acidic soils with weatherable minerals caused by vegetation uptake and nitrification, and less than 2.5 kmolc ha–1 yr–1 in podzolic soils caused by vegetation uptake and dissociation of organic acids. The soil acidification rate in the whole soil compartment of TG (2.5 kmolc ha−1 yr−1) was a value intermediate between the soil acidification rates of podzolic soils
182
TABLE 5.5 Uptake of Cations and Anions by Vegetation OM Production
Na
K
Ca
Mg
Fe
Al
Cl
S
P
(Cation)bio
(kg ha−1 yr−1)
(Anion)bio
NPGBio
(Mg C ha−1 yr−1)
(kmolc ha−1 yr−1)
NG
Wood increment Litterfall
1.5 1.7
2.0 1.5
1.3 11.7
10.6 35.1
1.2 6.6
0.6 4.2
0.1 1.0
0.1 0.2
0.2 1.4
2.0 3.5
0.78 2.92
0.08 0.20
0.71 2.71
TG
Wood increment Litterfall
2.5 2.1
4.4 2.9
5.5 9.2
15.8 21.2
5.2 4.7
0.6 2.2
0.1 0.4
0.1 0.2
0.4 2.0
0.1 0.8
1.57 1.93
0.03 0.15
1.55 1.78
KT
Wood increment Litterfall
10.1 2.9
18.1 2.2
29.1 18.9
29.1 26.7
6.2 9.2
4.1 1.8
1.8 9.0
0.7 0.4
1.9 2.7
1.1 2.4
3.84 3.73
0.17 0.24
3.67 3.49
World Soil Resources and Food Security
Horizon
Pedogenetic Acidification in Humid Asia
183
and acidic soils with weatherable minerals. Soil acidification rates in the whole soil compartment of NG (3.1 kmolc ha−1 yr−1) and KT (5.9 kmolc ha−1 yr−1) were similar to those of acidic soils with weatherable minerals. In the present study, taking together the data at NG and TG and the findings of Binkley [1992], the low soil acidification rates mainly result from low cation excess uptake by vegetation, caused by low wood increment (0.7 and 1.5 kmolc ha−1 yr−1 at NG and TG, respectively, and 0.5 kmolc ha−1 yr−1 on average) (Table 5.5). In the case of Binkley’s [1992] study, the low cation excess accumulation in wood might also be attributed to low cation contents in plant materials due to a low cation pool of soils derived from parent materials such as glacial till, as well as lower cation contents in plant materials of coniferous vegetation and low primary production due to low air temperatures and the old age of forests. Shibata et al. [1998] suggested that the higher contribution of internal proton generation by basic cation accumulation in vegetation to soil acidification in Japanese volcanogenous regosols was related to a larger basic cation pool of soils. The latter was considered to be responsible for higher cation content in plant materials. The higher cation excess accumulation in wood at NG and TG compared to those reported by Binkley [1992] was considered to be related to a higher cation pool of soils, as well as to the younger age of forests. In contrast, cation excess accumulation in wood was higher at KT (3.7 kmolc ha−1 yr−1) than at NG and TG; in this case, it was considered to be related to higher primary production under the warmer climate. Pedogenetic soil acidification was considered to include cation leaching by proton generation through the dissociation of organic acids and nitrification, and subsequent cation excess accumulation in wood in the growth stage of forests. Although the contribution of the temporary acids such as organic acids and nitric acid, which were produced in the O horizon and then consumed in the A and B horizons, to soil acidification in the whole soil compartment was low. Translocation of the temporary acids, as well as distribution of root biomass, contributed to the temporal and spatial heterogeneity of proton generation and consumption in the soil profile, resulting in different pedogenetic soil acidification.
5.2.4 Pedogenetic Implication of the Proton Budget The soil solution studies by Ugolini et al. [1988, 1990] and Ugolini and Sletten [1991] demonstrated the essential role of proton donors including organic acids and carbonic acid on pedogenesis, i.e., podzolization, andosolization, and brunification. The pedogenetic soil acidification processes of NG, TG, and KT can be described in relation to andosolization, podzolization, and brunification, respectively. At NG, the distribution pattern of Alo and Alp in soil with depth suggests that an organo-mineral complex was dominant in surface soil horizons (Table 5.3). On the other hand, the occurrence of allophane was suggested in subsoil by the ratio of (Alo – Alp)/(Sio – Sip), which was similar to the ideal Al/Si ratio of Al-rich allophane, i.e., 2 (Table 5.3). The low concentrations of DOC and Al in soil solution indicated that in situ weathering was a major weathering process at NG (Table 5.4). The low soil acidification rate and low NPGOrg contributed to incongruent dissolution, resulting in the formation of amorphous Al silicates (Figure 5.3). These processes were in accordance with the concept of andosolization suggested by Ugolini and Sletten
184
World Soil Resources and Food Security
[1991]. Acidity originating from organic acids was the determining factor of pedogenesis of ando and podzolic soils derived from volcanic parent materials. Lower acidity was responsible for andosolization, while higher acidity due to the dissociation of organic acids was responsible for podzolization [Shoji et al. 1982]. Lower acidity, together with volcanic parent material, was considered to be responsible for andosolization at NG. At TG, the distribution pattern of Alo and Alp in soil with depth suggested eluviation of Al from the A horizon and illuviation in the BA and B horizons (Table 5.3). High fluxes of DOC and Al leached from the O and A horizons and their illuviation in the B horizon at TG coincides with the concept of podzolization (Figure 5.2). In podzolic soils derived from volcanic ash, organic acids were found to be responsible for the eluviation of Al by congruent dissolution in surface soil and carbonic acid was responsible for the accumulation of Fe and Al oxides by incongruent dissolution in subsoil [Ugolini and Dahlgren 1987]. In the present study, it was considered that organic acids were the dominant anions, and that bicarbonate was negligible throughout the soil profile at TG due to lower pH of soil derived from acidic parent materials (Table 5.4). In addition, it was shown that cation excess uptake by vegetation due to concentrated fine root biomass in the O horizon resulted in intensive soil acidification in surface soil and subsequent high proton efflux to subsoil (Figure 5.3). This is consistent with the report by Nielsen et al. [1999] for the change in vegetation from heath to spruce, the root biomass of which concentrated in the surface soil and accelerated podzolization. At KT, higher Alo and Feo contents in surface soil horizons suggested that Al and Fe oxides were immobilized in metal-humus complexes (Table 5.3), which is in accordance with the concept of brunification. The immobilization of organic matter by oxides was supported by the lower Fe concentrations in soil solution and higher saturation of the negative charge of organic acids by Al in soil solution throughout the soil profile at KT, compared to TG (Table 5.4). Ugolini et al. [1990] suggested that incongruent dissolution by carbonic acid results in accumulation of Fe oxides. However, judging from the fact that the dominant anions in soil solution were organic acids and those of carbonic acid were low at KT (Table 5.4), pedogenesis of brown forest soils appears to be variable in terms of anion contribution [Ugolini et al. 1990]. On the other hand, at KT, high fluxes of Al leached from the O, 0.55 kmolc ha−1 yr−1, and A horizons, 0.63 kmolc ha−1 yr−1, and subsequent decrease of Al flux in the B horizon, 0.27 kmolc ha−1 yr−1, suggested that pedogenesis of brown forest soil includes incipient podzolization, as also suggested by Hirai et al. [1988] (Figure 5.2). However, judging from the higher Al input by litterfall (1.00 kmolc ha−1 yr−1) (Table 5.4), biological cycling of Al as well as weak podzolization resulted in Al mobilization in the O and A horizons. The difference in the intensity of podzolization between KT and TG was caused by 1) distribution patterns of the vegetation root system that determined the soil horizons subjected to intensive acidification by NPGBio, 2) the different contributions of strong acids to soil acidification, and 3) the capacity of acid neutralization. Net proton generation by cation excess uptake by vegetation was concentrated in the O horizon at TG; it was also distributed in the A and B horizons at KT in proportion to the distribution of the fine root biomass. This is consistent with
Pedogenetic Acidification in Humid Asia
185
the report by Nielsen et al. [1987] for the change in vegetation from heath to oak, the latter root biomass distributed throughout the soil profile converted podzolic soil to brown forest soil, i.e., depodzolization. In relation to the second point above, in spite of KT showing the highest soil acidification rate, the contribution of strong acids, e.g., NPGOrg and NPGNtr, to soil acidification was low, since both decomposition and adsorption of organic acids by oxides were high and nitrate uptake by vegetation was also sufficiently high to consume the net proton generation by nitrification in the O horizon (Figure 5.3). As for the third point, the high net proton efflux to the A horizon, 0.71 kmolc ha−1 yr−1, accelerated net Al leaching from the A horizon, 0.55 kmolc ha−1 yr−1, by mineral weathering at TG, while at KT net Al leaching from the A horizon was low, 0.08 kmolc ha−1 yr−1, due to low net proton efflux to the A horizon, 0.33 kmolc ha−1 yr−1 (Table 5.4; Figure 5.2). High acid neutralization through the release of basic cations by a high rate of organic matter decomposition in the O horizon of KT might be responsible for the low proton efflux to subsoil (Figure 5.2).
5.3 F ATES OF SOIL ACIDITY AND DISSOLVED ORGANIC MATTER DURING PEDOGENETIC SOIL ACIDIFICATION OF FOREST SOILS IN JAPAN In the previous section, DOC leaching toward subsoils under highly acidic conditions was observed, resulting in Al translocation through the soil profiles. These metal ions and migrated DOC often form organo-mineral complexes in subsoils, which is substantially stable against biodegradation. The carbon sequestration in soil is today receiving global concern, in relation to the countermeasure against global warming. The process and fate of metal ions and DOC should also be clarified from this context. It is reported that, in the pedogenesis of acidic forest soils, Al is considered to play an important role through, for example, dissolution from primary minerals, hydrolysis and formation of Al hydroxides, formation of organo-mineral complexes, and translocation through the soil profile [e.g., Driscoll 1989]. Among them, the Al hydroxides can work as one of the major components of the active ANC of soils through direct dissolution as well as indirectly through anion adsorption [Davis and Hem 1989; Funakawa et al. 1993; Parfitt 1978]. Funakawa et al. [1993] described the soil profiles from nonvolcanic origin in the cool temperate zone forests of northern Kyoto as being a sequence of different stages of soil acidification with lower horizons representing earlier stages. The first stage is characterized by the accumulation of amorphous Al compounds as a result of Al illuviation (weak podzolization), and is typically observed in the B horizons; successive monomerization of these amorphous Al hydroxides due to an increasing acid load forms the second stage; and a clear decrease in amorphous Al content, weakened structure, and consequent eluviation of Fe under reducing conditions composes the third stage. On the other hand, it is known that the sesquioxidic properties of the B-horizon soils are variable according to the soil temperature regimes [Hirai et al. 1991]. It is worthwhile to compare corresponding pedogenetic processes among soils under different bioclimatic conditions. In this section, we compared the pedogenetic
186
World Soil Resources and Food Security
acidification processes in forest soils under cool and warm temperate climatic conditions in southwestern Japan (focusing on soil solution composition dynamics and soil acidity/alkalinity of soils) to further understand the long-term processes of SOM accumulation and soil acidification.
5.3.1 Study Soils A total of 20 sites were selected from the eastern slope of Mt. Odaigahara (alti tude 1695 m) in southern Mie Prefecture and Ashiu Experimental Forest of Kyoto University in northern Kyoto Prefecture, Kinki District, Japan (Figure 5.1; Table 5.6). Elevation ranged from 110 to 1600 m and both thermic and mesic soil temperature regimes (TSTR and MSTR, respectively) were included. All the soils were located on upper convex slopes and the influence of volcanic ejecta could be small. Using these soils, acid- and alkali-titration data were obtained by adding 0.10 mol L−1 HCl or NaOH to the soil suspension using a potentiometric automatic titrator (soil to solution ratio of 1:10; 0.10 mol L−1 NaCl as the supporting electrolyte). Titratable acidity was determined as OH− consumption at pH 8.3, and titratable alkalinity as H+ consumption at pH 3.0 [Funakawa et al. 1993; Kinniburgh 1986]. In addition to general physicochemical properties and the titration data of the soils, in situ soil solution composition—including Al speciation into inorganic monomeric, organic monomeric, and strongly complexed species—was determined after collecting soil solution using a porous cup method. Detailed data were presented in Funakawa et al. [2008].
5.3.2 Physicochemical Properties and Titratable Alkalinity and Acidity of the Soils The general physicochemical and mineralogical characteristics of the soils studied are summarized as follows (Table 5.7):
1. The pH(H2O) of the soils, especially in the surface horizons, was very low, normally less than 5. The decrease in soil pH from the C horizon toward the surface reflects pedogenetic soil acidification processes. Among exchangeable cations, Al usually dominated and the base saturation of the soils was below 5% in most cases, suggesting that acid neutralization through cation exchange between H+ supplied and exchangeable bases of the soils was limited under the present situation. 2. Most of the soils showed a rather fine texture, the clay contents usually exceeding 40%. The dominant clay mineral species were mostly hydroxy-Al interlayered vermiculite (HIV) and vermiculite occasionally appeared in the highly acidified—pH(H2O) below 4.2—upper horizons. The Sio content of nearly zero suggests that amorphous Al–Si compounds were virtually absent, unlike Al hydroxides and/or organo-mineral complexes. 3. B horizon soils in MSTR contained higher amounts of amorphous compounds (Alo + 1/2Feo) and organic matter (Cp) than those in TSTR (Figure 5.4), consistent with earlier results reported by Hirai et al. [1991].
187
Pedogenetic Acidification in Humid Asia
TABLE 5.6 Brief Description of Selected Soils Studied in Kinki District, Japan USDAa
MD1
Typic Dystrudepts
MD2
Typic Dystrudepts
KD1
Lithic Dystrudepts
KD2
Humic Dystrudepts
TT
Andic Dystrudepts
OT
Alic Hapludands
H2
Andic Dystrudepts
S3
Alic Hapludands
760 m
mesic
Cryptomeria japonica, Fagus crenata, Quercus mongolica var. crispula, Benzoin umbellatum, Sasa palmata
N4b
Andic Haplorthods
790 m
mesic
Fagus crenata, Quercus mongolica var. crispula, Sasa palmata
H1b
Andic Haplohumods
800 m
mesic
Magnolia salicifolia, Hydrangea paniculata, Quercus mongolica var. crispula, Fagus crenata, Cryptomeria japonica, Daphniphyllum humile
a
Altitude
Soil Temperature Regime
Site
Major Vegetation
Mie Soils (Mt. Ohdaigahara) 110 m thermic Cleyera japonica, Neolitsea aciculata, Quercus glauca, Camellia japonica, Meliosma rigida 180 m thermic Quercus glauca, Eurya japonica, Meliosma rigida, Castanopsis cuspidata, Myrsine seguinii, Camellia japonica 840 m thermic/mesic Camellia japonica, Illicium anisatum, Hydrangea paniculata, Leucosceptrum stellipilum, Quercus mongolica var. crispula, Acer spp., Sasamorpha borealis var. borealis 830 m thermic/mesic Camellia japonica, Illicium anisatum, Hydrangea paniculata, Quercus mongolica var. crispula, Acer spp., Sasamorpha borealis var. borealis 1210 m mesic Schizophragma hydrangeoides, Pourthiaea villosa, Fagus japonica, Stuartia monadelpha, Sasamorpha borealis var. borealis 1220 m mesic Enkianthus cernuus, Quercus mongolica var. crispula, Betula grossa, Clethra barvinervis, Stuartia monadelpha, Acer spp. Kyoto Soils (Ashiu Experimental Forest) 880 m mesic Cryptomeria japonica, Fagus crenata, Acer sieboidianum, Sasa palmate
According to Soil Taxonomy [Soil Survey Staff 2006].
188
TABLE 5.7 Physicochemical Properties of Soils in Japanese Plots pH
Particle Size Distribution
Exchangeable
Depth
Total C
Bases
Al
CEC
Sand
(cmolc kg )
Silt
Clay
Feo
Sio
Fep
Cp
Horizon
(cm)
(H2O)
(KCl)
(g kg )
O A BA Bw BC
2–0 0–8 8–23 23–40 40–50+
3.9 4.5 4.6 4.6 4.7
3.2 3.9 4.0 3.9 4.0
312.7 67.7 21.8 13.4 7.4
3.4 0.7 0.3 0.3 0.2
14.6 9.5 7.2 8.3 7.6
77.8 25.4 23.7 19.0 16.0
6 11 11 13 40
13 24 23 24 18
82 65 66 63 42
4.8 8.3 6.5 4.2 3.3
6.3 7.8 5.2 4.1 3.1
0.1 0.1 0.1 0.1 0.1
6.3 13.4 8.5 3.7 1.3
6.3 7.8 5.2 4.1 3.1
43.4 17.4 6.3 4.2 2.1
MD2
A BA Bw1 Bw2 BC
0–8 8–20 20–45 45–62 62–90+
4.7 4.6 4.7 4.7 4.8
4.1 3.9 3.9 3.9 3.9
121.3 36.8 15.2 8.7 6.5
1.0 0.5 0.3 0.3 0.3
7.5 10.9 9.2 10.9 11.1
34.5 25.3 22.5 22.2 32.9
9 8 8 9 11
27 27 24 28 25
64 65 69 63 64
6.9 9.4 6.9 5.2 4.0
11.3 6.7 5.1 4.5 3.9
0.1 0.0 0.0 0.0 0.0
9.7 12.5 7.4 6.3 2.3
11.3 6.7 5.1 4.5 3.9
33.6 10.7 2.2 3.3 2.3
KD1
A BA Bw BC
0–3 3–12 12–32 32–44
4.6 4.7 4.8 4.9
3.4 3.7 3.7 4.0
124.4 44.6 37.0 25.8
2.4 0.6 0.4 0.3
16.6 22.5 17.6 9.6
34.9 34.6 31.6 17.9
56 25 20 47
18 27 24 19
26 48 56 35
7.0 11.6 16.9 8.7
5.7 9.2 9.4 6.3
0.1 0.0 0.0 0.1
3.8 5.0 7.2 3.6
5.7 9.2 9.4 6.3
28.2 17.0 15.6 6.0
KD2
A(o) AB Bw1
0–4 4–16 16–29
4.0 4.2 4.6
3.2 3.4 3.7
208.3 94.5 26.6
1.4 0.5 0.2
17.4 18.5 9.3
55.9 38.9 23.6
23 24 29
26 28 28
51 47 42
8.9 12.1 20.2
7.2 8.4 7.9
0.0 0.0 0.0
6.9 6.0 4.3
7.2 8.4 7.9
42.1 29.2 11.5
−1
(g kg )
Alp
MD1
−1
(%)
Alo
Site
a
(g kg )
−1
−1
World Soil Resources and Food Security
29–45 45–60+
4.7 4.7
3.8 3.9
13.0 9.6
0.2 0.2
7.4 6.1
22.5 12.7
33 46
25 22
41 33
8.8 5.8
5.6 4.0
0.0 0.0
3.3 3.4
5.6 4.0
6.3 3.8
TT
A(o) BA Bw1 Bw2 BC
0–5 5–20 20–40 40–60 60–70
3.6 4.1 4.6 4.7 4.7
3.1 3.5 3.9 4.0 4.0
265.0 52.2 26.5 19.3 8.1
2.4 0.6 0.3 0.2 0.2
17.7 18.0 7.6 4.7 5.4
54.2 29.1 18.5 12.9 10.9
24 28 26 22 35
27 24 26 26 22
49 48 48 52 43
5.5 12.7 12.6 10.2 5.5
4.8 6.6 5.8 6.0 3.9
0.0 0.0 0.0 0.1 0.1
4.3 8.5 7.4 4.3 1.6
4.8 6.6 5.8 6.0 3.9
42.6 19.9 11.3 7.5 3.2
OT
A(o) BA Bw1 Bw2 BC C
0–8 8–24 24–38 38–52 52–62 62–70+
3.8 4.1 4.5 4.7 4.8 4.8
3.2 3.6 3.9 4.1 4.2 4.3
223.7 76.9 61.3 55.9 32.4 16.3
1.7 0.6 0.3 0.2 0.1 0.1
15.7 23.9 10.6 6.8 4.6 2.8
29.7 40.4 26.9 21.9 17.6 12.7
10 14 20 23 34 43
17 26 32 37 19 16
73 60 48 40 47 41
6.4 21.1 19.6 20.8 16.4 7.9
4.6 9.4 9.0 10.8 11.3 9.6
0.3 0.0 0.0 0.1 0.2 0.6
5.0 16.2 13.5 10.4 8.7 2.5
4.6 9.4 9.0 10.8 11.3 9.6
43.9 32.3 25.2 20.3 16.0 8.2
H2
A AB BA Bw BC
0–6 6–14 14–28 28–48 48–60
4.0 4.2 4.4 4.6 4.8
3.3 3.5 3.7 3.9 3.9
115 75 49 18 13
3.0 1.2 0.7 0.5 0.5
12.2 12.1 9.2 6.4 5.3
37.3 31.5 23.5 15.8 13.4
16 12 11 15 20
31 33 35 33 33
53 55 54 52 46
16.6 18.2 19.5 11.5 6.7
7.9 8.3 9.0 7.7 6.6
0.2 0.4 0.3 0.6 0.7
11.2 13.7 18.3 10.0 6.4
5.7 6.8 7.9 7.1 5.7
33.7 31.1 19.9 10.3 6.7
S3
A1(o) A2 AB BA Bw BC
0–8 8–21 21–34 34–48 48–70 70–95
3.5 4.3 4.6 4.7 4.7 4.7
2.9 3.6 3.9 4.0 4.0 4.0
226 88 61 46 17 11
4.0 0.8 0.5 0.4 0.9 0.5
13.4 10.8 6.4 4.5 4.5 4.3
63.7 38.6 30.0 26.9 20.1 16.0
6 9 11 8 9 12
40 39 46 43 42 38
54 52 43 48 49 51
11.8 20.6 20.8 16.8 10.1 8.8
6.9 11.4 15.1 16.8 11.4 9.8
0.2 0.2 0.8 1.4 1.1 1.5
10.7 22.0 15.6 12.5 4.9 5.4
189
6.1 64.5 11.6 40.4 13.8 31.4 13.8 23.9 7.2 10.6 5.9 6.0 (continued)
Pedogenetic Acidification in Humid Asia
Bw2 BC
190
TABLE 5.7 (Continued) Physicochemical Properties of Soils in Indonesian Plots pH
Particle Size Distribution
Exchangeable
Depth
Total C
Bases
Al
CEC
Sand
Clay
Feo
Sio
Fep
Cp
Horizon
(cm)
(H2O)
(KCl)
(g kg )
0–6 6–12 12–15 15–30 30–40 40–52 52–83 83–100+
3.6 3.8 4.1 4.2 4.5 4.6 4.6 4.7
2.9 2.9 3.2 3.5 3.8 4.0 4.0 4.1
213 102 58 53 21 15 15 13
3.3 1.3 0.7 0.6 0.4 0.4 0.4 0.3
8.7 14.6 14.6 12.4 6.9 6.7 5.8 4.7
62.2 40.2 34.4 30.8 20.6 19.8 19.5 19.4
11 11 12 9 15 8 7 16
47 46 44 36 37 46 41 37
43 43 44 55 48 46 52 47
5.7 14.4 21.6 34.8 20.5 10.8 9.9 11.0
3.9 5.1 6.1 10.6 10.1 9.8 10.1 11.2
0.2 0.6 0.3 0.4 0.6 1.0 0.6 1.5
6.0 11.8 14.4 14.2 8.4 5.8 5.1 4.2
2.4 3.7 4.0 7.2 5.9 5.5 5.9 5.7
43.5 37.7 26.5 30.6 11.0 10.4 10.1 7.5
H1b
A(o) E Bh Bw BC
0–9 9–19 19–22 22–36 36–49
3.7 3.8 4.0 4.3 4.4
2.9 3.0 3.3 3.6 3.7
281 29 54 42 33
3.8 0.7 0.5 0.6 0.5
6.5 7.9 16.7 8.9 7.5
71.2 17.8 33.7 27.5 19.0
20 27 31 33 37
40 40 29 27 27
40 33 40 40 36
4.4 1.4 22.0 16.0 7.9
3.7 1.9 6.7 8.8 7.2
0.2 0.3 0.2 0.5 0.3
3.9 1.2 12.6 8.9 7.0
3.6 1.6 5.4 6.6 6.6
63.6 8.7 35.1 27.1 21.2
−1
(g kg )
Alp
A(o) AB BA Bs Bw1 Bw2 BC C
−1
(%)
Alo
N4b
(g kg )
−1
−1
A(o) horizon described here contains >200 g C kg−1 soil and is often regarded as a deposited organic layer in some soil classification systems.
World Soil Resources and Food Security
a
(cmolc kg )
Silt
Site
a
191
Pedogenetic Acidification in Humid Asia Thermic Intermediate Mesic Mesic, podzolic
30
Cp (g kg–1)
25 20 15 10 5 0
0
5
10
15
20
25
Alo + 1/2Feo (g kg–1)
30
FIGURE 5.4 Contents of (Alo + 1/2Feo) and Cp in the B-horizon soils (at around 30 cmlayers of soils).
MD1
8
BC Bw2
BA
pH 8.3
7
pH
pH
Examples of acid and alkali titration curves of the selected soils are presented in Figure 5.5. In general, the titratable alkalinity required to acidify the soils to pH 3.0 was higher in the B or C horizons than in the surface horizons for soils in the MSTR (OT), whereas it was at a maximum at or near the surface for soils in the TSTR (MD1). The titratable acidity required to neutralize soils to pH 8.3 increased from the C horizon toward the surface, presumably because of the increasing influence of organic matter dissociation. As this process may mask other important relationships between titration data and physicochemical properties, we refer only to soils with a total carbon content of less than 100 g kg−1 in the following discussion on the titration data. Table 5.8 shows correlation coefficients between the titration data and the selected chemical properties of the soil (n = 78). The titratable alkalinity required to acidify soils to pH 3.0 is highly correlated with some extractable Al or Fe compounds (Alo [r = 0.88**], Ald [r = 0.89**], Alp [r = 0.76**], and Fep [r = 0.76**]) and also with pH(NaF) (r = 0.78**). These correlations indicate the importance of amorphous
OT
C
8
Bw2
BA pH 8.3
7
6
pH 5.5
5
pH 4.5
4 3 pH 3.0 10 5 0 Addition of 0.1 M HCl (mL)
5
10 15 20 Addition of 0.1 M NaOH (mL)
6
pH 5.5
5
pH 4.5
4 3 pH 3.0 10 5 0 Addition of 0.1 M HCl (mL)
5
FIGURE 5.5 Examples of acid and alkali titration curves of the soils.
10 15 20 Addition of 0.1 M NaOH (mL)
192
TABLE 5.8 Correlation Coefficients among the Physicochemical Properties and Titration Data of the Soils (n = 78) pH(H2O) Initial pH Titratable alkalinity by pH 3.0 Titratable acidity by pH 8.3 OH−consumption in the range of pH 4.5 to 5.5 OH− consumption in the range of pH 5.5 to 8.3
Exch. Bases
Exch. Al
ECEC
CEC
Clay
0.94**
0.88**
–0.68**
–0.72**
–0.78**
–0.66**
0.09
0.37 –0.63** –0.67**
0.54** –0.52** –0.70**
0.78** –0.35 –0.68**
–0.18 0.54** 0.50**
–0.43* 0.63** 0.91**
–0.45* 0.67** 0.94**
–0.07 0.82** 0.81**
0.23 0.14 0.20
0.12 0.82** 0.51**
–0.45**
−0.28
−0.07
0.44*
0.34
0.38
0.68**
0.11
0.84**
Cp
Alo
Feo
Ald
Fed
Alp
Fep
Alo + 1/2Feo
Alp + 1/2Fep
−0.27 0.28 0.83** 0.40
0.45** 0.88** 0.26 −0.20
0.02 0.43* 0.65** 0.23
0.56** 0.89** 0.13 −0.29
−0.11 0.33 0.55** 0.33
0.26 0.76** 0.46** –0.03
−0.17 0.31 0.72** 0.36
0.26 0.73** 0.50** 0.01
0.05 0.58** 0.63** 0.18
0.93**
0.50**
0.80**
0.39
0.62**
0.69**
0.82**
0.72**
0.81**
Note: Significant at *5% and **1% level.
pH(KCl)
Total C –0.40
World Soil Resources and Food Security
Initial pH Titratable alkalinity by pH 3.0 Titratable acidity by pH 8.3 OH– consumption in the range of pH 4.5 to 5.5 OH– consumption in the range of pH 5.5 to 8.3
pH(NaF)
0.84**
Pedogenetic Acidification in Humid Asia
193
Al, including organo-mineral complexes and, probably, Fe compounds as an active acid buffering component in soils with low base saturation through protonation of oxides and/or partial monomerization of Al as also reported for acidic forest soils by Kamoshita et al. [1979] and Funakawa et al. [1993]. Between Alo (cmol kg−1) and titratable alkalinity to pH 3.0 (Alk3 in cmol kg−1), the following relationship was obtained:
Alk3 = 0.29Alo + 0.57 (r 2 = 0.76).
(5.1)
The value of the slope, 0.29, indicates that only a small portion of the Alo fraction can contribute to the acid neutralization reaction. Thus, the titratable alkalinity derived from amorphous Al oxide was higher in soils under MSTR than in soils under TSTR. Some titration curves display a conspicuous pH buffer zone in the pH range 4.5 to 5.5 (Figure 5.5). Based on the fact that OH− consumption in this zone (OH45 in cmol kg−1) is highly correlated with exchangeable Al content (exAl in cmolc kg−1) (Table 5.8; r = 0.91**), the buffering effect can be attributed to the hydrolytic reactions of exchangeable Al:
OH45 = 0.49exAl + 0.61 (r 2 = 0.83).
(5.2)
This regression has a slope of 0.49, indicating that the amount of OH− consumed in the pH range 4.5 to 5.5 is equivalent to half the exchangeable Al content. One reason for this discrepancy may derive from the incomplete hydrolysis of the Al3+ ion under relatively high rates of OH− addition. No clear relationship was observed between the extent of this buffering and the soil temperature regime of the samples. The OH− consumption in the pH range 5.5 to 8.3 is correlated with total carbon content, Cp and Alp + 1/2Fep (Table 5.8; r = 0.84**, 0.93** and 0.81**, respectively), indicating that the acidity in this pH range originates from organo-mineral complexes via the dissociation of the acidic functional groups of soil humus and/or the deprotonation of oxide surfaces, as follows:
R–COOH + OH− = R–COO − + H2O,
(5.3)
M+– −OOC–R + OH– = M–OH + −OOC–R,
(5.4)
M–OH + OH− = M–O − + H2O,
(5.5)
where M− represents metal ions (such as Fe or Al) at the surface of soil particles and R–COOH represents the acidic functional groups of organic matter. Following regression equation was obtained between Cp content and OH− consumption in the pH range 5.5 to 8.3 (OH58):
OH58 = 0.86Cp + 6.95 (r 2 = 0.87).
(5.6)
194
TABLE 5.9 Concentrations of Ions in Soil Solution from Forest Soils in Kinki District, Japan Alb DOC
Orgn− a NO3– Cl–+SO42−
Horizon pH (mg C L−1)
MD2
KD1
KD2
TT
(mmolc L−1)
Na++K++Mg2++Ca2+
Fe3+
Aln+
(mmolc L−1)
Si
Inorganic Monometric
(mmol L−1)
Organic Monomeric
StronglyComplexed
(×10−3 mmol L−1)
15
5.2
13.5
0.102
0.000
0.176
0.007
Mie Soils 0.215
0.001 0.054
0.117
4.6
(26)
10.3
(57)
3.0
(17)
40
5.2
6.9
0.061
0.000
0.150
0.006
0.165
0.001 0.038
0.097
6.3
(49)
5.9
(46)
0.6
(5)
15
5.5
3.2
0.029
0.000
0.098
0.003
0.112
0.000 0.012
0.043
1.9
(48)
1.9
(48)
0.2
(4)
40
5.5
2.2
0.028
0.000
0.100
0.003
0.115
0.000 0.009
0.049
2.4
(79)
1.2
(39)
0.0
(0)
15
5.1
3.1
0.023
0.000
0.177
0.008
0.158
0.000 0.032
0.074
8.5
(80)
1.6
(15)
0.6
(5)
40
5.0
1.3
0.016
0.000
0.190
0.009
0.164
0.000 0.033
0.054
8.5
(78)
0.9
(8)
1.5
(13)
15
5.0
4.4
0.048
0.003
0.105
0.010
0.109
0.000 0.036
0.060
5.6
(46)
5.3
(44)
1.2
(10)
40
5.2
1.2
0.021
0.004
0.101
0.006
0.105
0.000 0.013
0.035
3.3
(74)
1.0
(22)
0.1
(3)
15
4.7
4.0
0.042
0.124
0.096
0.018
0.173
0.000 0.069
0.037
16.1
(70)
4.8
(21)
2.0
(9)
40
4.9
2.3
0.023
0.144
0.061
0.013
0.167
0.000 0.046
0.033
11.9
(77)
2.0
(13)
1.5
(10)
World Soil Resources and Food Security
MD1
H+
H2
H1b
S3
N4b
a b
15
4.9
3.7
0.033
0.012
0.081
0.012
0.073
0.001 0.040
0.085
7.0
(53)
4.2
(32)
2.0
(15)
40
5.2
1.1
0.018
0.000
0.059
0.006
0.060
0.001 0.011
0.031
2.5
(67)
1.1
(29)
0.1
(4)
10
4.5
3.0
0.231
0.192
0.435
0.035
Kyoto Soils 0.488
0.001 0.135
0.041
41.0
(91)
2.5
(6)
1.5
(3)
35
4.7
1.3
0.228
0.174
0.265
0.018
0.426
0.000 0.088
0.042
26.4
(90)
0.6
(2)
2.4
(8)
55
5.4
2.5
0.172
0.049
0.190
0.004
0.300
0.000 0.021
0.039
5.1
(73)
0.2
(3)
1.7
(24)
10
3.9
23.7
0.177
0.315
0.306
0.125
0.302
0.009 0.189
0.085
42.7
(68)
12.9
(21)
7.3
(12)
20
4.6
20.6
0.223
0.060
0.275
0.028
0.342
0.012 0.074
0.068
13.7
(55)
7.7
(31)
3.4
(14)
40
4.7
10.8
0.145
0.102
0.252
0.022
0.299
0.002 0.083
0.050
20.5
(74)
4.6
(17)
2.7
(10)
15
4.5
8.7
0.229
0.317
0.366
0.030
0.509
0.001 0.183
0.065
51.5
(84)
4.5
(7)
5.1
(8)
40
4.8
4.1
0.244
0.343
0.307
0.015
0.619
0.000 0.092
0.047
25.2
(82)
1.0
(3)
4.7
(15)
60
5.1
6.1
0.270
0.229
0.371
0.007
0.644
0.000 0.037
0.058
7.8
(63)
1.3
(10)
3.4
(27)
10
4.5
6.8
0.150
0.070
0.219
0.033
0.220
0.003 0.090
0.023
22.8
(76)
5.0
(17)
2.2
(7)
20
4.7
1.8
0.150
0.084
0.192
0.019
0.239
0.003 0.067
0.032
18.5
(82)
1.8
(8)
2.3
(10)
45
5.1
2.7
0.139
0.018
0.192
0.008
0.247
0.000 0.024
0.031
7.2
(89)
0.5
(6)
0.4
(5)
Pedogenetic Acidification in Humid Asia
OT
Orgn– represents anion deficit, the negative charge of organic acids. Parentheses denotes the percentage of total Al.
195
196
World Soil Resources and Food Security
As the carboxylic functional group content of fulvic acid in cool temperate zone soils is reported to be 0.61–0.85 cmol g−1 organic matter [Stevenson 1982], the slope obtained in the regression, 0.86 cmol g−1 C (equivalent to 0.51 cmol g−1 organic mat ter), seemed to be reasonable if most of the carboxylic groups of the Cp fraction react with added OH− as in the first and second reactions shown above. In early research relating to titration, the buffer region between pH 5.5 and 8.3 was often attributed to hydroxy-Al interlayer components [Schwertmann and Jackson 1963, 1964]. The third reaction above may also involve edges of layer silicates. Titratable acidity in this pH range was generally higher in MSTR rather than TSTR B horizon soils.
5.3.3 Soil Solution Composition Table 5.9 summarizes the soil solution composition for each soil profile. The concentration of DOC was generally lower than 1 mmol L−1, except for MD1 and H1b, in which it sometimes exceeded 2 mmol L−1. In most of the soils, the concentration of Al decreased in the subsoils with a concomitant increase in pH and decrease in DOC (Table 5.9), suggesting precipitation of Al through hydrolysis and/or coprecipitation of Al–organic matter complexes in the subsoils. The organic monomeric and strongly complexed Al fractions are relatively small (less than 50% of total dissolved Al except for MD1), and the inorganic monomeric Al is the major fraction. The concentration of inorganic monomeric Al drastically decreases as the pH approaches 5. In contrast, the concentration of organic monomeric Al generally has a positive correlation with DOC concentration in each of soil profiles, suggesting that DOC supplied from the upper soil layers and/or litter layers is primarily responsible for mobilizing Al in the form of organo-mineral complexes. Although there were significant concentrations of Al in the soil solution, Fe was rarely present except for Spodosols (N4b and H1b), indicating that translocation of Fe did not occur in these soil profiles.
5.3.4 E luvi-Illuviation of Inorganic Al and Subsequent Adsorption of DOC Figure 5.6 summarizes the soil solution composition and characteristics of the soil solid phase in each soil profile with the reaction of inorganic Al in the left figure and that of organic Al and DOC in the right figure. Three patterns of acid neutralization, that is, surface neutralization, subsurface neutralization, and limited neutralization, were observed judging from titratable alkalinity and the concentration of inorganic monomeric Al and the pH of the soil solution. In the surface layers of the MD1 and MD2 profiles, the solution pH was higher than 5, inorganic monomeric Al below 5 × 10 −3 mmol L−1, and titratable alkalinity higher than 10 cmol kg−1. Most of the acidity supplied by the soil surface, if any, is probably neutralized in these layers (surface neutralization). In the surface horizons of the other profiles, the titratable alkalinity of the surface soils is lower than 10 cmol kg−1, solution pH is 5 or below, and larger concentrations of inorganic monomeric Al (>5 × 10 −3 mmol L−1) are dissolved. In the KD2, OT, H2, S3, and N4b profiles the concentrations of inorganic monomeric Al in the soil solution decreased at or below the AB or Bw horizon,
197
Pedogenetic Acidification in Humid Asia
(a)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
A
20 40 Depth (cm)
5
20
MD1 A
60
6
0
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Alo ( 10
b A
60
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Solution pH ( )
A
40
Typic dystrudepts Depth (cm)
a
(b)
) (g kg–1) 20
a
a
4
Alo ( 10
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
20
A
KD2
40
5
6
Solution pH ( )
(c)
Depth (cm)
4
60
0
40 b
Humic dystrudepts
B
Depth (cm)
b
60
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Alo ( 10
A
20
B
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
40
A
MD2 Typic dystrudepts
60 Depth (cm)
4 5 6 Solution pH ( )
0
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
40 Depth (cm)
a
20
20 A
A a A
60
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
FIGURE 5.6 Chemical properties and titration data of the soil profiles superimposed with the soil solution composition. Error bars are standard error. Values with the same letter are not significantly different (p < 0.05). Alo, Feo, extraction in the dark with acid (pH 3) 0.2 mol L−1 ammonium oxalate; Alp, Fep, extraction with 0.1 mol L−1 pyrophosphate at pH 10; DOC, dissolved organic carbon.
198
World Soil Resources and Food Security
(d)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
Alo ( 10
a a
20
TT
40
A
Andic dystrudepts
60
5
6
Solution pH ( )
(e)
Depth (cm)
4
0
Alo ( 10
B
60
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
A a
40 Depth (cm)
a
20
A
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
20 A 40
5
6
0
Alo ( 10
A
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
A
40 b 60
60
Solution pH ( )
(f )
KD1 Lithic dystrudepts
Depth (cm)
4
A
Depth (cm)
a
20
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
20
A
40
5
60
6
Solution pH ( )
0
60
1 2 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
FIGURE 5.6 (Continued)
A
40 b Depth (cm)
4
Alic hapludands
B
Depth (cm)
b
OT
20
B
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
199
Pedogenetic Acidification in Humid Asia
(g)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Alo ( 10
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
20 a
5
6
Solution pH ( )
60
4
5
6
0
2 4 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Solution pH ( )
(i)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Alo ( 10
a B
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a
a
20
A
40
5
60
6
Solution pH ( )
S3
Alic hapludands
40 60
B 0
20
Depth (cm)
c
AB
Depth (cm)
b
4
60
A
B
40 Depth (cm)
Andic haplorthods
B
a
a
20
N4b
40
B
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
AB
a Ba
0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
A
Depth (cm)
c
60
Alo ( 10
20
b
40
2 4 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Titratable alkalinity ( ) (cmol kg–1) 20 10 0 a
Andic dystrudepts
B
0
A
20
H2
Depth (cm)
60 Depth (cm)
4
AB
40 b
(h)
A
2 4 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
FIGURE 5.6 (Continued)
A a B
a
B 0 0.5 1 DOC ( ) (mmol L–1) or organic monomeric Al ( ) (*0.01 mmoL L–1) in soil solution
200
World Soil Resources and Food Security
(j) Titratable alkalinity ( ) (cmol kg–1) 20 10 0
Alo ( 10
Titratable acidity initial pH–pH 5.5 ( ) Alo + 1/2Feo ( ) pH 5.5–8.3 ( ) or Alp + 1/2Fep ( ) (g kg–1) (cmol kg–1) 40 20 0 10 20
) (g kg–1) 20
a a
A
20
b
40
20
Andic haplohumods
40
60
4
5
6
Solution pH ( )
0
2 4 Inorganic monomeric Al in soil solution ( ) (*0.01 mmoL L–1)
Depth (cm)
B
Depth (cm)
b
H1b
B
A AB
a
a
B
60
0 0.5 1 1.5 2 2.5 DOC ( ) (mmol L–1) or organic –1 monomeric Al ( ) (*0.01 mmoL L ) in soil solution
FIGURE 5.6 (Continued)
where the titratable alkalinity increases to over 10 cmol kg−1, along with a pH rise to approximately 5.2 (subsurface neutralization). In contrast, the titratable alkalinity of the KD1, TT, and H1b soils never exceeded 10 cmol kg−1 at any horizons, solution pH remains at or below 5, and inorganic monomeric Al concentrations are high even in the B horizons (limited neutralization). These results show that a large part of the soil solution acidity can be removed, and inorganic monomeric Al in solution precipitated through hydrolysis by soil layers with high titratable alkalinity, that is >10 cmol kg−1 in the condition of our study. Although such an acidity-neutralizing layer was found at or near the soil surface under TSTR (MD1 and MD2), it was occasionally formed in the subsoil under MSTR, at which soil solution acidity is often neutralized, resulting in accumulation of solution phase Al through hydrolysis and precipitation. Dissolved organic matter in the soil solution, still possessing a certain concentration of dissociated acidic functional groups, can be adsorbed on these amorphous compounds of Al. Thus, as organo-mineral complexes accumulated in B horizons are considered to also comprise the active fraction of titratable acidity, the process described here also contributes to the increase in titratable acidity in the B horizon soils.
5.3.5 Comigration of Al and DOC in the Soil Profiles The concentration of organic monomeric Al in the soil solution substantially decreases with depth in KD2, TT, OT, and all of Kyoto soils (Table 5.9; Figure 5.6). As DOC also decreases to some extent with depth, precipitation of complexes of Al and dissolved organic matter complexes may have occurred in the subhorizon, as discussed by DeConinck [1980] for the formation of spodic horizons. As fairly large amounts of amorphous compounds (Alo + 1/2Feo) already exist in the subhorizons, adsorption of Al-dissolved organic matter complexes onto amorphous compounds may also be possible. This process may be involved in soil formation under MSTR. It seems, to some extent, similar to podzol formation, at least in terms of Al translocation, and enhances the accumulation of titratable acidity in subsoils.
201
Pedogenetic Acidification in Humid Asia
5.3.6 Dynamics of Titratable Alkalinity and Acidity in the Pedogenetic Processes Figure 5.7 summarizes the amounts of titratable acidity, titratable alkalinity, and total C in the soil profiles on a kmol ha−1 basis using all the data from the present study (20 profiles) and data from previous studies (4 profiles from northern Kyoto:
Titratable alkalinity in the 15–45 cm layers of soil (kmol ha–1)
(b)
1000 Thermic Intermediate Mesic Mesic, podzolic
800 600 400 200 0
50 40 30 20 Cp in the 15 10 of soil –45 cm lay s (Mg – ers ha 1)
0
0
Ex ch lay Al in ers th (km e 1 ol 5–4 c h a –1 5 cm )
Titratable acidity in the 15–45 cm layers of soil (kmol ha–1)
(a)
500 400 300 200 100
Thermic Intermediate Mesic Mesic, podzolic
500 400 300 200 100 0
0
50
100
150
Titratable alkalinity in the 0–15 cm layers of soil (kmol ha–1) (c) Total C in the 15–45 cm layers of soil (Mg ha–1)
150
100
Thermic Intermediate Mesic Mesic, podzolic
50
0
0
50
100
150
Total C in the 0–15 cm layers of soil (Mg ha–1)
FIGURE 5.7 Amounts of (a) titratable acidity, (b) total C, and (c) titratable alkalinity on a kmol ha−1 basis. Cp, pyrophosphate extractable C; Exch Al, exchangeable Al. (Data presented here include data cited from Funakawa, S. et al., Soil Sci. Plant Nutr. 49, 387–396, 2003; Funakawa, S., et al., Soil Sci. Plant Nutr., 39, 677–690, 1993.)
202
World Soil Resources and Food Security
ON1 and ON2 in TSTR, and KT1 (podzolic soil) and KT2 in MSTR) [Funakawa et al. 2003]. As discussed previously, pedogenetic acidification in forest soils with MSTR is characterized by an accumulation of acidity in the form of amorphous compounds and/or organo-mineral complexes in the B horizons, originally supplied from the overlying horizon. Figure 5.7a shows that the titratable acidity in subsoil layers (15–45 cm) amounts to 292–816 kmol ha−1 and is higher among soils in MSTR than in TSTR because of a higher contribution of organo-mineral complexes (represented by Cp). As the in situ annual acid load on acid soils (i.e., low and intermediate rates of soil acidification) are reported to mostly range from 1 to 7 kmol ha−1 yr−1 [van Breemen et al. 1984], the amounts of titratable acidity observed in the present study are equivalent to hundreds of times the annual acid load. The cumulative acid load that accompanies DOC adsorption and/or precipitation enhances the accumulation of total C in the subsoil layers (Figure 5.7b). As the DOC flux that can accumulate between the 15-cm and 40- to 45-cm layers of soils is calculated to be 0.10 (MD2) to 1.29 (H1b) Mg ha−1 yr−1 (average: 0.40) based on the soil solution composition in the present study assuming the annual amount of percolating water to be 1,000 mm, the amounts of organic C accumulated in the subsoil layers (i.e., 25–112 Mg ha−1; Figure 5.7c) also corresponding to hundreds of times the annual DOC flux. Thus, the pedogenetic processes, including acid transfer and C migration to the subsoil layers are considered to be prolonged at least for hundreds of years. According to Figure 5.7b, the amounts of titratable alkalinity in the 15- to 45-cm layers of soils tend to be higher compared with those in the surface 15-cm layers of soils in MSTR. As amorphous Al hydroxides (Alo) compose major parts of the titratable alkalinity, they play an important role in retarding further acidification through protonation and/or partial monomerization. In this context, amorphous Al hydroxides, which would otherwise be leached out directly from the soil profile, can be regarded as making a temporary contribution to the acid-neutralizing capacity of the soil. This process delays the outflow of acids (H+ or Al3+) from soils. Both stepwise processes of precipitation of inorganic Al and a subsequent adsorption of DOC, and comigration and precipitation of Al–soluble organic matter complexes are involved in the accumulation processes of organo-mineral complexes in the B horizons. In contrast, for soils under TSTR, the acid-buffering reactions, if any, occur mostly at or near the soil surface in the present study. One possible explanation for such an apparent deep penetration of soil acidification in MSTR is that, under MSTR, the amount of percolating water is higher than under TSTR because the precipitation is higher and evapotranspiration is lower at high elevations in the region, and that such a difference in the degree of leaching possibly brings more intensive soil acidification under MSTR. The fact that the titratable alkalinity required to acidify soils to pH 3.0 and the titratable acidity required to neutralize soils to pH 8.3 in B horizon soils is apparently lower in TSTR soils than MSTR soils reflects the lower accumulation of amorphous and/or organo-mineral complexes in the B horizons of the former.
Pedogenetic Acidification in Humid Asia
203
5.4 C HANGES IN THE BLOCKAGE EFFECT OF HYDROXY-AL POLYMERS ON THE FRAYED EDGE SITE OF ILLITIC MINERALS DURING THE PROCESS OF PEDOGENETIC ACIDIFICATION IN JAPAN As discussed in the preceding sections, the pedogenetic acidification process largely modifies soil properties relating to sesquioxidic properties and organic materials. Pedological research and mineralogical analyses of many soil profiles also revealed that HIV-dominated B horizons lie below eluvial, Al-, or Fe-leached horizons rich in vermiculite or smectite in Spodosols and the environmentally associated Inceptisols [Maes et al. 1999; Kitagawa 2005], strongly suggesting that the pedogenetic acidification process brings the sequential transformation from HIV to vermiculite or smectite as soils are podzolized. In the present section, in order to detect gradual transition of 2:1 minerals during the pedogenetic acidification, the RIP methodology introduced previously [see Chapter 4] is applied to discuss the fate of illitic minerals under podzolization. We first described the vertical distribution of HIV, vermiculite, and the frayed edge site in profiles of forest soils with different degrees of podzolization, and then explored the interactions between the frayed edge site and hydroxy-Al polymers by directly comparing the amount of the frayed edge site before and after extracting the hydroxy-Al polymers from soil clays using Tamura’s method [1958].
5.4.1 Soils Studied For this experiment, we collected samples from ten soil profiles in mountainous forest areas with an elevation of 240–1000 m in southwestern Japan (Figure 5.1). The soils were mostly derived from sedimentary rocks and overall chemical and mineralogical properties are similar to those analyzed in the previous sections. The soil moisture and temperature regimes are, respectively, similarly udic and mesic. Common soil types in this region are Udepts (in the USDA soil taxonomy). Spodosols and Andisols are also sparsely distributed in this region. Analytical items relating to chemical and mineralogical properties in this study were the RIP experiment as well as several soil characteristics including the acid-oxalate-extractable Fe (Feo) [Mckeague and Day 1966], dithionite-citrate-bicarbonate (DCB) extracted-Fe (Fed) [Mehra and Jackson 1960], and total Fe (Fet) of the soils (for <2-mm fractions) and hot sodium citrate extracted Al and Si (Alcit and Sicit, respectively) [Tamura 1958], total K content (TKclay), and Cs-fixed capacity of the clay fraction as indicators of interlayered materials of 2:1 minerals, clay mica content, and the vermiculitic layer, respectively [Komarneni and Roy 1980].
5.4.2 Soil Types and Clay Mineralogy We estimated the types of 2:1 phyllosilicates from the x-ray diffraction patterns for the soil clays. Vertical variations in the XRD patterns in a profile are shown in Figure 5.8 for JP9 (Andic Dystrudepts) and JP3 (Typic Haplorthods). Magnesiumsaturated specimens exhibited a prominent diffraction peak at approximately 6°2θ,
204
World Soil Resources and Food Security
(a) JP9; Andic dytrudepts A
AB 1.4 1.0 0.72
1.4 1.0 0.72
C 1.4 1.0 0.72 K-550ºC K-350ºC K-25ºC Mg-gly Mg-25ºC
3
6
9
12 15 3
6 9 12 15 º2θ (Cu Kα)
3
6
9
12 15
(b) JP3; Typic haplorthods E
EB
1.7 1.4 1.0 0.72
Bs
1.7 1.4 1.0 0.72
BC1
1.4 1.0 0.72
1.4 1.0 0.72 K-550ºC K-350ºC K-25ºC Mg-gly Mg-25ºC
3
6
9
12 15
3
6
9
12 15 3 6 º2θ (Cu Kα)
9
12 15
3
6
9
12 15
FIGURE 5.8 X-ray diffraction patterns of soil clays extracted from different soil horizons of JP9 (Andic Dystrudepts) and JP3 (Typic Haplorthods).
corresponding to the presence of 2:1 phyllosilicates with 1.4 nm d-spacing. Compared with the 1.4 nm diffraction peak, the 1.0 nm peak (if any) appearing at 8–9°2θ was negligible in Mg-saturated specimens, which often is attributed to the minor amount of illitic minerals in soil clays [Harris et al. 1988]. Magnesium-saturation and glycerol solvation shift the peak from 1.4 nm to 1.7 nm: this was particularly clear for the E soil clay and partly so for EB soil clays in the JP3, suggesting that these clays included expansible 2:1 phyllosilicates, such as smectite. The 1.4 nm diffraction peak shown by the Mg-saturated JP9 specimen did not collapse and move to 1.0 nm after the K-saturation, although it did do so when K-saturation was followed by 550°C heating. This pattern indicates that HIV predominates in these soil clays. From the bottom to the surface horizon in JP3, however, the collapse of 1.4 nm and shift to 1.0 nm gradually increased, indicating an increasing amount of vermiculite or
Pedogenetic Acidification in Humid Asia
205
smectite toward the surface horizons. Thus, the XRD analysis represents well the sequential decrease in HIV as podzolization proceeds, in contrast to the concomitant increase in vermiculite or smectite [Carnicelli et al. 1997; Brahy et al. 2000b]. All these features were also commonly observed among the Mie and Ashiu soils in the preceding section. The amount of kaolinite in the soils was generally small, compared with that in common tropical soils, a finding consistent with other general understandings [Velde 2001; Thiry 2000]. The proportion of the illitic minerals was estimated from the total K content of clay (TKclay), assuming the K content of pure illite, which is 85.4 g kg−1 for Montana illite [Thompson and Ukrainczyk 2002]. Based on this supposition, for example, the TKclay of 24.1 g kg clay−1 (JP3-Bs) corresponds to about 282 g kg clay−1 of illitic minerals provided that negligible amounts of K-feldspar are included in the soil clays. However, JP3-Bs apparently did not have a 1.0 nm peak in XRD analysis (Figure 5.8). Harris et al. [1992] confirmed the presence of a large amount of illitic minerals within HIV clays via electron microprobe analysis. We note that the amount of illitic minerals in HIV-dominated clays may be underestimated greatly when basing the evaluation only on the XRD peaks.
5.4.3 RIP Variation and the Sequential Transformation of HIV to Vermiculite Figure 5.9 depicts the vertical distribution of RIP, Cs-fixed capacity, and hot-citrate Al in each profile. All the profiles, more or less, showed a decreasing trend in the amount of hot-citrate Al toward the surface horizon, and an increasing trend in the Cs-fixed capacity; since these properties are associated, respectively, with the degree of hydroxy-Al interlayering, and the vermiculite content, these vertical trends in the profile strongly indicate the sequential transformation of HIV into vermiculite at the upper horizons. Our observations agree with the XRD patterns that we derived and with previous studies on the relationships between soil development and clay mineralogy in this region [Hirai et al. 1989; Funakawa et al. 1993; Funakawa et al. 2003]. As soils become acidified and humic substances accumulate, hydroxy-Al polymers can be released from the HIV interlayers via organo-mineral complexation with organic acids [Carnicelli et al. 1997]. Thus, in soils located in upper horizons and therefore subjected to more intensive podzolization, HIV transforms into vermiculite or further into smectite [Barnhisel and Bertsch 1989; Kitagawa 2005]. Similar to the distribution of the Cs-fixed capacity, the trend in the vertical distribution of RIP also increased toward the surface horizon for most of the Inceptisols. However, in the podzolic profiles (i.e., JP1 and JP3), RIP firstly increased in parallel with the Cs-fixed capacity toward the surface, but then declined in the eluvial surface horizons (Figure 5.9). This trend resembles Maes et al.’s observation [1999] that in soils with a podzolic character, the ability to fix 137Cs was smaller at their surface horizon. The causation of the RIP reduction seems not to be the small amount of illitic minerals because the eluvial clays contained larger amounts of illitic minerals than did the other clays at lower horizons. Maes et al. [1999] also mentioned the weak relationship between the amount of illitic minerals and trace 137Cs fixation in Belgian forest soils. Pure illites not subjected to pedogenetic weathering have average RIP
206
Depth (cm)
0
World Soil Resources and Food Security 0
10 20 30 40 50 60
10 20 30 40 50 60
0
20
20
40
40
40
0
0
5
10 15 20 25 30
JP2 Andic dystrudepts
80 0
0
5
10 15 20 25 30
0
20
20
40
40
40
80 0
0
10 20 30 40 50 60
JP5 Typic dystrudepts
80 0
0
5
10 15 20 25 30
0
20
20
40
40
40
60 JP7 Ruptic-ultic dystrudepts
80 0
0
5
80
JP3 Typic haplorthods 0
10 20 30 40 50 60
JP6 Typic dystrudepts
80
20
60
10 15 20 25 30
60
60 JP4 Andic dystrudepts
5
80
20
60
0
60
60 JP1 Lithic haplorthods
80
Depth (cm)
0
20
60
Depth (cm)
0
0
5
10 15 20 25 30
60 JP8 Typic dystrudepts
80
JP9 Andic dystrudepts
10 15 20 25 30
Depth (cm)
20 40 60 80
JP10 Andic dystrudepts
FIGURE 5.9 Vertical distribution of Cs-fixing capacity (∇; cmolc kg clay−1), hot-citrate Al (○; g kg clay−1), and RIP (- - -; mol kg clay−1).
207
Pedogenetic Acidification in Humid Asia
values of ca 10–15 mol kg−1 [Delvaux et al. 2001; de Koning et al. 2007; Nakao et al. 2008], i.e., much lower than did the vermiculitic soil clays we studied, though much higher than did the HIV-dominated soil clays. We suggest that although the presence of illitic minerals primarily controls the amount of the frayed edge site in soil clays, its amount per mass of illitic minerals might vary orderly, depending on the degree of vermiculitization, hydroxy-Al interlayering, or the charge reduction in the vermiculitic layers closely associated with the frayed edge site.
5.4.4 Blockage Effect of Hydroxy-Al Polymers on the Frayed Edge Site The RIP values were determined before and after hot-citrate extraction for soil clays from two soil profiles (JP1 and JP4) and for the HIV-dominated soil clays from four different soil profiles (JP2-Bw2, JP3-BC1, JP9-AB, and JP10-AB) to elucidate the effect of hydroxy-Al polymers on the frayed edge site (Figure 5.10). Since the DCB and hot-citrate extractions, respectively, result in loss of soil materials corresponding to approximately 100 g kg clay−1 of SiO2, Al2O3, and Fe2O3 (DCB-oxides) and 10 g kg clay−1 of Al-oxides (hot-citrate Al) on average, the RIP values presented in Figure 5.10 have been corrected to the value per 1 kg of the original samples. In the JP1 soil profile, the soil clays in the E, EB, and Bs horizons registered no increase in RIP after hot-citrate extraction, whereas the soil clay in the CB horizon showed a significant increase (Figure 5.10). Similarly, in the JP4 soil clay’s profile, RIP (mol kg clay–1) 1
10
100 JP1
E
RIP (mol kg clay–1)
EB
JP2 Bw2
B
1
JP3 BC1
RIP (mol kg clay–1) 1
10
1
JP4
10
100
RIP (mol kg clay–1)
100
JP9 AB
1
AB
10
100
RIP (mol kg clay–1)
Bw BC
100
RIP (mol kg clay–1)
CB
A
10
JP10 AB
1
10
100
FIGURE 5.10 RIP before and after hot-citrate extraction. Bars indicate standard error (n = 3). □ Untreated samples; ■ after hot-citrate extraction.
208
World Soil Resources and Food Security
there were no significant changes in the RIP in the A and AB horizon after hotcitrate extraction, whereas the subsurface clays (Bw, BC) revealed a large increase in RIP similar to, or more than, that of the upper soil clays after treatment. This pattern in the RIP was evident in the subsurface clays from other profiles (Figure 5.10). The increase in RIP is associated with the HIV-dominated clay mineralogy, and is consistent with the findings of Maes et al. [1998], who investigated the large increase in both 137Cs net retention and the distribution coefficient of 137Cs in HIV-dominated soil clays after hot-citrate extraction. In contrast, the limited increase in RIP values for the upper soil clays was associated with vermiculitic- or smectitic-clay mineralogy. These findings signify that in strongly acidic surface soils, the hydroxy-Al polymers already have been released from the 2:1 interlayers, and therefore, a large number of the frayed edge sites could be available for 137Cs adsorption. They also strengthen the conclusion that the loss of hydroxy-Al polymers from the HIV layers through podzolization exposes the weathering front of illitic minerals. We can deduce the fate of the exposed illitic minerals from the sequential variations of RIP values with different mineralogical stages under podzolization. Figure 5.11 illustrates the results of Cs-fixed capacity and RIP against hot-citrate Al. Along with the decreasing amount of hot-citrate Al, the RIP increases first consistent with the pattern of Cs-fixed capacity. However, RIP decreases for soil clays with a hotcitrate Al content of less than 10 g kg clay−1, which is inconsistent with the case for Cs-fixed capacity (Figure 5.11). Apparently, the decrease starts after the frayed edge site is exposed almost completely because of the release of hydroxy-Al polymers, as evidenced by the findings that soil clays exhibiting the decreasing phase did not display a RIP increase after hot-citrate extraction (Figure 5.11). The progressive decrease in the RIP indicates that the frayed edge site is more susceptible to the accumulative acid loads under podzolization, compared with the bulk of the vermiculitic charges. Hydroxy-Al polymers may play a role in protecting frayed edge sites from weathering. Le Roux and Rich [1969] suggested that the reduction of frayed edge (b)
RIP (mol kg clay–1)
100
Cs-fixed capacity (cmolc kg clay–1)
(a)
10
1
0.1 0.1
1
10
Hot-citrate Al (g kg clay–1)
100
100
10
1
0.1 0.1
1
10
Hot-citrate Al (g kg clay–1)
100
FIGURE 5.11 Scattered diagrams of (a) Cs-fixing capacity or (b) RIP and hot-citrate Al. Open circles mean the samples were not extracted with hot-citrate solution; plus signs represent the treated samples. The amount of hot-citrate Al in the horizontal axis for the hot-citrate samples (plus) is the same as that of the original samples.
Pedogenetic Acidification in Humid Asia
209
sites resulted from the interlayer opening completely through the particles. On the other hand, Maes et al. [1999] considered that it was because of charge degradation in the vermiculitic layers closely associated with the illitic minerals. Both explanations might apply to our results, but we cannot determine which one is predominant. Even though the micromorphological feature of the illite–vermiculite interfaces was unclear, the correspondence between the increasing pattern of RIP and Cs-fixed capacity with releasing hydroxy-Al polymers was definitive, while the distinctive difference between them at the late stage of podzolization afforded a strong clue to understanding the weathering of 2:1 phyllosilicates during podzolization.
5.5 C HARGE CHARACTERISTICS OF FOREST SOILS DERIVED FROM SEDIMENTARY ROCKS IN JAPAN IN RELATION TO PEDOGENETIC ACIDIFICATION PROCESSES In the preceding sections, pedogenetic acidification processes both in mesic and thermic soil temperature regimes were analyzed. The clay mineralogy including amorphous and/or organo-mineral complexes could largely modify the surface chemistry of soil colloids. In turn, the analysis of charge characteristics of the soils having different mineralogical characteristics would give an insight on the pedogenetic processes analyzed so far. Charge characteristics determine the physicochemical behavior of soil colloids through mechanisms such as cation retention, ion selectivity, dispersion/flocculation, etc. It is widely recognized that soils rich in sesquioxides exhibit variable charge characteristics. Such soils include volcanic soils dominated by amorphous alminosilicates, highly weathered soils rich in Fe oxides, and spodic horizons of podzolic soils [Okamura and Wada 1983; Tessens and Zauyah 1982; Laverdiere and Weaver 1977]. Similarly, the interlayered materials of 2:1 minerals behave as variable charge components [Inoue and Satoh 1992; Helmy et al. 1994], bringing several changes to the physicochemical properties of soils originally dominated by expansible 2:1 minerals (e.g., vermiculites and smectites). Their influence includes a reduction in CEC in the pH range below the zero point of charge (ZPC) of the interlayered materials, changes in cation exchange selectivity, and an increase in the specific adsorption of metal ions [Barnhisel and Bertsch 1989]. In the present section, the main objectives were set to be: 1) to analyze the charge characteristics of forest soils derived from sedimentary or metamorphic rocks in the Kinki District, in which HIV are predominant, compared to other soils formed under similar humid climatic conditions; and 2) to examine the pedogenetic processes that can affect these charge characteristics with special reference to the differences in bioclimatic conditions and processes of soil acidification.
5.5.1 Soil Samples In total, 28 soil samples were collected from 20 soil profiles, including 9 profiles (MD1, MD2, FD, OM, OT, TT, KD1, KD2, and MT) in the eastern part of Mt. Odaigahara, Mie Prefecture, 7 profiles (SW2, H2, S3, N1b, N4b, SW1, and H1b) in the Ashiu Experimental Forest of Kyoto University in Kyoto Prefecture, and
210
TABLE 5.10 Physicochemical Properties of Soils in Japanese Plots Used for Experiments Relating to Charge Characteristics pH
MD1
Dystrudepts
MD2 FD ON1 ON2
Dystrudepts Dystrudepts Dystrudepts Dystrudepts
OM OT
Dystrudepts Dystrudepts
TT
Dystrudepts
Depth Horizon
(cm)
Sample Categoryb
BA Bw2 BC Bw1 Bw2 Bw2 Bw2
2–10 25–42 42–50+ 20–45 14–40 36–63 38–60
E/uB B-thermic B-thermic B-thermic B-thermic B-thermic B-thermic
BC BA Bw2 C Bw1
33–45 8–24 38–52 62–70+ 20–40
B-mesic E/uB B-mesic B-mesic B-mesic
(H2O)
Total C (g kg−1)
4.52 4.62 4.69 4.70 4.80 5.00 4.84
Bases
Al
CEC
Sand
(cmolc kg−1)
Dystrudepts, Thermic 67.7 0.7 9.5 13.4 0.3 8.3 7.4 0.2 7.6 15.2 0.3 9.2 19.1 0.2 6.7 4.9 0.6 7.8 10.6 0.4 5.9
25.4 19.0 16.0 22.5 16.1 21.2 14.6
Dystrudepts and Hapludands, Mesic 4.70 48.5 0.3 2.1 14.4 4.12 76.9 0.6 23.9 40.4 4.65 55.9 0.2 6.8 21.9 4.80 16.3 0.1 2.8 12.7 4.55 26.5 0.3 7.6 18.5
Silt
Clay
Feo
(%)
Alo
Sio
(g kg−1)
Clay Mineralogical Compositionc
11 13 40 8 32 15 45
24 24 18 24 23 33 14
65 63 42 69 45 52 41
7.8 4.1 3.1 5.1 4.7 3.4 2.5
8.3 4.2 3.3 6.9 3.0 2.3 1.3
0.1 0.1 0.1 0.0 0.1 0.0 0.0
HIV >> Kao HIV >> Kao HIV >> Kao, Mica HIV >> Kao HIV >> Kao HIV, Kao >> Mica HIV, Kao > Mica
57 14 23 43 26
21 26 37 16 26
23 60 40 41 48
9.3 9.4 10.8 9.6 5.8
14.7 21.1 20.8 7.9 12.6
0.2 0.0 0.1 0.6 0.0
HIV, Kao >> Mica HIV >> Kao HIV >> Kao HIV > Mica, Kao HIV > Kao > Mica
World Soil Resources and Food Security
Site
Soil Classificationa
Particle Size Distribution
Exchangeable
Dystrudepts Dystrudepts Hapludands Dystrudepts Dystrudepts
S3 N1b KT2
Fulvudands Hapludands Dystrudepts
N4b SW1
Haplorthods Haplorthods
H1b KT1
Haplorthods Haplorthods
a b
c
BC Bw2 Bw2 Bw BA Bw BC Bw Bw Bw
32–44 29–45 35–55 25–50 14–28 28–48 48–57 48–70 19–55 15–36
B-mesic B-mesic B-mesic B-mesic E/uB B-mesic B-mesic B-mesic B-mesic B-mesic
4.92 4.65 4.85 4.62 4.19 4.63 4.79 4.65 4.77 4.88
25.8 13.0 35.0 41.0 75.0 18.0 13.0 17.0 33.0 44.3
0.3 0.2 0.2 0.6 1.2 0.5 0.5 0.9 0.3 0.7
9.6 7.4 3.0 6.1 12.1 6.4 5.3 4.5 3.1 6.7
Bw1 E Bs C Bh E2
30–40 2–10 23–45 60–73+ 19–22 6–11
B-mesic E/uB Bs-pod B-mesic Bs-pod E/uB
4.51 3.74 4.20 4.88 3.99 4.24
Haplorthods, mesic 21.0 0.4 6.9 57.0 0.8 13.4 37.0 0.5 15.6 12.0 0.4 2.5 54.0 0.5 16.7 42.7 0.5 10.6
17.9 22.5 15.6 28.2 31.5 15.8 13.4 20.1 19.9 32.6
47 33 22 10 12 15 21 9 16 21
19 25 19 49 33 33 33 42 32 34
35 41 59 41 55 52 46 49 53 46
6.3 5.6 13.4 9.2 8.3 7.7 6.6 11.4 20.6 16.0
8.7 8.8 17.1 13.4 18.2 11.5 6.7 10.1 15.9 25.6
0.1 0.0 0.4 0.6 0.4 0.6 0.7 1.1 2.9 0.0
HIV >> Kao HIV >> Kao HIV >> Kao HIV > Kao > Mica HIV > Kao HIV > Kao HIV > Kao HIV > Kao HIV > Kao HIV > Kao
20.6 32.0 33.9 12.8 33.7 21.7
16 31 19 67 31 16
37 36 31 16 29 57
48 33 50 17 40 28
10.1 3.0 8.2 8.3 6.7 2.4
20.5 4.4 22.8 4.0 22.0 9.2
0.6 0.3 0.3 1.2 0.2 0.0
HIV > Kao Mica-Sm >> Kao Mica-Vt > Kao HIV > Mica > Kao Vt, Sm, Mica, Kao Vt > Kao
Pedogenetic Acidification in Humid Asia
KD1 KD2 MT SW2 H2
According to Soil Taxonomy [Soil Survey Staff 2006]. E/uB: E or upper B horizon soils; Bs-pod: Bs horizon of podzolic soils; B-mesic: Bw and lower B horizons of forest soils with MSTR; B-thermic: Bw and lower B horizons of forest soils with TSTR. Kao, kaolin minerals; Mica, clay mica; Sm, emectite; Vt, vermiculite; HIV, hydroxy-Al interlayered vermiculite.
211
212
World Soil Resources and Food Security
4 profiles (ON1, ON2, KT2, and KT1) in the Kuta and Ohno Experimental Forests of Kyoto Prefectural University in Kyoto Prefecture (see also Tables 5.6 and 5.7). Soil samples were collected from either the E or upper B horizon (mostly above 20-cm depths) or lower B to C horizons (mostly below 30-cm depths). Since the main objective of this study was to analyze the charge characteristics of soil mineral components along with pedogenetic processes, we excluded soil samples from the surface few centimeters, which were expected to be considerably affected by organic matter accumulation. The general physicochemical properties of the soils studied are presented in Table 5.10, along with their taxonomic classification. The soils from the high elevation area (above 600 m) of Mt. Odaigahara and all the soils from the Ashiu and Kuta Experimental Forests were subjected to cool temperate forest conditions with a mesic soil temperature regime (MSTR), while the soils from the low elevation area (below 600 m) of Mt. Odaigahara and those from the Ohno Experimental Forest were subjected to warm temperate forest conditions with a thermic soil temperature regime (TSTR). The soils tested included Dystrudepts, Hapludands, and Haplorthods and all the soils were derived from sedimentary or metamorphic rocks, which were not appreciably affected by volcanic ejecta. In most cases, crystalline HIV predominated with small amounts of mica and kaolin minerals in the clay fraction, whereas expansible 2:1 minerals such as vermiculite and smectite sometimes predominated in the upper layers of podzolized soils. Soil samples collected were air-dried, passed through a 0.2-mm mesh sieve, and analyzed for charge characteristics by the ion adsorption method.
5.5.2 Analytical Methods Determination of cation and anion retention, or cation and anion exchange capacities (CEC and AEC), of the soils at different ionic concentrations and pH values was performed according to the method of Schofield [1950]: In a 50-mL centrifugation tube, 2 g of soils were washed with a 1 mol L−1 NaCl solution 3 times and then with a 0.2, 0.1, 0.02, or 0.005 mol L−1 NaCl solution at pH values ranging between 4.5 and 8. The samples were equilibrated in the NaCl solutions at each concentration or pH value for 2 days. The solution pH was maintained at the objective pH using dilute HCl or NaOH solution, if necessary. After centrifugation, Na and Cl ions adsorbed on the soil were extracted with 0.5 M KNO3. The content of Na was determined by atomic absorption spectrophotometry (AAS) [Shimadzu, AA-840-01] and the Cl content by mercury (II) thiocyanate colorimetry at 460 nm [Frankenberger et al. 1978]. The amounts of Na and Cl in the occluded solution were corrected by subtraction based on their concentration in the supernatant and the weight of the occluded solution. The negative and positive charges of the soil were calculated from the amount of adsorbed Na and Cl, respectively. In addition, to analyze the effect of organic matter on the charge characteristics of the soils, the cation retention of the soils was measured after the destruction of organic matter with H2O2, as follows: 25 mL of deionized water and 5 mL of a 30% H2O2 solution were added to 2 g of soil and heated at 80°C until the soil organic matter and H2O2 were completely decomposed. After centrifugation, the samples were washed twice with 80% ethanol and then air-dried
Pedogenetic Acidification in Humid Asia
213
and weighed. The ion adsorption method described above was then applied, using a 0.1 mol L−1 NaCl solution as the electrolyte. The pH of the equilibrated solution was adjusted within the range of 5.5 to 6.9. In addition, in order to evaluate the effect of the sesquioxidic components on the charge characteristics of the soils, two representative B horizon soils (OT-Bw2 and MD1-Bw2) under MSTR or TSTR, which contained certain amounts of free oxides and interlayered materials, were used to determine the CEC, with a 0.1 mol L−1 sodium acetate buffer solution adjusted to pH 5 or 6 by acetic acid as electrolyte, after successive removal of organic matter by H 2O2 treatment with heating at 80°C, Al and/or Fe oxides by dithionite-citrate-bicarbonate (DCB) treatment with heating at 80°C [Mehra and Jackson 1960], and interlayered materials of 2:1 clay minerals by 0.33 M sodium citrate treatment with heating at 100°C for 8 hours [Tamura 1958]. Data obtained were analyzed using SYSTAT 8.0 software [SPSS Inc. 1998].
5.5.3 General Charge Characteristics of the Soils The CEC at different pH values and ionic concentrations is shown in Figure 5.12 for selected soils. Since anion retention was always less than 0.5 cmolc kg−1 even for a pH value below 5, the contribution of the sesquioxidic components to anion retention through a positively charged surface was limited for the soils studied, in contrast to volcanic or highly weathered soils, which showed fairly large amounts of positive charges within low pH ranges below 5 [Espinoza et al. 1975; Van Raij and Peech 1972]. It is assumed that the variable positive charge of the soils in the present study was neutralized by the permanent negative charge of the soils, which were derived from 2:1 clay minerals, unlike in the above-cited volcanic or highly weathered soils. Hence, the following discussion will focus on cation retention. In order to analyze the charge characteristics on a quantitative basis, the cation retention curves were fitted to the following equation and the coefficients a, b, and c were calculated by multiple regression analysis:
log CEC = a pH + b log C + c,
(5.7)
where C is the ionic concentration expressed in mol L−1. Higher values for a and b indicate a larger contribution of the pH and/or concentration-dependent negative charges to the total charge of the soils. In contrast, the value of c is a parameter that represents log CEC at pH 0 in a 1 mol L−1 solution. For example, based on Figure 5.13, the contribution of the variable negative charge to the total charge was especially low (a = 0.09 and b = 0.12) in the SW1-E sample (Haplorthods) compared to most of the B horizon soils. There was a positive correlation between a and b with a slope of almost unity (Figure 5.13a), suggesting that the negative effects of H+ concentration and the positive effects of the ionic strength of the supporting electrolyte on the appearance of the variable negative charge were similar. The negative correlation between (a + b)/2 and c indicates that the development of variable charge characteristics contributed to the decrease of the CEC in a low pH range (Figure
0
4
5
7
8
CEC (cmolc kg–1)
log CEC = 0.14pH + 0.17 log C + 0.78 50 R2 = 0.93 40 30 20 10 0
4
5
6
pH
7
8
30
30 20
20
10
10 0
4
OT Bw2
30 20 10
0
4
5
6
pH
7
8
7
8
0
4
5 OT C
6 pH
7
8
40 30
30 20
20
10
10 0
4
5
6
pH
7
8
0
4
SW1 Bs
5
6
pH
7
8
SW1 C
30
log CEC = 0.14pH + 0.20 log C + 0.65 log CEC = 0.27pH + 0.29 log C – 0.53 40 R2 = 0.87 R2 = 0.88 30
20
20
40
CEC (cmolc kg–1)
CEC (cmolc kg–1)
CEC (cmolc kg–1)
log CEC = 0.09pH + 0.12 log C + 0.93 R2 = 0.90
6 pH
log CEC = 0.26pH + 0.40 log C + 0.00 log CEC = 0.34pH + 0.37 log C – 0.89 40 R2 = 0.93 R2 = 0.91
SW1 E 40
5
10
0
10
4
5
6
pH
7
8
0
4
5
6
pH
7
8
FIGURE 5.12 CEC of the soils determined at different pH values and electrolyte concentrations. ○ 0.2 M, △ 0.1 M, □ 0.02 M, × 0.005 M.
World Soil Resources and Food Security
CEC (cmolc kg–1)
OT BA
6 pH
40
CEC (cmolc kg–1)
10
MD1 BC
log CEC = 0.14pH + 0.24 log C + 0.47 log CEC = 0.17pH + 0.15 log C + 0.10 40 R2 = 0.87 R2 = 0.79
CEC (cmolc kg–1)
CEC (cmolc kg–1)
CEC (cmolc kg–1)
20
MD1 Bw2
214
MD1 BA
log CEC = 0.24pH + 0.21 log C – 0.13 40 R2 = 0.91 30
215
Pedogenetic Acidification in Humid Asia
0.5
b
0.4
(b)
y = 1.02x + 0.04 R2 = 0.67
0.4
0.3 0.2
0.2 0.1
0.1 0 0
y = –0.13x + 0.24 R2 = 0.72
0.3
(a + b)/2
(a)
0.1
0.2
a
0.3
0.4
0.5
–1
–0.5
0
0
c
0.5
1
FIGURE 5.13 Relationships between coefficients a and b (a) and c and (a + b)/2 (b) values in the regression equation, log CEC = a pH + b log C + c, which represents the charge characteristics of the soils. △ E or upper B horizon soils, ○ Bs horizon of Haplorthods, ▲ Bw and lower B horizons of Dystrudepts and Hapludands under MSTR, ● Bw and lower B horizons of Dystrudepts under TSTR.
5.13b). These results are consistent with earlier reports [Okamura and Wada 1983]. The values of (a + b)/2 and c of the E and B horizons of the Haplorthods and the upper B horizons of the Dystrudepts ranged from 0.11 to 0.23 (mean: 0.17) and from −0.13 to 0.93 (mean: 0.57), respectively. Some of these soils showed especially high c values. The values of (a + b)/2 and c of the B horizons of Dystrudepts with TSTR ranged from 0.14 to 0.25 (mean: 0.19) and from 0.01 to 0.47 (mean: 0.24), respectively. These values were similar to those of Ultisols or Oxisols in Thailand, which were reported by Wada and Wada [1985]. Corresponding values for the B horizons of Dystrudepts or Hapludands with MSTR ranged from 0.23 to 0.39 (mean: 0.29) and from −0.89 to 0.17 (mean: –0.28), respectively, which were comparable to those of the B horizons of Andisols [Okamura and Wada 1983; Wada and Wada 1985]. These soils were dominated by a variable negative charge with a rather small apparent permanent negative charge.
5.5.4 C ontribution of Each Component to the Charge Characteristics of the Soils—Statistical Analysis Table 5.11 lists the correlation coefficients between the physicochemical properties and several parameters relating to the charge characteristics of the soils. Acid ammonium oxalate-extractable Al (Alo) showed a positive correlation and exchange acidity a negative correlation with coefficients a and b, while properties relating to soil acidity (such as pH(H2O), pH(KCl), and the content of exchangeable Al) showed significant correlations with c. These relationships indicate the importance of amorphous compounds and/or soil acidity for the charge characteristics of these soils. To analyze the effect of physicochemical properties on charge characteristics, principal component analysis was conducted for the soils, followed by stepwise multiple linear regression. Variables employed included pH(H2O), pH(KCl), sum of exchangeable bases, levels of exchangeable Al, CEC, total C, pyrophosphateextractable C, Al, and Fe (Cp, Alp, and Fep, respectively), acid ammonium oxalate
216
TABLE 5.11 Correlation Coefficients between Physicochemical Properties and Charge Characteristics of the Soils
c
pH (H2O)
pH (KCl)
Exch. Bases
Exch. Al
CEC
Total C
Alo
Feo
Sio
Alo + 1/2Feo
0.58** 0.41* 0.51**
0.75** 0.66** 0.74**
−0.29 −0.37 −0.35
−0.63** −0.56** −0.62**
−0.39* −0.34 −0.38*
−0.10 −0.08 −0.10
0.67** 0.68** 0.71**
0.14 0.27 0.22
0.54** 0.51** 0.55**
0.47* 0.54** 0.53**
−0.73**
−0.83**
0.31
−0.52**
0.09
−0.48**
0.38*
0.78**
Note: *Significant at 5% level, **significant at 1% level, n = 28.
0.59**
−0.26
Clay Content −0.08 0.08 0.01 0.18
World Soil Resources and Food Security
a b (a + b)/2
217
Pedogenetic Acidification in Humid Asia
(pH 3)-extractable Al, Fe, and Si (Alo, Feo, and Sio, respectively), DCB-extractable Fe (Fed), Alo + 1/2Feo, Alp + 1/2Fep, and clay content. Table 5.12 shows the factor pattern for the first three principal components after varimax rotation, which accounted for 81% of the total variance. High positive coefficients were given to Cp, Alo, Feo, Sio, Alp, Alo + 1/2Feo, and Alp + 1/2Fep for the first component. These variables corresponded to the properties derived from amorphous and/or organo-mineral complexes and, hence, the first component was referred to as the amorphous materials factor. Based on Figure 5.14, B horizon soils under TSTR showed relatively low scores for the first component, while those under MSTR showed high scores. The second component exhibited high coefficients, positive or negative, with pH(H2O), pH(KCl), exchangeable Al, CEC, total C, Cp, and Fep, indicating the existence of a close relation with the soil acidity and organic matter content. The scores of this component were especially high among upper layer soils (Figure 5.14), suggesting that the component reflected soil acidification accompanied by the accumulation of organic matter from surface horizons. This factor was referred to as the acidity factor. The third component showed high coefficients with Fed and the clay content and was considered to be a weathering factor. Most of the variables were closely related
TABLE 5.12 Factor Pattern for the First Three Principal Components Relating to Soil and Environmental Variables Variable pH(H2O) pH(KCl) Exch. Bases Exch. Al CEC Total C Cp Alo Feo Sio Alp Fep Fed Alo + 1 / 2Feo Alp + 1 / 2Fep Clay content Eigenvalue Proportion
PC1
PC2
PC3
0.05 0.48 0.11 −0.17 0.18 0.42 0.62 0.95 0.74 0.61 0.94 0.51 0.34 0.94 0.80 0.06
−0.91 −0.80 0.58 0.85 0.85 0.81 0.72 −0.17 0.42 −0.31 0.08 0.64 0.24 0.11 0.41 –0.01
0.08 0.03 0.08 0.32 0.33 −0.01 0.04 0.05 0.29 −0.33 0.20 0.40 0.78 0.18 0.34 0.83
7.38 0.46
4.11 0.26
1.42 0.09
Amorphous materials
Acidity
Weathering
218
World Soil Resources and Food Security 3 2
Factor 2
1
–2
–1
0
0
1
2
–1 –2
Factor 1
FIGURE 5.14 Scattergram between the first and second principal component scores determined for each sampled soil. △ E or upper B horizon soils, ○ Bs horizon of Haplorthods, ▲ Bw and lower B horizons of Dystrudepts and Hapludands under MSTR, ● Bw and lower B horizons of Dystrudepts under TSTR.
to only one component with high coefficients above 0.6, except for Cp, which was affected by both the accumulation of organo-mineral complexes in the B horizons and percolation of organic matter from the soil surface. In the next step, stepwise multiple regression analysis was conducted to examine the contribution of each factor to (a + b)/2 or c, which represents the charge characteristics of the soils. The following equations were obtained:
(a + b)/2 = 0.238 + 0.049 × (amorphous materials factor) − 0.041 × (acidity factor),
(5.8)
c = 0.043 − 0.210 × (amorphous materials factor) + 0.346 × (acidity factor) + 0.115 × (weathering factor).
(5.9)
These equations clearly indicated that:
1. The value of (a + b)/2, which represents the variable charge characteristics of the soils, increased in parallel with the increase in the amorphous material content, while it decreased during the process of soil acidification. 2. The value of c was contributed negatively by the amorphous materials factor and positively by the acidity and weathering factors. The negative contribution of the amorphous materials factor to the value of c indicates that such components prevented the appearance of the permanent negative charge of 2:1 minerals by blocking it through the variable positive charges of the oxide surface. Soil acidification is considered to release such blocked negative charge sites through dissolution of amorphous materials. The weathering factor may be related to the direct increase in the colloidal surface area through the increase in the clay fraction.
219
Pedogenetic Acidification in Humid Asia
5.5.5 Contribution of Each Component to the Charge Characteristics of the Soils —Analysis by Successive Removal of Soil Components
CEC after H2O2 treatment (cmolc kg–1)
(a)
(b) 40
1:1
30 20 10 0
10
20
30
CEC of original sample (cmolc
40 kg–1)
20 15 10 5 0 2
1
0 –1 –2
1 r) or to c t fac Fa ity d ci (A
0
CEC decrease after H2O2 treatment (cmolc kg–1)
In the last part of this experiment, several chemical treatments were applied to remove each component of the soil in order to detect changes in charge characteristics. Since such treatments could, however, have affected the nature and/or composition of the remaining components of the soil, the following considerations may not be conclusive. Figure 5.15 shows the relationship between the CEC of the H2O2-treated samples (determined in a 0.1 mol L−1 NaCl solution with pH adjusted in the range of 5.5 to 6.9) and the CEC calculated for untreated samples (at the same pH and ionic concentration). The decomposition of organic matter scarcely affected the CEC of forest soil B horizons, suggesting that the direct contribution of the acidic functional groups of soil organic matter to the variable negative charge was not appreciable among these B horizon soils. This can be attributed to the fact that most of the acidic functional groups were already bound to oxide surfaces and the few R-COOH ligands that remained were dissociated along with the pH increase. However, there were several exceptions among the E or upper B horizon soils, which showed a significant decrease of the CEC after H2O2 treatment (Figure 5.15a). Since the contribution of variable charge characteristics, or (a + b)/2, was low for these soils (Figure 5.13a), such a decrease of CEC after H2O2 treatment might be attributed to a decrease in the permanent negative charge through the destruction of 1.4 nm minerals by the treatment. Figure 5.15b shows that these soils were already affected by extensive soil acidification, i.e., higher scores for the acidity factor. These soils may have been somewhat unstable due to partial destruction of their structure. Figure 5.16 depicts the changes in the CEC (at pH 5.0 and 6.0 in 0.1 M sodium acetate buffer solutions) of the OT-Bw2 (MSTR) and MD1-Bw2 (TSTR) soils by
0.1
0.2
0.3
0
(a + b)/2
FIGURE 5.15 CEC of the soils before and after H2O2 treatment for decomposing organic matter (a) and the relationship between the charge decrease, variable charge characteristics ((a + b)/2), and soil acidity (scores of acidity factor) (b). △ E or upper B horizon soils, ○ Bs horizon of Haplorthods, ▲ Bw and lower B horizons of Dystrudepts and Hapludands under MSTR, ● Bw and lower B horizons of Dystrudepts under TSTR.
220
World Soil Resources and Food Security
60
Interlayered materials removed
CEC/clay (cmolc kg–1)
50 40 30
Free oxides removed
20 10
Organic matter removed
0 OT OT Bw2, Bw2, MD1 pH5 pH6 Bw2, pH5
Original soil MD1 Bw2, pH6
FIGURE 5.16 Changes of CEC after successive removal of organic matter, free oxides, and interlayered materials in the 2:1–2:1:1 intergrades.
successive treatments with H2O2, DCB, and sodium citrate solution with heating (to remove organic matter, free oxides of Al and/or Fe, and interlayered materials, respectively). The CEC significantly increased after the DCB treatment, suggesting that the DCB-extractable fraction of free oxides had prevented the appearance of the permanent negative charge of 2:1 minerals by blocking it through the variable positive charge of the oxide surface. Extraction of interlayered materials by a 0.3 mol L−1 sodium citrate solution with heating resulted in a further increase in the permanent negative charge of the soils, presumably due to the removal of Al hydroxides and K+ ions in the interlayer space of 2:1 minerals. Thus the free oxides and/or interlayered materials were considered to have neutralized the layer charge of 2:1 minerals.
5.5.6 C hanges in the Charge Characteristics of the Soils through Pedogenetic Acidification Figure 5.17 plots the relationship between (a + b)/2 and the CEC calculated for low pH (5.0) and low electrolyte (0.01 mol L−1) levels, at which the contribution of the acidic functional groups of soil organic matter to the CEC is limited, if any. The B horizons of the forest soils were characterized by the predominance of a variable negative charge with a CEC/clay less than 10 cmolc kg−1 (at pH 5, 0.01 mol L−1) and (a + b)/2 values above 0.1. Generally the B horizon soils with MSTR gave higher (a + b)/2 values (normally in the range of 0.2 to 0.4) than the B horizon soils with TSTR [(a + b)/2 of 0.1–0.25]. This difference was attributed to the contribution of amorphous compounds, mainly of Al, with ZPC values around 7.0 to 7.7 [Parks 1965]. These compounds neutralized the permanent negative charge in the pH range below the ZPC and developed a variable negative charge above the ZPC. In contrast, the soils from the upper B horizons of the Dystrudepts and the E to B horizons of the Haplohumods were characterized by relatively high CEC/clay values, i.e., mostly above 15 cmolc kg−1 (at pH 5, 0.01 mol−1), and low (a + b)/2 values in the
221
Pedogenetic Acidification in Humid Asia E and Bs horizons of haplorthods and upper B horizons of dystrudepts under MSTR
CEC/clay at 0.01 M, pH 5 (cmolc kg–1)
40
Bw horizons of dystrudepts under TSTR
30 20
Bw and lower B horizons of dystrudepts and hapludands under MSTR
10 0
0
0.1
0.2
0.3
(a + b)/2
0.4
0.5
FIGURE 5.17 Relationship between the (a + b)/2 values and CEC/clay determined in 0.01 M electrolyte at pH 5. △ E or upper B horizon soils, ○ Bs horizon of Haplorthods, ▲ Bw and lower B horizons of Dystrudepts and Hapludands under MSTR, ● Bw and lower B horizons of Dystrudepts under TSTR.
range of 0.1 to 0.2. These findings indicate the predominance of a permanent negative charge, with only a small decrease in the CEC, even for a low pH value around 5. As reported in Sections 5.3 and 5.4, most of the amorphous and/or interlayered materials were lost in these soils during the process of pedogenetic acidification. A large part of the negative charge thus exposed could be occupied by exchangeable Al. In conclusion, the processes of pedogenetic acidification directly affected the charge characteristics of the forest soils as follows. During the formation of the B horizons, accumulation of amorphous compounds and/or interlayered Al hydroxides resulted in the development of a variable charge. Amorphous compounds, which were preferentially formed in the B horizon of soils with MSTR, contributed significantly to the development of variable charge characteristics. During extensive soil acidification, the E and upper B horizons of the forest soils occasionally lost amorphous and/or interlayered compounds. As a result, the permanent negative charge of the expansible 2:1 minerals, which can retain high amounts of monomeric Al, became predominant in these soils.
5.5.7 Conclusions of Sections 5.2 to 5.5 The soil solution study, which was carried out at 3 plots in Japan, revealed that protons were produced by the dissociation of organic acids and nitrification mainly in the O horizon, while being consumed by adsorption and decomposition of organic acids or nitrate uptake by vegetation in the deeper soil horizons. Cation excess uptake by vegetation was highest among the proton sources in the whole soil compartment and hence was responsible for pedogenetic soil acidification in the growth stage of forests. Pedogenetic soil acidification was thus closely related to biological processes; they commonly included cation leaching by proton generation through the dissociation of organic acids and nitrification and subsequent cation excess accumulation in wood in the growth stage of forests. Next, the fates of DOC and acidity (i.e., Al and
222
World Soil Resources and Food Security
protons) supplied from overlying horizons were traced based on the soil solution composition and titratable alkalinity and acidity in the soil profiles. Two processes were postulated for pedogenetic acidification, that is, eluvi-illuviation of inorganic Al followed by subsequent adsorption of DOC and comigration of Al and DOC in the form of organo-mineral complexes. Both processes were conspicuous in MSTR soils and significantly contributed to SOM storage in the subsoil layers. Pedogenetic acidification in forest soils with MSTR was characterized by an accumulation of acidity in the form of amorphous compounds and/or organo-mineral complexes in the B horizon. It seems, to some extent, similar to podzol formation, at least in terms of Al translocation. Amorphous Al hydroxides protect against further acidification through protonation and/or partial monomerization and can thus be regarded as a temporary storage of acid-neutralizing capacity of the soil, which would otherwise be leached out directly from the soil profile. In contrast, the acid-buffering reactions of TSTR soils seemed to occur, if at all, mostly at or near the soil surface and the contribution of the B-horizon soils was limited. Thus, the pedogenetic processes in MSTR and TSTR were different in terms of the dynamics of acidity and Al. The pedogenetic acidification-modified soil properties relate not only to the organomineral complexes, but also to crystalline illitic minerals. The analysis using XRD and RIP methodologies revealed that soils located in the upper horizons were subjected to more intensive weathering and the content of HIV decreased in contrast to the increase in vermiculite. An increase in the vermiculitic nature upward in a profile was consistent with both an increase in the Cs-fixed capacity and a decrease in the amount of hot-citrate-extracted Al. However, the behavior of RIP was unique in that it first rose in parallel with the Cs-fixed capacity toward the surface, but then Under thermic soil temperature regime Organo-mineral complexes are present in small amounts and usually decrease with depth.
After intensive acidification, expandable 2:1 minerals are dominated. Under mesic soil temperature regime High concentrations of organo-mineral complexes are present, composing an active ANC. The layer gradually moves to downward along with acidification.
FIGURE 5.18 Alteration of illitic minerals along with pedogenetic acidification in forest soils in Japan.
Pedogenetic Acidification in Humid Asia
223
declined in the upper layers of podzolic soils. This progressive decrease in RIP at the late stage of podzolization suggested the frayed edge site was more susceptible to the accumulative acid loads compared to the bulk of the vermiculitic charges. Last, the change in the charge characteristics of the soils well demonstrated these pedogenetic acidification processes, including both transition of crystalline clay minerals and amorphous and/or organo-mineral complexes. The main conclusions in these four sections are the unique alteration process of illitic minerals and associated organo-mineral complexes along with pedogenetic acidification in Japan, as illustrated in Figure 5.18.
5.6 F LUXES OF DOC UNDER TROPICAL FORESTS UNDER DIFFERENT GEOLOGICAL CONDITIONS IN EAST KALIMANTAN, INDONESIA In forest ecosystems, most of the organic matter supplied to the organic (O) horizon mineralizes to CO2, but a portion is leached as DOC, as soil water percolates [McDowell and Likens 1988; Zech and Guggenberger 1996]. DOC transported into the mineral soil horizons may be mineralized, leached, or adsorbed onto mineral surfaces. The DOC fluxes from the O horizon, as well as root litter, are an important C source for mineral soils, and therefore may contribute to SOM formation over a long timescale. In boreal and temperate forests, the importance of this DOC flux to the soil C cycle and SOM formation has been quantified [Michalzik et al. 2001; Kleja et al. 2008; Sanderman and Amundson 2008; Section 5.2 in this chapter]. On the other hand, in tropical forests, the role of DOC in the soil C cycle and its contribution to SOM formation have not been fully understood because of limited data [McDowell 1998; Tobón et al. 2004a; 2004b; Schwendenmann and Veldkamp 2005]. The aims of this study were to: (1) quantify the DOC fluxes under tropical forest serpentines and mudstones; (2) evaluate the role of these DOC fluxes in the soil C cycle and SOM formation; and (3) evaluate the influence of parent materials (serpentine, mudstone, and sandstone) on DOC fluxes.
5.6.1 Experimental Plots and Experimental Design Experiments were carried out in the natural secondary forest, slightly damaged by the fires in 1982–1983 and 1997–1998, in Bukit Soeharto (BS plot) from September 2004 to October 2005, the pristine forest in Bukit Bankirai (BB plot) from October 2005 to October 2006, and tropical secondary forests in Kuaro (KR1, KR2, and KR3 plots) from August 2006 to August 2007. All the plots were located in East Kalimantan Province, Indonesia (Figure 5.19). The parent materials of this area are largely sedimentary rocks, but there are patches of serpentine (ultramafic) intrusions. Soils at both BS and BB are derived from sedimentary rocks, whereas KR1, KR2, and KR3 were located along a traverse across serpentine-sedimentary rock (mudstone) (Figure 5.19). The site description is given in Table 5.13. At these plots, soil respiration rates were quantified once or twice per month in five replicates using a closed-chamber method together with the monitoring of soil
224
World Soil Resources and Food Security
E116º
E116º
E117º BB
Serpentine
Kuaro
KR3 KR2
S1º
Balikpapan
belt
0º
BS
KR1
20 km
S2º
FIGURE 5.19 Location of the experimental plots in East Kalimantan, Indonesia.
temperature at a depth of 5 cm and volumetric water contents of soils at depths of 5, 15, and 30 cm using data-loggers. Soil solutions were collected in five replicates using tension-free lysimeters beneath the O, A, and B1 horizons (0-, 5-, and 30-cm depths, respectively). Throughfall was collected using a precipitation collector in five replicates. These samples were collected once or twice per month for 1 year in each plot. Then, fluxes of DOC from each horizon were calculated by multiplying the water fluxes by the DOC concentrations in throughfall and soil solutions. The water fluxes of throughfall were measured using precipitation collectors, whereas those of soil water percolating at depths of 5, 15, and 30 cm were estimated by applying Darcy’s law to the unsaturated hydraulic conductivity and the gradient of the hydraulic heads at each depth, as described in the Appendix of Chapter 4 (4.A3). To collect litterfall, circular litter traps of 60-cm diameter were used. The fine root biomass, both in the O horizon and mineral soils, was estimated. The aboveground biomass was estimated by applying the diameters of stems at breast height (DBH), to the regression equations obtained by Yamakura et al. [1986]. The C and N contents, as well as phosphorus and Klason-lignin concentrations of the plant materials or foliar litters, were also determined [Allen et al. 1974].
5.6.2 Physicochemical Properties of Soils and C Stock in Soils and Ecosystems The data are presented in Tables 5.14 and 5.15. The soil pH values were highest in the KR1 soil from serpentine (6.2–6.4), followed by the KR2, and KR3 soils from mudstone (4.6–5.6 and 4.5–4.6, respectively). Soil pH at the BS and BB plots was consistently low. The contents of DCB-extractable Fe oxides were highest in KR1 (176–216 g kg−1), followed by KR2 (67–78 g kg−1), KR3 (30–38 g kg−1), BB (9–18 g kg−1), and BS (7–15 g kg−1). The total C contents were higher in the KR1 and KR2 soils (73 g kg−1)
East Kalimantan, Indonesia Coordinates Mean annual air temperature (°C) Mean annual precipitation (mm) Elevation (m) Soil typea Parent materials Vegetation
a
BS
BB
KR1
KR2
KR3
S00°51′, E117°06′ 27
S01°01′, E116°52′ 27
S01°51′, E116°02′ 27
S01°49′, E115°59′ 27
S01°49′, E115°56′ 27
2187
2427
2256
2256
2256
99 Typic Paleudults Sedimentary rocks
80 Typic Paleudults Sedimentary rocks Shorea leavis Dipterocarpus cornutus
204 Typic Paleudults Sedimentary rocks Serpentine Harpullia arborea Durio spp.
167 Typic Paleudults Sedimentary rocks
Shorea leavis Dipterocarpus cornutus
92 Rhodic Eutrudox Sedimentary rocks Serpentine Harpullia arborea Bauhinia purpurea
Pedogenetic Acidification in Humid Asia
TABLE 5.13 Site Description of Indonesian Plots
Harpullia arborea Artocarpus lanceolata Durio spp.
Soils were classified according to Soil Taxonomy [Soil Survey Staff 2006].
225
226
TABLE 5.14 Physicochemical Properties of Soils in Indonesian Plots
Depth Site
(H2O)
Particle Size Distribution
Exchangeable (KCl)
Total C
Total N
Bases
(g kg )
Al
CEC
Sand
(cmolc kg )
Silt
Clay
(%)
Feo
Alo
Ald
BS (Indonesia)
A BA B1 Bt
0–5 5–25 25–40 40+
4.0 3.8 4.0 4.3
3.9 3.8 3.8 3.8
22.9 4.2 3.5 2.5
1.6 0.5 0.5 0.4
2.2 0.8 0.8 1.0
3.0 3.9 4.8 7.0
8.5 6.2 5.0 5.0
52 49 43 34
25 27 30 35
23 24 27 31
3.6 1.4 0.9 0.6
0.7 0.7 0.8 1.0
6.6 9.1 11.3 14.6
1.0 1.4 1.7 2.2
309 413 502 643
BB (Indonesia)
A BA B1 Bt BC
0–5 5–20 20–37 37–70 70+
4.2 4.1 4.1 4.1 4.2
3.4 3.7 3.8 3.8 3.8
36.1 10.7 7.1 3.7 3.3
2.3 0.9 0.6 0.4 0.4
1.2 0.8 1.1 0.8 0.9
6.7 5.5 5.7 4.6 6.7
13.0 8.3 8.5 9.0 13.6
42 45 43 39 34
31 31 33 34 25
27 24 24 27 40
2.5 2.8 1.9 1.6 1.8
1.2 1.3 1.3 1.2 1.2
9.0 10.7 11.4 15.0 18.2
1.7 1.9 2.1 2.6 3.1
400 369 440 531 630
−1
(g kg )
ANC(s)
m
(cm)
−1
(g kg )
Fed
Horizon
−1
−1
(cmolc kg−1)
World Soil Resources and Food Security
pH
A1 A2 B1 B2 B3
0–5 5–20 20–35 35–50 50–67+
6.3 6.2 6.4 6.4 6.4
5.8 5.7 6.1 5.8 6.2
72.7 23.8 7.5 5.7 4.8
6.4 2.6 0.8 0.7 0.6
19.1 5.0 2.0 1.8 2.7
0.0 0.0 0.0 0.0 0.0
24.8 11.4 7.2 7.0 7.7
6 14 19 16 38
39 45 37 37 33
55 41 44 47 28
3.5 4.4 6.1 6.2 6.7
1.4 2.2 1.8 1.7 1.7
175.7 196.2 208.4 217.1 215.9
8.5 9.9 10.7 11.2 10.9
1358 1608 1817 1853 1725
KR2 (Indonesia)
A BA Bt1 Bt2
0–5 5–20 20–45 45–70
5.6 4.6 4.8 5.1
5.2 3.9 4.1 4.2
73.1 24.9 14.2 10.0
5.5 2.9 1.8 1.4
14.5 1.4 1.0 0.8
0.2 2.9 2.3 1.4
19.8 13.7 17.0 24.8
6 5 4 7
15 18 10 11
79 77 86 82
4.0 4.5 2.5 1.9
5.0 2.8 2.6 2.5
66.6 70.9 71.3 78.1
12.0 10.9 10.7 11.8
1837 2012 2011 1996
KR3 (Indonesia)
A BA Bt1 Bt2 Bt3
0–4 4–15 15–30 30–45 45–50+
4.5 4.5 4.5 4.6 4.6
3.7 3.7 3.7 3.8 3.8
38.7 17.9 11.1 7.4 7.5
3.5 1.7 1.2 0.9 1.0
1.8 1.0 1.2 1.4 0.9
4.9 4.9 6.3 7.0 7.7
20.8 17.2 19.7 23.3 27.1
28 29 28 25 25
24 15 15 15 9
48 56 57 60 66
3.3 2.4 1.9 1.3 1.4
2.2 1.9 1.7 1.7 1.7
29.7 31.6 34.0 36.2 37.5
4.9 5.1 5.1 5.3 5.7
1163 1245 1293 1430 1459
Pedogenetic Acidification in Humid Asia
KR1 (Indonesia)
227
228
TABLE 5.15 Stock and Annual Flow of Carbon in the Indonesian Plots BS
C flow (Mg C ha–1 yr –1) Soil respiration Root respiration Decomposition of soil organic matter Litterfall Wood increment
292.6
(−)
346.0
KR1
(−)
134.2
KR2
(−)
259.8
KR3
(−)
282.1
(−)
1.0 1.2 2.1
(0.4) (0.3) (0.4)
2.3 1.3 2.0
(0.3) (0.2) (0.5)
– 1.5 3.1
(0.2) (0.5)
– 0.6 0.4
(0.2) (0.2)
0.3 0.4 1.1
(0.1) (0.1) (0.2)
3.5 26.6
(0.3) (0.8)
4.5 51.4
(1.1) (4.6)
4.1 65.8
(1.1) (2.2)
2.9 74.7
(0.4) (3.2)
3.6 55.0
(0.2) (2.1)
7.5 2.0 5.4
(0.5) (0.2) (0.5)
6.7 2.4 4.2
(0.4) (0.2) (0.5)
6.4 2.2 4.2
(0.3) (0.4) (0.3)
6.9 2.2 4.7
(0.3) (0.4) (0.2)
8.0 3.8 4.2
(0.4) (0.4) (0.1)
4.1 10.6
(0.5) (–)
3.6 11.1
(0.1) (–)
4.8 7.2
(0.3) (–)
4.0 9.8
(0.7) (–)
4.5 5.1
(0.3) (–)
Note: The figures in parentheses represent standard errors. a Organic carbon in soil at 0- to 30-cm depths was counted.
World Soil Resources and Food Security
C stock (Mg C ha−1) Aboveground biomass Fine root biomass O horizon A horizona B horizona Soil organic matter O horizon Mineral soil horizonsa
BB
229
Pedogenetic Acidification in Humid Asia
than the other soils (23–39 g kg−1). The O horizons in KR1 and KR2 consisted of the Oi layer only, while in the other acidic soils (KR3, BS, and BB), there were Oea and Oi. The aboveground biomass was lower in KR1 (134 Mg C ha−1) than in the others (260–346 Mg C ha−1). The C stock in the O horizon was consistently low at all plots (2.6–4.5 Mg C ha−1). The C stock in the mineral soil was highest in KR2 (74.7 Mg C ha−1). The higher aboveground biomass and the lower C stock in the O horizon and C flux (kg C ha–1 month–1) 1000
C flux (kg C ha–1 month–1) 1000
BS
800
800
600
600
400
400
200
200
0 Sep-04 Dec-04 Feb-05 Apr-05 Jul-05 Sep-05 Nov-05
Date
C flux (kg C ha–1 month–1) 1000
0 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07
Date
C flux (kg C ha–1 month–1) 1000
BB
800
800
600
600
400
400
200
200
0 Oct-05 Dec-05 Feb-06 Apr-06 Jun-06 Aug-06 Oct-06
Date
Whole soil respiration OM decomposition Root respiration
KR1
KR2
0 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07
Date
C flux (kg C ha–1 month–1) 1000
KR3
800 600 400 200 0 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07 Date
FIGURE 5.20 The seasonal fluctuations of soil respiration, OM decomposition, and root respiration. Bars indicate standard errors (n = 5).
230
Water Flux
DOC
HCO3−
Orgn− a
NO3−
Cl− + SO24−
H+
NH+4
Fe2+
Aln+ b
(mmolc L−1)
Si
Horizon
(mm)
pH
(mg C L−1)
BS
TF O A2 BA-B1
2031 1619 1196 545
5.23 4.44 4.22 4.39
9.0 34.7 17.2 9.9
0.009 0.001 0.000 0.003
0.101 0.257 0.196 0.119
0.036 0.038 0.044 0.032
0.104 0.136 0.100 0.116
0.006 0.037 0.061 0.041
0.038 0.088 0.035 0.018
0.197 0.256 0.197 0.176
0.003 0.019 0.009 0.005
0.007 0.040 0.038 0.027
0.017 0.070 0.068 0.051
BB
TFc O A BA-B1
2068 1914 1639 893
5.54 4.08 3.97 4.07
4.7 24.6 19.1 6.0
0.003 0.000 0.001 0.000
0.086 0.261 0.211 0.132
0.024 0.137 0.161 0.103
0.098 0.174 0.161 0.170
0.005 0.118 0.131 0.087
0.030 0.106 0.057 0.030
0.171 0.283 0.269 0.219
0.001 0.016 0.013 0.009
0.004 0.053 0.066 0.061
0.008 0.069 0.075 0.072
c
(mmolc L–1)
Na+ + K+ + Mg2+ + Ca2+
(mmol L−1)
World Soil Resources and Food Security
TABLE 5.16 Water Flux and Annual Volume-Weighed Mean Concentrations of Ions in Throughfall and Soil Solution
TFc O A1 A2
2240 1907 1071 566
6.34 6.12 5.74 6.20
7.3 31.6 13.4 5.4
0.047 0.082 0.030 0.160
0.143 0.272 0.165 0.098
0.020 0.098 0.466 0.308
0.067 0.230 0.169 0.192
0.000 0.001 0.002 0.001
0.017 0.028 0.042 0.031
0.257 0.643 0.770 0.691
0.000 0.004 0.001 0.000
0.002 0.003 0.004 0.000
0.015 0.140 0.275 0.627
KR2
TFc O A BA-Bt
2211 1922 967 553
6.18 6.52 5.89 5.14
6.2 8.6 5.7 1.7
0.048 0.183 0.026 0.002
0.070 0.120 0.053 0.028
0.027 0.091 0.071 0.062
0.069 0.113 0.040 0.041
0.001 0.000 0.001 0.007
0.012 0.017 0.008 0.010
0.198 0.470 0.167 0.098
0.000 0.000 0.003 0.002
0.003 0.006 0.009 0.008
0.018 0.021 0.046 0.065
KR3
TFc O A BA-Bt1
2205 1594 645 416
6.07 5.57 4.97 5.11
5.7 16.4 10.1 2.8
0.036 0.039 0.009 0.000
0.075 0.216 0.180 0.036
0.013 0.035 0.060 0.019
0.060 0.135 0.101 0.040
0.001 0.003 0.011 0.008
0.013 0.017 0.033 0.008
0.168 0.385 0.284 0.067
0.000 0.004 0.003 0.001
0.002 0.016 0.016 0.011
0.015 0.075 0.055 0.027
a b c
Pedogenetic Acidification in Humid Asia
KR1
Orgn– represents anion deficit, the negative charge of organic acids. The total charge equivalent of Al ions was calculated as the equivalent sum of Al3+, AlOH2+, and Al(OH)+2 . TF represents throughfall.
231
232
World Soil Resources and Food Security
mineral soil, compared to the temperate forests (Table 5.2), are consistent with previous reports [Nakane 1980].
5.6.3 Organic Matter Decomposition The seasonal fluctuations of soil respiration rates at all plots are presented in Figure 5.20. Soil temperature varied little over the year at all plots, while the volumetric water content in the soils increased gradually after the dry period (September– November). The rates of soil respiration, organic matter decomposition, and root respiration varied over the year. However, these fluctuations were independent of soil temperature and moisture. The annual rates of soil respiration, organic matter decomposition, and root respiration, which were calculated using the average rates of CO2 emission measured (Figure 5.20), are presented in Table 5.15. The annual rates of organic matter decomposition were similar (4.2–5.4 Mg C ha−1 yr−1) among all plots (Table 5.15). These values are similar to the lowest values reported for other tropical rain forests (4.8–8.9 Mg C ha−1 yr−1), but higher than in temperate forests [Bond-Lamberty et al. 2004]. Organic matter decomposition was almost balanced by C inputs via litterfall (4.0–4.8 Mg C ha−1 yr−1) (Table 5.15).
5.6.4 C oncentrations and Fluxes of DOC in Throughfall and Soil Solution Soil solutions were moderately acidic to neutral (5.0–6.5) in all plots (Table 5.16). Along the gradient of soil pH, soil solution pH was lowest in KR3 (5.0–5.6), followed by KR2 (5.2–6.5), and KR1 (5.7–6.2). The DOC concentrations in soil solutions were highest in the O horizon, and there was a decrease with increasing depth in each plot (Figure 5.21). The DOC concentrations in the O horizon solutions varied seasonally in KR1 (19.9–70.5 mg C L−1) and KR3 (11.4–45.9 mg C L−1) (Figure 5.21). The DOC concentrations were highest during the dry periods, followed by a gradual decrease during the rainy season and in the O horizon solution. They were consistently low in KR2 (4.6–16.8 mg C L−1) (Figure 5.21). The DOC fluxes in throughfall and B1 horizon (30-cm depth) vary within a narrow range of 97–182 and 10–54 kg C ha−1 yr−1, respectively, at all plots (Figure 5.22). The DOC (i.e., Orgn−) fluxes in the soil profiles increased markedly in the O horizon, and there was a decrease with increasing depth in each plot. The DOC fluxes from the O horizon were highest at all the plots. However, these fluxes varied between 166–603 kg C ha−1 yr−1. The DOC fluxes from the O horizon corresponded to 13.8%, 13.0%, 12.6%, 4.1%, and 5.8% of C input in BS, BB, KR1, KR2, and KR3, respectively, and to 10.3%, 11.1%, 14.3%, 3.5%, and 6.2% of organic matter decomposition, respectively. The concentrations and fluxes of DOC were the largest in the O horizon (Figure 5.22), and are consistent with previous reports [Michalzik et al. 2001]. The annual volume-weighted mean concentrations and fluxes of DOC in the O horizon solutions vary widely between the five plots studied. The DOC fluxes from the O horizon in BS and KR1 are comparable with the highest values reported for tropical forests, 166–603 kg C ha−1 yr−1 in the present study, and those in temperate forests, 100–482
233
Pedogenetic Acidification in Humid Asia DOC concentration (mg C L–1) 80
BS
DOC concentration (mg C L–1) 80
60
60
40
40
20
20
0
Sep-04 Nov-04 Jan-05 Mar-05 May-05 Jul-05 Sep-05 Nov-05
Date
DOC concentration (mg C L–1) 80
BB
0
Aug-06 Oct-06 Dec-06 Feb-07 Apr-07 Jun-07 Aug-07
Date
DOC concentration (mg C L–1) 80 KR2
60
60
40
40
20
20
0
Oct-06 Dec-06 Feb-06 Apr-06 Jun-06 Aug-06 Oct-06
Date
KR1
0
Aug-06 Oct-06 Dec-06 Feb-07 Apr-07 Jun-07 Aug-07
Date
DOC concentration (mg C L–1) Throughfall O horizon A horizon B1 horizon
80
KR3
60 40 20 0
Aug-06 Oct-06 Dec-06 Feb-07 Apr-07 Jun-07 Aug-07
Date
FIGURE 5.21 The seasonal fluctuation of concentrations of DOC in throughfall and soil solution. Bars indicate standard errors (n = 5).
kg C ha−1 yr−1 from Michalzik et al. [2001] and Yano et al. [2004] and 690–836 kg C ha−1 yr−1 from Moore [1989] and Moore and Jackson [1989]. In contrast, DOC fluxes in KR2 are close to the lowest reported values. Among the Ultisol soils from sedimentary rocks in our study, DOC fluxes from the O horizons vary between 166–261 kg C ha−1 yr−1 for the Ultisol soils from mudstone (KR2 and KR3) to 470–562 kg C ha−1 yr−1 for those from sandstone (BS and BB). Comparing the DOC fluxes from the O horizon on a basis of C input, the DOC fluxes correspond to 4.1%–13.8% of C input (Figure 5.22). The magnitude of DOC
234
World Soil Resources and Food Security The annual fluxes of C (kg C ha–1 yr–1) The stock of C (kg C ha–1) KR1 KR2 Aboveground biomass 134 × 103 260 × 103
KR3
BS
BB
282 × 103
293 × 103
346 × 103
Throughfall litterfall 4783
164
136
OM decomposition 4231 O horizon Mineral soila Soil solution
4 090 603
65.4 × 103 31
Serpentine
4008
4475
126
4701
4236
2 881
3 605
74.7 × 103
55.0 × 103
166
4084
182
261
10
12 Mudstone
3618
97
4249
5446
3 454
4 493
562
470
26.6 × 103
Sedimentary rocks
51.4 × 103
54
54 Sandstone Acidic
FIGURE 5.22 The C stock and the annual C fluxes via litterfall, organic matter (OM decomposition, throughfall, and soil solution) in East Kalimantan, Indonesia. aThe C stock in soils at the depths of 0–30 cm was counted.
production in the O horizon is variable among the soils studied. According to published data and those from our study, the proportion of DOC flux relative to C input increases with decreasing soil pH (Figure 5.23), with the exception of KR1. The substantial DOC translocation from the O horizon may be common to the highly acidic soils (soil pH < 4.3; Spodosols and acidic Ultisols) under a humid climate. This is consistent with the significant contribution of DOC, the sources of organic acids, to podzolization and soil acidification [Ugolini and Dahlgren 1987; Do Nascimento et al. 2008]. Lower soil pH may be favorable for substantial DOC production due to: (1) enhanced litter solubilization by fungal activities [Kalbitz et al. 2000] and inhibited mineralization [Saggar et al. 1999; Kemmitt et al. 2006]; and (2) recalcitrance of DOC produced from litter rich in polyphenol and lignin [Kalbitz et al. 2003, 2006]. This is partly supported by our study, in which DOC production from lignin-rich organic materials in KR3 (42%) was higher than in KR2 (28%). However, the variation of DOC concentration in the O horizons between KR1, KR2, and KR3 could not be entirely accounted for by lignin concentration in foliar litter. In our study in East Kalimantan, Indonesia, the local-scale variation of DOC concentration in the O horizons among five soils could be accounted for by litter P availability (Figure 5.24). The DOC concentrations in the O horizon solution increased with decreasing P concentration in the foliar litter. This relationship also is confirmed for data in Amazonian tropical rain forest soils from different parent materials (Figure 5.24) [Tobón et al. 2004a, 2004b]. The P concentration, as well as the lignin concentration, in foliar litter is an important factor regulating DOC
235
Pedogenetic Acidification in Humid Asia Proportion of DOC flux relative to C input (%)
Published data Sandstone Mudstone Serpentine
30 25
r = –0.63 p < 0.01, n = 27
20 15
BB
BS
10
KR1
5 0
KR3 3
KR2
4
5
6
Soil pH
FIGURE 5.23 Relationship between soil pH and proportion of DOC flux relative to C input in the O horizon. (Data sources include Moore, T.R., Water Resource Res., 25, 1321–1330, 1989; Moore, T.R., and R.J. Jackson, Water Resource Res., 25, 1331–1339, 1989; Yavitt, J.B. and T.J. Fahey, J. Ecol., 74, 525–545, 1986.)
DOC concentration (mg C L–1) 40
Published data Sandstone Mudstone Serpentine
BS
30
KR1
BB
20 KR3 10 KR2
Columbia 0 0.00
0.02
0.04
0.06
0.08
0.10
Foliar P (%)
FIGURE 5.24 Relationship between P concentration in foliar litter and the concentration of DOC in the O horizon solution. (Data sources include Qualls, R.G. et al., Ecology 72, 254–266, 1991; Johnson, C.E. et al., Ecosystems, 3, 159–184, 2000; Tobón, C. et al., Biogeochem., 69, 315–339, 2004; Tobón, C. et al., Biogeochem., 70, 1–25, 2004; Kleber, M. et al., Geoderma, 138, 1–11, 2007; Raich, J.W. et al., For. Ecol. Manag., 39, 128–135, 2007.)
236
World Soil Resources and Food Security
biodegradability [Wieder et al. 2008]. The DOC mineralization (consumption) rates could be constrained by P limitation in tropical forest soils [Cleveland et al. 2002]. Consequently, litter P availability may determine whether DOC released from litter is rapidly mineralized or leached from the O horizon. Litter P availability is strongly constrained by the P content in parent materials in tropical soils [Kitayama et al. 2000]. In our study, the lower litter P availability derived from P-poor parent materials (serpentine) or acidic conditions (sandstone) is considered to constrain DOC consumption and result in substantial DOC fluxes from the O horizon of KR1 and the highly acidic Ultisol soils from sandstone.
5.6.5 Influence of Parent Rocks on DOC Dynamics The relative importance of parent materials on DOC dynamics is poorly understood, compared to the importance of climate and vegetation [Nambu et al. 2008]. Judging from the fact that C fluxes of organic matter production and decomposition in the five tropical forests studied are similarly high (Figure 5.22), organic matter production and decomposition are considered to largely depend on climate and vegetation, rather than parent material. Similarly, the DOC fluxes from the O horizon are considered to depend primarily on climate [Zech et al. 1997; Kalbitz et al. 2000], and they are reported to increase with increasing amounts of precipitation, substrate supply, and its decomposition at regional to global scales [Kleja et al. 2008]. However, in our study, the DOC fluxes from the O horizon varied among the soils from different parent materials despite similar amounts of precipitation and C input (Figure 5.22). These local-scale variations of DOC fluxes also are reported between the clayey and sandy soils under similar climate and vegetation conditions [Don and Schulze 2008]. Our data indicate the importance of parent materials in regulating the DOC fluxes through their effects on soil pH and litter P availability (Figures 5.23 and 5.24), although the effects of parent materials on the DOC fluxes are not unclear under different climate and vegetation conditions [Dosskey and Bertsch 1997]. To date, parent materials generally have been considered to influence stabilization of organic matter (e.g., DOC adsorption) through effects on soil texture and clay mineralogy [Sollins et al. 1996]. In addition, our data suggest that parent materials play an important role in the translocation, as well as stabilization, of organic matter at a local scale among tropical forest soils.
5.7 C ONTRIBUTION OF DIFFERENT PROTON SOURCES TO SOIL ACIDIFICATION UNDER TROPICAL FORESTS UNDER DIFFERENT GEOLOGICAL CONDITIONS IN EAST KALIMANTAN, INDONESIA In the humid tropics, the vast majority of soils are highly weathered and acidified because of intensive leaching over long periods of time [Eyre 1963]. Soil-acidifying processes vary with parent materials and vegetation, as well as climate (leaching intensity) [Ugolini and Sletten 1991]. Parent materials influence soil-acidifying processes through effects on soil texture, acidity, and the acid-neutralizing capacity
Pedogenetic Acidification in Humid Asia
237
(ANC) of soils [Ugolini and Sletten 1991]. For example, the soil-acidifying processes (e.g., mobility of organic acids) vary depending on parent material texture and they appear to determine the dominant pedogenetic processes (ferralitization vs. podzolization), which occurred in a sequence of Oxisol–Ultisol–Spodosol under tropical forests [Do Nascimento et al. 2008]. Parent materials determine the distribution of soil types (Ultisols and Oxisols) in East Kalimantan [Petersen 1991]. To understand the dominant pedogenetic processes in this region, the effects of parent materials on soil-acidifying processes and their impacts on weathering reactions need to be clarified for different parent materials. The dominant soil-acidifying processes can be identified by quantifying the relative importance of the individual proton sources to soil acidification using proton budgets in soil vegetation systems and including solute leaching and vegetation uptake [van Breemen et al. 1983, 1984]. To date, only a few studies have dealt with the dominant acidifying processes in tropical soils [Johnson et al. 1983; van Breemen et al. 1984]. The objectives of our study were to: (1) analyze the dominant acidifying processes by quantifying proton budgets in five Indonesian soils derived from different parent materials; and (2) evaluate the effects of parent materials on pedogenetic soil acidification. The analysis was carried out using the same dataset as given in the previous section (Section 5.6).
5.7.1 Chemical Properties of the Soils Studied General physicochemical properties of the soils were introduced in the previous section. The ANC of the soils increased with soil pH and clay content, varying from 309–643 cmolc kg−1 in the BS and BB soils to 1163–2012 cmolc kg−1 in the KR soils (Table 5.14). The higher ANC values are mainly contributed by Mg and Fe in the KR1 soil, but by Al in the KR2 and KR3 soils. The O horizons had only an Oi layer in the KR1 and KR2 soils, while the acidic BS, BB, and KR3 soils (pH < 4.5) had an Oea layer as well as an Oi layer. In the BB and BS soils, the O horizons are acidic (pH 4.5 to 5.0), consistent with lower contents of basic cations (17–63 cmolc kg−1). The higher contents of basic cations in the O horizons of the KR soils (87–129 cmolc kg−1) are probably a consequence of parent materials rich in basic cations.
5.7.2 Soil Solution Composition Soil solutions were strongly acidic in the BS and BB soils (pH 4.0–4.4), while they were moderately acidic at the KR sites (pH 5.0–6.2) (Table 5.16). DOC concentrations in the O horizon solution were higher in the BS, BB, and KR1 soils (24.6– 34.7 mg C L−1) compared to the KR2 and KR3 soils (8.6–16.4 mg C L−1). Organic acids were the dominant anions in the O horizon solution in all plots (0.22–0.27 mmolc L−1) except for the KR2 soil. Concentrations of organic acids and DOC in the soil solution decreased with depth. From linear regression analysis between the concentrations of DOC and organic acids in soil solutions, the negative charge per 1 mole of DOC (0.09–0.17 molc) corresponds to one dissociated acidic functional group for 5.9–11.5 C atoms. The high DOC to charge ratios in the soil solution suggests the
238
World Soil Resources and Food Security
presence of high molecular weight fulvic acids, which contain 7 C atoms for each acidic functional group [Thurman 1985]. Bicarbonate is present in moderately acidic soil solutions of the KR sites, while it is negligible in acidic soil solutions of the BS and BB sites. In the O horizon of the KR2 soil, bicarbonate was the dominant anion (0.18 mmolc L−1) owing to the relatively high solution pH (6.5), while concentrations decreased with depth (Table 5.16). Nitrate concentrations were low (0.02–0.16 mmolc L−1) at all plots except for the A1 and A2 horizons of the KR1 soil (0.31–0.47 mmolc L−1), where the understory vegetation was the nitrogen-fixing Bauhinia purpurea. The major accompanying cations were K+, Mg2+, and Ca2+ in the KR1, KR2, and KR3 soils, while they were H+, NH +4 , and Aln+, as well as basic cations, in the BS and BB soils. The highest concentrations of Si in the soil solutions were measured in the KR1 soil (0.14–0.63 mmol L−1) compared to the other soils from sedimentary rocks (0.02–0.08 mmol L−1).
5.7.3 Fluxes of Ions in Solute Leaching and Vegetation Uptake and Proton Budgets in Soils Cation contents exceeded anion contents in litter and wood materials at all plots (Table 5.17). The excess cation charge was compensated for by the net proton load to the soil as NPGBio. NPGBio in each of the soil horizons was calculated by distributing it based on the distribution of the fine root biomass in the soil profiles (Table 5.15), according to Shibata et al. [1998]. Based on the fluxes of solutes entering and leaving the soil horizon compartment (Figure 5.25) and vegetation uptake (Table 5.17; NPGBio) in each of the soil horizons, net proton generation and soil acidification rates were calculated based on the proton budget theory (Figure 5.26). In the entire soil profiles, NPGBio was highest among the proton sources at all plots (Figure 5.26). NPGBio was present mainly in the A and B horizons of soils at all plots (1.7–10.8 kmolc ha−1 yr−1), while it was also present in the O horizons of the KR3, BS, and BB soils (1.5–3.1 kmolc ha−1 yr−1) (Figure 5.26). In the O horizons, proton sources include NPGOrg, NPGCar, and NPGNtr, as well as NPGBio (Figure 5.26). NPGOrg is the largest proton source in the O horizons (1.8–3.2 kmolc ha−1 yr−1) except for the KR2 soil, where NPGCar is higher than NPGOrg (0.8 kmolc ha−1 yr−1). In the moderately acidic O horizons of the KR1 and KR2 soils, protons were produced by the dissociation of carbonic acid (NPGCar: 0.5–2.5 kmolc ha−1 yr−1) (Figure 5.26). In the A and B horizons, protons were consumed by the mineralization and adsorption of organic acids and by the protonation of carbonic acid (Figure 5.26). Protons were produced by nitrification in the O horizons of the KR1, KR2, KR3, and BB soils (NPGNtr: 0.3–1.3 kmolc ha−1 yr−1), while they were consumed by nitrate uptake by vegetation or microorganisms in the A and B horizons. In the acidic O horizons of the BS and BB soils, an increase in the NH +4 flux (Figure 5.25; 0.6 kmolc ha−1 yr−1) indicated that protons were mainly consumed by mineralization of organic + N to NH 4 (NPGNtr: –0.7 kmolc ha–1 yr–1) (Figure 5.26). In the A horizons of the BS and BB soils, protons were released owing to the excess uptake of NH +4 over NO3− by biomass or adsorption of NH +4 on clays (NPGNtr: 0.9–1.1 kmolc ha−1 yr−1) (Figure 5.26). Exceptionally, in the A1 horizon of the KR1 soil, where the understory
OM Production
Na
K
Ca
(Mg C ha−1 yr−1)
Mg
Fe
Al
Cl
S
P
(Cation)bio
(kg ha−1 yr−1)
(Anion)bio
NPGBio
(kmolc ha−1 yr−1)
BS
Wood increment Litterfall
10.6 4.1
14.7 3.6
33.0 45.2
29.3 35.7
15.8 18.5
10.0 1.7
0.5 3.3
1.0 0.7
8.6 5.0
2.2 2.5
4.66 5.05
0.64 0.41
4.02 4.63
BB
Wood increment Litterfall
11.1 3.6
15.5 5.4
34.8 22.9
30.9 15.0
16.7 17.1
1.9 1.9
5.3 4.3
1.1 0.7
9.1 5.4
2.4 3.2
5.13 3.52
0.67 0.46
4.46 3.06
KR1
Wood increment Litterfall
7.2 5.7
3.3 5.3
52.6 63.7
59.4 82.3
8.7 26.4
1.6 0.2
0.8 2.1
4.4 1.1
3.6 7.5
5.6 4.1
5.32 8.37
0.53 0.63
4.79 7.74
KR2
Wood increment Litterfall
9.8 3.6
3.4 2.5
55.4 31.1
132.5 81.1
9.8 33.7
0.5 0.9
1.2 18.9
1.9 0.5
5.4 9.1
5.4 5.0
9.13 9.85
0.57 0.75
8.57 9.10
KR3
Wood increment Litterfall
5.1 4.5
3.0 4.7
23.3 27.2
42.2 57.3
2.9 33.3
0.9 2.3
2.1 1.9
1.0 1.0
3.4 8.3
2.3 4.4
3.34 6.80
0.31 0.69
3.02 6.11
Pedogenetic Acidification in Humid Asia
TABLE 5.17 Uptake of Cations and Anions by Vegetation
239
240
World Soil Resources and Food Security Horizon
SO42–
TF O
Orgn–
Horizon HCO3– Na+ Na+ 4
Cl–
HCO3– Na+ Mg2+Ca2+ Na+ SO42– NO3– 4
TF
K+
Mg2+
O
Ca2+
Orgn–
K+
Cl–
Aln+ Fe3+
NO3–
A1
A
A2 –15
BA, B1
KR1 –10
–5
0
5
10
–15
15
(kmolc ha–1 yr–1) Fluxes of cations (+) and anions (–)
Horizon
–10
–5
0
5
10
15
(kmolc ha–1 yr–1) Fluxes of cations (+) and anions (–)
Horizon + – SO42– NO3 Cl– Na Na4+
TF
Orgn–
O
BB
H+
SO42–
TF
2+ Ca2+ + HCO– 3 K Mg
Na+ Mg2+ Ca2+
Orgn–
O
NO3– Cl– H+ Na+ K+ 4
Aln+ Fe3+
A
A
KR2
BA, Bt –15
–10
–5
0
5
10
15
(kmolc ha–1 yr–1) Fluxes of cations (+) and anions (–)
Horizon
BS
BA, B1 –15
–10
–5
0
5
10
15
(kmolc ha–1 yr–1) Fluxes of cations (+) and anions (–)
HCO– 3 – Na+ Na4+ SO42– NO3
TF
Aln+ Orgn–
O
Cl– K+ Mg2+ Ca2+
A BA, Bt
KR3
–15
–10
–5
0
5
10
15
(kmolc ha–1 yr–1) Fluxes of cations (+) and anions (–)
FIGURE 5.25 Fluxes of solutes at each horizon. TF represents throughfall. O, A, A1, A2, BA, B1, Bt1, and Bt represent soil horizons.
vegetation was the nitrogen-fixing Bauhinia purpurea, protons were produced by nitrification (NPGNtr: 3.2 kmolc ha−1 yr−1) (Table 5.16). In the O horizons of the KR soils, acid loads contributed mainly by NPGNtr, NPGOrg, and NPGCar (3.5–4.3 kmolc ha−1 yr−1) was completely neutralized by basic cations. On the other hand, in the O horizons of the BS and BB soils, the intensive
241
Pedogenetic Acidification in Humid Asia Horizon
Horizon
O
O
KR1
A1
A
A2
BA, B1
Total
Total
–10
–5
0
5
10
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–) O
KR2
–10
O A
BA, Bt
BA, B1
Total
Total –5
0
5
10
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–) O
0
5
10
–10
–5
0
5
10
BB
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
KR3 (H+)in – (H+)out
A
NPGNtr
NPGOrg NPGBio
BA, Bt1
NPGCar ∆ANC
Total –10
–5
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
A1
–10
BS
–5
0
5
10
(kmolc ha–1 yr–1) Net proton generation (+) or consumption (–)
FIGURE 5.26 Net proton generation and consumption in the soil profiles. TF represents throughfall. O, A, A1, A2, BA, B1, Bt1, and Bt represent soil horizons.
acid loads contributed mainly by NPGOrg and NPGBio (3.2–7.0 kmolc ha−1 yr−1) were largely neutralized by basic cations but a portion of protons were transported downward. The protons transported from the O horizon ((H+)in − (H+)out: 0.5–1.4 kmolc ha−1 yr−1) are neutralized in the B horizons of the BS and BB soils (Figure 5.26) or are leached further downward.
242
World Soil Resources and Food Security
5.7.4 Dominant Soil Acidification Processes in Tropical Forests NPGBio is a dominant proton source in all the soil profiles (Figure 5.26). The soil acidification rates in tropical forests are higher than those for temperate forests (0.1–4.6 kmolc ha−1 yr−1 from Bredemeier et al. [1990], Binkley [1992], Shibata et al. [2001] and Section 5.2) (Figure 5.26). The higher acid loads in tropical regions is considered to be caused by higher biomass production and resulting higher NPGBio in tropical regions (Tables 5.15 and 5.17). Since NPGBio attributable to litter production would be neutralized by cations released from the fallen litter, soil acidification is mainly caused by excess cation accumulation in wood (3.0–8.6 kmolc ha−1 yr−1) during forest growth. Considering the complete biomass turnover of wood (cation release from coarse woody debris) in a steady state forest, NPGBio attributable to wood production might also be a neutral process for proton budgets on pedogenetic timescales. In our study, net-leaching losses of cations are typically small on an annual basis (Figure 5.25), but could be significant over pedogenetic timescales. Although the impacts of these processes on pedogenetic soil acidification are difficult to quantify and remain to be studied, our data quantitatively support the idea that higher rates of cation cycling through biomass and soils result in consistently high acid loads to soils under tropical forests (Figure 5.26) compared to those under temperate forests (Figure 5.3).
5.7.5 Proton Generation and Consumption in Soil Profiles In all the soil profiles, the contributions of NPGNtr, NPGCar, and NPGOrg to soil acidification are minor (Figure 5.26). This is consistent with the fact that complete cycles of C and N are balanced with no net proton fluxes in forest ecosystems [Binkley and Richter 1987]. However, translocation of the temporary acids (carbonic, organic, and nitric acids), as well as distribution of root biomass, contributed to heterogeneity of proton generation and consumption throughout the soil profiles, which varied from soil to soil (Figure 5.26). Organic acid dissociation is a common proton-generating process in the O horizons of the soils studied. DOC-associated proton generation accounts for 18%–77% of total proton generation in the O horizons (Figure 5.26). The large contribution of organic acids to soil acidification in the KR1, BS, and BB soils arises from the substantial fluxes of DOC production in the O horizon, which in turn is primarily caused by the greater fluxes of precipitation and C input and the quality of the foliar litter. On the other hand, carbonic acid dissociation is also a proton-generating process in the less acidic O horizons of the KR1 and KR2 soils (Table 5.16; pH > 5) because of their weakly acidic nature. This is consistent with substantial proton generation by carbonic acid dissociation in soils at neutral pH reported by Johnson et al. [1983], van Breemen et al. [1984], and Gower et al. [1995]. Although NPGCar associated with active root and microbial respiration has generally been recognized as a dominant acidifying process in tropical regions, the process is dominant only in moderately acidic and neutral soils.
Pedogenetic Acidification in Humid Asia
243
Proton generation by nitrification is generally the dominant process involved in NPGNtr in the O horizons, while proton consumption by mineralization of organic N + to NH 4 is also involved in NPGNtr in the highly acidic O horizons of the BS and BB soils (Figure 5.26). These differences are dependent on the balance between mineralization, nitrification, and NH +4 and NO3− uptake by vegetation and microorganisms (Figure 5.26). NPGBio is the dominant proton-generating process in the mineral soil horizons of the KR1 and KR2 soils, while NPGBio is present in both the organic and mineral horizons of the KR3, BS, and BB soils (Figure 5.26). Soil acidity and vegetation type could be the factors controlling the distribution of fine roots, and thus NPGBio. In acidic soils, fine root and ectomycorrhizal systems are developed in the O horizons [Fujimaki et al. 2004]. Dipterocarps, which are the dominant vegetation on the BS and BB soils, have fine roots and ectomycorrhizal systems developed in the O horizons of acidic Ultisols [Ashton 1988]. The high NPGBio in the O horizons of the KR3, BS, and BB soils arises from the presence of a fine root mat (Table 5.15; 0.3–2.3 Mg C ha−1 yr−1), which is related to soil acidity (pH < 4.5) and ectomycorrhizal associations of Dipterocarps in the BS and BB soils. Soil acidity and vegetation have a strong influence on the intensity and distribution of acids.
5.7.6 Acid Neutralization in Soils The release of cationic components is the principal mechanism of acid neutralization in organic and mineral soil horizons [van Breemen et al. 1984]. In the O horizons, the extent of acid neutralization varies with basic cation contents. The higher basic cation contents in the O horizons of the KR soils (87–129 cmolc kg−1) are considered to result in complete acid neutralization. In the O horizons of the BS and BB soils, lower basic cation contents (17–63 cmolc kg−1), as well as the intensive acid loads, probably result in incomplete acid neutralization and net eluviation of protons, Al, and Fe. In mineral soil horizons, the extents of acid neutralization depend on the ANC of soils and their parent materials. Based on both published data and those from our study, soil ANC is variable, depending on parent materials and the extent of soil acidification or weathering and clay migration. Our data show that parent materials have a strong influence on soil ANC with the BS and BB soils from sandstone having less ANC and a lower pH than the KR soils from serpentine and mudstone (Figure 5.27). In the KR soils from serpentine or mudstone, their high ANC suggests that their acidity is completely neutralized by basic cation release (Figures 5.26 and 5.27). In the BS and BB soils from sandstone, acidity is not completely neutralized due to their low ANC (Figures 5.26 and 5.27). Thus, parent materials have a strong influence on acid neutralization processes through their effects on basic cation contents in the O horizons and on soil ANC.
5.7.7 Implication of Proton Budgets for Pedogenetic Soil Acidification Acid loads are consistently higher in tropical regions than in temperate regions (Figures 5.3 and 5.26). This supports the presence of strongly weathered soils in the
244
World Soil Resources and Food Security
Soil pH 6.5
KR1
6
KR2
5.5
KR3
5
BS
4.5
BB
4
Oxisols on andesitic PM (Costa Rica)
3.5 3
Spodosols on sandstone (Malaysia) 0
500
1000
1500
2000
2500
3000
Acid neutralizing capacity in soil (mANCs) (cmolc kg–1)
FIGURE 5.27 Relationship between soil pH and the acid neutralizing capacity of soil (mANCs). PM represents parent materials. (Data sources include Andriesse, J.P., Geoderma, 2, 201–227, 1969; Kleber, M. et al., Geoderma, 138, 1–11, 2007.)
humid tropics from the viewpoint of acidification [Eyre 1963]. However, since the kind, intensity, and distribution of acid loads vary with parent materials and vegetation, weathering reactions and pedogenetic soil acidification could also differ among tropical soils. The effects of parent materials on the dominant acidifying processes and pedogenesis can be characterized using the fluxes of Si, Al and Fe, and proton budgets throughout the soil profiles. Judging from the low concentrations of Al and Fe in the moderately acidic and neutral soil solutions of the KR1 and KR2 soils (Table 5.16), the accumulated Al and Fe oxides appear to arise from in situ weathering rather than eluviation or illuviation processes. In the KR1 soil, the substantial fluxes of Si (2.67–3.55 kmol ha−1 yr−1, calculated from the data in Table 5.16) and the high contents of Fe oxides throughout the soil profiles (Table 5.14) support the concept of ferralitization, which implies an absolute loss of Si (desilication) and a relative accumulation of Al and Fe oxides [Cornu et al. 1998]. Dissolution of olivine by acids (Mg1.6Fe0.4SiO4 (olivine) + 4H+ = 1.6Mg2+ + 0.4Fe2+ + H4SiO4) and Si leaching are considered to result in desilication, shown by a decrease of Si content from 45% for serpentine to 9% in the KR1 soil [Effendi et al. 2000]. The fluxes of Si leaching from the KR1 soil (3.55 kmol Si ha−1 yr−1) are higher than those of Oxisols from sedimentary rocks under Amazonian forests (1.1 kmol Si ha−1 yr−1) because of the higher dissolution rates of serpentine (olivine) than of quartz and kaolinite [Cornu et al. 1998]. This is consistent with the rapid formation of Oxisols (ferralitization) from easily weatherable serpentine, as compared to sedimentary rocks [Pfisterer et al. 1996]. In the KR2 and KR3 soils from sedimentary rocks, no net loss of Si occurs. In the KR2 soil, the high rates of NPGCar and the minor contribution of NPGOrg to soil acidification have contributed to incongruent dissolution of Fe-rich parent materials, which results in the accumulation of Al and Fe oxides throughout the profile.
Pedogenetic Acidification in Humid Asia
245
This process in the KR2 soil is similar to brunification, which implies accumulation of Al and Fe oxides owing to incongruent dissolution by weak acids (e.g., carbonic acid) in brown forest soils formed under temperate forests [Ugolini et al. 1990; Section 5.2]. In the highly acidic BS and BB soils, the intensive acid loads contributed by NPGOrg and NPGBio in the O horizons results in net eluviation of protons, Al, and Fe (Figure 5.26). These acidification processes are similar to podzolization [Cronan and Aiken 1985; Guggenberger and Kaiser 1998; Sections 5.2 and 5.3], which involves the complexing of Al and Fe with organic acids and their translocation downward. However, translocation of Al and Fe in the BS and BB soils is different from podzolization because of the absence of spodic B horizons. The degree of podzolization is considered to be controlled by the ANC and Fe contents in the parent materials [Duchaufour and Souchier 1978]. The higher ANC and Fe contents in the BS and BB soils (309–643 cmolc kg−1 and 3.6%–3.9% Fe2O3, respectively), as compared to the typical values for the tropical Spodosols (av. 291 cmolc kg−1 and < 2% Fe2O3, respectively) (Figure 5.27) are considered to reduce the mobility of organic acids and, thus, the degree of podzolization.
5.8 R ELATIONSHIP BETWEEN CHEMICAL AND MINERALOGICAL PROPERTIES AND THE RAPID RESPONSE TO ACID LOADS OF SOILS IN HUMID ASIA As discussed so far, pedogenetic acidification is one of major ecosystem processes under humid climate. Soil response to an external or internal acid load is important: soil acidification causes declines in nutrient levels and Al toxicity, and geochemical processes or soil mineral weathering due to acid load and accompanying nutrient cation release compose the essential part of a biogeochemical process in ecosystems [Likens and Bormann 1995]. There are various acid-neutralizing reactions in soil including 1) exchange reactions or cation desorptions; 2) acid and anion adsorptions; 3) secondary mineral dissolutions; and 4) primary mineral dissolutions. These reactions are quantitatively expressed in terms of the proton budget or changes in acid-neutralizing capacity (ΔANC) [van Breemen et al. 1983, 1984]. In Japan, acid neutralizations due to cation exchange [Sato and Ohkishi 1993; Shibata et al. 2001; Takahashi et al. 2001], secondary mineral dissolution [Funakawa et al. 1993; Baba and Okazaki 2000; Section 5.3], and primary mineral dissolution [Sato and Takahashi 1996] were reported. However, in Southeast Asia, fundamental information on soil acid-neutralizing reactions is still scarce, though high sensitivity to acid loads in the region is assumed [Kuylenstierna et al. 2001]. The rate of acid neutralization may differ among the neutralizing reactions, and the buffering capacity of each reaction may differ among soils. Kinetic analysis, explained in detail by Sparks [1989], must be useful in characterizing the complex and simultaneous reactions in soil. The objective of the present study is to clarify the dominant process of acid neutralization in soils from humid Asia regions such as Japan, Thailand, and Indonesia. To achieve the objective, we investigated the soil chemical and mineralogical properties and rapid soil response to acid load by acid
246
World Soil Resources and Food Security
titration and column experiments. In the column experiment, we kinetically analyzed cation releases to gain information on the source of soil alkalinity. Then, we investigated the relationship between soil chemical and mineralogical properties and the soil response to acid load.
5.8.1 Soils Studied We selected 46 subsoil samples in Japan (16 samples), Thailand (14 samples), and Indonesia (16 samples), which were formed under a natural or secondary forest and well-drained residual soils. Three Andisol soils were included in the Japanese soils, whereas the remaining 13 were Inceptisols or Spodosols. Most of Thai and Indonesian soils were Ultisols, except for 1 Alfisol and 2 Inceptisols from Thailand, and 2 Alfisols and 4 Inceptisols from Indonesia. They have representative clay mineral composition and are derived from various parent materials in each region being used in our previous study on clay mineral distributions [Sections 4.2 and 4.4]. We used soils from a subsurface horizon with a low total carbon content to mitigate the effect of organic matter. In addition to the analyses on general physicochemical properties, acid titration was conducted to investigate rapid soil response to acid load under the same experimental condition as Section 5.3. After titration, the concentrations of Na, K, Mg, Ca, Al, Mn, Fe, and Ti were determined to identify the acidneutralizing reactions in each of soils. The eight soils were selected from the 46 soil samples for the column experiment (Table 5.18), which are different in the degree of acidification, have representative chemical and mineralogical properties (e.g., pH, CEC, clay mineral species, and oxalate extractable Al), and are widely distributed soil types in each region. The four soils from Japan included a volcanic soil (JP-V) and three forest soils derived from sedimentary rocks in cool and warm temperate and subtropical zones, which were denoted JP-S1, JP-S2, and JP-S3, respectively. TH-S was from a sedimentary rock area in Thailand. ID-L, ID-I, and ID-S from Indonesia were formed on representative geological rocks, i.e., limestone and intermediate–basic igneous, and noncalcareous sedimentary rocks, respectively. Detailed information on the profiles was reported by Mori et al. [2005] and Watanabe et al. [2007]. The experimental procedure was roughly described as follows. Because the clay fraction dominates cation and anion exchange reactions and secondary mineral dissolution, soils were diluted with quartz sand to obtain a clay content of 30.0 % for the comparison of soils with different clay contents. Then 5.0 g of each sample (soil plus quartz sand) was packed in a column. Extraction with 0.01 mol L−1 HCl was conducted at a flow rate of 1 mL min−1 for 60 or 120 min, until the effluent pH reached a value near that of the extractant. HCl (0.01 mol L−1) was used to enhance mineral dissolution and exchange reactions and to detect the potential capacity of the reactions. Effluent was collected every 10 min and the concentrations of Na, K, Mg, Ca, Al, Fe, Mn, and Si were determined. To investigate the release manner of each element, the time course of element release was simulated using first-order kinetics, Y = A(1 − exp(−kt)), where Y is the cumulative amount of ions released, A is the convergence value, t is the extraction time, and k is the rate constant of the reaction.
Exchangeable
Sample JP-V JP-S1 JP-S2 JP-S3 ID-L TH-S ID-I
ID-S
a
Soil Classification Parent Rock Acrudoxic Melanudands Typic Fulvudands Andic Dystrudepts Typic Dystrudepts Typic Dystrustepts Ustic Haplohumults Typic Dystrudepts
Typic Paleudults
Volcanic ejecta Sandstonemudstone Sandstonemudstone Sedimentary rock Limestone Sandstonemudstone Andesiticbasaltic volcanic breccia Sandstonemudstone
Temp.
pH
a
Total C
(H2O) (g kg )
CEC
Bases
Al
Clay
Alo
(%)
Feo
Sio
Fed
Ca
Exch./Total
Mg
Mg
Horizon Bw1
5.0
23.2
17.3
0.7
n.d.
43
48.6 25.4 22.6 42.0 69.5 223.9
0.4
0.1
9.3
Bw2
4.7
55.9
21.9
0.3
6.8
40
13.7 26.1
15.4
Bw2
4.6
13.4
19.0
0.3
8.3
63
5.1
22.3
Bt
4.7
5.6
14.0
0.8
7.7
37
25.5
Bw2
5.8
15.7
29.3
17.2
n.d.
24.3
BA
5.6
11.0
10.5
2.7
24.9
Bw2
4.7
10.2
30.8
26.6
Bt
4.2
6.2
20.5
−1
(cmolc kg )
Ca
6.2
−1
(g kg )
Total
(°C)
−1
(cmolc kg )
Selective Dissolution
−1
(%)
0.3 30.4
4.5
40.4
1.1
0.2
5.7
0.3 32.1
2.5
62.5
1.1
0.2
1.7
0.7
0.1 56.3
0.5
27.0
3.9
2.0
76
5.0
5.7
1.3 63.6 22.3
16.3
65.3 14.4
0.7
51
1.8
4.1
0.2 35.3
4.6
26.4
20.8
5.8
5.7
9.4
66
2.8
5.3
0.3 31.8
7.1
32.8
46.0
6.7
0.5
9.9
49
2.1
1.3
0.1 37.3
2.4
27.8
2.7
Pedogenetic Acidification in Humid Asia
TABLE 5.18 Physicochemical Properties of Soils Used for the Column Experiment
0.7
Temperature was calculated from the mean annual temperature at the nearest meteorological stations and elevations at the sites and stations with a lapse rate of 5.5°C km−1.
247
248
World Soil Resources and Food Security
5.8.2 Acid Titration Experiment and Source of Soil Alkalinity The amount of consumed H+ during the acid titration experiment, i.e., titratable alkalinity by a solution at pH 3.0, differed in each region and ranged from 2 to 11 cmolc kg−1. In Japan, the amount of consumed H+ was large for the Andisol samples (Table 5.19; 6–9 cmolc kg−1) and small for the soils with low clay content derived from felsic rocks (≤3 cmolc kg−1). Thai soils had low titratable alkalinity, mostly less than 3.5 cmolc kg−1, in spite of their higher pH(H2O) values, except for the soils at high altitude locations, where Alo content was high. Indonesian soils differed in their alkalinity reflecting parent materials; alkalinity was low in the soils from noncalcareous sedimentary rocks, and high in the soils from limestone or volcanic rocks. The proportion of sources of soil alkalinity in the acid titration experiment is shown in a ternary diagram for released base (Ca, Mg, K, Na) and metal (Al, Mn, Fe, Ti) ions and acid and anion adsorption (Figure 5.28). The alkalinity of Japanese soils was mainly caused by the release of metal ions, especially Al (Table 5.19), which correlated with Alo content (Figure 5.29a). In the two Andisol samples, not much Al was released in spite of a high Alo content. The reason for the low Al release rate was considered to be the dominance of the other H+ consuming reaction, i.e., acid and anion adsorption. The source of alkalinity in tropical soils was base ions (Figure 5.28), especially Ca and Mg (Table 5.19). The amount of released bases was strongly correlated with that of exchangeable bases (Figure 5.29b). The proportion of released metal ions in soil alkalinity were statistically larger in Japanese soil samples than in Thai and Indonesian soil samples (p < 0.001), while the proportion of released base ions was smaller (p < 0.001). In all the regions, acid and anion adsorption was more important for soils with lower alkalinity (Rs = −0.72, p < 0.001; Figure 5.29c). In the Andisol samples having high alkalinity, the proportion of acid adsorption was large relative to the amount of consumed H+ because of the large amount of variable charge sites in the soils. The amount of H+ consumed by acid adsorption weakly correlated with clay content (Rs = 0.39, p < 0.01) and pH (Rs = 0.35, p < 0.05). The ratio of exchangeable amount to total amount of Ca and Mg in some soils was high (Table 5.18), whereas the ratio of exchangeable amount to total amount of K and Na was usually low (less than 1%). Ca- and/or Mg-containing minerals, such as calsite, Ca-feldspar (anorthite), and pyroxene, typically have low resistance to weathering. They must weather rapidly and release Ca and Mg into soils, which would be retained in clay minerals. In contrast, K- and/or Na-containing minerals, such as K/Na-feldspars (e.g., microcline and albite) and muscovite have high resistance to weathering, and the monovalent ions are weakly retained in soils, which may result in the low ratio. The exchangeable/total ratios of Ca and Mg in soils were correlated with mean annual temperature (Rs = 0.52, p < 0.001; Rs = 0.43, p < 0.01) and pH (Rs = 0.43, p < 0.01 for Mg). Temperature was thought to enhance the weathering of primary minerals, which releases the base cations; the exchangeable bases were leached out by acidification. The ratios indicate that a high proportion of total amounts of the bases were labile in some Thai and Indonesian soils (Table 5.18), which means that the bases are easily utilized by plants but exposed to leaching by an acid load caused by acid precipitation, continuous cropping, etc.
Consumed H+
Samples
Number of Samples
AVE
SE
Released Cations Ca2+ AVE
Mg2+ SE
Average values in samples from different countries Indonesia 16 3.99 1.52 a 1.06 1.34 Thailand 14 3.45 1.00 a 0.31 0.57 Japan 13 3.84 1.05 a 0.10 0.34 Japan, 3 7.99 1.33 b 0.17 0.33 Andisols Samples used for the column experiment JP-V 8.95 0.28 JP-S1 9.09 0.06 JP-S2 4.34 0.03 JP-S3 2.47 0.03 ID-L 10.85 7.09 TH-S 3.27 0.44 ID-I 4.05 1.18 ID-S 2.71 0.05
AVE
K+
SE
AVE
Na+ SE
AVE
Al3+ SE
Adsorbed H+
Mn4+
AVE
SE
0.51 0.70 2.17 4.59
0.68 0.93 1.13 1.43
AVE
SE
0.15 0.26 0.05 0.04
0.55 0.59 0.39 0.20
AVE
SE
1.36 1.52 1.23 2.83
0.50 0.60 0.53 0.72
(cmolc kg–1)
a a a a
0.78 0.51 0.16 0.10
0.15 0.10 0.11 0.45 1.34 0.85 0.95 0.17
1.13 0.62 0.40 0.25
a a a a
0.07 0.14 0.05 0.06
0.06 0.05 0.03 0.04 0.04 0.06 0.06 0.14
0.18 0.38 0.16 0.13
ab b a ab
0.05 0.01 0.05 0.04
0.04 0.06 0.03 0.13 0.10 0.01 0.04 0.01
0.28 0.07 0.19 0.16
a a a a
5.53 5.99 2.62 0.37 0.15 0.05 0.25 1.09
a a b c
0.08 0.00 0.00 0.00 0.64 1.04 0.01 0.01
a a a a
a a a b
Pedogenetic Acidification in Humid Asia
TABLE 5.19 Consumed H+, Released Cations, and Adsorbed H+ during the Acid Titration Experiment
2.33 2.80 1.50 1.43 1.49 0.81 1.52 1.20
Note: AVE, average; SE, standard error. The values with the same letters are not significantly different by Tukey test (p < 0.05).
249
250
World Soil Resources and Food Security Japan Thailand Indonesia
0 100
Anion and acid adsorption 50
Base ions (Ca, Mg, K, Na)
50
100 0
0 100
50 Metal ions (Al, Mn, Fe, Ti)
FIGURE 5.28 Ternary diagram outlining the source of alkalinity (base and metal ions and acid adsorption) in the acid titration experiment.
(b)
10
Japan Thailand
8
Indonesia
6 4 2 0
0
10
20
Alo
Acid adsorption (%)
(c)
Released bases (cmolc kg–1)
Released Al (cmolc kg–1)
(a)
30
40
(g kg–1)
50
10
Japan Thailand
8
Indonesia
6 4 2 0
0
5
10
15
20
25
Sum of exchangeable bases (cmolc
80
30
kg–1)
Japan Thailand
60
Indonesia
Andisols
40 20 0
0
2
4
6
8
10
Consumed H+ (cmolc kg–1)
12
FIGURE 5.29 Contribution of the different sources of alkalinity of soils in the acid titration experiment: (a) Oxalate-extractable Al (Alo); (b) exchangeable bases; and (c) anion and acid adsorption.
251
Pedogenetic Acidification in Humid Asia
5.8.3 Column Experiment for Eight Selected Soils The eight soils selected for the column experiment varied in soil pH value, the amount and source of alkalinity in the acid titration experiment, and the mean annual temperature under which they formed (Table 5.18). All the soils were already acidified under humid climate conditions. ID-L and TH-S had relatively high pH values because they formed under an ustic moisture regime, and ID-L was derived from limestone. ID-L, TH-S, and ID-I were rich in exchangeable Ca and Mg. A large amount of exchangeable Al was retained in strongly acidified soils with pH values below 5. According to Watanabe et al. [2007], x-ray diffractograms showed that JP-V has a small amount of crystalline clay minerals, whereas the other Japanese soils were determined to have an appreciable amount of 2:1-type clay minerals (1.4 nm), which were HIV. Gibbsite was clearly detected in all the Japanese soils, whereas no or only small amounts of gibbsite was present in the four tropical soils. For the tropical soils, clay mineral compositions were characterized by a large amount of kaolin minerals, and the 2:1-type clay minerals of the tropical soils were mica in TH-S, smectite in ID-I, and vermiculite in ID-S. The amounts of Alo and Feo were large in JP-V and JP-S1 (Table 5.18). In JP-V, the large amount of Alo and Sio were attributed to the presence of allophane/imogolite. Figure 5.30 shows that releases of base cations, especially Ca and Mg, contributed to acid neutralization in ID-L, TH-S, and ID-I, whereas Al release contributed to acid neutralization in the other soils. The relative contribution of cation releases in each soil sample were similar to that in the titration experiments (Figure 5.30; Table 5.19), though Al release was enhanced due to lower endpoint pH value. All the regression curves for Ca, Mg, K, Na, and Al releases in Figure 5.31 had low p-values (<0.001).
0
Released cations (cmolc kg–1) 10
20
Released Si (cmol kg–1)
30
40
0.0
0.5
1.0
1.5
JP-V (a)
JP-S1
(b)
JP-S2 JP-S3
Ca
ID-L
Mg
TH-S ID-I ID-S
K Na Al Other cations
FIGURE 5.30 Amounts of released (a) cations and (b) Si during the column experiment.
JP-V 4
(1.2)
2
K
Na
Al
pH
0
0.4
(1.6)
0.2 0
1.0 (1.1)
0.5
(3.8)
5
(1.9)
0.2
(0.8)
(0.9)
0.5
(20.0†)
2
4 2
0
4
0
2
5
ID-I
(1.7)
5
(1.1)
5
10 (55222†)
20
20
(8.9)
5
0
0
0
0
0
10
10
10
10
10
5
(5.5)
5
(9.4)
5
(4.0)
5
5 0
0
0
(8.8)
(25.2)
(12.7)
20
0
5
5
0
0
4
0 (11.0)
4
0
0 100
50
50
20
20
2
5
20
20
0
0
0
0
0
0
0
0
7
7
7
7
7
7
7
7
4
4
4
4
4
4
4
4
1
0
50 100
1
0
50 100
40
1
(1.3)
0
30
60
40
1
4
(1.0)
0
30
60
1
0
Time (min)
30
60
1
(7.9†) 40
0
30
60
1
(9.3)
2
0
(5.8)
(1.0)
2
100
10
(1.0)
0.5 0.0
40
10
(1.3)
0.2
1.0 (1.1)
20
0
0.4
0.0
40
10
(15.0)
(1.1)
0
0
40
40
ID-S
0
10 (1.1)
20
(20.5)
TH-S
10
0
40
(1.5)
10 (4.9)
(1.2)
50
0
4
ID-L
100
(16.2)
2
0.0
4
JP-S3
0
1.0
0.4
0
10
0.4
0.0
0.2
0.0
JP-S2
(0.6)
0
30
60
40
1
(0.8)
0
30
60
World Soil Resources and Food Security
Cumulative amount of released cations (mmol kg–1)
Mg
JP-S1
252
Ca
Pedogenetic Acidification in Humid Asia
253
The time courses of Ca and Mg releases are shown in Figure 5.30, together with the regression curves of first-order kinetics. The ratio of the convergence value (A) to the exchangeable amount indicates that the A-value is close to the amount of exchangeable bases, except for JP-S3. The high ratio of Ca in JP-S3 was presumably caused by the rapid dissolution of Ca-feldspar, which is present in small amounts, and thus may be present as small particles. Ca release from Ca-feldspar may be one reason why the ratio of Ca is larger than that of Mg. For K and Na, the released amounts exceeded the exchangeable amounts in the first 10 min in most cases (Figure 5.31), and the ratio of the A-value to the exchangeable amount was much larger than one. These results indicate that the dissolution of minerals is much more important than the exchange reaction for K and Na releases. Because the amount of released Si was not sufficiently large to assume the congruent dissolution of mica or feldspars, the preferential releases of K and Na from minerals were supposed to occur. The most possible sources of K and Na were mica and K-feldspar and Na-feldspar, respectively. The dissolution manner with such preferential releases of alkali metals from mica was explained by Fanning et al. [1989], and those from feldspars by Nahon [1991] and Blum and Stillings [1995]. K is released from mica with the opening of 2:1 layers. Na and K releases from feldspars are caused by an exchange reaction in feldspar surfaces occurring as an initial dissolution [Blum and Stillings 1995]. Feldspar surfaces are in a quasisteady state under natural conditions forming transient residual layers in which alkali ions are depleted. Under experimental acidic conditions, the alkali ion-depleted layers become deeper as alkali ions are released [Chou and Wollast 1984]. Mica and feldspars in the silt and sand fractions are commonly found, which may contribute to the releases as well as those in the clay fraction. No relationship was observed between Na in feldspars and Na released beyond the exchangeable amount, presumably due to variations in the freshness and particle size distribution of the feldspars. Both K and Na releases from mica and feldspars in this study must be diffusion-controlled [Chou and Wollast 1984; McLean and Watson 1985] and promoted by a low K/Na activity in HCl solution in the column experiment. For Al release, we assumed a time lag for JP-V, ID-L, and TH-S, for which the release was clearly delayed (Figure 5.31). For JP-V, a large amount of Al was released continuously. The dissolution of allophane/imogolite contributed to the release, judging from a large amount of Alo and Sio, and released Si (Figure 5.30). The dissolution of Al(OH)3 also contributed to the release, because the ratio of Al to Si released was higher than the ratio of Alo to Sio. For JP-S1, JP-S2, JP-S3, ID-I, and ID-S, the relative importance of mineral dissolution against the exchange reaction was considered to be higher in the soils with higher Alo content, judging from the A-values for these soils (Figure 5.31; Table 5.18). For ID-L and TH-S with high pH values, the amount
FIGURE 5.31 Cumulative amount of cations released together with the regression curve assuming first-order kinetics and pH in the column experiment. The dotted lines represent the amount of exchangeable cation. The ratio of the convergence value (A) to the exchangeable amount is enclosed in parentheses, where daggers (†) indicate p > 0.05. JP-V and ID-L have no exchangeable Al. The pH value at 0 min indicates pH(H2O).
254
World Soil Resources and Food Security
ID-S
ID-I
TH-S
ID-L
JP-S3
JP-S2
JP-S1
JP-V
of Al released was small. They had either small amounts of exchangeable A1 or none at all (Figure 5.31); thus, Al release depended on slow mineral dissolution. For the soils except for JP-V and ID-L, the low release rate of Si indicates a weak silicate mineral dissolution (Figure 5.30). In JP-V and ID-L, Si was assumed to be released by Sio dissolution (Table 5.18). In JP-V, Sio was attributed to allophane/ imogolite, but in ID-L the mineral contributing to Sio content was not apparent. The effluent pH values of JP-V, JP-S1, and JP-S2 decreased gradually, owing to Al release (Figure 5.31). For ID-L, the exchange reaction was extended and pH did not decrease in the first 10 min because of the large amount of exchangeable bases. For the other soils, pH decreased rapidly because of the small amount of exchangeable bases and low Alo content. The rate constants (k) of element releases are shown in Figure 5.32, where some releases with high p-values (>0.05) (marked with daggers) seem to be zero-order reactions. Generally, the variation of k-values was consistent with the acid-neutralizing reactions assumed before. The k-value of Ca and Mg releases, which was assumed to be caused by exchange reactions, were as large as 10 −1 min−1. Meanwhile, the k-values of Al, K, Si, and Na releases were small. The k-values of Al releases from ID-I, ID-S, and JP-S3 were smaller than those of Ca and Mg, though both releases were mainly caused by exchange reactions, implying that exchangeable Al is more strongly retained in the soils. The k-values of Al releases from JP-V, JP-S1, ID-L, and TH-S, and those of K and Si, which were mainly released by mineral dissolution, were smaller than those of the exchange reactions. Na release was mainly caused by the exchange reaction occurring beneath the residual layer of feldspars, and the k-value of Na release was intermediate between those of Ca and Mg releases and mineral dissolution.
0 –1 –2
log k (min–1)
†
–3
†
–4 –5 –6 –7
†
† †
Ca Mg K Na Al Si
FIGURE 5.32 Rate constants (k) of the element released in the column experiment expressed in logarithms. *p > 0.05.
Pedogenetic Acidification in Humid Asia
255
5.8.4 Interpretation of Acid-Neutralizing Reactions under Laboratory and Field Conditions The results of the acid titration and column experiments indicated that the following reactions were mainly assumed for rapid soil response to internal and external acid loads in each region: exchange reactions of bases in the tropical soils with large amounts of exchangeable bases (especially Ca and Mg), Al dissolution in the Japanese soils with high Alo content, and anion adsorption in both the Japanese and tropical soils with small amounts of exchangeable bases and low Alo content and, thus with low alkalinity. Because of the low pH and high addition rates of HCl over a short time in the titration and column experiments, all the acid-neutralizing reactions were enhanced and some differences in acid neutralization between the experimental and the field conditions were assumed. Differences in the relative contributions of the acid-neutralizing reactions between the experimental and the field conditions is not revealed by the experiments, because the degree of enhancement is not apparent for each reaction. However, the regional trend in the relative contribution to acid neutralization is considered to be similar between the experimental and the field conditions, because Al dissolution coincided with Alo distribution. On the other hand, slow dissolution of primary minerals occurs during long times of weathering under the field conditions, and it releases Al, Ca, Mg, etc., which results in acid neutralization, formation of secondary minerals, and replenishment of exchange bases. This long-term effect of primary mineral dissolution must be considered in the field. The exchange reaction of bases and the dissolution of Al are important in the field conditions [van Breemen et al. 1983; Chadwick and Chorover 2001]. The pH range of the exchange reaction of bases is approximately 5.5 to 4.0 [Chadwick and Chorover 2001; Brady and Weil 2002]. Al dissolution is important at low pH values (3.0–5.0) [van Breemen et al. 1983]. Acid and anion adsorption are also assumed to occur in response to an acid load [van Breemen et al. 1983; Brady and Weil 2002]. In this study, soils with low total carbon content were used, which decreased the amounts of acid adsorption to organic matter. Al dissolution in the Japanese soils is considered to be important even under the field conditions compared with the tropical soils, although its relative contribution to acid neutralization is lower than in the experiments under low pH conditions. Japanese soils have low pH values of less than 5 where Al dissolution is important [van Breemen et al. 1983]. At progressive stages of the soil acidification process, monomerization of amorphous Al(OH)3 and eluviation of Al is reported by Funakawa et al. [1993]. Alo dissolution is quantitatively important for acid neutralization because of the large buffering capacity: 10 g kg−1. Alo has an acid-neutralizing capacity comparable to that of 111 cmolc kg−1 exchangeable cations, being assumed to be present as Al(OH)3 and to dissolve as Al3+. The contribution of gibbsite dissolution to acid neutralization [Chadwick and Chorover 2001], which occurs at pH (KCl) values below 4.2 [Ulrich 1989], was also assumed in the Japanese soils, where gibbsite is usually present. In the Thai and Indonesian soils, rapid acid neutralization due to mineral dissolution was limited, because the amount of Alo and gibbsite content were generally small, and kaolinite, which is a dominant clay mineral, dissolves more slowly than gibbsite [Nagy 1995; Lasaga 1998].
256
World Soil Resources and Food Security
The Al3+ activity in soil, which is ecologically significant because of its toxicity to vegetation, is mainly governed by Al(OH)3 solubility and pH [Chadwick and Chorover 2001], although in O/A horizons organic matter decreases Al3+ activity [Cronan et al. 1986]. Low-crystalline minerals in soil are assumed to have both positive and negative effects on Al3+ activity. At a given pH, Al3+ activity is higher in soils where poor-crystalline minerals control the activity compared with soils where crystalline minerals do; for example, when poor- and well-crystalline minerals control Al3+ activity at a pH of 4, the activities are 123 and 3 mg L−1, respectively [Lindsay 1979]. In this case, poor-crystalline minerals have a more adverse effect on the biota. In contrast, low-crystalline minerals were more reactive because they dissolved rapidly in the titration and column experiments, and these acid-neutralizing reactions and mineral dissolution are expected to keep the soil pH high, resulting in lower Al3+ activity. The low pH and high Alo content in some Japanese soils in the present study are considered to result in high Al3+ activity, exchange reactions of Al3+ with base ions, and low base saturation, which explains the low contribution of base ion releases to acid neutralization in these soils. K release from mica is predominant in soils with a pH(H2O) value below 5.0– 5.5, resulting in vermiculite formation as discussed in the previous sections. K and Na releases from feldspars as an initial dissolution were expected to occur easily under lower pH conditions, but to subside after a deeper surface residual layer forms. Because the releases are diffusion-controlled reactions [Chou and Wollast 1984; McLean and Watson 1985] and depend on the K/Na activity in soil, the release of these monovalent ions was thought to be enhanced in a specific zone, such as the rhizosphere. In the rhizosphere, low K/Na activity and strong acidic conditions are possible [Campbell and Greaves 1990] and K and Na may be easily released, even from the minerals.
5.9 G ENERAL DISCUSSION ON THE PEDOGENETIC ACIDIFICATION PROCESS The natural soil acidification processes vary from ecosystem to ecosystem, depending on climate, vegetation, and parent materials [van Breemen et al. 1983, 1984; Ugolini and Sletten 1991]. According to van Breemen et al. [1984], the dominant acidification processes were carbonic acid dissociation in soils at neutral pH, vegetation uptake and nitrification in acidic soils with weatherable minerals, and vegetation uptake and dissociation of organic acids in podzolic soils. Johnson [1977] and Johnson et al. [1983] suggested the importance of carbonic acid to soil acidification in the tropical forests. In the present study, soil solution composition and proton budget in each of the soil horizons were quantitatively evaluated in temperate and tropical forests, and the factors controlling proton generation and consumption were assessed in relation to climate and soil types (Sections 5.2, 5.6, and 5.7). At the same time, the fates of the acids as well as H+ and Al translocated were traced (Section 5.3), and the resulting modification of soil minerals was also analyzed using different methodologies, e.g., analyzing RIP (Section 5.4), charge characteristics (Section 5.5), and assessing the rapid response of minerals against acid addition (Section 5.8). In the present section, these data—obtained in different regions and from the preceding works for the
Pedogenetic Acidification in Humid Asia
257
forest in northern Thailand, i.e., RP [Fujii et al. 2008; Section 4.7]—are comparatively analyzed.
5.9.1 Organic Matter Dynamics Since natural soil acidification induced by the internal acid load is associated with the organic matter cycles in the ecosystems [Binkley and Richter 1987; Devries and Breeuwsma 1987], the amounts of production and decomposition of organic matter and the balance between them have strong influences on proton generation and consumption, and thus the proton budget. The stock and flow of C in ecosystems and soils analyzed in the present study are summarized in Table 5.20. The aboveground biomass increases with air temperature from 78 to 346 Mg C ha−1 yr−1. The wood increment in the warm temperate and tropical forests (5.11–11.12 Mg C ha−1 yr−1) is higher than those in the cool temperate forests (1.50–2.50 Mg C ha−1 yr−1). The litterfall also increases with air temperature, ranging from 1.70–2.90 Mg C ha−1 yr−1 in the temperate forests to 3.62–4.78 Mg C ha−1 yr−1 in the tropical forests. Assuming that the C input to the soil is the sum of litterfall, root litter, and DOC leached as throughfall, the C budget between C input and its decomposition was almost balanced in each of the soils.
5.9.2 Soil Acidification Rate in the Entire Soil Profile In the present study, NPGBio was a dominant proton source in the entire soil profile. NPGBio attributable to the wood increment ranged from 0.7–8.6 kmolc has−1 yr−1, while NPGBio attributable to litter production ranged from 1.8–9.1 kmolc ha−1 yr−1. NPGBio attributable to the wood increment and litter production increased with the respective biomass production (Figure 5.33). Litter production and wood increment contributed to net proton generation at the rates of 0.010–0.030 molc and 0.004– 0.010 molc for production of 1 mol organic C, respectively. Judging from the balance between C input and its decomposition (Table 5.20), NPGBio attributable to litter production could be neutralized by cation release from the fallen litter. Therefore, soil acidification (0.5–6.1 kmolc ha−1 yr−1) is contributed mainly by excess cation accumulation in wood in the growth stage of the forests (Figure 5.33). The complete proton cycles in the processes of carbon and nitrogen cycles (e.g., mineralization, nitrification, and nitrate uptake by vegetation, dissociation and protonation of carbonic acid, and dissociation, mineralization, and adsorption of organic acids) result in the minor contribution of NPGNtr, NPGCar, and NPGOrg to soil acidification in the entire soil profiles [Binkley and Richter 1987]. However, translocation of the temporary acids (carbonic, organic, and nitric acids) and distribution of the root biomass contribute to the heterogeneity of proton generation and consumption throughout the soil profiles, and result in different processes of natural soil acidification (e.g., podzolization). In TG, KT, BS, and BB, NPGBio and NPGOrg concentrated in the O horizon are responsible for intensive acid loads (sum of NPGs) in the O horizon (3.1–7.0 kmolc ha−1 yr−1) (Figure 5.34). While the intensive acid load (3.1–7.0 kmolc ha−1 yr−1) could be largely neutralized by cation release from organic matter and mineral weathering
258
TABLE 5.20 Stock and Flow of Organic Carbon in Ecosystems Studied in Section 5.3
Site
a b c
Mean Annual Precipitation
Aboveground Biomass
DOC Flux
(°C)
(mm)
(Mg C ha−1)
7 11 16 25
1422 1782 1490 2084
78 83 115 169
1.50 2.50 10.10 5.79
1.70 2.10 2.90 3.99
0.40 0.59 0.84 1.25
3.40 5.50 5.10 5.46
−1.64 −3.32 −2.12 −1.44
58 78 83 40
35 344 112 32
2 16 4 1
27
2187
293
10.56
4.08
0.88
5.45
−1.18
182
562
13
27
2427
346
11.12
3.62
1.14
4.25
−0.53
97
470
13
27
2256
134
7.23
4.78
0.93
4.23
0.72
164
603
12
27
2256
260
9.85
4.00
0.18
4.70
−0.57
136
166
4
27
2256
282
5.11
4.50
0.36
4.24
0.39
126
261
6
Wood Increment
Litterfall
Root Littera
Decomposition of OMb
C Budget
(Mg C ha−1 yr−1)
Throughfall
O Horizon
(kg C ha−1 yr−1)
The annual rates of root litter incorporation were assumd to be 20% of the fine root biomass in the forest [Nakane 1980]. OM represents organic matter. C input is the sum of litterfall and DOC in throughfall.
DOC/C Inputc (%)
World Soil Resources and Food Security
NG, Japan TG, Japan KT, Japan RP, Thailand BS, Indonesia BB, Indonesia KR1, Indonesia KR2, Indonesia KR3, Indonesia
Mean Annual Air Temperature
259
Pedogenetic Acidification in Humid Asia
NPGBio (kmolc ha–1 yr–1) 10
KR2
KR2
8
Wood increment
KR3
6
KR1
BS
4
KT RP NG
2 TG 0
Litter production
KR1
0
BS KT
KR3
BB
BB
TG NG
RP 5
10
15
Biomass production (Mg C ha–1 yr–1)
FIGURE 5.33 Relationship between the biomass production of litter and wood and NPGBio.
(2.0–4.8 kmolc ha−1 yr−1), some excess protons (0.5–2.2 kmolc ha−1 yr−1) are transported into the A and B horizons in these soils (Figure 5.34). In the mineral soil horizons, the acid load contributed mainly by NPGBio, as well as the proton load transported from the overlying horizon, could be compensated for by proton consumption associated with the mineralization of organic anions and nitrate uptake by vegetation. In some Japanese plots, acid neutralization by the reactions with amorphous and/or organo-mineral complexes could be emphasized (Sections 5.3 and 5.8). In the other plots (NG, RP, KR1, KR2, and KR3), the acid load (1.7–4.3 kmolc ha−1 ∆ANC (kmolc ha–1 yr–1)
1:1
–8
–6
–4
–2
0
0
Moderately acidic soils (soil pH: 4.5–6.4)
KR1 KR2 KR3 RP
NG
BB
Highly acidic soils (soil pH: 3.8–4.3)
TG BS
KT 2
4
Acid load (kmolc
6
ha–1 yr–1)
8
FIGURE 5.34 Relationship between the acid load and the soil acidification rate in the O horizon.
260
World Soil Resources and Food Security Proportion of NPGBio in the O horizon relative to NPGBio in the entire soil profile (%) 100 80
TG
60 BB
40
KT
20 0
BS 3
4
KR3 NG
RP 5 Soil pH
KR2
KR1 6
7
FIGURE 5.35 Relationship between soil pH (0–5 cm) and the proportion of NPGBio in the O horizon relative to NPGBio in the entire soil profile.
yr−1) could be completely neutralized by basic cations in the same horizons (Figure 5.34). The surface neutralization observed for some Japanese soils in Section 5.3 (MD1 and MD2) might be included into this category. These differences in acid neutralization are considered to be caused by those of acid-neutralizing capacity in the O horizon. The contents of basic cations in the O horizon decreased with decreasing soil pH (Figure 5.35). In the moderately acidic soils, acid load could be completely neutralized by basic cations in the O horizon. In the highly acidic soils of TG, KT, BS, and BB, intensive acid load by NPGBio and NPGOrg in the O horizon, as well as the lower contents of basic cations in the O horizon, results in incomplete acid neutralization and net eluviation of protons and Al (Figure 5.34).
5.9.3 Factors Controlling Proton Generation and Consumption in Relation to Organic Matter Cycles The NPGBio contributes to intensive acidification in the O horizon of the highly acidic soils (pH < 4.3) in TG, KT, BS, and BB, while NPGBio distributed evenly in the A and B horizons of the moderately acidic soils in the other plots. The proportion of NPGBio distributed in the O horizon relative to that in the entire soil profiles increased with decreasing soil pH (Figure 5.35). The increased distribution of NPGBio in the O horizon with decreasing soil pH is caused by the presence of a fine root mat in the highly acidic soils. In the highly acidic soils, a fine root and ectomycorrhiza system was reported to develop in the O horizon to enhance NH +4 mobility [Aber et al. 1985]. Further, Fagaceae and Dipterocarpaceae are known to have a fine root and ectomycorrhiza system developed in the O horizons of the acidic soils [Ashton
Pedogenetic Acidification in Humid Asia
261
1988]. Judging from this, vegetation species are also considered to contribute to the increased distribution of NPGBio in the O horizon. NPGOrg also substantially contributes to acidification of the O horizon. The NPGOrg in the present study (1.8–3.2 kmolc ha−1 yr−1), except for RP and KR2, is comparable with the higher values reported for the Spodosols in the temperate forests (0.8–3.7 kmolc ha−1 yr−1) [Cronan and Aiken 1985; Guggenberger and Kaiser 1998]. The higher NPGOrg in the O horizon is caused by the higher fluxes of DOC, which has one dissociated acidic functioning group per 5.9–12.2 C atoms in TG, KR1, BS, and BB. The fluxes of DOC leached from the O horizon are dependent on the fluxes of C input and the proportion of DOC leaching relative to C input. The higher DOC fluxes in the highly acidic soils of TG, BS, and BB could be accounted for by the increased proportion of DOC leaching relative to C input with decreasing soil pH. In KR1, the limited mineralization of DOC of the litter with the low phosphorus content, as well as the high fluxes of C input, is considered to result in the high fluxes of DOC leaching from the O horizon (Table 5.20), as discussed in Section 5.6. Owing to the weak acid nature of carbonic acid, NPGCar contributes to acidification of the O horizons only in KR1 and KR2. This is consistent with the high NPGCar in the soils at neutral pH reported by Johnson et al. [1983], van Breemen et al. [1984], and Gower et al. [1995]. NPGCar is dependent on the soil solution pH. As compared to NPGNtr associated with SOM loss in the cropland plots (1.5–5.0 kmolc ha−1 yr−1) [see Section 4.7], NPGNtr was lower (<1.1 kmolc ha−1 yr−1) owing to the balanced mineralization and assimilation of nitrogen in all the forest plots except for the O and A horizons of KR1 (1.3–3.2 kmolc ha−1 yr−1), where the understory vegetation was nitrogen-fixing vegetation. In conclusion, the distribution of fine roots and dissociation of organic and carbonic acids contribute to the heterogeneity of proton generation and consumption throughout the soil profiles, depending on the factors relating to soil pH and vegetation species.
5.9.4 Pedogenetic Implication for Proton Budget in Soil The differences in the amounts and sources of acid loads contribute to different mineral weathering and pedognetic soil acidification [Ugolini and Sletten 1991]. Typically, organic acids contribute to congruent mineral dissolution and Al eluviation caused by complexation in Spodosols [Lundström 1993; van Hees et al. 2000]. In the present study, the intensive acidification of the O horizon caused by vegetation uptake as well as organic acid dissociation is shown to be responsible for eluviation of Al and Fe from the O and A horizons and their illuviation in the B horizon in Spodosols of TG. However, since intensive acidification of the O horizon and Al eluviation were also observed in the acidic Ultisols of BS and BB, the high acid load by organic acid dissociation and vegetation uptake and Al eluviation are considered to be the common pedogenetic processes in the highly acidic soils of TG, KT, BS, and BB, both in temperate and tropical zones. On the other hand, the essential roles of carbonic acid in andosolization, brunification and ferralitization have been recognized [Stevenson 1982; Ugolini et al. 1988, 1990; Ugolini and Sletten 1991; Patel-Sorrentino et al. 2007]. The carbonic acid
262
World Soil Resources and Food Security
contributes to incongruent dissolution and concentration of Al and Fe oxides. In the present study, judging from the lower contribution of NPGOrg to soil acidification and lower mobilization of Al and Fe in NG, RP and KR2, incongruent mineral dissolution by weak acids is considered to result in accumulation of Al and Fe oxides or hydroxides. In KR1 and KR3, despite the high rates of acid load and the high contribution of NPGOrg to soil acidification, the high soil ANCs could contribute to complete acid neutralization and accumulation of Al and Fe oxides.
5.9.5 Fates of Soil Acidity and Dissolved Organic Matter during Pedogenetic Soil Acidification of Forest Soils A soil solution study revealed that, under certain geological and climatic conditions, Al dissolves organic matter and excess protons migrate downward in soil profiles. The fates of these materials are considered to be different depending on geological and climatic conditions. One of the unique pedogenetic processes in Japan is the supply of Al in high activities, which contributes to forming HIV as well as temporal storage of amorphous Al and organo-mineral complexes in subsoils (Sections 5.3 and 5.8). It is worthwhile to explore the origin of Al in order to understand the pedogenetic conditions of Japan, i.e., whether Al could originate from volcanic materials or from other weatherable primary minerals. The organo-mineral complexes thus accumulated largely modify the charge characteristics of the soils (Section 5.5). In contrast, in similarly highly acidic soils in Indonesia, the migrated Al and DOC did not result in high accumulation of organo-mineral complexes in subsoils. Higher soil temperature and resulting high decomposition rates of SOM may be a possible explanation.
5.9.6 Response of Soil Minerals against Further Acid Load Rapid soil response to the acid load differed among the soils and three types of responses were indicated in Section 5.8. In the tropical soils with large amounts of exchangeable bases, the exchange reactions of Ca and Mg were mainly responsible for the acid neutralization. Some tropical soils had a high ratio of exchangeable to total amounts for Ca and Mg, which means that a high proportion of the bases are easily utilized by plants but will be leached out from soil with the acid load. In the Japanese soils, the dissolution of secondary minerals, i.e., allophane and imogolite, amorphous Al hydroxides, or gibbsite, also contributed to acid neutralization. Secondary mineral dissolution, which occurred rapidly in the acid titration and column experiments, results in the large acid-neutralizing capacity. The organo-mineral complexes, which are substantially accumulated in B horizon soils under a northtemperate zone (Section 5.3) might be involved in these components. In the soils with small amounts of exchangeable bases and low Alo content, and thus with low alkalinity, acid adsorption is relatively important for acid neutralization and the soils more easily decrease in pH. The generation and consumption of ecosystem-internal acids are thus variable depending on climatic and geological conditions and have essential roles both in pedogenetic and ecosystem processes.
Pedogenetic Acidification in Humid Asia
263
REFERENCES Aber, J.D., J.M. Mellio, K.J. Nadelhoffer, C.A. McClaugherty, and J. Paster. 1985. Fine root turnover in forest ecosystems in relation to quantity and form of nitrogen availability: A comparison of two methods. Oecologia 66:317–321. Allen, S.E., H.M. Grimshaw, J.A. Parkinson, and C. Quarmby. 1974. Chemical analysis of ecological materials. New York: Wiley. Andriesse, J.P. 1969. A study of the environment and characteristics of tropical podzols in Sarawak (East-Malaysia). Geoderma 2:201–227. Ashton, P.S. 1988. Dipterocarp biology as a window to the understanding of tropical forest structure. Annual Reviews in Ecology and Systematics 19:347–370. Baba, M., and M. Okazaki. 2000. Change in aluminum pools of Andisols due to soil acidification. Soil Sci. Plant Nutr. 46:797–805. Barnhisel, R.I. and P.M. Bertsch. 1989. Chlorites and hydroxy-interlayered vermiculite and smectite. In Minerals in soil environments, ed. J.B. Dixon and S.B. Weed, 729–788. Madison: Soil Science Society of America. Binkley, D., and D. Richter. 1987. Nutrient cycles and H+ budgets of forest ecosystems. Advances in Ecological Research 16:2–51. Binkley, D. 1992. Proton budgets. In Atmospheric deposition and forest nutrient cycling. Vol. 91 of Ecological studies, ed. D.W. Johnson and S.E. Lindberg, 450–466. New York: Splinger-Verlag. Blum, A.E., and L.L. Stillings. 1995. Feldspar dissolution kinetics. In Chemical weathering rates of silicate minerals. Vol. 31 of Reviews in Mineralogy, ed. A.F. White and S.L. Brantley, 291–351. Washington, DC: Mineralogical Society of America. Bond-Lamberty, B., C. Wang, and S.T. Gower. 2004. A global relationship between the heterotrophic and autotrophic components of soil respiration? Global Change Biology 10:1756–1766. Brady, N.C., and R.R. Weil. 2002. Soil acidity. In The nature and properties of soils, 13th ed., 363–411. Upper Saddle River: Prentice Hall. Brahy, V., H. Titeux, and B. Delvaux. 2000a. Incipient podzolization and weathering caused by complexation in a forest Cambisol on loess as revealed by a soil solution study. Eur. J. Soil Sci. 51:475–484. Brahy, V., H. Titeux, A. Iserentant, and B. Delvaux. 2000b. Surface podzolization in Cambisols under deciduous forest in the Belgian loess belt. Eur. J. Soil Sci. 51:15–26. Bredemeier, M., E. Matzner, and B. Ulrich. 1990. Internal and external proton load to forest soils in Northern Germany. J. Environ. Quality 19:469–477. Campbell, R., and M.P. Greaves. 1990. Anatomy and community structure of the rhizosphere. In The rhizosphere, ed. J. M. Lynch, 11–57. New York: John Wiley and Sons Inc. Carnicelli, S., A. Mirabella, G. Cecchini, and G. Sanesi. 1997. Weathering of chlorite to a low-charge expandable mineral in a Spodosol on the Apennine mountains, Italy. Clays Clay Miner. 45:28–41. Chadwick, O.A., and J. Chorover. 2001. The chemistry of pedogenic thresholds. Geoderma 100:321–353. Chou, L., and R. Wollast. 1984. Study of the weathering of albite at room temperature and pressure with fluidized bed reactor. Geochim. Cosmochim. Acta 48:2205–2217. Cleveland, C.C., A.R. Townsend, and S.K. Schmidt. 2002. Phosphorus limitation of microbial processes in moist tropical forests: Evidence from short-term laboratory incubations and field studies. Ecosystems 5:680–691. Cornu, S., Y. Lucas, J.P. Ambrosi, and T. Desjardins. 1998. Transfer of dissolved Al, Fe and Si in two Amazonian forest environments in Brazil. Eur. J. Soil Sci. 49:377–384. Cronan, C.S., and G.R. Aiken. 1985. Chemistry and transport of soluble humic substances in forested watersheds of the Adirondack Park, New York. Geochimica et Cosmochimica Acta 49:1697–1705.
264
World Soil Resources and Food Security
Cronan, C.S., W.J. Walker, and P.R. Bloom. 1986. Predicting aqueous aluminum concentrations in natural waters. Nature 324:140–143. Davis, J.A., and J.D. Hem. 1989. The surface chemistry of aluminum oxides and hydroxides. In The Environmental chemistry of aluminum, ed. G. Sposito, 185–219. Boca Raton: CRC Press. DeConinck, F. 1980. Major mechanisms in formation of spodic horizon. Geoderma 24:108–128. Devries, W., and A. Breeuwsma. 1987. The relation between soil acidification and element cycling. Water Air Soil Pollut. 35:293–310. de Koning, A., A.V. Konoplev, and R.N.J. Comans. 2007. Measuring the specific caesium sorption capacity of soils, sediments and clay minerals. Appl. Geochem. 22:219–229. Delvaux, B., N. Kruyts, E. Maes, and E. Smolders. 2001. Fate of radiocesium in soil and rhizosphere. In Trace elements in the rhizosphere, ed. G.R. Gobran, W.W. Wenzel, and E. Lombi, 61–91. London: CRC Press. Do Nascimento, N.R., E. Fritsch, G.T. Bueno, M. Bardy, C. Grimaldi, and A.J. Melfi. 2008. Podzolization as a deferralitization process: Dynamics and chemistry of ground and surface waters in an Acrisol-Podzol sequence of the upper Amazon Basin. Eur. J. Soil Sci. 59:911–924. Don, A., and E.D. Schulze. 2008. Controls on fluxes and export of dissolved organic carbon in grasslands with contrasting soil types. Biogeochem. 91:117–131. Dosskey, M.G., and P.M. Bertsch. 1997. Transport of dissolved organic matter through a sandy forest soil. Soil Sci. Soc. Amer. J. 61:920–927. Driscoll, C.T. 1989. The chemistry of aluminum in surface waters. In The environmental chemistry of aluminum, ed. G. Sposito, 241–277. Boca Raton: CRC Press. Duchaufour, P., and B. Souchier. 1978. Roles of iron and clay in genesis of acid soils under a humid, temperate climate. Geoderma 20:15–26. Effendi, S., S. Miura, N. Tanaka, and S. Ohta. 2000. Serpentine soils on catena in the southern part of East Kalimantan, Indonesia. In Rainforest ecosystems of East Kalimantan. Vol. 140 of Ecological studies 140, ed. E. Guhardja, M. Fatawi, M. Sutisna, T. Mori, and S. Ohta. Tokyo: Springer-Verlag Tokyo. Espinoza, W., R.G. Gast, and R.S. Adams. 1975. Charge characteristics and nitrate retention by two Andepts from south-central Chile. Soil Sci. Soc. Am. Proc. 39:842–846. Eyre, S.R. 1963. Vegetation and soils: A world picture. London: Edward Arnold. Food and Agriculture Organization of the United Nations (FAO). 2001. Lecture notes on the major soils of the world. Report 94 of World soil resources, ed. P.M. Driessen, J.A. Deckers, O.C. Spaargaren, and F.O. Nachtergaele, 334. Rome: FAO. FAO/Unesco. 1971. Soil map of the world. Volume IV, South America. Paris: FAO/Unesco. FAO/Unesco. 1977. Soil map of the world. Volume IX, Africa. Paris: FAO/Unesco. Fanning, D.S., V.Z. Keramidas, and M.A. El-Desoky. 1989. Micas. In Minerals in soil environments, ed. J.B. Dixon and S.B. Weed, 551–634. Madison: Soil Science Society of America. Forest Soil Division. 1976. Classification of forest soil in Japan (1975). Bull. Gov. For. Exp. Sta. 280:1–28 (in Japanese). Frankenberger, Jr., W.T., M.A. Tabatabai, D.C. Adriano, and H.E. Doner. 1978. Bromine, chlorine, and fluorine. In Methods of soil analysis, Part 3—Chemical methods, ed. D.L. Sparks, 833–867. Madison: Soil Science Society of America, Inc. and American Society of Agronomy, Inc. Fujii, K., S. Funakawa, C. Hayakawa, S. Sukartiningsih, and T. Kosaki. 2008. Quantification of proton budgets in soils of cropland and adjacent forest in Thailand and Indonesia. Plant Soil 316:241–255. Fujimaki, R., R. Tateno, M. Hirobe, N. Tokuchi, and H. Takeda. 2004. Fine root mass in relation to soil N supply in a cool temperate forest. Ecol. Res. 19:559–562. Funakawa, S., M. Ashida, and K. Yonebayashi. 2003. Charge characteristics of forest soils derived from sedimentary rocks in Kinki District, Japan, in relation to pedogenetic acidification process. Soil Sci. Plant Nutr. 49:387–396.
Pedogenetic Acidification in Humid Asia
265
Funakawa, S., H. Hirai, and K. Kyuma. 1992. Soil-forming processes under natural forest north of Kyoto in relation to soil solution composition. Soil Sci. and Plant Nutr. 38:101–112. Funakawa, S., K. Nambu, H. Hirai, and K. Kyuma. 1993. Pedogenetic acidification of forest soils in northern Kyoto. Soil Sci. Plant Nutr. 39:677–690. Funakawa, S., K. Hirooka, and K. Yonebayashi. 2008. Temporary storage of soil organic matter and acid neutralizing capacity during the process of pedogenetic acidification of forest soils in Kinki District, Japan. Soil Sci. Plant Nutr. 54:434–448. Geological Survey of Japan. 2004. Digital geological map of East and Southeast Asia. 1:2,000,000, 2nd ed. In Geological survey of Japan, Japan: AIST. Gower, C., D.L. Rowell, S. Nortcliff, and A. Wild. 1995. Soil acidification: Comparison of acid deposition from the atmosphere with inputs from the litter/soil organic layer. Geoderma 66:85–98. Griffiths, R.P., J.E. Baham, and B.A. Caldwell. 1994. Soil solution chemistry of ectomycorrhizal mats in forest soil. Soil. Biol. Biochem. 26:331–337. Guggenberger, G., and W. Zech 1994. Dissolved organic carbon in forest floor leachates: Simple degradation products or humic substances? Sci. Total Environ. 152:37–47. Guggenberger, G., and K. Kaiser. 1998. Significance of DOM in the translocation of cations and acidity in acid forest soil. Z. Pflanzenernähr Bodenkd. 161:95–99. Harley, J.L., and S.E. Smith. 1983. Mycorrhizal symbiosis, 201–230. New York: Academic Press Inc. Hallbäcken, L., and C.O. Tamm. 1986. Changes in soil acidity from 1927 to 1982–1984 in a forest area of south-west Sweden. Scad. J. For. Res. 1:219–232. Harris, W.G., A.A. Morrone, and S.E. Coleman. 1992. Occluded mica in hydroxy-interlayered vermiculite grains from a highly-weathered soil. Clays Clay Miner. 40:32–39. Harris, W.G., K.A. Hollien, T.L. Yuan, S.R. Bates, and W.A. Acree. 1988. Nonexchangeable potassium associated with hydroxy-interlayered vermiculite from coastal plain soils. Soil Sci. Soc. Am. J. 52:1486–1492. Helmy, A.K., E.A. Ferriro, and S.G. De Bussetti. 1994. Cation exchange capacity and condition of zero charge of hydroxy-Al montmorillonite. Clays Clay Miner. 42:444–450. Hirai, H., S. Araki, and K. Kyuma. 1988. Characteristics of brown forest soils developed on the Paleozoic shale in northern Kyoto with special reference to their pedogenetic process. Soil Sci. Plant Nutr. 39:157–170. Hirai, H., S. Araki, and K. Kyuma. 1989. Clay mineralogical properties of brown forest soils in northern Kyoto with special reference to their pedogenetic process. Soil Sci. Plant Nutr. 35:585–596. Hirai, H., K. Yoshikawa, S. Funakawa, and K. Kyuma. 1991. Characteristics of brown forest soils developed under different bio-climatic conditions in the Kinki district with special reference to their pedogenetic processes. Soil Sci. Plant Nutr. 37:639–649. Inoue, K., and C. Satoh. 1992. Electric charge and surface characteristics of hydroxyaluminosilicate- and hydroxyaluminum-vermiculite complexes. Clays Clay Miner. 40:311–318. Johnson, D.W. 1977. Carbonic acid leaching in a tropical temperate, subalpine, and northern forest soil. Arctic Alpine Research 9:329–343. Johnson, D.W., D.D. Richter, H. Van Miegroet, and D.W. Cole. 1983. Contributions of acid deposition and natural processes to cation leaching from forest soils: A review. J. Air Pollution Control Association 33:1036–1041. Johnson, C.E., C.T. Driscoll, T.G. Siccama, and G.E. Likens. 2000. Element fluxes and land scape position in a northern hardwood forest watershed ecosystem. Ecosystems 3:159–184. Kalbitz, K., S. Solinger, J.H. Park, B. Michalzik, and E. Matzner. 2000. Controls on the dynamics of dissolved organic matter in soils: A review. Soil Sci. 165:277–304. Kalbitz, K., D. Schwesig, J. Schmerwitz, K. Kaiser, L. Haumaier, B. Glaser, R. Ellerbrock, and P. Leinweber. 2003. Changes in properties of soil-derived dissolved organic matter induced by biodegradation. Soil Biol. Biochem. 35:1129–1142.
266
World Soil Resources and Food Security
Kalbitz, K., K. Kaiser, J. Bargholz, and P. Dardenne. 2006. Lignin degradation controls the production of dissolved organic matter in decomposing foliar litter. Eur. J. Soil Sci. 57:504–516. Kamoshita, Y., M. Kojima, Y. Hasegawa, and Y. Satoh. 1979. “Ko-Sanenki-Hi” An experimental numeral to learn soil buffer against acid and base. J. Soil Sci. Manure Jpn. 50:249– 254 (in Japanese). Kemmitt, S.J., D. Wright, K.W.T. Goulding, and D.L. Jones. 2006. pH regulation of carbon and nitrogen dynamics in two agricultural soils. Soil Biol. Biochem. 38:898–911. Killham, K. 1990. Nitrification in coniferous forest soils. Plant Soil 128:31–44. Kinniburgh, D.G. 1986. Towards more detailed methods for quantifying the acid susceptibility of rocks and soils. J. Geol. Soc. 143:679–690. Kitagawa, Y. 2005. Characteristics of clay minerals in Podzols and Podzolic soils. Soil Sci. Plant Nutr. 51:151–158. Kitayama, K., N. Majalap-Lee, and S. Aiba. 2000. Soil phosphorus fractionation and phosphorus-use efficiencies of tropical rainforests along altitudinal gradients of Mount Kinabalu, Borneo. Oecologia 123:342–349. Kleber, M., L. Schwendenmann, E. Veldkamp, J. Robner, and R. Jahn. 2007. Halloysite versus gibbsite: Silicon cycling as a pedogenetic process in two lowland neotropical rain forest soils of La Selva, Costa Rica. Geoderma 138:1–11. Kleja, D.B., M. Svensson, H. Majdi, et al. 2008. Pools and fluxes of carbon in three Norway spruce ecosystems along a climatic gradient in Sweden. Biogeochem 89:7–25. Komarneni, S., and D.M. Roy. 1980. Hydrothermal effects on cesium sorption and fixation by clay minerals and shales. Clays Clay Miner. 28:142–148. Krug, E.C., and C.R. Frink. 1983. Acid rain on acid soil: A new perspective. Science 221:520–525. Kuylenstierna, J.C.I., H. Rodhe, S. Cinderby, and K. Hicks. 2001. Acidification in developing countries: Ecosystem sensitivity and the critical load approach on a global scale. Ambio 30:20–28. Lasaga, A.C. 1998. Rate laws of chemical reactions. In Kinetic theory in the earth sciences, 3–151. Princeton: Princeton University Press. Laverdiere, M.R., and R.M. Weaver. 1977. Charge characteristics of spodic horizons. Soil Sci. Soc. Am. J. 41:505–510. Le Roux J., and C.I. Rich. 1969. Ion selectivity of micas as influenced by degree of potassium depletion. Soil Sci. Soc. Am. Proc. 33:684–690. Likens, G.E., and F.H. Bormann. 1995. Weathering. In Biogeochemistry of a forested ecosystem, 95–102. New York: Springer-Verlag. Lindsay, W.L. 1979. Aluminum, silica, aluminosilicate. In Chemical equilibria in soils, 34–77. New York: Wiley. Lundström, U.S. 1993. The role of organic acids in the soil solution chemistry of podzolized soil. J. Soil Sci. 48:115–120. Maes, E., B. Delvaux, and Y. Thiry. 1998. Fixation of radiocesium in an acid brown forest soil. Eur. J. Soil Sci. 49:133–140. Maes, E., L. Vielvoye, W. Stone, and B. Delvaux. 1999. Influence of the nature of clay minerals on the fixation of radiocaesium traces in an acid brown earth-podzol weathering sequence. Eur. J. Soil Sci. 50:107–115. McDowell, W.H. 1998. Internal nutrient fluxes in a Puerto Rican rain forest. Journal of Tropical Ecology 14:521–536. McDowell, W.H., and G.E. Likens. 1988. Origin, composition and flux of dissolved organic carbon in the Hubbard brook valley. Ecological Monographs 58:177–195. McKeague, J.A., and J.H. Day. 1966. Dithionite- and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. Can. J. Soil Sci. 46:13–22.
Pedogenetic Acidification in Humid Asia
267
McLean, E.O., and M.E. Watson. 1985. Soil measurements of plant-available potassium. In Potassium in agriculture, ed. R.D. Munson, 277–308. Madison: American Society of Agronomy. Mehra, O.P., and M.L. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite– citrate system buffered with sodium bicarbonate. Clays Clay Miner. 7:317–327. Michalzik, B., K. Kalbitz, J.H. Park, S. Solinger, and E. Matzner. 2001. Fluxes and concentrations of dissolved organic carbon and nitrogen—a synthesis for temperate forests. Biogeochem. 52:173–205. Moore, T.R. 1989. Dynamics of dissolved organic carbon in forested and disturbed catchments, Westland, New Zealand. 1. Maimai. Water Resource Res. 25:1321–1330. Moore, T.R., and R.J. Jackson. 1989. Dynamics of dissolved organic carbon in forested and disturbed catchments, Westland, New Zealand. 2. Larry River. Water Resource Res. 25:1331–1339. Mori, K., H. Shinjo, A. Kato, and T. Kosaki. 2005. Comparison of different soil classification systems using 5 profiles from different forest ecosystems in Japan. Pedologist 49:10–21. Nagy, K.L. 1995. Dissolution and precipitation kinetics of sheet silicates. In Chemical weathering rates of silicate minerals. Vol. 31 of Reviews in Mineralogy, ed. A.F. White and S.L. Brantley, 173–233. Washington, DC: Mineralogical Society of America. Nahon, D.B. 1991. Introduction to major geochemical processes of weathering. In Introduction to the petrology of soils and chemical weathering, 1–49. New York: John Wiley and Sons Inc. Nakane, K. 1980. Comparative studies of cycling of soil organic carbon in three primeval moist forests. Jpn. J. Ecol. 3:155–172. Nakao, A., Y. Thiry, S. Funakawa, and T. Kosaki 2008. Characterization of the frayed edge site of micaceous minerals in soil clays influenced by different pedogenetic conditions in Japan and northern Thailand. Soil Sci. Plant Nutr. 54:479–489. Nambu, K., P.A.W. van Hees, D.L. Jones, S. Vinogradoff, and U.S. Lundström. 2008. Composition of organic solutes and respiration in soils derived from alkaline and nonalkaline parent materials. Geoderma 144:468–477. Nielsen, K.E., K. Dalsgaard, and P. Nornberg. 1987. Effects on soils of an oak invasion of a Calluna heath, Denmark, I. Morphology and chemistry. Geoderma 41:79–95. Nielsen, K.E., U.L. Ladekarl, and P. Nornberg. 1999. Dynamic soil processes on heathland due to changes in vegetation to oak and Sitka spruce. For. Ecol. Manage. 114:107–116. Okamura, Y., and K. Wada. 1983. Electric charge characteristics of horizons of Ando (B) and Red-Yellow B soils and weathered pumices. J. Soil Sci. 34:287–295. Patel-Sorrentino, N., Y. Lucas, F. Eyrolle, and A.J. Melfi. 2007. Fe, Al and Si species and organic matter leached off a ferallitic and podzolic soil system from Central Amazonia. Geoderma 137:444–454. Petersen, L. 1991. Soils of Kalimantan, Indonesia. Folia Geographica Danica 19:173–187. Parfitt, R.L. 1978. Anion adsorption by soils and soil materials. In Advances in agronomy, Vol. 30. New York: Academic Press. Parks, G.A. 1965. The isoelectric points of solid oxides, solid hydroxides and aqueous hydroxo complex systems. Chem. Rev. 65:177–198. Pfisterer, U., H.P. Blume, and M. Kanig. 1996. Genesis and dynamics of an Oxic Dystrochrept and a Typic Haploperox from ultrabasic rock in the tropical rain forest climate of southeast Brazil. Z. Pflanzenernähr Bodenkd. 159:41–50. Qualls, R.G., B.L. Haines, and W.T. Swank. 1991. Fluxes of dissolved organic nutrients and humic substances in a deciduous forest. Ecology 72:254–266. Raich, J.W., A.E. Russell, and R. Bedoya-Arrieta. 2007. Lignin and enhanced litter turnover in tree plantations of lowland Costa Rica. For. Ecol. Manag. 39:128–135.
268
World Soil Resources and Food Security
Rustad, I.E., and C.S. Cronan. 1995. Biogeochemical controls on aluminum chemistry in the O horizon of a red spruce (Picea rubens Sarg.) stand in central Maine, USA. Biogeochem. 29:107–129. Saggar, S., A. Parshotam, C. Hedley, and G. Salt. 1999. 14C-labelled glucose turnover in New Zealand soils. Soil Biol. Biochem. 31:2025–2037. Sanderman, J., and R. Amundson. 2008. A comparative study of dissolved organic carbon transport and stabilization in California forest and grassland soils. Biogeochem. 89:309–327. Sato, K., and H. Ohkishi. 1993. Rapid-neutralizing capacity of surface soils in Japan. Ambio 22:232–235. Sato, K., and A. Takahashi. 1996. Acidity neutralization mechanism in a forested watershed in central Japan. Water Air Soil Pollut. 88:3–4. Schofield, R.K. 1950. Effect of pH on electric charges carried by clay particles. J. Soil Sci. 1:1–8. Schwendenmann, L., and E. Veldkamp. 2005. The role of dissolved organic carbon, dissolved organic nitrogen, and dissolved inorganic nitrogen in a tropical wet forest ecosystem. Ecosystems 8:339–351. Schwertmann, U., and M.L. Jackson. 1963. Hydrogen-aluminum clays: A third buffer range appearing in potentiometric titration. Science 139:1052–1054. Schwertmann, U., and M.L. Jackson. 1964. Influence of hydroxy aluminum ions on pH titration curves of hydronium–aluminium clays. Soil Sci. Soc. Am. Proc. 28:179–183. Shibata, H., M. Kirikae, Y. Tanaka, T. Sakuma, and R. Hatano. 1998. Proton budgets of forest ecosystem on canogenous Regosols in Hokkaido. Water Air Soil Pollut. 105: 63–72. Shibata, H., F. Satoh, K. Sasa, et al. 2001. Importance of internal proton production for the proton budget in Japanese forested ecosystems. Water Air Soil Pollut. 130:685–690. Shinjo, H., A. Kato, K. Fujii, K. Mori, S. Funakawa, and T. Kosaki. 2006. Carbon dioxide emission derived from soil organic matter decomposition and root respiration in Japanese forests under different ecological conditions. Soil Sci. Plant Nutr. 52:233–242. Shoji, S., Y. Fujiwara, I. Yamada, and M. Saigusa. 1982. Chemistry and clay mineralogy of ando soils, brown forest soils, and podzolic soils formed from recent Towada ashes, northeastern Japan. Soil Sci. 133:69–86. Soil Survey Staff. 1999. Soil Taxonomy. A basic system of soil classification for making and interpreting soil surveys. 2nd ed. Washington, DC: US Government Printing Office. Soil Survey Staff. 2006. Keys to soil taxonomy. 10th ed. Washington, DC: US Government Printing Office. Sollins, P., P. Homann, and B.A. Caldwell. 1996. Stabilizatioon and destabilization of soil organic matter: mechanisms and controls. Geoderma 74:65–105. Sparks, D.L. 1989. Kinetics of soil chemical processes. San Diego: Academic Press Inc. SPSS Inc. 1998. SYSTAT 8.0. Statistics. Chicago: SPSS Inc. Stevenson, F.J. 1982. Reactive functional groups of humic substances. In Humus chemistry: Genesis, composition, reactions, 221–243. New York: John Wiley and Sons. Takahashi, M., T. Sakata, and K. Ishizuka. 2001. Chemical characteristics and acid buffering capacity of surface soils in Japanese forests. Water Air Soil Pollut. 130:727–732. Tamura, T. 1958. Identification of clay minerals from acid soils. J. Soil Sci. 9:141–147. Tessens, E., and S. Zauyah. 1982. Positive permanent charge in Oxisols. Soil Sci. Soc. Am. J. 46:1103–1106. Thiry, Y., N. Kruyts, and B. Delvaux. 2000. Respective horizon contribution to cesium-137 soil-to-plant transfer: A pot experiment approach. J. Environ. Qual. 29:1194–1199. Thompson, M.L., and L. Ukrainczyk. 2002. Micas. In Soil mineralogy with environmental applications, ed. J.B. Dixon and D.G. Schulze, 431–466. Madison: Soil Science Society of America.
Pedogenetic Acidification in Humid Asia
269
Thornthwaite, C.W. 1948. An approach toward a rational classification of climate. Geographical Review 38:55–94. Thurman, E.M. 1985. Humic substances in groundwater. In Humic substances in soil, sediment, and water, ed. G.R. Aiken, D.M. McKnight, R.L. Wershaw, and P. MacCarthy, 87–103. New York: Wiley. Tobón, C., J. Sevink, and J.M. Verstraten. 2004a. Litterflow chemistry and nutrient uptake from the forest floor in northern Amazonian forest ecosytems. Biogeochem. 69:315–339. Tobón, C., J. Sevink, and J.M. Verstraten. 2004b. Solute fluxes in throughfall and stemflow in four forest ecosystems in northwest Amazonia. Biogeochem. 70:1–25. Ugolini, F.C., and R.A. Dahlgren. 1987. The mechanism of podzolization revealed by soil solution studies. In Podzols and podzolization, ed. D. Righi and A. Chauvel, 195–203. Paris: Assoc. Franc. Etude Sol. INRA, Plaisir et Paris. Ugolini, F.C., R.A. Dahlgren, S. Shoji, and T. Ito. 1988. An example of andosolization and podzolization as revealed by soil solution studies, southern Hakkoda, northeastern Japan. Soil Sci. 145:111–125. Ugolini, F.C., R.S. Sletten, and D.J. Marrett. 1990. Contemporary pedogenic processes in the Artic: Brunification. Science du Sol 28:333–348. Ugolini, F.C., and R.S. Sletten. 1991. The role of proton donors in pedogenesis as revealed by soil solution studies. Soil Sci. 151:59–75. Ulrich, B. 1989. Effects of acidic precipitation on forest ecosystems in Europe. In Biological and ecological effects. Vol. 2 of Advances in environmental science. Acidic precipitation, ed. D.C. Adriano and A.H. Johnson, 189–262. New York: Splinger-Verlag. van Breemen, N., J. Mulder, and C.T. Driscoll. 1983. Acidification and alkalization of soils. Plant Soil 75:283–308. van Breemen, N., C.T. Driscoll, and J. Mulder. 1984. Acidic deposition and internal proton in acidification of soils and waters. Nature 307:599–604. van Hees, P.A.W., U.S. Lundström, and R. Giesler. 2000. Low molecular weight organic acids and their Al-complexes in soil solution-composition, distribution and seasonal variation in three podzolized soils. Geoderma 94:173–200. Van Raij, B., and M. Peech. 1972. Electrochemical properties of some Oxisols and Alfisols of the tropics. Soil Sci. Soc. Am. Proc. 36:587–593. Velde, B. 2001. Clay minerals in the agricultural surface soils in the Central United States. Clay Minerals 36:277–294. Wada, S.-I., and K. Wada. 1985. Charge characteristics and exchangeable cation status of Korean Ultisols and Alfisols and Thai Ultisols and Oxisols. J. Soil Sci. 36:21–29. Watanabe, T., S. Funakawa, and T. Kosaki. 2007. Profile description, properties, and classification of seven upland soils formed under different conditions in Japan, Thailand, and Indonesia. Pedologist 51:24–34. Wieder, W.R., C.C. Cleveland, and A.R. Townsend. 2008. Tropical tree species composition affects the oxidation of dissolved organic matter from litter. Biogeochem. 88:127–138. Yamakura, T., A. Hagihara, S. Sukardjo, and H. Ogawa. 1986. Aboveground biomass of tropical rain forest stands in Indonesia Borneo. Vegetation 68:71–82. Yano, Y., K. Lajtha, P. Sollins, and B.A. Caldwell. 2004. Chemical and seasonal controls on dynamics of dissolved organic matter in a coniferous old-growth stand in the Pacific Northwest, USA. Biogeochem. 71:197–223. Yavitt, J.B., and T.J. Fahey. 1986. Litter decay and leaching from the forest floor in Pinus contorta (lodgepole pine) ecosystems. J. Ecol. 74:525–545. Zech, W., and G. Guggenberger. 1996. Organic matter dynamics in forest soils of temperate and tropical ecosystems. In Humic substances in terrestrial ecosystems, ed. A. Piccolo, 101–170. Amsterdam: Elsevier. Zech, W., N. Senesi, G. Guggenberger, et al. 1997. Factors controlling humification and mineralization of soil organic matter in the tropics. Geoderma 79:117–161.
Resources Affecting 6 Soil Food Security and Safety in South Asia Tapan J. Purakayastha, Bal Ram Singh, R. P. Narwal, and Promod K. Chhonkar CONTENTS 6.1 Introduction................................................................................................... 272 6.1.1 Magnitude of Food Insecurity........................................................... 272 6.1.2 Linking Soils and Food Production................................................... 272 6.1.3 Current Situation and Trends............................................................. 273 6.1.4 Food Safety Issues............................................................................. 275 6.2 Soil Resources............................................................................................... 275 6.2.1 Total Arable Lands............................................................................ 275 6.2.2 Per Capita Land Availability and Future Land Availability Trends.....276 6.3 Causes of Food Insecurity............................................................................. 278 6.3.1 Soil Degradation................................................................................ 278 6.3.1.1 Current Trends in Soil Degradation in South Asia............. 281 6.3.2 Urban Encroachment......................................................................... 290 6.3.3 Water Supply...................................................................................... 291 6.3.4 Poor Management and Limited Nutrient Supply............................... 292 6.4 Food Production and Supply......................................................................... 293 6.4.1 Current Production: Total and per Capita.......................................... 293 6.4.1.1 India.................................................................................... 293 6.4.1.2 Bangladesh.......................................................................... 294 6.4.2 Progress and Future Projections........................................................ 295 6.4.3 Strategies for Improving Food Security............................................ 296 6.4.3.1 India.................................................................................... 297 6.4.3.2 Bangladesh.......................................................................... 301 6.4.3.3 Nepal................................................................................... 303 6.4.3.4 Sri Lanka............................................................................. 305 6.4.3.5 Pakistan...............................................................................306 6.5 Food Safety and Quality................................................................................307 6.5.1 Elements of Food Safety....................................................................307 6.5.1.1 Malnutrition and Micronutrients........................................307 6.5.1.2 Toxic Agents.......................................................................308 6.5.2 Food Safety in Production and Postharvest.......................................308 271
272
World Soil Resources and Food Security
6.5.2.1 Conventional Breeding.......................................................308 6.5.2.2 Fertilizer Use......................................................................309 6.5.2.3 Bioavailability Issue............................................................309 6.5.2.4 Postharvest and Storage...................................................... 310 6.5.3 Safety Legislations............................................................................. 310 6.6 Strategies for Improving Security and Safety of Food.................................. 311 6.7 Conclusion and Future Perspectives.............................................................. 312 References............................................................................................................... 312
6.1 INTRODUCTION 6.1.1 Magnitude of Food Insecurity Food security refers to the availability of food and one’s access to it. A household is considered food secure when its occupants do not live in hunger or fear of starvation. Worldwide around 1 billion people are chronically hungry due to extreme poverty, while up to 2 billion people lack food security intermittently due to varying degrees of poverty [FAO 2003]. For many of these people, food security depends on income from agriculture, and thus on the quality and productivity of agricultural inputs such as land and labor. Six million children die of hunger every year—17,000 every day. In 2007, increased farming demand for biofuels, world oil prices at more than $100 a barrel, global population growth, climate change, loss of agricultural land to residential and industrial development, and growing consumer demand in China and India had pushed up grain prices. In India, 30 million people have been added to the ranks of the hungry since the mid-1990s and 46% of all children are underweight. It is not because India does not have sufficient food in the buffer stock but because of the poor purchasing power of the ~30% of its population who live below the poverty line. Not only India but other south Asian countries have also been facing the food insecurity problem and some of them even with greater intensity. An alternative view takes a collective approach to achieving food security. This view states that, globally, enough food is produced to feed the entire world population at a level adequate to ensure that everyone can be free of hunger and that no one should live without enough food because of economic constraints or social inequalities. This approach is often referred to as “food justice,” and it views food security as a basic human right. It advocates fairer distribution of food, particularly grain crops, as a means of ending chronic hunger. The core of the food justice movement is the belief that what is lacking is not food, but the political will to fairly distribute food regardless of the recipient’s ability to pay. More than half of the planet’s population, approximately 3.3 billion people, live in urban areas as of November 2007. Any disruption to farm supplies may precipitate a uniquely urban food crisis in a relatively short time. The ongoing global credit crisis has affected farm credits, despite a boom in commodity prices.
6.1.2 Linking Soils and Food Production Soils are fundamental to supporting plant, animal, and human life on earth. For sustainable crop production, it is necessary to maintain soils in situ and in a healthy
Soil Resources Affecting Food Security and Safety in South Asia
273
and fertile state. Food production is directly linked with a soil’s ability to produce crops. Degradation of soil in terms of physical, chemical, and biological properties could adversely influence crop production. If a soil is degraded, it cannot produce crops, even if all other inputs are optimum. With proper management, it can produce a bumper crop. Judicious soil management is proving to be an increasingly difficult and challenging task in many parts of the world. Although nutritious and adequate food is the focus of ensuring food security and safety, soils also support the production of other essential materials and services such as fiber for clothing and bedding and timber for fuel and housing. In essence, soil is the thin skin of planet earth which must be protected from damage by soil erosion, contamination, pollutants, and nutrient depletion by extractive farming. Misuse of soils has threatened ecosystems and the ability of agricultural ecosystems to meet global food needs. Soil is a finite natural resource limited by the area of land. Moreover, many of the soils cannot be used for agriculture because they are too dry, too wet, too cold, too shallow, or too toxic. Only 10%–12% of earth’s soils have no natural limitations in terms of agricultural production. Soil degradation caused by erosion, waterlogging, and salinity are the major contributors for lower productivity. As a matter of fact, the plants that are waterlogged are very susceptible to salinity, especially in their early growth stages [Barrett-Lennard 2002]. The development of extensive areas of secondary salinization and waterlogging has been a feature of agricultural lands in parts of western and southwestern Australia for over 100 years [Bennett and Macpherson 1983; McFarlane and Williamson 2002]. Thus, waterlogging and salinity problems pose a serious threat to the world’s productive agricultural land. In south Asia, vast areas of land are prone to chemical, physical, and biological degradation of soil resources leading to deterioration in quantity and quality of food production.
6.1.3 Current Situation and Trends In addition to their limited area, soils are also subject to degradation and loss due, principally, to erosion by wind and water and also physical (e.g., compaction, loss of structure) and chemical (e.g., salinity, toxic contamination) degradation. Erosion is a part of the natural geological process, but the rate and extent of soil erosion and degradation have increased as a result of human activities. These include overgrazing, excessive cultivation, farming of unsuitable land, and contamination through industrial activity. It has been estimated that, globally, as much as two billion hectares of arable and grazing land is affected by moderate to severe degradation, while around seven million hectares of farmland are permanently lost every year [Mackenzie and Mackenzie 1995]. The South Asian region is undergoing rapid industrialization and economic growth, leading to increases in emission of CO2 and other greenhouse gases (GHGs) into the atmosphere. Furthermore, the urgency of meeting increased demands for agricultural produce (food, feed, fiber, and fuel) is rapidly degrading soil quality and exacerbating degradation. Most agricultural soils have low soil organic matter (SOM) reserves due to fertility-mining practices (residue removal, uncontrolled and excessive grazing, imbalance in application of plant nutrients, etc.) and widespread problems of soil degradation.
274
World Soil Resources and Food Security
Although new soil is formed by natural weathering, the process is slow and it takes millennia for a new soil to develop. In many areas the rate of erosion greatly exceeds the rate of soil formation. Moreover, large areas of highly fertile soils are removed from production as new cities are created and urban ecosystems expand due to bourgeoning populations. Many new global threats and demands have emerged over the past decade to aggravate stresses on soil resources. These include increases in population, climate change, production of biofuels, sequestration of carbon by soils, increases in the demand for animal protein, and rising competition for water. A further complication is that the amount of land suitable for cultivation is not evenly distributed around the world, not only with respect to area but particularly with respect to population density. The impact of this uneven distribution is evidenced by the large areas of land being cleared and burnt in South Asia due to the heavy pressure on land resulting from the increasing demand for food by a rapidly increasing population. With changes in agronomic and cultivation practices, soil degradation can be controlled and reversed so that the loss of productive soils can be minimized. The price of food needs to be maintained at a level so that soils are not overexploited and farmers are able to maintain the health and fertility of their soils. Cities should be built on agriculturally marginal rather than the prime soils. It is worth noting that agriculture has made massive advances in efficiency and productivity over the past century driven largely by the use of fertilizers, improved machinery and mechanization, and agrochemicals. However, there is an upper biological limit to what can be achieved by these means. Modeling suggests that the same operating paradigm will enable agriculture to feed a population of 8.5 billion by 2025, but it is less clear that increased food production could be continued further without a major change from the current production paradigm or, perhaps more importantly, a limit on human population growth. Physical and chemical properties of soils vary widely depending on drainage, topography, native vegetation, and past management practices. Geographical information systems (GIS) have given farmers and soil scientists the ability to treat microecosystems according to their production attributes, spatial variability, and soil quality. Soil quality parameters can be assessed and improved in small-scale agroecosystems. Adoption of these procedures should improve environmental quality and protect soil for future generations. The modern era of agriculture has witnessed a major change in the way soils are treated. Despite Leopold speaking convincingly of the land ethic, land has continued to be degraded in its ability to function in ecosystems. The advent of fossilfuel-driven food production, increasing demands for food and fiber, and economic pressures that drive agriculture to produce more without considering the long-term effects are among the leading problems. Many statistics point out the problems and challenges facing agriculture. Agricultural sustainability concerns are being driven by limited land resources, increasing population in developing countries of the tropics and subtropics, conversion of land to urban and industrial uses, and the continued failure to address hunger and malnutrition in several regions of the world.
Soil Resources Affecting Food Security and Safety in South Asia
275
6.1.4 Food Safety Issues Soil degradation directly influences the quality of the crop produced. If a soil is deficient in nutrients and it is not properly managed, it not only adversely affects crop production, but also causes nutrient deficiency in the produce. In south Asia, vast areas of soils are contaminated with heavy metals and metalloids like arsenic and selenium. Thus, the agricultural products of these areas are a matter of concern with respect to food safety. Meanwhile, concerns remain about the impact of agricultural intensification on the qualities of soil, water, food produced, and other environmental resources. In order to address these challenges, it is extremely essential to understand the links between soil quality, soil degradation, agricultural productivity, and food security. This review focuses on soil resources in terms of arable lands available in the South Asian countries, describes causes and magnitude of soil degradation, and explains how these resources affect the current food security and safety in the regions. Furthermore, strategies for enhancing food security and safety are proposed for the South Asian region.
6.2 SOIL RESOURCES 6.2.1 Total Arable Lands An assessment [Eswaran et al. 1999] adds to the general belief [Greenland et al. 1998] that the world’s resources are adequate to produce enough food for its population in the next few decades. The Indo-Gangetic Plain (IGP) is one of the cradles of human civilization, blessed with relatively good land resources, and have evolved a culture and heritage unrivaled in the world. With time, the population has spread to exploit much of the 4.78 M km2 land comprising South Asia. Today, the region comprises Afghanistan, Bangladesh, Bhutan, India, Nepal, Pakistan, and Sri Lanka, with a combined population of about 1.5 billion people, which is 38% of the combined population of South, Southeast, and East Asia, and about 15% of the global population. The constraint-free land area [Dent 1990] that must support this region is only about 1.7% of the global land area or 3.4% of the Asian region. India’s geographical area is 328.7 million hectares (Mha). The net sown area, which increased from 119 Mha in 1950–1951 to 140 Mha in 1970–1971, has been ~141 Mha during the past four decades. Nevertheless the gross cropped area (GCA) has been consistently increasing, owing to double and multiple cropping systems. The GCA has increased from 132 Mha in 1095–1951 to 193 Mha in 2005–2006, with cropping intensity going up from 111% to 136%. The intensity increased mainly due to expansion of irrigation facilities, contributed to by public and farmer investment in irrigation. Since 1951, the net sown area has increased by 18% while the population has nearly tripled. The cropping intensity increased from 111% to 135%, and the area under nonagricultural use increased from 9 Mha to 25 Mha. South Asia covers a large land area (Table 6.1) with diverse ecoregions, land uses, and management practices. Over 22% of the world’s population lives on less than 5% of the world’s land area [FAO 1994]. The cropland area represents 34% of the
276
World Soil Resources and Food Security
TABLE 6.1 Area (M ha) under Different Land Use in Selected Countries of South Asia Country
Total Land Agricultural Arable Permanent Permanent Forest & Irrigated Area Area Area Land Crops Pasture Woodland Cropland
Bangladesh 14.4 13.0 Bhutan 4.5 4.7 India 328.7 297.3 Nepal 14.7 14.3 Pakistan 79.6 77.1 Sri Lanka 6.6 6.5 Total 678.5 641.7
9.1 0.6 180.8 4.9 27.2 2.4 323.6
8.1 0.15 161.8 3.1 21.5 0.9 217.8
0.4 0.02 8.1 0.09 0.7 1.0 12.7
0.6 0.4 10.9 1.8 5.0 0.4 93.8
1.3 3.0 64.1 3.9 2.5 1.9 85.4
4.4 0.04 54.8 1.1 17.8 0.6 88.6
Source: FAO, 2001.
total land area, and the high proportion is a reflection of the high population density. In accord with the land and population statistics, the per capita land area in some countries is <0.1 ha and decreasing over time. There is no possibility of expansion of cropland area, and the potential of expanding irrigated land area is also limited.
6.2.2 Per Capita Land Availability and Future Land Availability Trends The ratio of population to constraint-free land has increased steadily in almost all the South Asian countries since 1960 (Table 6.2). The increase was higher between 1960 and 2000 than between 2000 and the projected ratio in 2025. The ratio has increased because of the reclamation of problem soils to constraint-free land. Eswaran et al. [1999] indicated that constraint-free land, when managed adhering to principles of sustainable management, can support about 6.2 billion people under low-input systems, 8.7 billion under medium-input, and about 19 billion under high-input systems. Realistically, there will be an adequate supply of grains to feed the 9 billion people who will inhabit the earth in the year 2025; this does not imply that all will be fed adequately. The study also indicated that the better endowed areas for food production occur in the temperate regions of the world with the tropics having proportionately more constraints. Lal [1989] suggests that to sustain human population at an acceptable level, about 0.5 ha of cropland per capita is needed. Thus, each hectare of land must have the capacity to provide enough food for at least two persons. In Table 6.2, the estimated population at three times (1960, 2002, and 2025) is related to the constraint-free land by calculating a ratio between population and constraint-free land, assuming that all this land is used for grain production. Today, each hectare of land must support at least 5 persons in India, 10 persons in Bangladesh, and 20 persons in Pakistan. Based on the grain needs of an individual and the current scientific technology, the expected productivity of the land to meet these requirements is unrealistically high. The burgeoning population is making the people to land ratio increasingly unfavorable. Per capita, the total availability of land has decreased drastically from 0.91
Population (millions)
Countries
Total Area (km2)
1960
2000
2025
Constraint-Free Land (km2)
Bangladesh Bhutan India Nepal Pakistan Sri Lanka Total
133,910 47,000 2,973,190 136,800 778,720 64,740 4,781,860
51.41 0.85 442.34 9.44 49.96 9.89 574.66
130.50 1.10 1100.00 25.00 150.00 20.20 1452.80
196.13 3.14 1600.00 40.69 284.83 25.03 2195.08
13,0387 27,925 2,310,986 91,552 79,356 62,992 2,734,872
Ratio of Population to Constraint-Free Land (persons/hectare manageable land) 1960
2000
2025
3.9 0.3 1.9 1.0 6.3 1.6 0.2
10.0 0.4 4.8 2.7 18.9 3.2 0.5
15.0 1.1 6.9 4.4 35.9 4.0 0.8
Source: Eswaran, H., et al., International Conference on Managing Natural Resources for Sustainable Agricultural Production in the 21st Century, Feb. 14–18, 2000, 97–109.
Soil Resources Affecting Food Security and Safety in South Asia
TABLE 6.2 Estimates of Land Resources and Population Attributes for the Countries of South Asia
277
278
World Soil Resources and Food Security
ha in South Asia in 1951 to about 0.32 ha in 2001 against a world average of 2.19 ha. Similarly per capita net-sown area availability reduced from 0.33 ha in 1951 to 0.14 ha in 2001, and is expected to decline further to 0.09 ha by 2050. A minimum economic holding size of 2 ha of unirrigated land and 1 ha of irrigated land has been suggested in India for sustaining a family of 5 to 6 persons [ADB 2007]. Against 0.32 ha, a minimum of 0.5 ha of land is required for producing a diverse diet similar to that of the United States and Europe [Pimentel and Pimentel 2008]. The decrease in per capita land area is a serious issue, but the decline in soil quality is even a bigger challenge.
6.3 CAUSES OF FOOD INSECURITY Food security refers to the availability of food and access to it. A household is considered food secure when its occupants do not live in hunger or fear of starvation. Two commonly used definitions of food security come from the United Nation’s Food and Agriculture Organization (FAO) and the United States Department of Agriculture (USDA): • Food security exists when all people, at all times, have physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life. [FAO 2003] • Food security for a household means access by all members at all times to enough food for an active, healthy life. Food security includes at a minimum: (1) the ready availability of nutritionally adequate and safe foods and (2) an assured ability to acquire acceptable foods in socially acceptable ways (that is, without resorting to emergency food supplies, scavenging, stealing, or other coping strategies). [USDA 2008]
6.3.1 Soil Degradation Land is a limited and nonrenewable resource and experience has shown that with continuous utilization, even with the best technologies and skills, its quality is deteriorating. Soil degradation is defined as “a process which lowers the current and/or the potential capability of soil to produce (quantitatively and/or qualitatively) goods or services” [FAO 1979]. Soil degradation has also been defined as “the decline in soil quality caused through its misuse by humans” [Lal et al. 2005]. It is a broad and vague term; however, it refers to decline in the soil’s productivity through adverse changes in nutrient status and SOM, structural attributes, and concentrations of electrolytes and toxic chemicals. In other words, it refers to a diminution of the soil’s current and/or potential capability to produce quantitative or qualitative goods or services as a result of one or more degradative processes [UNEP 1994]. Soil degradation has been defined in many ways, most often referring to (the agro) productive function of the soil [Barrow 1991]. In a general sense, soil degradation could be described as the deterioration of soil quality or in other words the partial or entire loss of one or more functions of the soil [Blum 1988]. Hence, it is imperative to not only preserve soil quality but also to adopt measures to ameliorate it while keeping pace with increasing demands from various sectors.
Soil Resources Affecting Food Security and Safety in South Asia
279
Soil degradation can be classified in three distinct categories: (1) physical degradation, (2) chemical degradation, and (3) biological degradation. Each of these types has different processes. Physical degradation includes soil erosion, and waterlogging etc. Chemical degradation includes sodicity, alkalinity and acidity of soil, and accumulation of toxic heavy metals (Zn, Cu, Pb, Cd, Hg, Ni, etc.) and metalloid (As, Se) in soil and biological degradation is erosion of diversity of soil flora and fauna. Soil degradation results from natural hazards, direct causes, and underlying causes. The natural hazards are environmental conditions, which lead to high susceptibility to erosion such as steep slopes, high intensity rains, and high speed winds. The direct causes include deforestation, over-cutting of vegetation, shifting cultivation, over-grazing, non-adoption of proper soil conservation measures, extension of cultivation into fragile or marginal lands, improper crop rotation, and faulty management practices. These underlying causes are basically due to high population, lack of land tenure, and open access to resources, economic pressure, poverty, and poor infrastructure. Irrigated agriculture, especially through canal systems, has resulted in degradation due to waterlogging and secondary salinization. India faces a major challenge of producing 310 million Mg of food grains and 190 million Mg of fibers and edible and nonedible oils by 2050 from 140 Mha of land. According to the Directorate of Economics and Statistics (DES), India would require an agronomic productivity of 3.3 Mg/ha against an average productivity of all food grains of 1.7 Mg/ha in 2006–2007, which is an increase of 94% over 43 years [DES 2007]. Hence, there is an urgent need to enhance productivity of arable and nonarable lands to meet the requirements of a growing population with shrinking soils and other resources. The Global Assessment of Human-Induced Soil Degradation (GLASOD) indicated that 15% of land is degraded [Bai et al. 2008]. The highest proportions were reported for Europe (25%), Asia (18%), and Africa (16%); the least for North America (5%). By the same measure, as a proportion of the degraded area, soil erosion affects 83% of the global degraded area—ranging from 99% in North America to 61% in Europe; nutrient depletion affects 4% globally, but 28% in South America; salinity affects less than 4% worldwide, but 16% in West Asia; chemical contamination affects about 1% globally, but 8% in Europe; soil physical problems affect 4% globally, but 16% in Europe. Recent estimates by Bai et al. [2008] state that 1964 Mha of the world’s soils are degraded, and in Asia (excluding West Asia) alone, the degraded soils occupied an area of 747 Mha (38%) (Figure 6.1). Intensive farming often leads to a vicious cycle of exhaustion of soil fertility and decline of agricultural yields. Approximately 40% of the world’s agricultural land is seriously degraded. Over the past 25 years, 24% of the land area has been degrading, directly affecting the livelihoods of 1.5 billion people; this is on top of the legacy of thousands of years of mismanagement in some long-settled areas. GLASOD estimated that 15% of the land was degraded (Table 6.3), much of which does not overlap with the areas highlighted by the new analysis; land degradation is cumulative—this is the global issue [Bai et al. 2008]. It is estimated that about 1.964 billion ha in the world is affected by various forms of human-induced land degradation [Bai et al. 2008] with erosion by water being the chief contributor (1094 Mha), followed by wind erosion (548 Mha),
280
World Soil Resources and Food Security 1200
1094
World Asia
800 548
600 440 400
222
th e
r
10 3 O
al
12
Ph ys ic
at io in
lin
am
nt rie
79
22 2
N
ut
Sa
n de p
le
tio
os io er d
W in
W at e
re
ro
sio
n
n
0
76 53 ity
15
n
135
nt
200
Co
Degraded soil (Mha)
1000
FIGURE 6.1 GLASOD estimates of human-induced soil degradation. (Extracted from Bai, Z.G., et al., Soil Use Manag., 24, 223–234, 2008.)
nutrient depletion (135 Mha), salinity (76 Mha), contamination (22 Mha), and physical (79 Mha) and other (10 Mha) degradation. About 5–7 million hectares of arable land of the world is lost annually through land degradation [Lal and Stewart 1990]. Rich governments and corporations are buying up the rights to millions of ha of agricultural land in developing countries in an effort to secure their own long-term food supplies.
TABLE 6.3 Extent of Soil Degradation and Net Primary Productivity (NPP) Loss in South Asian Countries (1981–2003) Country India Bangladesh Pakistan Nepal Sri Lanka Bhutan Total
Degradation (Mha)
% Global Degradation Area
Total NPP Loss (tons C/23 yr)
59.25 6.84 2.06 5.47 2.11 2.70 78.43
1.751 0.199 0.073 0.182 0.060 0.073 2.33
16.5 49.12 3.58 48.93 25.62 66.5 210.25
Source: Bai, Z.G., et al., Soil Use Manag., 24, 223–234, 2008.
Soil Resources Affecting Food Security and Safety in South Asia
281
6.3.1.1 Current Trends in Soil Degradation in South Asia The principal causes of soil degradation and land desertification in the region are erosion by water, followed by wind and biophysical and chemical degradation. Arid and semiarid regions of Afghanistan (85%), India, and Pakistan are prone to desertification problems. In India, approximately 57% of the land is prone to some form of degradation. In Bhutan, because of its low population density, soils have not yet been affected by degradation. However, deforestation, which is often the initial cause of soil degradation, is accelerating and 10% of Bhutan’s agricultural land is already affected by soil erosion. Four countries with humid climates (Bangladesh, Nepal, Sri Lanka, and the greater part of India) are severely affected by water erosion. In addition, the rain-fed lands are affected by soil fertility decline and by deforestation [SACEP 2010]. In parts of the hill and mountain areas of Nepal, deforestation and water erosion have reached an extreme degree. Wind erosion is extensive in India and Pakistan, affecting about 25 Mha of land. The most devastating form of degradation in Bangladesh is riverbank erosion. It is a serious problem in the active floodplains of the Ganges, the Brahmaputra-Jammua, the Tista, and the Meghna rivers. Waterlogging and salinization affect between 2 and 3 Mha of land in India and Pakistan, respectively. Particularly, large parts of the IGP in India and Pakistan and, to some extent, Bangladesh are already affected by soil salinity and water logging [Joshi et al. 2002]. In Pakistan, salt build-up in the soil is known to reduce crop yields by 30%. In Bangladesh, over 30% of the land available for cultivation is situated in the coastal belt, and most of the land is not utilized for crop production due to soil salinity. Chemical soil degradation in the region is mainly caused by agricultural mismanagement and industrial pollution. 6.3.1.1.1 India Of India’s total land area of 328.7 Mha, 59.3 Mha (18%) land are degraded [FAO AGL 2005]. The intensity of degradation is light (0.08 Mha), moderate (1.91 Mha), severe (34.32 Mha), and very severe (22.94 Mha) [FAO AGL 2005]. Estimates of the land area affected by degradation in India vary widely among agencies and ranges from about 53 Mha to 188 Mha. Variations in estimates are attributed to different approaches and methodologies being adopted in defining degraded lands and/or differentiating criteria used. These data sets have now been harmonized to 121 Mha of degraded lands, which respond to application of restorative amendments. Further, 8.4 Mha of the irrigated lands are affected by soil salinity and alkalinity; of which about 5.5 Mha are also waterlogged [IDNP 2002]. Adverse impacts of degradation include more landlessness, lower and less reliable food production, increased labor requirements, and lower incomes. Chemical degradation is the result of environmental conditions or anthropogenic activities. The physical and chemical degradation together adversely affect soil erosion and biological activities of flora and fauna. The former degradation is the cause and the latter is the effect. A major factor responsible for degradation of the natural resource is soil erosion. Accelerated soil erosion has irreversibly destroyed some 430 Mha of land area
282
World Soil Resources and Food Security
covering 30% of the present cultivated area in different countries of the world [Brown 1984]. Over population, harsh climatic conditions, overexploitation, and unwise use of soil resources, deforestation, etc.—resulting in soil-food population imbalance— have rendered most of the tropical and subtropical ecosystems extremely vulnerable to soil erosion and erosion-induced degradation. Dhruva Narayana and Babu [1983] analyzed the existing soil loss data from India and concluded that soil erosion was taking place at an average rate of 16.35 Mg/ha/yr totaling 5334 million Mg/yr. The annual water erosion rate values ranged from less than 5 Mg/ha/yr (for dense forests, snow-clad cold deserts, and the arid regions of Western Rajasthan) to more than 80 Mg/ha/yr in the Shiwalik hills. Ravines along the banks of the rivers Yamuna, Chambal, Mahi, Tapti, and Krishna, and shifting cultivation regions of Orissa and the northeastern states experience soil erosion rates exceeding 40 Mg/ha/yr. The annual erosion rate in the Western Ghats coastal regions ranges from 20 to 30 Mg/ha/yr. Of this, 29% is lost permanently to sea, 10% is deposited in reservoirs, decreasing their capacity by 1%–2% every year, and the remaining 61% is redistributed on land. Among different land resource regions, the highest erosion occurs in black soils (24–112 Mg/ha), followed by the Shiwalik region (80 Mg/ha), the northeastern region with shifting cultivation (27–40 Mg/ha), and least in the North Himalayan forest region (2 Mg/ha). Permissible soil loss in India varies from 2.5 to 12.5 Mg/ha/yr, depending upon the soil order and soil depth [Mandal et al. 2008]. Wind erosion is a serious problem in the arid and semiarid regions, including the states of Rajasthan, Haryana, Gujrat, and Punjab. Removal of natural vegetative cover by excessive grazing and the extension of agriculture to marginal areas are the major human-induced factors leading to accelerated wind erosion. Wind erosion is also prevalent in coastal areas where sandy soils predominate and in the cold desert regions of Leh in the extreme northwest covering an area of 28,600 km2, of which 68% is covered by sand dunes and sandy plains [Gupta et al. 1990]. However, active wind erosion is observed in the extreme western sectors of the country. The dune soils are single-grained, noncoherent, and structureless. Owing to low silt and clay contents, the unstabilized dunes have hardly any crust formation. The expansion of irrigation has been one of the key strategies in achieving selfsufficiency in food production. The net irrigated area in India has increased from about 22 Mha in 1950 to about 87.26 Mha in the year 2007–2008 [DES 2010]. When the groundwater table reaches within 2 m of the surface, it contributes significantly to evaporation from the soil surface and causes secondary salinization. In India, about 7 Mha are salinized [Abrol and Bhumbla 1971], of which 2.5 Mha represent alkali soils in the IGP. Nearly 50% of the canal-irrigated areas are suffering from salinization and/or alkalization due to inadequate drainage, inefficient use of available water resources, and sociopolitical reasons [Abrol and Bhumbla 1971]. Typical examples of salinization caused by the rise in groundwater are observed in Uttar Pradesh, Haryana, Rajasthan, Maharastra, and Karnataka. In the Sultanpur and Etah districts of Uttar Pradesh, groundwater (brackish in nature) has risen to within one m of the surface, aggravating salinity and forcing the farming community to abandon the affected areas. In many coastal regions, especially Gujarat, exploitation of groundwaters has caused the intrusion of sea water, resulting in a rise in the groundwater table and soil salinity problems.
Soil Resources Affecting Food Security and Safety in South Asia
283
The term waterlogging refers to a condition of short- or long-term water stagnation caused by changes in the hydrologic regime, landscape, developmental activities, silting up of river beds, etc. Repeated flooding is yet another cause of waterlogging in coastal and flood-plain areas of major rivers. The introduction of canal irrigation in India has resulted in almost 7 Mha of cultivated land becoming affected by soil salinity and waterlogging [Joshi and Tyagi 1994]. In India, the National Commission for Irrigation [1972], National Commission on Agriculture [1976], and the Ministry of Water Resources [1991] have estimated the extent of the waterlogged area as 4.84, 6.00, and 2.46 Mha, respectively. Bhattacharya [1992] reported that the total area suffering from waterlogging in India is estimated to be about 3.3 Mha and the state of Bihar alone constitutes an area of nearly 0.9 Mha. Hoffman and Durnford [1999] reported how soil salinity and waterlogging problems have developed worldwide, and the speed with which they are advancing at present. In recent years, several thousand hectares of land are also degraded by waterlogging by sea water during the severe natural calamities like Tsunami, Aila, Ogni, Baaz, Mala, etc. [DES 2010]. Since about 70% of land has a soil-loss tolerance limit of 10 Mg/ha/yr, that rate has been used for computation of degraded lands in the country covering 120.7 Mha. This analysis has revealed that about 39% of land area has erosion exceeding the permissible rate, thereby resulting in reduced agricultural productivity. About 11% of the area falls in the very severe category with erosion rates of >40 Mg/ha/yr. Some of the states in the northwest and northeast Himalayas are the worst-affected, with more than one-third of their geographical area affected by very severe soil erosion (40–80 Mg/ha/yr) category. In India, total wastewater generated per annum from industry and the domestic front from 200 cities is about 2600 Mm3. Thus the long-term applications of sewage effluents and sludge (both industrial and domestic) have been reported to increase the concentrations of trace metals (copper, zinc, lead, cadmium, nickel, chromium, arsenic, selenium, etc.) significantly in large areas under periurban agriculture around Delhi, Kolkata, Ludhiana, Kanpur, Varanasi, Hyderabad, Patna, Maduarai, and Coimbatore. In India more than 1000 ha of selenium-contaminated soils exist in the Hoshiarpur and Nawansahar districts of northwestern Punjab [Sanyal and Dhillon 2005]. In the Bengal Basin of Bangladesh and West Bengal, India [Bhattacharya et al. 2004, 2006], Aresenic (As) concentration in groundwater has emerged as the largest environmental health disaster, putting at least 100 million people at risk of cancer and other As-related diseases. The geogenic As occurs in the Central Gangetic Plains of Uttar Pradesh, Bihar, Jharkhand, and the Brahmaputra valley in Assam, and several regions of Madhya Pradesh and Chattisgarh, India [Chakraborti et al. 2004; Mukherjee et al. 2006]. Total and available As in affected soils of West Bengal ranges from 2.9 to 24.3 mg/kg and the boro (winter) rice grown on such soils exhibits an elevated concentration of As in the grain (10 mg/kg). Rattan et al. [2005] reported that soils in several villages of Delhi that had been receiving sewage irrigation for 20 years exhibited significant build-ups of Zn (208%), Cu (170%), Fe (170%), Ni (63%), and Pb (29%) as compared to adjacent soils under tube well irrigation. In periurban agriculture around Delhi, more than 1000 ha of land in Madanpur Khadar that had been irrigated with sewage effluents over four decades showed 184%, 106%, 160%, 117%, 108%, 58%, and 83% increases in available Zn, Cu, Fe, Mn, Cd, Pb, and Ni
284
World Soil Resources and Food Security
compared to those in adjacent tube well irrigated soils [Purakayastha 2008]. In soils irrigated by sewage effluent around Kolkata, there had been severe increases in the concentrations of Zn (80 times), Cd (75 times), Pb (25 times), Cu (16 times), and Cr (4 times) over nonsewage-irrigated soils [Barman and Lal 1994]. Further, the contents of heavy metals in sewage-irrigated radish, gourd, spinach, and cauliflower around Kolkata were comparatively 2 to 40 times higher than those in the nonsewageirrigated vegetables. The Cd concentration in the soils of Ludhiana, Jalandhar, and Malerkotla has increased by 6, 4, and 3 times due to irrigation with sewage effluents and, consequently, the Cd concentration in Brassica has increased by 8, 6, and 4 times [Sikka et al. 2009]. 6.3.1.1.2 Bangladesh Bangladesh consists of 13.77 Mha of land area, of which 2.62 Mha (19%) are degraded [FAO AGL 2005]. The intensity of degradation is moderate (0.1.67 Mha) to severe (0.96 Mha) [FAO AGL 2005]. Degraded soils impose a severe limitation on successful crop production due to the unfavorable effects of certain chemical and/or physical properties. Acid sulfates soils, saline and alkali soils, peat soils, soils with nutrient toxicity (very minor area), and nutrient deficiency are examples of chemical soil problems. In comparison, steeply sloping soils, coarse-textured soils, shallow soils, poorly drained soils, heavy-textured soils, and soils with plough pan are examples of soils with physical problems [Rahman 2000]. Acid sulfate soils occur near the coast on the Chittagong Coastal Plains and in the southwest of the Ganges (tidal floodplain areas where the mangrove forest has been cleared for cultivation) [Rahman 1990]. These soils are also called Kosh soils in Chittagong. High acidity (pH 3.2–4.0) in surface soils, salinity (14–34 dS/m), Al toxicity (8–103 ppm), low Zn and exchangeable Ca2+ status, adverse physical and biological conditions with medium SOM content, and poor tilth and activity of microorganisms limit crop growth. Embanking to prevent intrusion of saline or brackish water, improvement of drainage, applying lime to increase soil pH, and flushing the embanked area with brackish water can reclaim these soils and make them suitable for producing more than one crop as is presently grown. However, the methods are costly and sometimes may not be cost-efficient. The electrical conductivity (EC) of a saturated extract of a saline soil is >4 dSm−1, the exchangeable Na2+ % (ESP) is <15, the pH is usually less than 8.5, and the sodium adsorption ratio (SAR) is <13. Saline soils occur mainly in the Ganges tidal floodplain, young Meghna estuarine floodplain, and in the tidal floodplain areas of the Chittagong coastal plains and offshore islands. High salinity, low fertility with respect to SOM, nitrogen, zinc and copper, scarcity of quality irrigation water during winter, variability in rainfall from year to year, short winter season, heavy texture (silty clay to clay), and perennial waterlogging due to inadequate drainage hamper crop growth. Cultivation of salt-tolerant crops, mulching in the summer, improvement of drainage, and prevention of intrusion of saline water from sea the can improve these soils [Rahman 2000]. In alkali soils, ESP is >15, the EC of a saturated extract is <4 ds/m, the pH is commonly 8.5 or less because of the presence of neutral salts, and the SAR is at least 13.
Soil Resources Affecting Food Security and Safety in South Asia
285
Soils with dense and alkaline topsoils (pH > 8.5) occupy very small areas among calcareous loamy ridge soils in widely scattered areas in the west of the Ganges floodplains. Topsoil alkalinity is the main problem. Plowing the topsoil followed by the cultivation of deep rooting grasses for several years and application of sufficient organic manure can restore soils for crop growth. Peat soil occurs in the Gopalganj-Khulna Beels, deep depressions in the Sylhet Basin and in the northern and eastern hills. The bulk density of peat soil is only 0.20–0.30 gm cm−3 [Rahman 1990]. Low bearing capacity when wet, strong acidity, low nutrient status, and perennial wetness are constraints to high crop production. Drying the peat soils eventually leads to shrinkage and acidification. Control of drainage, allowing the soil not to dry, and making raised cultivation beds and embankments may make the soils suitable for vegetable crops. Phosphate-fixing soils include some terraced soils developed on the Madhupur and Barind tracts. Strongly acid soils rich in iron and aluminum, and calcareous soils rich in calcium are prone to phosphate fixation. Appropriate methods of placing phosphatic fertilizers, addition of farmyard manure (FYM) or compost, and foliar application of phosphatic fertilizers may reduce harmful effects. Iron, zinc, and sulfur deficiencies occur only locally but may be widespread on the calcareous soils of the Ganges floodplain and sandy soils, peat soils, high pH saline soils, and light-textured piedmont soils [Rahman 1990]. Sulfur deficiencies in Bangladesh are acute and widespread in light-textured and irrigated soils, where high-yielding varieties are cultivated. The main causes are 1) increased use of high analysis sulfur-free fertilizers like Urea and TSP (triple super phosphate); 2) increased crop yield through high-yielding varieties (HYV); and 3) cultivation and continuous wetting of the soil. Application of zinc and sulfur-containing fertilizers can easily mitigate zinc and sulfur problems. Steeply sloping soils occur widely in hill areas and in a small proportion of side valleys in the Barind and Madhupur Tracts. Unfavorable slopes (quite often steep to very steep), shallowness of the soils profile, severe doughtiness in the dry season, moderate to rapid permeability, and susceptibility to serious erosion pose problems for crop cultivation [Rahman 1990]. Growing trees is an appropriate strategy for very steep to steeply sloping soils. Growing tea, coffee, rubber, bananas, etc. on low hills with gentle slopes and deep loamy soils, and jackfruits in similar soils are possible options. Attempts to increase SOM content, cultivation of cover crops wherever possible, and leaving the soil surface barren as little as possible are strategies for minimizing risks of soil degradation. Coarse-textured soils mainly occur on the flood-free piedmont plain, active floodplains, the hills, and river chars. Low structural stability, low moisture-holding capacity, low SOM content, and zinc and sulfur deficiencies limit crop production. Drought-tolerant crops including tree crops on the soils of highlands, cultivation of fruits, spices, or vegetables with mulching, controlled irrigation, use of FYM, and slow release fertilizers are recommended management practices. Shallow soils (<50-cm deep) occur in eroded parts of the hills summits. Low fertility, low moisture status especially in the dry season, insufficient space for proper root development, and vulnerability to erosion limit crop production. Some soils are
286
World Soil Resources and Food Security
used for the cultivation of rice and others for dry land crops. Such lands should be kept under shrub, grass, and dwarf bamboo to minimize soil degradation. Soils with poor drainage occur in most of the active and meander floodplains and basins. Poor drainage leads to unfavorable physical, chemical, and biological conditions. Soils become devoid of oxygen, with consequent evolution of other gases such as carbon dioxide, methane, and nitrous oxides, which affects upland crop production adversely. These soils are, however, suitable for rice cultivation and need drainage improvement for upland crops [Rahman 2000]. Plow pan is a dense soil layer occurring below the cultivated topsoil, and is caused by pressure from the sole of the plow. It occurs in most soils, specifically in silt loam soils, which are cultivated in a wet condition and are puddled for transplanted rice cultivation. It is advantageous for rice cultivation but poses limitations for upland crops. Thus, plow pans must be eliminated by subsoiling. However, for cultivation of rice and upland crops on the same field in rotation, a pan could be created at a deeper depth using deep plowing equipment. In conclusion, low fertility (northern and eastern hill soils, and terraced soils), strong acidity (piedmont apron, northern and eastern piedmonts, basin soils in the dry season, and acid sulfate soils), droughtiness (basin soils in the dry season, and northern and eastern piedmont plains), wetness (peat soils, hill soils, and basin soils in the early dry season), zinc efficiency (basin soils, peat soils, and estuarine floodplain soils), phosphate fixation (terraced soils), plough pan (Tista floodplain, old Brahmaputra floodplain, Barind tract, Meghna estuarine floodplain), high permeability (Ganges floodplain soil, and friable red clay terraced soils), low moisture status (Ganges floodplain soils, Barind tract, and northern and eastern piedmont soils), and heavy consistence (terraced soils) and low-bearing capacity are the major soil constraints to crop production in Bangladesh. Though the constraints referred to here are not mutually exclusive in their occurrence, the total area is quite significant and needs careful study to identify sustainable management options because Bangladesh is a densely populated country and is barely self-sufficient in grain production [Rahman 2000]. 6.3.1.1.3 Nepal Nepal consists of 14.73 Mha of land area, of which 2.07 Mha (14%) are degraded [FAO AGL 2005]. The intensities of degradation are light (0.11 Mha), moderate (0.54 Mha), and severe (1.41 Mha) [FAO AGL 2005]. Degradation is caused by soil erosion and soil deterioration and natural erosional processes are inherent because of the topography of Nepal, tectonic activities, monsoons, and river flows. Landslides and floods are also common and are accelerated by human-induced changes in landscapes due to agriculture, human settlements, and infrastructure. An increase in population has continued the reliance on natural resources, thereby causing deforestation, intensified cultivation, and expansion of agriculture to marginalized areas and steep slopes. Further overgrazing has increased soil erosion, depleted soil fertility, and reduced the productive capacity of the land. Soil erosion has further led to the deposition of sediment in tanks and reservoirs that has reduced storage capacity and increased flooding and landslides, and loss of arable land in low-lying areas.
Soil Resources Affecting Food Security and Safety in South Asia
287
The direct and primary effects of soil erosion include soil loss, nutrient-leaching, and reduction of land productivity. A study in the mid-hills of Nepal revealed a soil loss of 20 Mg/ha/year from rain-fed marginal land, with a nutrient loss of 300 kg of SOM, 15 kg of nitrogen (N), 20 kg phosphorus (P), and 40 kg potash (K) [Carson 1992]. Similarly, in the western hills [Tripathi et al. 1999], erosion caused soil loss of 20 Mg /ha/year including 12 kg of N, 66 kg of P, and 24 kg of K. More than 50% of these losses occurred during the premonsoon (May–June) season when ground covers are low, and the losses occurred mostly through leaching rather than through surface run-off. Analysis of soil samples from the western hills [Tripathi 1999] indicated that soils are mostly acidic (3.7–7.5 pH) and that 32.6% of soils have low organic carbon content (<1.49%), 5.8% have low total N (<0.1%), 8.4% have low available P (<6.4 mg/ kg), and 35.35% have low exchangeable K+ (<0.2 cmols/kg). As much as 10%–20% of soils are low in zinc (<0.5 mg/kg), manganese (10 mg/kg), and copper (0.5 mg/ kg), while 87% are deficient in Boron (1 mg/kg). These deficiencies call for an appropriate soil fertility management program in the mountainous tracks of Nepal. Forests constitute the second largest resource base (after water). At present only 29% of the total land area in Nepal is forested. The annual deforestation rate is 2.3% in the hills and 1.3% in the Terai plains [Tripathi 1999]. High population growth is the principal cause of deforestation due to conversion of land for agriculture, settlements, infrastructure, and the grazing of livestock. The reliance on fuel wood for energy and extraction of forest products (fodder, leaf, litter, etc.) has resulted in degradation of forests. Illicit felling and the transboundary timber trade are also causes of deforestation. Deforestation has increased soil degradation, reduced forest stock, and reduced habitats and biodiversity. It has also increased vulnerability to natural disasters and increased the concentrations of CO2 in the atmosphere. Nepal has the highest rate of urbanization (5% per annum) in South Asia. Urbanization is a recent trend in Nepal and is occurring in an uncontrolled and haphazard manner. It has resulted in the addition of new issues to Nepal’s environmental and developmental problems. The problem is most severe in Nepal’s cities, which must deal with increased amounts of solid waste generated by its expanding populations, overcrowded cities, changing consumption patterns, use of new materials (like plastics), and increased industrialization. This trend has resulted in air and water pollution, health risks, and adding to the general deterioration of the cities. As much as 85% of Nepal’s total waste is solid waste, generated mainly by households, while waste from hospitals, agriculture (pesticides, fertilizer), and industries make up the rest. Hazardous waste is mainly disposed of with all other waste (dumping or incinerators) or stockpiled without adequate safety measures, thereby adding to the problems. The availability of vast amounts of marginal land including waste land, grazing land, range land, shrub land, and unclassed forest in the Hindukush Himalayan (HKH) can provide a ray of hope to farmers in their quest to provide support to farming and livelihood in the hills and mountains [Pratap 2001]. This strategy requires identification and promotion of suitable technologies with necessary logistic supports to bring about impacts and economic transformation. During the past few years, As
288
World Soil Resources and Food Security
has also been detected in the groundwaters of the sedimentary aquifers of the Terai Belt in southern Nepal [Bhattacharya et al. 2003; Tandukar et al. 2006]. 6.3.1.1.4 Sri Lanka Sri Lanka consists of 6.55 Mha of land area, of which only 50% are arable. The remainder is not cultivatable because of unsuitable terrain, inland water bodies, and forest reservations [Mapa et al. 2002]. As much as 2.17 Mha (33%) of land is degraded. The intensities of degradation are light (0.027 Mha), moderate (0.34 Mha), severe (0.26 Mha), and very severe (1.55 Mha) [FAO AGL 2005]. Land degradation is one of the most critical problems affecting the economic development of Sri Lanka. The high demand for the limited land resources has caused pressures on the island’s land resources and these, in turn, have resulted in a high level of environmental degradation. About one-third of the land in Sri Lanka is affected by soil erosion, the erodible proportion ranging from less than 10.0% in some districts to over 50.0% in others [UNCCD 2000]. Other important manifestations are heavy soil losses, high sediment yields, declining soil fertility, and reduction of crop yields, marginalization of agricultural lands, salinization, landslides and deforestation, and forest degradation. Severe erosion takes place in the hill country on sloping lands under cash crops like potatoes, tobacco, vegetables, and poorly managed seedling tea and gram cultivation. Management of solid and liquid wastes is a critical issue particularly in urban areas and around industrial sites. The daily collection of waste is about 2500 million Mg, of which the western province accounts for 57%. A large quantity of hazardous and nonhazardous waste is generated at industrial and hospital sites, and the waste generation is increasing at a rate of 1.2% per year. The present method of solid waste disposal is open dumping in low-lying lands, leading to pollution and contamination. Sri Lanka ‘s mean annual rainfall ranges from 900 mm to 6000 mm, with an island-wide average of about 1900 mm, which is about two-and-a-half times more than the world annual mean of 750 mm. The surface water is carried radially from the central hills through 103 distinct river basins covering 90% of the island. Sri Lanka’s inland waters include manmade lakes and ponds and marshes, constituting one of the highest densities in the world. The area under water bodies is 2905 km2 or 4.43% of the total land area. The major intentional (direct) causes for inland water pollution are agriculture, urbanization, and industrialization, which change land-use patterns. Excessive use of agrochemicals, release of industrial effluents, domestic waste, and sewage, and dumping of solid waste into waterways are the unintentional (indirect) causes. Sri Lanka has the highest biodiversity per unit area of land among the Asian countries in terms of flowering plants and all vertebrate groups except birds. The major threats to biodiversity are the ever-increasing demand for land to cater to increasing population, poor land-use planning, indiscriminate exploitation of biological resources, weak enforcement of legislation, and the absence of an integrated conservation management approach. Being an island country, Sri Lanka is endowed with a coastline of 1585 km. The coastal region includes a terrestrial ecosystem, mangroves, lagoons and estuaries,
Soil Resources Affecting Food Security and Safety in South Asia
289
a shoreline ecosystem, coral reefs, etc. Coastal erosion has been a severe problem over the years, especially in the south, west, and northwestern coasts. In severe cases, the recorded net erosion rate has been up to 1 m per year. On the other hand, accretion rates have not exceeded 0.1 m per year, except in the north and east where the rate is 0.3 m per year. The average net mean rate of erosion for the entire country ranges from 0.2 to 0.35 m per year [UNCCD 2000]. The major causes of depletion of the island’s coastal resources are concentration of population in coastal areas, prawn-farming practices, and collection of nonedible aquarium species. 6.3.1.1.5 Bhutan Bhutan consists of 4.01 Mha of land area, of which 0.34 Mha (8.39%) are degraded [FAO AGL 2005]. The intensities of degradation are light (.07 Mha), moderate (0.17 Mha), and very severe (0.10 Mha) [FAO AGL 2005]. Land degradation in Bhutan is a manmade as well as natural phenomenon. Urbanization and industrialization are exerting pressure on the environment and on the natural resources of the country. Land degradation in the country is mostly manifested in displacement of soil material through water erosion and internal biophysical and chemical deterioration. Human-induced activities trigger soil erosion mainly in the mountainous terrain. Loss of vegetation due to deforestation, overcutting beyond silviculturally permissible limits, unsustainable fuel wood extraction, shifting cultivation, encroachment into forest land, forest fire, overgrazing, extension of cultivation onto lands of low potential or high natural hazards, nonadoption of adequate soil conservation measures, and improper crop rotation are some of the important factors contributing to land degradation in Bhutan. Waste disposal is an emerging problem in almost all urban centers in Bhutan. The increase in waste is primarily attributed to factors such as rapid rates of urbanization, rural–urban migration, changing consumption patterns, and a high population growth rate. While the magnitude of the problem is relatively small and manageable in rural areas, it appears to be growing significantly in urban areas in recent times. 6.3.1.1.6 Pakistan Pakistan consists of 79.8 Mha of land area, of which 23.99 Mha (31%) are degraded (FAO AGL, 2005). The intensities of degradation are light (0.13 Mha), moderate (0.13.13 Mha), severe (9.62 Mha), and very severe (1.11 Mha) [FAO AGL 2005]. About 96% of the arable land has inadequate SOM content resulting in low yields. The major causes of soil degradation and low productivity in the country are water and wind erosion, waterlogging, and salinity intrusion. Per capita water availability in the country has been decreasing at an alarming rate, and is now 12,000 m3, which is nearing the water scarcity level. Industrial and domestic wastes are the major causes of water pollution. Recent studies indicate that Pakistan’s woody biomass is declining at a rate of 4%–6% per year and the principal cause is the consumption of fuel wood and timber. It is feared that the country is experiencing the world’s second highest rate of deforestation. In recent years, arsenic concentration in drinking water of Pakistan is also reported [Nickson et al. 2005].
290
World Soil Resources and Food Security
6.3.2 Urban Encroachment The South Asian region is undergoing rapid industrialization and economic growth, leading to increases in emission of CO2 and other GHGs into the atmosphere. Furthermore, the urgency of meeting the increased demand for agricultural produce (food, feed, fiber, and fuel) is rapidly degrading soil quality and exacerbating degradation. There is huge population pressure on the big cities of south Asia due to migration of people from villages. There is no space available in these cities to cater to the housing needs of the ever-growing population. Therefore huge chunks of fertile lands are being utilized for housing purposes. In India, unprecedented population growth coupled with unplanned developmental activities has led to urbanization, which lacks infrastructure facilities. This also has posed serious implications on the resource base of the region. The urbanization takes place either in radial directions around a well-established city or linearly along the highways. This dispersed development along highways, or surrounding the city and in the rural countryside is often referred as sprawl [Theobald 2001]. Some of the causes of the sprawl include population growth, the economy, and the proximity to resources and basic amenities. Patterns of infrastructure initiatives like the construction of roads and service facilities (such as hotels, etc.) also often encourage regional development, which eventually leads to urbanization. The direct implication of such urban sprawl is the change in land use and land cover of the region. The ability to service and develop land heavily influences the economic and environmental quality of life in towns [Turkstra 1996]. In cities like Delhi, the Indian Government has created national capital regions (NCR), which already brought huge areas of agriculturally important land surrounding the city into urban development. Because of the greater demand for housing, the price of the land has increased tremendously. The urban expansion of Delhi has led to engulfment of the surrounding village lands for the urban development process. Unlike the United States or European countries, the urban sprawl phenomenon in Delhi is separated from the city by rural areas, rather it shares common boundaries. Hence the distance factor is missing between the western sprawl and Indian sprawl. Also, due to urban expansion programs, all agricultural fields are acquired by government agencies for urban development. These areas are turned into planned residential or commercial areas. The problem is with the unplanned and haphazard urban growth and development in the villages. These villages are also known as urban villages. These are marked with high densities of lower economic populations—mainly migrants in the form of tenants, unplanned buildings, narrow roads and by-lanes, lack of proper sanitation and drinking water, low cost business activities, informal or unregistered business and manufacturing activities, etc. The main cause of urban sprawl in Delhi is the lack of a proper urban development vision by the Delhi Development Authority, along with a lack of awareness among villagers regarding urban laws. Also the neglect of urban villages by the municipality and other urban bodies are responsible for the deteriorating condition of peripheral villages of Delhi. There is an urgent need to look into the urban villages by the government, urban planning bodies, and urban planners in order to save these areas from becoming urban slums. This is also a lucrative proposition for the farmers in these
Soil Resources Affecting Food Security and Safety in South Asia
291
regions to sell their land at higher prices instead of continuing cultivation on these pieces of land. Therefore, the government has to develop the land-use classification map indicating marginal soils on which the industrial establishments and cities could be built rather than on the prime soils.
6.3.3 Water Supply Much of the world is heading for water shortages. Since agriculture is responsible for about 70% of all the water withdrawn for human use, it is feared that water scarcity will affect future food production [FAO 2002]. At a global level, there seems to be no cause for alarm. However, at the level of some localities, countries and regions, serious water shortages are major concerns. Groundwater mining also occurs at local levels in several other countries of the Near East and North Africa, South Asia, and East Asia. In large areas of India and China, groundwater levels are falling by 1 to 3 m per year, causing subsidence in buildings, intrusion of seawater into aquifers, and higher pumping costs [FAO 2002]. Agriculture is the principal user of all water resources taken together, i.e., rainfall (so-called green water) and water in rivers, lakes, and aquifers (so-called blue water). It accounts for about 70% of all withdrawals worldwide, with domestic use amounting to about 10% and industry using 20%. The annual groundwater use for the world as a whole is 750–800 km3 [Shah et al. 2000], which may appear modest compared to the overall groundwater reserves. However, only a fraction of the world’s groundwater reserves are economically viable for agriculture. It is estimated that about 30% of the world’s irrigation supply is made up of groundwater, but this input accounts for some of the highest yields and highest value crops [FAO 2003]. The number of tube wells providing water to irrigated land in India, China, Pakistan, Mexico, and many other countries has grown rapidly since the 1970s. For example, some 60% of irrigated cereal production in India depends on irrigation from groundwater. This has led to widespread and uncontrolled overabstraction of the resource and the creation of a bubble groundwater economy [Roy and Shah 2002]. The root of the Asian groundwater crisis alluded to by Shah et al. [2000] threatens millions of poor rural communities that lie in the open access nature of the resource. Paradoxically, it is precisely this feature of groundwater in shallow aquifers that has made it a powerful tool in the fight against poverty [Moench 2002]. Everyone who can afford to install a pump has free access to water. Irrigation with groundwater is generally more productive than canal irrigation because groundwater is produced close to where it is used with hardly any losses during transport. In addition, farmers are in control of the timing and amount of water extracted. Evidence in India suggests that crop yield per m3 on groundwater-irrigated farms tends to be 1.2 to 3 times higher than on farms irrigated with surface water [Shah et al. 2000]. FAO [2002] reported that the overall water withdrawal for irrigation in all developing countries may increase from 2128 km3 in the benchmark period of 1997–1999 to 2420 km3 in 2030, an increase of nearly 14%. Further, irrigated areas in developing countries as a whole may increase from 202 Mha in 1997–1999 to 242 Mha by 2030, an increase of nearly 20%. The largest increase is expected in sub-Saharan Africa with 44%, and the lowest in East Asia with 6%. The expected increase is 32% for Latin America, ~10% for the Near
292
World Soil Resources and Food Security
East/North Africa, and 14% for South Asia [FAO 2002]. The effectively cultivated irrigated area is expected to grow by 34% during the period under consideration because of higher cropping intensities. Much of the difference in the rates of increase of water withdrawal and irrigated area relates to the higher water productivity in irrigated agriculture that is expected to have occurred by 2030, with some effect also from a change from water-intensive rice to wheat production, especially in China. Secondary salinization, caused by excessive irrigation and lack of adequate drainage, is a problem in regions irrigated by canal water, and is accompanied by a rise of the water table and seasonal or permanent waterlogging. In contrast, groundwater depletion is occurring in regions intensively irrigated by tube wells.
6.3.4 Poor Management and Limited Nutrient Supply Nutrient mining is also a serious problem in South Asia. Farmers often use insufficient fertilizers to replace the N, P, and K harvested with their crops and lost through leaching, while trace elements, such as Fe or B, may also be deficient. Replenishment of nutrients is paramount in maintaining soil quality. Nutrient-depleted soils are highly erosive because of poor land cover and insufficient economic returns; hence, they tend to be neglected. Although fertilizer use in most of the South Asia appears to be inadequate, considerable improvement in proper fertilization application rates and nutrient balance can be made. Precision agriculture can make a big contribution to the nutrient status of Asian soils. Deforestation of tropical rain forests also contributes to carbon dioxide emissions and, hence, global warming. Most soils have extremely low levels of soil organic carbon (SOC) contents, ranging from 8 to 10 g/kg. Most of the agricultural soils have low SOM reserves due to fertilitymining practices (residue removal, uncontrolled and excessive grazing, imbalance in the application of plant nutrients), and widespread problems of soil degradation. Depletion of SOC pool is caused by fertility exploitative practices and soil degradation processes. Low external input of chemical fertilizers and organic amendments causes depletion of the SOC pool because nutrients harvested in agricultural products are not replaced, and are made available through the mineralization of SOM. In some cases, soil is burnt to release nutrients contained in SOM. Fuel for household use is limited, and crop residues and animal dung are used as fuel. Crop residues are also used as fodder for livestock. Reduction in the SOC pool sets in motion other degradation processes, including a decline in soil structure and aggregation, reduction in exchangeable bases, decrease in plant available nutrients, and a reduction in plant-available water capacity. Indeed, some argue that the sustainability of the rice–wheat system is threatened by the continuous decline in the SOC pool. Imbalanced fertilizer management is also one of the reasons for low agronomic productivity. Salinization occurs in irrigated areas, usually when inadequate drainage causes salts to concentrate in the upper soil layers where plant roots are concentrated. It is a problem mainly in the arid and semiarid zones, where 10% to 50% of the irrigated area may be affected. Salinization can decrease yields by 10% to 25% for many crops and may prevent cropping altogether when it is severe. About 3% of the world’s agricultural land is affected. The proportion is 6% in East Asia, and 8% in South
Soil Resources Affecting Food Security and Safety in South Asia
293
Asia [FAO 2002]. For the arid and semiarid tropics as a whole, 12% of agricultural land may be affected by salinization. For the problems in the South Asia region, restoring SOM and improving soil quality are key strategies to land restoration in severely degraded situations. Nutrient management in the rice–wheat ecosystems of South Asia also is critical for maintaining crop yields, which have been decreasing since 1990s. The decline in crop yields has been a consequence of decreasing SOM levels and inefficient N use. More use of knowledge-intensive management approaches, especially site-specific technologies, are needed. Long-term trials are key to developing new management techniques; analysis of existing trials shows that key soil quality issues, such as fertility, SOM buildup, salinity control, and prevention of waterlogging are critical.
6.4 FOOD PRODUCTION AND SUPPLY 6.4.1 Current Production: Total and per Capita 6.4.1.1 India The annual compound growth rates of the agricultural sector in India as a whole have been quite impressive during all decades after independence, except during the 1960s. It was 2.6% per annum during the 1950s but decelerated to 1.7% per annum in the 1990s. During 2000–2006, it was 2.9% per annum. The growth of the agricultural sector in all decades remained higher than the growth rate of population in the country. Divergence between agricultural growth rates of the Indian economy as a whole increased consistently, particularly since the 1980s, mainly due to the faster growth of nonagricultural sectors. With a target of 239.10 million Mg of food grain production in 2009–2010, India has achieved 91.3% of that target (218.2) and it was lower than the last year’s (2008– 2009) production of 234.5 million Mg [DES 2010]. The production of various food grain crops like rice, wheat, coarse grain, and pulses were 89.13, 80.17, 33.77, and 14.59 million Mg, respectively [DES 2010]. The crop subsector grew at an annual rate of 1.8% in the 1990s. The growth in the crop subsector was 2.7% per annum during 2000–2006. In the first decade of independence, except fruits and vegetables, all crop groups registered remarkable growth, partly attributed to the lower base. The growth rates of these crops, however, either decelerated or turned negative in the following decade, which by and large (except cereals) witnessed a similar trend in the 1970s. The cereals, dominated by rice and wheat, registered an annual growth rate of around 4% in the 1950s, which declined to 2.1% in the 1960s. Subsequently, the growth rate increased to 4.3% per annum in the 1980s but again decelerated to 2.6% in the 1990s, a trend that continued in the early years of the current decade [Abrol 2000]. The pulses witnessed a modest growth rate of 1.5% in the 1980s, but decelerated to 0.8% in the 1990s. However, pulse production improved substantially (2.8% per year) during 2000–2006. The technology mission on oilseeds (TMO) of the 1980s boosted oilseed production, resulting in their higher growth rate (5.2% per annum), but it plummeted to 1.9% in the 1990s. An upturn was observed (5.9%) in 2000–2006. The fiber crops (mainly cotton), registered reasonable growth rates of 2.3% to 2.6% during the 1970s, 1980s, and 1990s; however, it recorded a sharp
294
World Soil Resources and Food Security
acceleration in 2000–2006 of 15.2%, owing to large-scale adoption of Bt cotton. Fruits and vegetables also registered significant growth rates since the 1960s. A noteworthy aspect of performance of the crop subsector is that more than 90% of the increase in production is contributed by the improvement in productivity. The cereals production increased from 97 million Mg in 1970–1971 to 216 million Mg in 2007–2008. This increase was through improvement in average yield per hectare, which went up from 949 kg to 2142 kg during the period. The area under cereals, in fact, declined from 101.8 Mha in 1970–1971 to 100.6 Mha in 2007–2008. In oilseeds (nine oilseeds combined), production increased from 9.6 million Mg in 1970–1971 to 28.8 million Mg in 2007–2008, and in the incremental production of 19.2 million Mg, as much as 94% was contributed by improvement in the average yield per hectare, from 579 kg to 1086 kg/ha, while area increased from 16.6 Mha to 26.5 Mha. For cotton, production registered an increase from 4.8 MBa (million bales) in 1970–1971 to 25.8 MBa in 2007–1908, with yield improvement contributing as much as 93% to incremental production; and the average yield of raw cotton increasing from 106 to 466 kg/ha and the area increasing from 7.6 9.4 Mha [DES 2009]. 6.4.1.2 Bangladesh In a subsistence-oriented agrarian economy such as Bangladesh, domestic food production has an important role to play in the quest for food security. Major items in the food basket in Bangladesh are rice, wheat, pulses, potato, vegetables, and fish. These food items account for almost 85% of the total calorie and protein intake. Rice and wheat alone contribute to 74% and 57% of the total per capita calorie and protein intake, respectively [BBS 2010]. Rice occupies 71% of the gross cropped area and accounts for 94% of the food grain production. Most farmers with access to irrigation facilities grow two crops of rice during the year. The net cultivated area in Bangladesh is about 8.0 Mha, but the total cropped area of rice is about 11.0 Mha; such is the importance of rice in agriculture in Bangladesh. Rice production declined in absolute terms immediately after the independence in 1971 due to the destruction of infrastructure by the civil war and the consecutive natural disasters. Indeed, the country faced severe food insecurity and famine in 1974–1975 due to the shortfall in domestic production caused by floods, the government’s incapacity to import, and mismanagement in distribution, which led to skyrocketing rice prices [Alamgir 1980; Sen 1982; Sobhan 1979]. However, the growth of cereal production resumed in 1976 and had almost unhindered growth since then (except for a short period in the early 1990s). The growth in rice production kept pace with population growth in the 1980s, and surpassed population growth by a significant margin in the 1990s. The respectable growth in rice production was propelled by the adoption of high-yielding modern varieties of rice, facilitated by an expansion of irrigation infrastructure. Almost 56% of the cultivated land now has access to irrigation facilities, developed mostly by private investment on small-scale shallow tube wells and power pumps [Hossain 2003]. The adoption of modern rice varieties has reached 70% of the cropped area. Only in the deep-flooded areas in the depressed basins, and in the salinity-affected coastal areas, do farmers still grow low-yielding traditional varieties. Almost 90% of the growth in rice production came from the increase in
Soil Resources Affecting Food Security and Safety in South Asia
295
yields made possible by the technological progress in rice cultivation. Bangladesh does not have favorable agroclimatic environments for growing wheat because of the short and mild winter season and heavy soils. Wheat is grown mostly in the northwestern region of the country, which has a relatively longer winter period. Until the late 1960s, wheat was an unimportant crop, occupying <1% of the cropped area. The availability of high-yielding modern varieties in the late 1960s, however, induced farmers to grow more wheat. The food grain production target for the current 2009/10 fiscal year is set at 35.05 MMg, which is 9% higher than last year’s actual production and 31.3% higher than the five year (2002/03–2006/07) average [WFP 2010]. Hence, the area under wheat expanded exponentially from 126,000 ha in 1976 to 800,000 ha in 2000–2001, but the area under wheat decreased to 400,000 ha in 2009–2010 [WFP 2010]. The production increased from 117,000 Mg to 1.70 MMg in 2001, but it decreased to 0.9 MMg in 2009–2010. Wheat is an important crop since Atta (wheat flour) is the second staple for Bangladeshis. Due to insufficient domestic production, wheat is at the top of the food grain imports. The expansion halted over the next decade, but picked up again since 1996 due to a favorable trend in the price of wheat relative to rice. Over the past 3 decades, wheat production increased at a rate of 10% per year but wheat still accounts for only 7% of the total cereal production. In the context of food security, an important point to note is that cereal production has become more resilient to natural disasters over time because of the dramatic change in the seasonal composition of production. The area under the early-monsoon, low-yielding aus rice (April to July) has been reduced from 3.5 to 1.2 Mha, so the loss of production from the late arrival of the monsoon rains is now substantially lower than in the pre-Green Revolution period. The boro rice area has expanded from 0.5 Mha in the early 1970s to nearly 4.0 Mha by 2003. The boro (winter) rice, together with wheat, now makes up over half of cereal production during the March to June period; their share of the total cereal harvest was less than 10% in the early 1970s. The farmers can now recover the loss from the traditional monsoon season aman (wet) crop within 6 months, while they previously had to wait a year to recover the losses. This change in the seasonal composition of production also had a smoothing effect on the seasonal variation in rice prices and the ability of the country to cope with disastrous floods, such as in 1988 and 1998. However, the Green Revolution in cereal production has not been an unmixed blessing. The rapid expansion in the area of boro rice and wheat was achieved partly through reduction of area and production of pulses and oilseeds. These two crops are important sources of protein and micronutrients, particularly for the poor. The production of sugarcane and fruits has also remained stagnant. Among other food crops, the growth has been acceptable only for potatoes and vegetables, because of higher productivity and profitability compared to rice and wheat.
6.4.2 Progress and Future Projections Looking further ahead, slower population growth and the leveling off of food consumption in many countries will continue to dampen demand, the growth of which is expected to slow down to 1.2% a year over the period 2015 to 2030 [FAO 2002]. Nevertheless, the production task facing world agriculture is massive. By 2030, an
296
World Soil Resources and Food Security
extra billion tons of cereals will be needed each year. Unforeseeable events such as oil price booms, dramatic growth spurts, or crises could, of course, alter effective demand over short periods, but will not greatly change the big picture. In the future, 80% of increased crop production in developing countries will have to come from intensification: higher yields, increased multiple cropping, and shorter fallow periods. Increases in crop production derive from three main sources: expansion of arable lands, increases in cropping intensity (the frequency with which crops are harvested from a given area), and improvements in yield. The slower growth in production projected for the next 30 years means that yields will not need to grow as rapidly as in the past. Growth in wheat yields is projected to slow down to 1.1% a year in the next 30 years, while rice yields are expected to rise by only 0.9% per year. Nevertheless, increased yields will be required—so is the projected increase feasible? One way of judging this is to look at the difference in performance between groups of countries. Some developing countries have attained very high crop yields. In 1997–1999, for example, the top-performing 10% had average wheat yields more than six times higher than those of the worst-performing 10%, and twice as high as the average in the largest producers, China, India, and Turkey [FAO 2002]. For rice, the gaps were roughly similar. The global food and beverage industry was valued at US $3668 billion in the year 2005. In the global food processing industry, Asia-Pacific is accounting for 31.10% of the global market. India is the world’s second largest producer of food, next to China, and has potential to become the largest. In contrast, in South Asia and the Near East and North Africa, where almost all suitable land is already in use, there will be next to no expansion in area. By 2030, the Near East and North Africa will be using 94% of its suitable cropland, with a remaining surplus of only 6 Mha. In South Asia, the situation will be even tighter, with 98% already in cultivation. In South and East Asia, more than 80% of the increase in production will have to come from yield increases, since only 5% or 6% can come from expansion of the arable area.
6.4.3 Strategies for Improving Food Security The livelihoods of the preponderant majority of the people in the HKH countries (Nepal, Bangladesh, Bhutan, Pakistan, India, China, Myanmar, and Afghanistan) revolve around agriculture. In the HKH countries, while a population of 1.5 billion inhabiting an area of 3.4 million sq. km gives about 35 people per sq. km, the actual pressure on agricultural land is much higher, therefore the availability of cropland is too small to support the livelihood of rural households in the mountains. Management of the marginal lands is becoming an increasing priority with the increasing population pressure, poverty, soil erosion and degradation, and loss of natural resources for food security, improved livelihood, and environmental protection. Diversification of farm activities into high-value commercial crops, and processing of agricultural and other natural resource-based materials, while adequately maintaining soils, forests, and other natural resources, are the most logical steps to improving the livelihood of mountain people. Mountains provide an excellent avenue of diverse agroenvironments, thereby niches for horticulture, floriculture,
297
Soil Resources Affecting Food Security and Safety in South Asia
TABLE 6.4 Food Security Scenario in South Asia (in 1000 Mg) Country
Food Production
Food Exports
Food Imports
Food Balance
26,924 174,655 5839 24,936 1938
1.6 9490 11 2966 9.8
2827 56 39 288 1307
−4601 23,826 57 3818 252
Bangladesh India Nepal Pakistan Sri Lanka Source: FAO, 2004.
spices, and medicinal plants. Poverty, small-holding size, and food insecurity are the critical challenges in the mountain areas of HKH and call for a holistic approach to improvement and sustainability. Among the South Asian countries, India is the largest producer of food followed by Pakistan, Bangladesh, Nepal, and Srilanka (Table 6.4). Among these countries, India is the largest exporter of food, exhibiting the largest food balance, with Bangladesh being the smallest exporter of food, showing a negative food balance. 6.4.3.1 India In 2009–2010, with a target of food grain production of 239.10 million Mg, India missed the target by producing 218.20 million Mg [DES 2010]. The per capita availability of food grains is 444.0 g/day [DES 2010]. India’s initiatives to ensure food security for its citizens range from concerted efforts to boost agricultural production to far-ranging market interventions aimed at both income and price stabilization. Measures also have been introduced to improve the access to food of the really poor through public distribution and income-generating schemes. In the mid-1960s, the government ushered in what has widely come to be known as the Green Revolution by encouraging the use of high-yielding varieties of seeds, expansion of irrigation networks, and active encouragement of the use of chemical fertilizers in a bid to boost the productivity of Indian agriculture. Most of the agricultural subsidies such as those for power and fertilizers were introduced to improve farmers’ access to inputs that would help improve farm productivity. One of the oldest initiatives taken by the Indian government was the establishment of a public distribution system (PDS), with the objective of making basic food grains available to all at affordable prices. It set up the Food Corporation of India under an act of Parliament in 1964 to oversee its implementation. The corporation was entrusted with the task of procuring food grains at minimum support prices announced for 24 crops by the government and regulating supply to ensure that prices remained stable by building a buffer stock of food grains. In 1975, it launched the Integrated Child Development Scheme (ICDS) to provide nutrition and healthcare services to children and pregnant women. The Antyodaya Anna Yojana (AAY), launched in 2000, sought to provide affordable food to below poverty level (BPL) households. The objective of the scheme was to make the TDPS
298
World Soil Resources and Food Security
more focused and it targeted an identified 10 million of the poorest of the poor in different states. Wheat and rice at subsidized prices of Rs.2 and Rs.3 per kg, respectively, are provided to these households under the scheme. State governments are expected to meet the distribution costs of the program. Another scheme that addresses the nutritional requirements of a specified target group is the midday meal scheme. The scheme was launched in 1995 for the benefit of students in government-run and -aided primary schools and in schools run by local bodies. Under the scheme, the government supplies free food grains to schools in quantities determined on the basis of specific nutritional requirements for a specified minimum number of days in a year. In 2005, the government passed the National Rural Employment Guarantee Act to improve the livelihood security of rural households by providing them with guaranteed wage employment. Semiskilled and unskilled workers living below the poverty line in rural areas have been specifically targeted under the program. Works undertaken in the program aim to create enduring assets in rural areas and include projects such as land-leveling, bush-clearing, deep plowing, building of earthen bunds, flood control works, and horticulture. These contribute to improved farm productivity and higher farm incomes. An important initiative toward food security was the launch of the National Food Security Mission (2007) to increase the production of rice by 10 million Mg, wheat by 8 million Mg, and pulses by 2 million Mg by the end of the Eleventh Plan (2011– 2012). To ensure that previous mistakes in policy formulation and implementation are not repeated, the Mission has put in place strong monitoring and evaluation mechanisms that involve all implementing and line departments. M. S. Swaminathan Research Foundation and the World Food Program have brought out a report on food insecurity in urban India. The dimensions of the problem were studied across the urban areas of the States and Union Territories and across different size classes of towns—metropolitan cities, big towns, medium towns, and small towns. The analysis indicates that there are wide variations in the nature and extent of the problem of food security across different states in the country, across different size classes of towns in the states, and within different types of towns. None of the states are free from problems. However, the remarkable achievements of some states can provide guidance to others. Better food affordability achieved by Jammu and Kashmir, better livelihood access in Himachal Pradesh and Delhi, better sanitation and health created in Himachal Pradesh, Assam, and Kerala, and better nutritional standards achieved by Kerala and Karnataka provide clear examples. A detailed analysis of metropolitan cities has shown that, while urban problems are in general much less severe in the metros compared to other urban areas, the magnitude of the problem in metros is very high indeed. The findings also suggest that a decentralized and comprehensive policy approach should be adopted, as there is a great deal of variation in the nature of the problem of food security across states and towns [Murali 2003]. In India, the second-most populous country in the world, 30 million people have been added to the ranks of the hungry since the mid-1990s and 46% of children are underweight. The growth in food production has strengthened during the recent past while the consumption need of the growing population has increased. To meet
Soil Resources Affecting Food Security and Safety in South Asia
299
the growing food grain demands, the National Development Council has adopted a resolution to enhance the production of rice, wheat, and pulses by 10, 8, and 2 million Mg, respectively by 2011. The scheme is to be implemented through a mission mode and through a farmer centric approach. All the stakeholders are to be actively associated at district levels with achieving the set goals. Expansion must be in areas of pulses and wheat and not in rice. Bridging the yield gap between the potential and the present level of productivity can be accomplished through acceleration of seed production, integrated nutrient management (INM), and integrated pest management (IPM), promotion of new production technologies like hybrid rice, timely planting of wheat, and promotion of new improved variety of pulses—supplying their input and ensuring their timely availability. The following are the components of National Food Security Mission (NFSM): Rice (i) Production of improved technology including hybrids and a system of rice intensification (SRI) (ii) Promotion of mechanical weeders and other farm implements (iii) Awards given to the best-performing district Wheat (i) Providing subsidies on diesel pump sets and community generators for irrigation (ii) Promotion of micronutrient use in deficient areas (iii) Assistance for innovative intervention at the local level India has high population pressures on land and other resources to meet its food and development needs. The natural resource base of land and water and biodiversity is under severe pressure. The massive increase in population (despite the slowing down of the rate of growth) and substantial income growth demand an extra 2.5 million Mg of food grains annually, and significant increases in the supply of livestock, fish, and horticultural products. Under the assumption of 3.5% growth in per capita GDP (low-income growth scenario), demand for food grains (including feed, seed, wastage, and export) is projected in the year 2020 to be 255 MMg, comprising 112 MMg of rice, 82 MMg of wheat, 39 MMg of coarse grains, and 22 MMg of pulses. The demand for sugar, fruits, vegetables, and milk is estimated to grow to a level of 33 MMg, 77 MMg, 1 MMg, 36 MMg and 116 MMg, respectively. The demand for meat is projected at 9 MMg, fish 11 MMg, and eggs at 77.5 billion [Bhartiya et al. 2010]. During the Eleventh plan (2007–2008 to 2011–2012), a fund requirement of Rs. 48.8 billion (~US $1 billion) is estimated to enhance the production of rice, wheat, and pulses (Table 6.5). The total budget allocation during the five-year plan for enhancing production was highest in wheat, followed by rice and pulses. However, the NFSM of India provided greater emphasis on its activities under seed, soil amendments, local initiatives, pest management, mechanization, publicity, and others being 36%, 25%, 9%, 8%, 6%, 5%, and 11%, respectively (Figure 6.3). On the other hand, NFSM would distribute its fund maximum (Rs.13.6 billion) toward giving subsidies for distribution systems (Figure 6.2). The budget allocation would be Rs.4.2, 4.2,
300
World Soil Resources and Food Security
TABLE 6.5 Fund Allocation (Rs in Millions) in Different Crops during Current Five-Year Plan Year 2007–2008 2008–2009 2009–2010 2010–2011 2011–2012 Total
Rice
Wheat
Pulses
Total
708 3481 3663 4283 5089 17,223
2346 6827 2908 3415 3708 19,203
969 2859 2872 2864 2834 12,399
4023 13,168 9442 10,563 11,630 48,825
Source: Bhartiya, A., et al., Food Security in India, http://www.scribd.com/doc/18009143/Food-Securityin-India, 2010.
3.4, 1.3, 0.9, 0.8, and 5.84 billion under INM, district level projects, micronutrients, sprinklers, liming, gypsum, and others, respectively [Bhartiya et al. 2010]. In agriculture and animal husbandry, the Green Revolution popularized the use of conventional hybridization to increase yield by creating HYVs. Often, the handful of hybridized breeds originated in developed countries and was further hybridized with local varieties in the rest of the developing world to create high-yield strains resistant to local climate and diseases. Local governments and industry have been pushing hybridization, which has resulted in several of the indigenous breeds becoming extinct or threatened. Because of unprofitability and uncontrolled intentional and unintentional cross-pollination and cross-breeding (genetic pollution), formerly huge gene pools of various wild and indigenous breeds have collapsed causing widespread 11% 5% 36%
6%
Seeds Soil amendments Local initiatives
8%
Pest management Mechanization Publicity
9%
Others 25%
FIGURE 6.2 National Food Security Mission interventions in different activities to enhance production. (Redrawn from Bhartiya, A., et al., Food Security in India, http://www.scribd .com/doc/18009143/Food-Security-in-India, 2010.)
Soil Resources Affecting Food Security and Safety in South Asia
301
583.6 1330
900
800
Distribution subsidy INM
3400 13570
District level projects Micronutrients Sprinklers Liming
4210
Gypsum Others 4210
FIGURE 6.3 National Food Security Mission interventions allocating millions of Indian rupees for different activities to enhance production. (Redrawn from Bhartiya, A., et al., Food Security in India, http://www.scribd.com/doc/18009143/Food-Security-in-India, 2010.)
genetic erosion and genetic pollution. This has resulted in loss of genetic diversity and biodiversity as a whole. 6.4.3.2 Bangladesh Bangladesh has increased its food grain production over the past 28 years, from 11.8 MMg in 1974 to more than 39 MMg in 2003 [USAID 2010a]. Total food grain import in the fiscal year 2009/10 was 3.45 MMg, i.e., 14% higher than the previous year [WFP 2010]. Of the total food grain import, 84% of the total food grain import was conducted by the private sector, of which only 1.2% was rice and 98.7% was wheat. Aid import stood at 47.2 thousand Mg, while the government imported 508.62 thousand Mg, i.e., representing about 13% of the total food grain import. Public stock of food grains has significantly decreased from 1.36 MMg in July 2009 to 0.MMg in June 2010. Although food grain is more available in good harvest years, Bangladesh as a whole still has a very low level of nutrition. This means many households and individuals do not eat a balanced, nutritious diet, even in good years. According to the World Bank, approximately 33 million of the 150 million people in Bangladesh cannot afford an average daily intake of more than 1800 kilocalories (the minimum standard for nutrition as set by the World Food Program). For people in most developing countries, the daily calorie average is 2828, compared with an average of only 2190 in Bangladesh. Poverty is the major factor effecting food security in Bangladesh. Despite the impressive increases in food grains, around half of Bangladeshis remains below the established food-based poverty line. And, as many as one-third are living in extreme poverty and severely undernourished. Recent food price increases, regular natural disasters, and strains on the global economic market have caused additional destabilization. The very poor in Bangladesh simply do not have enough money for food, much less enough to eat nutritiously. About 49% of Bangladeshis fall below the poverty line and 42% of the total population survives on less than a dollar a day. Bangladesh is a disaster-prone area subject to flooding, mud slides, and cyclones. As much as 50% of the population lives in these disaster prone
302
World Soil Resources and Food Security
areas, which further complicates their already vulnerable situation. During natural disasters, families often lose whatever few possessions they may own. Feed The Future (FTF) Implementation Plan (IP) for FY 2010 will provide a framework for all USG funded programs to address the underlying causes of poverty, hunger, and undernutrition in Bangladesh [USAID 2010a]. Although FTF will receive targeted funding to support predominantly agricultural programs that promote food security, USAID/Bangladesh will also use resources from other initiatives and ongoing programs, such as the Global Health Initiative, the Climate Change Initiative, Global Engagement Initiative, and Food for Peace Title II resources, to fund other core investment areas such as family-planning, health approaches to nutrition, social safety nets, and to support relevant governance issues, as described in the IP. It is important to use ongoing programs as a foundation for FTF: 1) Deep Urea Placement Program; 2) Agricultural Biotechnology Support Program; 3) National Food Policy Capacity Strengthening Program; 4) Protected Area Co-Management Program; 5) Health and Nutrition Program; and 6) PL 480 Title II Program. These programs are described in Section 4.3. In addition, the USAID Mission will reengage with the CGIAR and Bangladesh Agricultural Research Institutes to explore areas of mutual interest [USAID 2010b]. To reduce poverty in Bangladesh, it is crucial to develop and improve the capacities of its most vulnerable populations and regions. For this, Bangladesh needs to accelerate the growth and productivity of its agriculture and nonfarm sectors, improve the quality of social services, ensure the proper functioning of its community and rural institutions, and expand the rural support infrastructures. The government of Bangladesh has, over the past three-and-a-half decades since 1970s, introduced reform measures and policies for agricultural development in its quest for food security for all. In the 1970s and early 1980s, Bangladesh pursued a policy of agricultural modernization by supplying modern agricultural inputs (seed, fertilizer, and irrigation) and technology (HYVs and machinery) through government agencies and parastatal organizations like the Bangladesh Agricultural Development Corporation (BADC), and Bangladesh Water Development Board (BWDB). The government liberalized the seed market, allowed import of improved germplasm for research and development, and developed its own facilities for producing foundation seeds (except for five notified crops—rice, wheat, sugarcane, potato, and jute) through the Seed Policy Act of 1992 and 1998. It also encouraged the involvement of the private sector and NGOs in the seed delivery system. More recently, the newly elected government has decided to extend subsidies to fertilizers other than urea in a bid to promote balanced fertilizer use. Since the 1970s, the Bangladesh government has also undertaken a series of measures to expand its irrigation network. Minor irrigation systems have been developed in the country at a rapid pace. To promote the use of irrigation facilities, electricity for irrigation purposes is subsidized. In the 1970s and early 1980s, irrigation was promoted through public agencies, but since the mid-1980s, the government has also involved the private sector. The government made output market-related interventions through the domestic procurement of rice and wheat, distribution of food grains through the public food grain distribution system, and through tariffs on imported rice and wheat.
Soil Resources Affecting Food Security and Safety in South Asia
303
Currently, a big project titled the National Agricultural Technology Project (NATP) is being implemented. The government has also encouraged private sector and NGO participation in the development and promotion of HYVs and hybrid rice. Four types of programs were implemented in 2009—cash transfer programs, food security programs, microcredit programs, and special funds and development sector programs. Over the three decades since 1980, food grain production in Bangladesh has more than doubled, where rice and wheat production has increased from around 10 MMg in the early1970s to 25 MMg by the early 2000s. However, nearly half of the population still cannot afford an adequate diet. Also, as much of the countryside lies in disaster-prone, largely floodplain areas, annual flooding and occasional flashflooding, together with other periodic natural disasters, often cause crop damage and food shortages for the vulnerable population. These risks and uncertainties lead to transitory food insecurity. The major food security problem is the fact that around half of the Bangladeshis remain below the established food-based poverty line and as many as one-third are in extreme poverty and severely undernourished despite the impressive increases in aggregate national food grain availability. Success in making staple foods available coexists with a very high prevalence of undernourishment (insufficient caloric intake) and malnutrition. According to the Bangladesh Bureau of Statistics, in 2000, the malnutrition problem was desperately serious for the poorest 14% of the rural population who were consuming fewer than 1600 kcal per day. Another 10% consumed between 1600 and 1800 kcal per day, and around 23% consumed between 1800 and 2122 kcal. Considering the minimum caloric requirement to be food secure, 45% of women had low (<18.5) body mass indices and 52% of children were underweight. Program foci for improving food availability, access, utilization, and risk management include closing the seasonal food gap through improved storage, smallscale postharvest transportation, crop diversification (but not at the expense of food crops), expanded market opportunities, enhanced knowledge of dietary diversity, and restoration of food value, and on overcoming household cash flow and liquidity constraints, enhancing local capacity for disaster-risk management, and improved coping mechanisms for disasters like floods and cyclones. The focus on food utilization places a major emphasis on changing critical nutrition (diet, care, and feeding) and health-enhancing behaviors. Improved food utilization includes such areas as basic education, maternal and child health, control of infectious diseases, crop and food diversification activities, and improvements in water and sanitation. In addition, disaster-risk management and enhanced coping mechanisms can help reduce life and asset losses or minimize distress sales and contingency plans can foster quick postdisaster recovery. 6.4.3.3 Nepal For the past 13 years since 1997, agricultural policy in Nepal has been shaped by the Agricultural Perspective Plan (APP), which covers the period 1995–2010. The APP strategy is to achieve economic growth and poverty reduction objectives through accelerated growth of agriculture. Perhaps reflecting the relative antiquity of the APP, food security receives scant mention, and what discussion does exist is largely confined to the improved access the poor will receive as a result of increased
304
World Soil Resources and Food Security
employment opportunities, and the lower prices that are a spin-off of increased production and efficiency. The National Agricultural Policy of 2004 added new food access provisions for vulnerable groups, which are more radical than those of the APP. The interim constitution of Nepal 2006–2007 recognized food sovereignty as a fundamental human right. Consequently, the government has approved a food security plan as part of the Three Year Interim Plan to ensure the right to sustainable food security for all. Diversification of farm activities into high-value commercial crops and the processing of agricultural and other natural resource-based materials are therefore the most logical steps toward improving the economic levels of the mountains people [Sharma 1998]. Certain agroforest species are found at high-altitude conditions. The diverse agroecological conditions prevailing in the mountain form niches for horticulture, floriculture, spices, and medicinal plants [Pratap 2001]. Poverty, smallholding size, and food insecurity are the critical challenging issues in the mountains and call for a holistic approach to economic growth and environmental protection [Koirala and Thapa 1997]. In recent years, the role of horticultural crops including fruits and vegetables has been found significant to affect livelihood in the mountains [Tulachan 1999]. Livestock, particularly buffaloes and goats, provide greater opportunities to generate cash income for mountain households. In the HKH region, the majority of the households operate mix-farming systems [Tulachan 1999]. Over the years, many changes have been taking place in terms of land resource allocation, production, and productivity of cereal food grain crops, horticultural crops, and livestock structure and composition, all generally influenced by forest and other natural resources, input supplies, marketing, and other socioeconomic infrastructures. The important conditions characterizing mountain agriculture are inaccessibility, fragility, marginality, diversity, and niche [Jodha 1993]. The critical challenges in mountain agriculture are crop land security, soil erosion and declining soil fertility, increasing poverty, and farming on marginal farms [Sah 2001]. The studies have indicated numerous indicators (Table 6.3) of unsustainable mountain agriculture in the HKH [Jodha and Shrestha 1994]. Interventions that focus on improving agricultural practices and technologies will have immediate and significant impact on the levels of agricultural production in Nepal. The adoption of appropriate crop, livestock, and fishery practices, INM and IPM systems, sustainable water and soil management systems, and appropriate postharvest management practices are imperative. Coupled with market price incentives, enhanced agricultural extension is a key component for adoption of improved agricultural practices in Nepal. Through increasing the effectiveness of agricultural extension systems, appropriate practices and methods can be disseminated and adopted. Interventions that focus on proper management of water resources and development of effective small-scale irrigation practices will increase productivity and production while helping to conserve limited water resources. Natural resources management must be at the core of any food security strategy as the majority of rural Nepalese rely on natural resources for food, incomes, and livelihoods. Through increasing the sustainability of natural resource use and management, natural resources can be maintained and effectively utilized. Assisting producers to adapt to changing
Soil Resources Affecting Food Security and Safety in South Asia
305
climactic or meteorological conditions is important to prevent lower yields and crop losses. Through the introduction and adoption of improved varieties of crops and livestock (water-stress- or temperature-resistant), and better climatic forecasting, some of the adverse effects of climate changes can be mitigated. Through increasing the capacity of the private sector, and facilitating an increase in private investment into the agricultural sector and promoting an enabling environment, value chains can be improved, and agricultural productivity enhanced [USAID 2010c]. 6.4.3.4 Sri Lanka Almost 90% of Sri Lanka’s poor reside in the rural areas. The vast majority of the rural poor derive their living from agriculture and off-farm employment. Successful rural development means revitalizing the rural areas to provide a foundation for accelerated income growth, abundant job opportunities in farm and nonfarm activities, sustainable use of natural resources and strong, positive spillovers between urban and rural activity [MOA 2001]. In September, the Global Information and Early Warning System (GIEWS) on Food and Agriculture predicted that Sri Lankan farmers will harvest a bumper crop of rice during the 2010 August to October harvest season [USAID 2010d]. The March to April 2010 harvest season produced a record level of 2.65 million Mg, 12% more than the March to April 2009 harvest season, due to increased paddy cultivation areas, fertilizer subsidies, and favorable weather conditions. The GIEWS reported that increased harvests have led to reduced rice prices and improved food security in Sri Lanka. USAID/OFDA assistance provides community-based livelihood recovery activities, supports home gardening and dairy production, and provides agricultural assets, such as seeds and farming tools, benefitting nearly 21,000 individuals in the Vavuniya, Mannar, Jaffna, and Mullaittivu districts. In August, USAID/FFP provided $2.5 million to ACTED to fund cash/voucher-for-work activities that focus on rehabilitating productive assets and infrastructure. This program will benefit approximately 7000 returning IDP households. There are three main food security-related programs in Sri Lanka targeted at special groups and the poor. These are the Samurdhi program, Thriposha Program, and the midday meal program in schools. None of these are aimed specifically at ensuring food security; these are food-based welfare programs with objectives that go beyond mere food security. Sri Lanka’s largest welfare program, the Samurdhi program, was launched in 1995 with the twin objectives of ensuring food security and reducing poverty. The program covers two million households. Of its outlay, 80% is accounted for by a food stamp program. Its biggest drawback has been poor targeting of beneficiaries. To provide more poor families with secure title to land, the government will reduce state control over land and will continue to provide for freehold ownership of alienated state lands by issuing Jayabhoomi grants to both men and women. Government will remove restrictions on farmland and remove restrictions on the sale, lease transfer, subdivision, and mortgage of state lands in rural areas. Appropriate amendments will be made to the Land Development Ordinance (LDO), the Land Grants Act, and the State Land Ordinance to accomplish this. All restrictions on farm size and sale, lease and transfer of LDO and Land Grant (special provision) rural lands will
306
World Soil Resources and Food Security
be removed. Immediately, some 1.2 million freehold land titles were issued by the government in 2002, covering 2 million acres of land. A policy to sell and allocate state-owned land on the basis of marketable title will be developed. Anomalies in the legal structure relative to land inheritance that discriminate against women shall be corrected. Future allocations of state-owned lands will be limited to outright sales except for transfers targeted to the poor. An integrated data management system for all of the agencies dealing with land will be created in order to make information on land tenure, land use, and land capability transparent and accessible. A more flexible landmark combined with increased commercial agricultural investment will contribute to poverty reduction by expanding productive rural employment opportunities. The government’s strategy in agriculture is based on the need to be competitive in production and marketing by increasing productivity, lowering production costs, and adding value to raw materials. Raising the rate of growth in agriculture (including crops, livestock, fisheries, and forestry) can make an important contribution to rural poverty reduction. While substantial efforts were made to bolster paddy output during the 1960s, 1970s, and 1980s, other aspects of agriculture received scant public support. The government will give priority to encouraging small holders to cultivate high-value horticulture crops, livestock and fisheries products, and low-income farm households can increase returns by diversifying production out of low-value crops into higher value agricultural commodities. For example, unemployed women with secondary educations could engage in downstream activities, such as fruit processing, to enhance incomes. For this, better technology and better marketing opportunities is the key. 6.4.3.5 Pakistan Pakistan does not have any national food policy except for policies at the regional level. A plethora of food laws exist, mainly dealing with quality standards. However, these standards are also not addressed properly and require a complete overhaul. The Pakistani government has neglected the issue of food security and focused only on measures to increase production. Though production did increase, it was not sufficient to meet the country’s consumption needs. In Pakistan, procurement, handling, marketing, storage, and supplies are handled by the four provincial food departments and the national agency, the Pakistan Agricultural Services and Supplies Corporation. The Corporation was established in 1973 as a public limited company, fully owned by the federal government and six public sector banks. The agency has been entrusted with the tasks of procuring wheat and other agricultural commodities, providing price support to farmers, ensuring adequate supplies in deficit provinces/ regions, intervening in the open market to stabilize prices of agricultural commodities, and, above all, maintaining strategic reserves to meet any emergent situation. In 2006, the rated storage capacity with these agencies (Pakistan Agricultural Storage and Services Cooperation (PASSCO) and provincial food departments) was 4.34 million Mg, of which 2.45 million Mg was with the Punjab food department, 0.71 million Mg with the Sindh food department, 0.16 million Mg with the North West Frontier Province (NWFP), 0.44 million Mg with the Balochistan food department, 0.45 million Mg with PASSCO, and about 0.13 million Mg with other agencies. Results concluded that a state of food insecurity prevailed in rural Pakistan. Even
Soil Resources Affecting Food Security and Safety in South Asia
307
wheat surplus provinces were found, in the net, to be food insecure in terms of availability because of growing population pressure and the resultant rise in the demand for food. It recommended that a national food security strategy be formulated. It also recommended that the integration of the institutional arrangements for implementing the strategy with those for other socioeconomic interventions such as food availability and nutritional security be at the national level. It also stressed that initiatives such as poverty alleviation should go hand in hand with those aimed at food security to make the development process responsive to the needs of the people. In 2008, the Planning Commission of Pakistan formulated a task force on food security to provide policy recommendations to ensure food security. It was also asked to look into issues regarding procurement, medium-term research priorities, water resources, and climate change. Two major initiatives taken by the Pakistan government to improve food access are the Benazir Income Support Program (BISP) and the Food Support Program (FSP). The BISP was initiated with an initial allocation of Rs.34 billion (US $425 million) for the year 2008–2009. The program was initiated to partially offset the impact of inflation on the purchasing power of the poor. Food prices have seen a sharp rise since 2005. In the years 2005–2007, inflation stood at almost 10% with food inflation in the range of 13%–15%. In 2007–2008, the sharp rise in oil prices and primary products in the international as well as domestic market resulted in double-digit inflation, which almost halved people’s purchasing power. The program aims at covering almost 15% of the entire population, which constitutes 40% of the population below the poverty line. A monthly payment of Rs.1000 per family would increase the income of a family earning Rs.5000 by 20%. The BISP will cover all four provinces. The FSP was launched in 2002–2003, targeting the poor to improve their living standard by providing them with financial support. A subsidy of Rs.2000 was given to around 1.25 million poor families in two biannual installments of Rs.1000 through post offices countrywide. The government increased the rate of annual subsidy to Rs.2400 in 2003–2004 and further to Rs.3000 in 2005–2006. The program’s annual budget was increased from Rs.4.38 billion to Rs.6 billion in 2007–2008 to cover a larger number of the country’s poor.
6.5 FOOD SAFETY AND QUALITY Food safety is the assurance that food will not cause harm to the consumer when it is prepared and/or eaten according to its intended use. Ensuring safe and healthy food is an important precondition of food security. It is essential for human life in both developed and developing countries [Othman 2006]. Food-borne diseases result in suffering. Loss of lives comes from both malnutrition, including nutrient deficiency, or the presence of potentially toxic elements in the food. Both aspects of food safety are described below.
6.5.1 Elements of Food Safety 6.5.1.1 Malnutrition and Micronutrients Globally, malnutrition—including both overt nutrient deficiencies as well as dietrelated chronic diseases (e.g., heart disease, cancer, stroke, and diabetes)—is responsible for more deaths than any other cause, accounting for >20 million mortalities
308
World Soil Resources and Food Security
annually [Kennedy et al. 2003; WHO/FAO 2003]. Malnutrition also contributes to increased morbidity, disability, stunted mental and physical growth, and reduced national socioeconomic development [WHO/FAO 2003]. Micronutrient malnutrition alone afflicts more than two billion people, mostly among resource-poor families in developing countries, with Fe, I, Zn, and vitamin A deficiencies most prevalent [Kennedy et al. 2003]. More than five million childhood deaths occur from micronutrient malnutrition every year [Lancet 2007]. Leading global economists have identified investing in strategies to reduce malnutrition as the most cost-effective investments governments can make [Copenhagen Consensus 2008]. 6.5.1.2 Toxic Agents There exist a wide range of potentially toxic agents causing food-borne diseases and posing problems with respect to food safety. These include bacteria and bacterial toxins, zoonotic parasites, fungi and fungal toxins, aquatic and plant toxins, pesticide residues, heavy metals, radionuclides, drug residues, food adulterants, and certain food additives such as nitrates, nitrites, and nitrosamines [Bhatt 2010]. Important microorganisms causing food-borne diseases are Staphylococcus aureus, Baccilus cereus, Salmonella, Escherichia coli, Vibrio parahaemolyticus, and Clostridium perfringens. Acute bacterial food poisoning due to contamination of food is quite common. However, bacteriological confirmation of the diagnosis is lacking in most instances. Chemical contaminants in foods are also responsible for causing foodborne diseases and affecting the quality of food. Among the chemical contaminants, pesticide residues are the most important. Several outbreaks of diseases due to pesticides have been reported in India. Heavy metal contaminations of foods also pose a problem. Contamination may occur through environmental pollution of the air, water, and soil, such as the case with toxic metals, PCBs, and dioxins, or through the intentional use of various chemicals, such as pesticides, animal drugs, and other agrochemicals. Food additives and contaminants resulting from food manufacturing and processing can also adversely affect health.
6.5.2 Food Safety in Production and Postharvest 6.5.2.1 Conventional Breeding The task of plant breeders attempting to biofortify staple food crops is to increase the micronutrient level in the edible product of a staple food crop to have measurable impact on improving the nutritional health of individuals at high risk of developing micronutrient malnutrition [Howarth et al. 2010]. For this to be accomplished, plant breeders must work closely with food scientists and nutritionists to develop target micronutrient levels for their breeding programs. Crops may be modified to increase resistance to pests and disease, increase adaptability to environmental conditions, improve flavor or nutritional profile, delay ripening, or increase shelf life. Rice can be enriched with beta-carotene through genetic engineering. Biotechnology has moved at such a rapid pace that the safety of genetically modified foods has become a concern. At this time, there are no long-term, large-
Soil Resources Affecting Food Security and Safety in South Asia
309
scale tests to prove their safety or lack thereof. Unforeseen consequences may arise from widespread genetic modification of the food supply including allergic reaction, increased toxicity due to production of toxins, and resistance to antibiotics. In South Asia, Bt-cotton is being cultivated extensively in India and other food crops like Bt-brinjal are in the pipeline. Bt rice is rice that has been modified, by means of biotechnology, with genes from Bacillus thuringienis (Bt); the gene induces the rice plant to produce toxins (proteins) against common insect pests, for instance, stem borers. Maize, potato, and cotton plants containing Bt genes are now grown by farmers in several countries, with maize covering an estimated 7 million ha as of 1998 [TAC 2001]. International Crop Research Institute for Semi Arid Tropics (ICRISAT) has recently developed rapid and high throughput antibody-based screening assays for sensitive aflatoxin detection based on mutagenized germplasm from low-aflatoxin genotypes of Arachis species. It is hoped that molecular markers can accelerate an extensive breeding and backcrossing program to identify and incorporate useful traits [TAC-FAO 2001]. 6.5.2.2 Fertilizer Use Both macronutrient fertilizers containing N, P, K, and S, and certain micronutrient fertilizers (e.g., Zn, Ni, I, Co, Mo, and Se) can have significant effects on the accumulation of nutrients in edible plant products [Allaway 1986; Grunes and Allaway 1985]. Other micronutrient fertilizers have very little effect on the amount of the micronutrients accumulated in edible seeds and grains when they are applied to soils or when used as foliar sprays [Welch 1986]. For certain essential micronutrient elements (e.g., Zn, Ni, I, and Se), increasing soil-available supplies to food crops can result in significant increases in their concentrations in edible plant products [Graham et al. 2007; Welch 1995]. 6.5.2.3 Bioavailability Issue Increasing the concentrations of micronutrients in staple food crops is only the first step in making these foods richer sources of nutrients for humans. As stated previously, this is because not all of the micronutrients in plant foods are bioavailable to humans who eat these foods. Plant foods can contain substances (i.e., antinutrients) that interfere with the absorption or utilization of these nutrients in humans [Welch and Graham 1999]. In general, staple food crop seeds and grains contain very low bioavailable levels of Fe and Zn (i.e., only about 5% of the total Fe and about 25% of the total Zn present in the seed is thought to be bioavailable). Increasing the bioavailable amounts of Fe from 5% to 20% would be equivalent to increasing the total Fe by fourfold. Using conventional breeding, it should be genetically much easier to greatly improve the bioavailability of Fe and Zn compared with increasing their total content by this magnitude. Already biofortified products of sweet potato, bean, pearl millet, cassava, maize, rice, and wheat enriched with provitamin A and micronutrients (Zn, Fe) with good agronomic traits to resist various biotic (resistant to diseases) and abiotic (resistant to drought, lodging) have been approved for release by the national governments after 2–3 years of testing (Table 6.6).
310
World Soil Resources and Food Security
TABLE 6.6 Schedule of Product Release for Biofortified Products Countries of First Release
Crop
Nutrient
Bean
Fe, Zn
Pearl millet
Fe, Zn
Rwanda, Democratic Republic of Congo, India India
Rice
Zn, Fe
Bangladesh, India
Wheat
Zn, Fe
India, Pakistan
Agronomic Trait
Release Yeara
Virus resistance, heat and drought tolerance Mildew resistance, drought tolerance Disease and pest resistance, sub mergence tolerance Disease resistance, lodging
2010 2011 2012–2013
2012–2013
Source: Bouis, H.E., and Welsh, R.M., Crop Sci., 50, 20–30, 2010. a Approved for release by national governments after 2–3 years of testing.
6.5.2.4 Postharvest and Storage The standard of postharvest operations (storage, drying, processing, packaging, transportation, and marketing) has an equally important bearing on the quality and safety of the food item. Postproduction operations account for more than 55% of the economic value of the agricultural sector in developing countries and up to 80% in developed countries [TAC 2001]. IRRI has designed and field-tested prototype rice-harvesting machines and grain-drying systems. ICRISAT has a special project on pigeon pea improvement in eastern and southern Africa, with a subproject on harvesting, storage, and processing. IITA and CIRAD have jointly developed prototype mechanized rootharvesting systems for cassava. Centers such as CIP, CIAT, and IITA have engaged in farm research to develop loss-reducing crop storage systems [TAC 2001].
6.5.3 Safety Legislations People’s awareness of food safety has grown over the years and today’s government must respond quickly to food safety crises. Governments should develop comprehensive food safety policies and establish effective partnership among relevant stakeholders. The components and priorities of a food control system vary from country to country. Most systems in Southeast Asia will typically face challenges in strengthening the following key components: food legislation, food control management, inspection services, laboratory services, and information, education, communication, and training. Establishing and updating food legislation is a necessary first step in establishing an effective food safety system. In addition, there is a need to identify areas of the food chain not covered by the existing legislation, such as gaps in some countries’ laws governing feed, imports and exports, and hygiene. National regulatory standards must be formulated and reviewed based on risk assessment and, thus, incorporate
Soil Resources Affecting Food Security and Safety in South Asia
311
available scientific evidence. Whenever possible, these standards must be harmonized with international standards, i.e., the Codex standards [Othman 2007]. A regional meeting jointly organized by the FAO and WHO in 2004 made the following recommendations for Asian countries on food safety issues: • The large majority of countries of the region must urgently give higher priority to capacity-building to respond to the unacceptable burden of illnesses caused by the consumption of unsafe food. • Countries are urged to adopt a well-coordinated, multi-sectoral approach to food safety risk analysis. • Governments should make better use of resources available in the region including, for example, specialized reference laboratories, established surveillance systems and training capacities. • FAO, WHO, and other concerned international agencies and donors are called upon to support initiatives to address food safety challenges. [Othman 2007]
6.6 S TRATEGIES FOR IMPROVING SECURITY AND SAFETY OF FOOD Modern biotechnology has the potential to speed up the development and deployment of improved crops and animals. Marker-assisted selection, for instance, increases the efficiency of conventional plant breeding by allowing rapid, laboratory-based analysis of thousands of individual specimens without the need to grow the plants to maturity in the field. The techniques of tissue culture allow the rapid multiplication of clean planting materials of vegetatively propagated species for distribution to farmers. Genetic engineering or modification—manipulating an organism’s genome by introducing or eliminating specific genes—helps transfer desired traits between plants more quickly and accurately than is possible in conventional breeding. The spread of genetically modified (GM) crops has been rapid. Considerable research to develop more GM varieties is under way in some developing countries. The number and type of crops and applications involved is limited: two-thirds of the GM area is planted with herbicide-tolerant crops. All commercially grown GM crops are currently either nonfood crops (cotton) or are heavily used in animal feeds (soybean and maize). The success of Bt cotton in China has paved the way for further expansion of GM crops in India, which has considerable potential for GM products. While the adoption rates for GM technologies in developing countries are likely to rise, they are expected to slow in the developed world. This mainly reflects the impressive growth of the past, which limits the remaining potential. GM soybeans, for instance, already account for two-thirds of the soybean area worldwide and for an even larger share of the area in developed countries. As the global area of such crops expands, other more sophisticated biotechnology applications may gain importance. Examples include GM-based nutraceuticals or cosmetic applications. As these new applications are likely to produce a broader range of benefits than merely cheaper foods and feeds, consumers in the developed countries may become more inclined to accept them.
312
World Soil Resources and Food Security
6.7 CONCLUSION AND FUTURE PERSPECTIVES It is a matter of fact that a large portion of the land in South Asia is prone to chemical, biological, and physical degradation, due to both anthropogenic activities and natural processes. Because of land degradation, food production and the quality of the produce is severely affected. There is an urgent need to check this degradation, and necessary corrective measures should be adopted to reclaim the degraded soils. The respective governments of affected countries, in consultation with the FAO, should develop strategic plans in this regard. The governments should develop land-use maps of their countries and implement necessary legislation to stop the conversion of fertile lands into urban sprawl. In order to increase the food production to feed the ever-increasing population, the land earmarked for food crops should not be diverted to growing other biofuel-producing crops. The INM strategy is the key to sustainable crop production and to maintaining or enhancing the quality of produce. Biofortification, in conjunction with modern tools like biotechnology, is an emerging area of research that could probably meet the growing nutrient deficiencies in food crops that cause malnutrition in vulnerable populations in South Asia.
REFERENCES Abrol, I.P., and D.R. Bhumbla. 1971. Saline and alkali soils in India—Their occurrence and management. Paper presented at Soil Fertility Research FAO/UNDP Seminar. FAO World Soil Resources Rep. 41:42–51. Abrol, I.P. 2000. Agriculture in India, p.1–27. (www.planningcommission.gov.in) Asian Development Bank (ADB). 2007. Capacity building for resettlement risk management. In Handbook on resettlement for highway projects in India. Manilla: ADB. Alamgir, M. 1980. Famine in South Asia. Cambridge, MA: Oelgeschlager, Gunn, and Hain. Allaway, W.H. 1986. Soil–plant–animal and human interrelationships in trace element nutrition. In Trace elements in human and animal nutrition, ed. W. Mertz, 465–488. Orlando, FL: Academic Press. Bai, Z.G., D.L. Dent, L. Olsson, and M.E. Sachaepman. 2008. Proxy global assessment of land degradation. Soil Use Manag. 24:223–234. Bangladesh Bureau of Statistics (BBS). 2010. Bangladesh Food Security Monitoring. Bull. World Food Program. Issue no. 1, May–July, 2010. Barrett-Lennard E. 2002. Restoration of saline land through revegetation. Agr. Water Manage. 53:13–26. Barman, S.C., and M.M. Lal. 1994. Accumulation of heavy metals (Zn, Cu, Cd and Pb) in soils and cultivated vegetables and weeds grown in industrially polluted fields. J. Environ. Biol. 15:107–115. Barrow, C.J. 1991. Land degradation development and breakdown of terrestrial environments. Cambridge, UK: Cambridge University Press. Bhatt, R.V. 2010. Ensuring food safety and quality—The present picture in India. http://nutri tionfoundationofindia.res.in/archives.asp?archiveid=21. Bennett, D., and D.K. Macpherson. 1983. A history of salinity in Western Australia—Important (and some unimportant) dates. Technical Memorandum 83/1. Wembley, Western Australia: CSIRO Division of Groundwater Research. Bhartiya, A., A. Ratnakaran, B.A. Mohan, B. Banga, D. Ray, S. Seth, and S. Wasim. 2010. Food security in India. http://www.scribd.com/doc/18009143/Food-Security-in-India.
Soil Resources Affecting Food Security and Safety in South Asia
313
Bhattacharya, A. K. 1992. Status paper on waterlogged/drainage affected areas. In Proc. on Drainage Congestion in Northern Bihar, Vols. I and II. Patna, Bihar, India: Water and Land Management Institute. Bhattacharya, P., A.H. Welch, K.M. Ahmed, G. Jacks, and R. Naidu. 2004. Arsenic in groundwater of sedimentary aquifers. Appl. Geochem. 9:163–167. Bhattacharya, P., K.M. Ahmed, S. Broms, J. Fogelström, G. Jacks, and O. Sracek. 2006. Mobility of arsenic in groundwater in a part of Brahmanbaria district, NE Bangladesh. In Managing arsenic in the environment: From soil to human health, eds. R. Naidu, E. Smith, G. Owens, P. Bhattacharya, and P. Nadebaum, 95–115. Melbourne: CSIRO Publishing. Blum, W.E.A. 1988. Problems of soil conservation. Nature of Environment Series No. 40. Strasbourg: Council of Europe. Bouis, H.E., and M.W. Ross. 2010. Biofortification—A sustainable agricultural strategy for reducing micronutrient malnutrition in the global south. Crop Sci. 50:20–30. Brown, L.R. 1984. Facing food security. In State of the world, eds. L.R. Brown, C. Flavin, and S. Postel. London: Earthscan Publications. Carson, B. 1992. The land, the farmer and the future: A soil fertility management strategy for Nepal. ICIMOD Occasional Paper No. 92/21. Kathmandu, Nepal: ICIMOD. Chakraborti, D., M.K. Sengupta, M.M. Rahaman, S. Ahamed, U.K. Chowdhury, and Md. A. Hossain. 2004. Groundwater arsenic contamination and its health effects in the Ganga– Megna–Brahmaputra Plain. J. Environ. Monit. 6:74–83. Copenhagen Consensus. 2008. The Copenhagen Consensus 2008—Results, 1–6. Frederiksberg, Denmark: The Copenhagen Consensus. http://www.copenhagenconsensus.com/default .aspx?ID=953. Dent, F.J. 1990. In Problem soils of Asia and the Pacific. FAO RAPA Report, 1990/6, 44–68. Directorate of Economics and Statistics (DES). 2007. Agricultural statistics at a glance. New Delhi: Ministry of Agriculture, Government of India. DES. 2009. Agricultural statistics at a glance. New Delhi: Ministry of Agriculture, Government of India. DES. 2010. Agricultural statistics at a glance. New Delhi: Ministry of Agriculture, Government of India. Dhruva Narayana, V.V., and Babu, Ram. 1983. Estimation of soil erosion in India. J. Irrig. Drainage Engin. ASCE 109:419–434. Eswaran, H., P. Reich, and F. Beinroth. 2000. Land resources of South Asia—Stresses, quality and population supporting capacity. In International Conference on Managing Natural Resources for Sustainable Agricultural Production in the 21st Century, February 14–18, 2000, 97–109. Eswaran, H., F. Beinroth, and P. Reich. 1999. Global land resources and population supporting capacity. Am. J. Alternative Agr. 14:129–136. Food and Agriculture Organization of the United Nations (FAO). 1979. A provisional methodology for soil degradation assessment, 84. Rome: FAO. FAO. 1994. The state of food and agriculture. FAO Agriculture Series, No. 27, ISBN 92-5103550-4. Rome: FAO. FAO. 2002. World agriculture: Towards 2015/2030. Summary report. Rome: FAO. FAO. 2003. Rethinking the approach to groundwater and food security. AGL Water Publication No. 24. Rome: FAO. FAO. 2004. The state of food insecurity in the World. FAO of the United Nations, ISBN 92-5105178, Via le delle Terme di Caracalla, 00100. Rome: FAO, Italy. FAO AGL. 2005. Land and Water Development Division. National soil degradation map. http://www.fao.org./landwater/agll. FAO/WHO. 2006. Probiotics in food. Health and nutritional properties and guidelines for evaluation. Rome: FAO/WHO.
314
World Soil Resources and Food Security
Graham, R.D., R.M. Welch, D.A. Saunders, I. Monasterio, H.E. Bouis, M. Bonierbale, S. de Hann, G. Burgos, G. Thiele, R. Liria, C.A. Meisner, et al. 2007. Nutritious subsistence food systems. Adv. Agron. 92:1–74. Greenland, D.J., P.J. Gregory, and P.H. Nye. 1998. Land resources: On the edge of the Malthusian precipice? New York: CAB International, 180. Grunes, D.L., and W.H. Allaway. 1985. Nutritional quality of plants in relation to fertilizer use. In Fertilizer technology and use, 589–619. Madison, WI: SSSA. Gupta, R.K., S.P. Bhardwaj, and S.P. Mittal. 1990. Management of degraded lands in the Himalayas. In Technologies for wasteland development, 196–213. New Delhi: ICAR. Hoffman, G.J., and D.S. Durnford. 1999. Drainage design for salinity control. In Agricultural drainage. Agronomy No. 38, eds. R.W. Skaggs and J. van Schilfgaarde, 579–614. Madison, WI: American Society of Agronomy. Hossain, M. 2003. Development of boro rice cultivation in Bangladesh: Trends and policies. In Boro rice, eds. R.K. Singh, M. Hossain, and R. Thakur. New Delhi: Fine Grains Private Ltd. Howarth E.B., and R.M. Welch. 2010. Biofortification—A sustainable agricultural strategy for reducing micronutrient malnutrition in the global south. Crop Sci. 50:S-20–S-32. doi: 10.2135/cropsci2009.09.0531. Jodha, N.S. 1993. Perspective on property generating process in mountain areas. In Proc. of the International Forum on Development of Poor Mountain Areas March 22–27, 1993, Beijing, China, 38–46. Jodha, N.S., and S. Shrestha. 1994. Towards sustainable and more productive mountain farming. In Proc. of the International Symposium of Mountain Environment and Development: Constraints and Opportunities. Kathmandu, Nepal: ICIMOD. Joshi, P.K., A. Gulati, L. Tiwari, and P.S. Birthal. 2002. Diversification of Agriculture in South Asian countries. MSSD discussion paper. Washington, DC: International Food Policy Research Institute. Joshi, P.K., and N.K. Tyagi. 1994. Salt affected and waterlogged soils in India: A review. In Strategic change in Indian irrigation, eds. Svendsen, M., Gulati, A., 237–252. New Delhi: ICAR and IFPRI. Kennedy, G., G. Nantel, and P. Shetty. 2003. The scourge of “hidden hunger”: Global dimensions of micronnutrient deficiencies. Food Nutr. Agric. 32:8–16. Koirala, G.P., and G.B. Thapa. 1997. Food security challenges. Where does Nepal stand? Research Report Series No. 36. Kathmandu, Nepal: HMG-Ministry of Agriculture and Winrock International. Lal, R. 1989. Land degradation and its impact on food and other resources. In Food and natural resources, eds. D. Pimental and C.W. Hall, 65–140. San Diego: Academic Press. Lal, R., and B.A. Stewart. 1990. Soil degradation: A global threat. In Soil degradation: Advances in Soil Science, Vol. 11 eds. R. Lal and B.A. Stewart, xiii–xvii. Hong Kong: Springer-Verlag. Lal, R., T.M. Sobecki, T. Ivari, and J.M. Kimble. 2005. Soil degradation in the United States: Extent, severity, and trends. Boca Raton, FL: CRC Press. Lancet (editorial). 2007. Global childhood malnutrition. Lancet 367:1459. McFarlane, D.J., and D.R.Williamson. 2002. An overview of water logging and salinity in southwest Australia as related to the ‘Ucarro’ experimental catchment. Agric. Water Manage. 53:5–29. Mandal, D., V.N. Sharda, and A. Tiwari. 2008. Soil loss tolerance limits in planning conservation measures in different districts of India, 112. Dehradun, India: CSWCRTI. Mackenzie, E.T., and Mackenzie, J.A. 1995. Our changing earth: An introduction to earth system science and global environmental change. Upper Saddle River, NJ: Prentice Hall.
Soil Resources Affecting Food Security and Safety in South Asia
315
Mapa, R.B., D. Kumaragamage, W.D.L. Gunarathne, and A.R. Dassanayake. 2002. Land use in Sri Lanka: Past, present and the future. 17th World Congress of Soil Science, August 14–21, 2002, Bangkok, Thailand, Symposium No. 31, Paper No. 974. Ministry of Agriculture. 2010. National Planning Department, Vision 2010. Update of six-year development program for preparation of poverty reduction strategy paper for 2001– 2005, Colombo: Ministry of Agriculture. Moench, M. 2002. Groundwater and poverty: Exploring the links. In Intensive use of groundwater: Challenges and opportunities, eds. R. Llamas and E. Custodio, 478. Abingdon, UK: Taylor & Francis. Mukherjee, A.B., P. Bhattacharya, G. Jacks, D.M. Banerjee,, A.L. Ramanathan, and C. Mahanta. 2006. Groundwater arsenic contamination in India: Extent and severity. In Managing arsenic in the environment: From soil to human health, eds. R. Naidu, E. Smith, G. Owens, P. Bhattacharya, and P. Nadebaum, 533–94. Melbourne: CSIRO Publishing. Murali, Kanta. 2003. Food security atlas of urban India. Frontline. 20:18–31. National Commission for Irrigation. 1972. Report of the National Commission for Irrigation, 1972. New Delhi: Government of India, Ministry of Irrigation and Power. National Commission on Agriculture. 1976. Report of the National Commission on Agriculture, 1976, New Delhi: Government of India, Ministry of Agriculture. Nickson, R.T., J.M. McArthur, B. Shrestha, T.O. Kyaw-Mint, and D. Lowry. 2005. Arsenic and other drinking water quality issues in Muzaffargarh District, Pakistan. Appl. Geochem. 20:55–68. Othman, N.M. 2006. Food Safety in Southeast Asia: Challenges facing the region. Asian J. Agricul. Develop. 4:83–92. Pimentel, D., and M. Pimentel. 2008. Food, energy and society, 3rd ed., Boca Raton, FL: CRC Press. Pratap, T. 2001. High value cash crop in mountain farming: Mountain development process and opportunities. Kathmandu, Nepal: ICIMOD. Purakayastha, T.J. 2008. Improvement in soil fertility of Periurban Delhi through long-term sewage irrigation. Indian Farming 57:19–22. Rahman, R. 2000. Problem soils of Asia and the Pacific, Bangkok, Thailand, 1990. In Agroecological aspects of agricultural research in Bangladesh, ed. H. Brammer, H. Dhaka: University Press, Limited. Rattan, R.K., S.P. Datta, P.K. Chhonkar, K. Suribabu, and A.K. Singh. 2005. Long-term impact of irrigation with sewage effluents on heavy metal contents in soils, crops and ground water—A case study. Agricul. Ecosys. Environ. 109:210–322. Roy, A.D., and Shah, T. 2002. Socio-ecology of groundwater irrigation in India. In Intensive use of groundwater challenges and opportunities, eds. R. Llamas and E. Custodio, 307– 335. Abingdon, UK: Taylor & Francis. South Asia Co-operative Environment Programme (SACEPO). 2010. South Asia Co-operative Environment Programme. http://www.sacep.org. Sah, R.P. 2001. Socio-economic issues and options of marginal farms in the hills of mountains of Nepal: An experience in the western hills. International Symposium on Mountain Agriculture in the HKH Region, ICIMOD, May 21–24, 2001, Kathmandu, Nepal. Sanyal, S.K., and K.S. Dhillon. 2005. Arsenic and selenium dynamics in water-soil-plant system: A threat to environmental quality. Indian Soc. Soil Sci. Proc. Int. Confer. Soil, Water Environ. Qual., 293–263. Sen, A. 1982. Poverty and famine: An essay on entitlement and deprivation. Oxford: Clarendon Press. Shah, T., D. Molden, R. Sakthivadivel, and D. Seckler. 2000. The global groundwater situation: overview of opportunities and challenges. Colombo: IWMI. Sharma, U. 1998. Development of micro-enterprises—Ilam and Bhojpur districts. Discussion paper No. MEI 98/4. Kathmandu, Nepal: ICIMOD.
316
World Soil Resources and Food Security
Sikka, R., V. Nayyar, and S.S. Sidhu. 2009. Monitoring of Cd pollution in soils and plants irrigated with untreated sewage water in some industrialized cities of Punjab, India. Environ Monit. Assess.154:53–64. Sobhan, R. 1979. Politics of food and famine in Bangladesh. Economic and Political Weekly, 14. TAC Secretariat/FAO. 2001 Report from the Standing Committee on Priorities and Strategies— A status on food safety, 31. TAC Secretariat. Rome: FAO. Tandukar, N., P. Bhattacharya, A. Neku, and A.B. Mukherjee. 2006. Extent and severity of arsenic poisoning in Nepal. In Managing arsenic in the environment: From soil to human health, eds. R.E. Naidu, E. Smith, G. Owens, P. Bhattacharya, and P. Nadebaum, 595–604. Melbourne, Australia: CSIRO Publishing. Theobald, D. M. 2001. Quantifying urban and rural sprawl using the sprawl index. Paper presented at the annual conference of the Association of American Geographers, on March 2nd, 2001, New York, NY. Tripathi, B.P. 1999. Soil fertility status in the farmers’ fields of the western hills of Nepal. Lumle Seminar Paper No. 99/4. SAS/N Convention, June 9–11, 1999, Kathmandu, Nepal. Tripathi, B.P., R. Gardner, K.J. Mawdesley, G.P. Acharya, and R.P. Sah. 1999. Soil erosion and fertility losses in the western hills of Nepal: An overview. Workshop on enhanced farming livelihood in the hills of Nepal through improved soil fertility. July 25–26, 1999, Dhulikhel, Kavre, Nepal. Tulachan, P.M. 1999. Trends and prospects of sustainable mountain agriculture in the Hindu Kust—Himalayan Region. Issues in Mountain Development, No. 2. Kathmandu, Nepal: ICIMOD. Turkstra, J. 1996. Urban growth and land use options for lower-income groups: a case study of Villavicencio, Columbia. ITC Journal 1: 57–63. UNCCD. 2000. National report on desertification/land degradation in Sri Lanka, p. 1–40. Bonn, Germany: UNCCD. (http://www.unccd.int). UNEP. 1994. Land degradation in South Asia: Its severity, causes and effects upon the people. World Soil Research Report No. 78, p. 100. Rome: FAO. United States Agency for International Development (USAID). 2010a. Bangladesh: Food security. Washington, DC: USAID. http://www.usaid.gov/bd/programs/food_sec.html. USAID. 2010b. Bangladesh FY 2010 implementation plan. Washington, DC: USAID. http:// www.usaid.gov/our_work/agriculture/pdfs/2010/Bangladesh_IP_100727.pdf. USAID. 2010c. Nepal: Economic growth, education and food security. Washington, DC: USAID. http://nepal.usaid.gov/our-work/program-area/economic-growth-education-afood-security.html. USAID. 2010d. Sri Lanka—Complex emergency. Fact Sheet #3, Fiscal Year 2010. Washington, DC: USAID. http://www.usaid.gov/our_work/humanitarian_assistance/disaster_assistance/ countries/sri_lanka/template/fs_sr/fy2010/srilanka_ce_fs03_09-30-2010.pdf. USDA. 2008. Food security in the United States: Measuring household food security. Washington, DC: USDA. http://www.ers.usda.gov/Briefing/FoodSecurity/measurement .htm (retrieved 23-02-2008). WHO/FAO. 2003. WHO/FAO release independent expert report on diet and chronic disease. Geneva: WHO. http://www.who.int/mediacentre/news/releases/2003/pr20/en/. Welch, R.M. 1986. Effects of nutrient deficiencies on seed production and quality. Adv. Plant Nutr. 2:205–247. Welch, R.M. 1995. Micronutrient nutrition of plants. Crit. Rev. Plant Sci. 14:49–82. Welch, R.M., and R.D. Graham. 1999. A new paradigm for world agriculture: Meeting human needs—Productive, sustainable, nutritious. Field Crops Res. 60:1–10. World Food Programme (WFP). 2010. Bangladesh food security monitoring bulletin. Issue No. 1, May–July, 2010.
and 7 Formation Management of Cracking Clay Soils (Vertisols) to Enhance Crop Productivity Indian Experience D. K. Pal, T. Bhattacharyya, and Suhas P. Wani CONTENTS 7.1 Introduction................................................................................................... 317 7.2 Vertisols in a Climosequence: Their Formation and Modifications in their Properties.............................................................................................. 321 7.2.1 Formation........................................................................................... 321 7.3 Modification in Soil Properties Vis-à-Vis the Pedogenic Threshold............. 324 7.4 Pedogenic Threshold and Soil and Crop Productivity.................................. 328 7.4.1 Pedogenic Threshold in Dry Climates.............................................. 328 7.4.2 Loss in Crop Productivity Due to Pedogenic Thresholds................. 328 7.4.3 Loss in Soil Productivity Due to Irrigation....................................... 330 7.5 Management Intervention in Vertisols Vis-à-Vis Enhancement of Crop Productivity................................................................................................... 332 7.6 Concluding Remarks..................................................................................... 338 Acronyms................................................................................................................ 339 References............................................................................................................... 339
7.1 INTRODUCTION Cracking clay soils (Vertisols and their intergrades) [Soil Survey Staff 2006] are important natural resources in a wide range of climatic zones, from humid tropical to arid dry areas of the world [Ahmad 1996]. The natural vegetation of these areas is dry deciduous forest. However, most of the soils are now under cultivation. Soils of these climates have a significant role in the agricultural resource inventory of countries such as Australia, India, the Caribbean Islands, and the United States [Ahmad 1996]. 317
318
World Soil Resources and Food Security
Traditionally, these soils were often difficult to cultivate, particularly for small farmers using handheld or animal-drawn implements. Subsoil porosity and aeration are generally poor and roots of annual crops do not penetrate deeply. Farmers faced with these difficulties allow these soils to lie fallow for one or more rainy seasons or cultivate them only the postrainy season [Pal et al. 2009a]. Vertisols and their intergrades have limitations that restrict their full potential to grow two crops in a year during the rainy season and winter [NBSS and LUP-ICRISAT 1991]. However, with the advent of technologies and the introduction of new crops in assured rainfall ecoregions, two crops are now grown on these soils [El-Swaify et al. 1985; Wani et al. 2003]. Rainfed agriculture is globally predominant (80%) and contributes 56% to the world’s food basket. Current productivity of farmers’ fields in the rainfed tropics are two- to four-fold lower than the achievable crop yields [Wani et al. 2003; 2009; Rockstroem et al. 2007]. In India, sorghum (Sorghum spp.), maize (Zea mays), and pearl millets (Pennisetum glaucum) are the main cereal crops under dryland farming. Pigeonpea (Cajanus cajan), mung bean (Vigna radiate), and chickpea (Cicer arietinum) are the main pulses, safflower (Carthamus tinctorius), soybean (Glycine max), and groundnut (Arachis hypogaea) are the main oilseed, and cotton (Gossypium spp.) is the main industrial crop. Growing sorghum in the soils of wetter climates has been a recent trend. Sugarcane (Saccharam spp.), paddy (Oryza sativa), wheat (Triticum aestivum), and cotton are grown under irrigation. Intercropping is very common under rainfed agricultural systems [Jodha 1980]. The major combinations are sorghum/pigeonpea, cotton/pigeonpea, and cotton/sorghum/pigeonpea. These kind of mixtures usually combine crops with different maturity lengths, drought-sensitive with drought-tolerant crops, cereals with legumes, and cash crops with food crops [Swindale 1989]. In Africa, Vertisols occupy an area of over 100 million hectare (Mha) of arable land. The landscape of African Vertisols is quite diverse and a relatively small portion of the total area of Vertisols is irrigated in Sudan. The most widely cultivated crop is cotton. Wheat is grown in northern Africa in rainfed conditions. In the Ethiopian Highlands, several food cereals, grain legumes, and oil seed crops are produced due to an adequate and predictable rainfall. The major forms of land use in African Vertisols, however, are small-scale, mixed rainfed and arable farming, and extensive rearing of livestock. In general, under rainfed conditions, crop yields and productivity are low [Ahmad 1996]. In Caribbean Vertisols, sugarcane is the dominant crop and grown in both rainfed and irrigated conditions. In Guyana, the soils are rejuvenated periodically by flood fallowing. In Jamaica, the crop is grown with full irrigation, and in Trinidad, under rainfed conditions, the cropping cycle is synchronized with the seasons. Rice is grown continuously throughout the year where irrigation is available [Ahmad 1996]. In Australia, Vertisols are extensively used for dryland agriculture in the east and south. Wheat, safflower, barley, and oats are winter crops, and sorghum, maize, cotton, soybean, sunflower, and millets are grown as summer crops. Under irrigated conditions, cotton, some grain and fodder crops, and rice are grown [Ahmad 1996]. More than half of Vertisols in the United States are confined to Texas, where mean annual precipitation ranges from 760 to 1150 mm [Puentes et al. 1988]. This precipitation is received mainly before or during the growing season and the favorable rainfall makes these soils productive. Wheat, oats, sorghum, and maize are important crops;
Formation and Management of Cracking Clay Soils (Vertisols)
319
Global soil regions
Soil orders Alfisols
Entisols
Inceptisols
Spodosols
Rocky land
Andisols
Gelisols
Mollisols
Ultisols
Shifting sand
Aridisols
Histosols
Oxisols
Vertisols
Ice/glacier
FIGURE 7.1 Global distribution of Vertisols indicating areal extent in India. (From staff of National Bureau of Soil Survey and Land Use Planning, 2002.)
320
World Soil Resources and Food Security
other crops are cotton and hay. The land-use pattern in the upper coast area is more diverse with grain sorghum, soybean, rice, maize, and cotton. With limited irrigation, cotton and sorghum are the main crops grown [McKee and Hajek 1973]. These agricultural land-use scenarios clearly highlight that even though Vertisols make up a relatively homogeneous major soil group, they occur in a wide range of climatic environments throughout the world and show a considerable variability in their land uses and crop productivities [Coulombe et al. 1996; Syers et al. 2001]. In view of their shrink-swell properties and stickiness, Vertisols are recognized by a number of local regional and vernacular names [Dudal 1965; Dudal and Eswaran 1988]. In India alone, they are known by at least 13 different names [Murthy et al. 1982] that are related either to their characteristic dark color or some aspect of their difficult workability or both. Thus, these soils exhibit spatial and temporal variability in their properties and remain difficult land resources to manage successfully. This variability, however, needs to be fully comprehended while developing technologies improve their performance. Soil is an open system within an ecosystem and it is strongly influenced by the external environment. Therefore, it becomes very necessary to understand the factors that cause the variability in their properties. A critical review of the recent developments TABLE 7.1 Distribution of Vertisols in Different States of India under a Broad Bioclimatic System States Uttar Pradesh Rajasthan Gujarat Madhya Pradesh Maharashtra Andhra Pradesh Karnataka Tamil Nadu Puducherry and Karaikal Jharkhand Orissa India
Bio-Climatea SAM, SHD AD AD, SAD, SAM SAM, SHD, SHMc SAD, SAM, SHD, SHMc SAD, SAM, SHD AD, SHD, SHM, H SAD, SAM, SHD, SHM, H SHM SHM, SHD SHM, SHD, H
Area (Mha)(%)b 0.41 (0.12) 0.98 (0.30) 1.88 (0.57) 10.75 (3.27) 5.60 (1.70) 2.24 (0.68) 2.80 (0.85) 0.91 (0.28) 0.011 (0.003) 0.11 (0.034) 0.90 (0.28) 26.62 (8.10)
Source: Bhattacharyya, T. et al., NBSSLUP Publication 143, 2009. a AD: arid dry, 100–500 mm MAR (mean annual rainfall); SAD: semiarid dry, 500– 700 mm MAR; SAM: semiarid moist, 700–1000 mm MAR; SHD: subhumid dry, 1000–1200 mm MAR; SHM: subhumid moist, 1200–1600 mm MAR; H: humid, 1600–2500 mm MAR. b Parentheses indicate percent of the total geographical area of the country. c In addition, Vertisols occur in HT climate (>2500 mm MAR) in Madhya Pradesh and Maharashtra, but they are not mapable in 1:250,000 scale [Bhattacharyya et al. 1993, 2005, 2009; Pal et al. 2009b].
Formation and Management of Cracking Clay Soils (Vertisols)
321
in Vertisols research is necessary, as agricultural productivity is firmly rooted in the physical, chemical, and biological properties of soils [van der Merwe et al. 2001]. As in several parts of the world, Vertisols occur in wide climatic zones in India (Figure 7.1). They are found in humid tropical (HT), humid (H), subhumid moist (SHM), subhumid dry (SHD), semiarid moist (SAM), semiarid dry (SAD), and arid dry (AD) climatic environments [Bhattacharyya et al. 1993, 2005, 2009; Pal et al. 2009a, 2009b] in eleven states of India (Table 7.1). These soils have attracted scientific attention since the nineteenth century [Leather 1898; Harrison and Sivan 1912], mainly from India where they were historically important agricultural soils for growing cotton. A review on an array of Vertisols in a climosequence of India and the modification of soil properties (if any) impairing or favoring the crop productivity appears to be an excellent model case study to address and understand the factors for the variability in Vertisols vis-à-vis their agricultural land uses and crop productivity. Development of such state-of-the-art information may guide stakeholders in better understanding the intricacies of Vertisols for their efficient use and management in varied climatic environments, not only in tropical India but in the other tropical parts of the world.
7.2 V ERTISOLS IN A CLIMOSEQUENCE: THEIR FORMATION AND MODIFICATIONS IN THEIR PROPERTIES 7.2.1 Formation The majority of Vertisols in India occur in the lower piedmont plains or valleys [Pal and Deshpande 1987], or in microdepressions [Bhattacharyya et al. 1993]. They are developed mainly in the alluvium of weathering Deccan basalt [Pal and Deshpande 1987; Bhattacharyya et al. 1993] mostly during the Holocene period [Pal et al. 2001, 2006]. Frequent climatic changes occurred during the Quaternary [Ritter 1996]. As a result, the soils of many places of the world witnessed climatic fluctuations, especially in the last postglacial period. Brunner [1970] reported evidence of tectonic movements during the Plio-Pleistocene transition, which caused the formation of different relief types and relief generation. With the formation of the Western Ghats during the Plio-Pleistocene crustal movement, the humid climate of the MiocenePliocene was replaced by the semiarid conditions that continue to prevail in central and southern Peninsular India. The Arabian Sea flanks the Western Ghats, which rise precipitously to an average height of 1200 m, the result of a heavy orographic rainfall all along the west coast. The leeside toward the east receives less than 1000 mm of rainfall and is typically rain-shadowed [Rajaguru and Korisetter 1987]. The current aridic environment prevailing in many parts of the world, including India [Eswaran and van den Berg 1992], may create adverse physical, chemical, and biological properties of soils [Pal et al. 2001]. A recent study [Pal et al. 2009b] indicates that the color of Vertisols in HT climates is dark brown (7.5YR 3/3) to dark-reddish (5YR 3/3) and yellowish brown (10YR 4/3), and dark (10YR 3/1) to dark-grayish brown (10YR 3/2) in soils of other climates (SHM, SHD, SAM, SAD, and AD). Structurally, they are small, weak wedge-shaped aggregates with pressure faces that break to weak, angular blocky structures in soils of HT climates. In soils of SHM, SHD, SAM, SAD, and AD climates, their structure is strong, medium subangular
322
TABLE 7.2 Properties of Vertisols in a Climosequence Climosequence Properties Soil color Subsoil structure
Microfabric
Classification
Subhumid Moist (SHM)
Subhumid Dry (SHD)
Semiarid Moist (SAM)
Semiarid Dry (SAD)
Arid Dry (AD)
7.5YR 3/3 to 5YR 3/3 ←--------------------------10YR 3/1 to 10YR 3/2----------------------→ Weak and small wedgeStrong, medium subangular blocky to strong, coarse angular blocky with pressure faces and slickensides that break shaped aggregates that into small angular peds break to weak subangular blocks with pressure faces >0.5 cm wide, extend down to the zones of sphenoids and wedge-shaped peds with smooth or Deep cracks cut through Bss horizons slickensided surface Poro/grano/reticulate-striated plasmic fabric Stipple/ mosaicMosaic/ crystallitic Mosaic/ stippleCrystallitic plasma indicating high shrink-swell activity speckled plasma plasma indicating speckled plasma indicating very low indicating moderate moderate indicating low shrink-swell activity shrink-swell activity shrink-swell activity shrink-swell activity Typic Haplusterts Typic/aridic Haplusterts Sodic Haplusterts and sodic Calciusterts
Source: Pal, D.K., et al., Quaternary International 209, 6–21, 2009b.
World Soil Resources and Food Security
Cracks
Humid Tropical (HT)
Formation and Management of Cracking Clay Soils (Vertisols)
323
blocky to strong, coarse angular blocky with pressure faces and slickensides that break into small angular peds (Table 7.2). In soils of HT, SHM, SHD, and SAM climates, cracks > 0.5 cm wide extend down to the zones of sphenoids and wide-shaped peds with smooth or slickensided surfaces, whereas cracks cut through the slickensided horizons (Bss) in soils of SAD and AD climates (Table 7.2, Figure 7.2). With the lowering of mean annual rainfall (MAR), the soils are more alkaline, calcareous, and sodic due to a progressive increase in pedogenic CaCO3 (PC) content in HT to AD climates (Figure 7.2). This demonstrates the fact that the aridity in the climate is the prime factor in the formation of calcareous sodic soils (showing the presence of calcium carbonate and exchangeable sodium percentage, ESP ≥ 5), as evidenced by the formation of typic Haplusterts in HT, typic/udic Haplusterts in SHM, SHD, and SAM, and sodic Haplusterts and sodic Calciusterts [Soil Survey Staff 2006] in SAD and AD climates, respectively (Table 7.2, Figure 7.2). Smectite clay minerals are ephemeral in the HT climate and they transform to kaolin [Pal et al. 1989; Bhattacharyya et al. 1993]. Thus, it is difficult to understand the formation of Vertisols in HT climates. Ca-zeolites in these soils provide sufficient bases to prevent the complete transformation of smectite to kaolin. The presence of smectites and zeolites make the formation of Vertisols possible in lower physiographic positions, even under HT climates. The formation and persistence of slightly acidic to acidic Vertisols, not only in western and central India [Bhattacharyya et al. 1993, 1999, 2005] but elsewhere [Ahmad 1983], can only be possible in the presence of soil modifiers (Ca-zeolites, gypsum) that maintain the base saturation of these soils well above 50% [Pal et al. 2003c, 2006].
00 cm
SHM MAR (1134 mm)
HT MAR (3647 mm)
SHD MAR (1071 mm)
SAM MAR (977 mm)
SAD MAR (764 mm)
AD MAR (533 mm)
50 cm
100 cm
Typic haplustert Typic haplustert > 500 yr BP to <65 million yr BP
Ca-zeolite
Typic haplustert Aridic haplustert Sodic haplustert Sodic calciustert
Non-pedogenic CaCO3 (NPC)
Late holocene period Pedogenic CaCO3 (PC)
Bss horizon
Pressure faces
Vertical cracks
FIGURE 7.2 Successive stages of pedogenic evolution of Vertisols in a climosequence from humid to arid climate. (Based on data from Pal, D.K. et al., Quaternary International 209, 6–21, 2009b.)
324
World Soil Resources and Food Security
In arid climatic environments, the weathering of primary minerals contributes very little toward the formation of smectite and it has been demonstrated that the formation of PC at the expense of nonpedogenic CaCO3 (NPC) is the prime chemical reaction that triggers the increase in pH, exchangeable Mg+2, and Na+ on the exchange complex [Pal et al. 2000, 2009b; Srivastava et al. 2002]. The fine clay smectite of these soils is fairly well crystallized, but is also partially hydroxy-interlayered [Pal et al. 2009c]. Hydroxy-interlayering in smectite can occur in an acidic environment at a pH well below 8.3 [Jackson 1964]. The Vertisols of subhumid, semiarid, and arid climates are calcareous and some are sodic. Thus, hydroxy-interlayering does not appear to be a contemporary pedogenic process in the prevailing dry climatic conditions. The crystallinity of smectite is, however, preserved in the nonleaching environments of the subhumid to dry climates. This suggests that smectite in Vertisols must have formed in an earlier more humid climate and been deposited in the lower Piedmont Plains, valleys, or in microdepressions where the majority of Vertisols of India occur. These Vertisols are thus developed during the drier climate of the Holocene [Pal et al. 2009b].
7.3 M ODIFICATION IN SOIL PROPERTIES VIS- À-VIS THE PEDOGENIC THRESHOLD The soils of the SHM and SHD do not generally contain PC in the first 50 cm of the profile (Figure 7.2) and have better drainage (saturated hydraulic conductivity, sHC > 20 mm hr−1, Table 7.3). The subsoils of the Vertisols in SAM, SAD, and AD climates, due to the accelerated rate of formation and accumulation of PC, are becoming sodic, impairing their sHC even in the presence of soil modifiers like Ca-zeolites (Table 7.3) [Pal et al. 2006, 2009b]. The initial impairment of the percolative moisture regime in the subsoils (caused either by exchangeable Mg (EMP), or exchangeable Na (ESP), or both) results eventually in a system where gains exceed losses. This self-terminating process [Yaalon 1983] leads to the development of aridic to sodic Haplusterts and, subsequently, to sodic Calciusterts [Pal et al. 2009b]. Vertisols have considerable amounts of water-dispersible clay (WDC) [Pal et al. 2006, 2009b], which increases with depth. This proves that an adequate dispersion of clay smectite is possible in a slightly acidic to moderately alkaline pH under a very low electrolyte concentration (ECe ≤ 1 me L−1) [Pal et al. 2006, 2009b]. This ensures a conducive pH condition higher than the zero point of charge required for full dispersion of clay [Eswaran and Sys 1979]. Thus, the movement of deflocculated fine clay smectite and its subsequent accumulation in the Bss horizons is possible in noncalcareous as well as calcareous Vertisols. The depth distribution of EMP and ESP and soluble Na+ ions in the majority of Vertisols of India [Pal et al. 2003c, 2006, 2009b] indicates clearly that the precipitation of CaCO3 as PC enhances the pH and also the relative abundance of Na+ ions, both in soil exchange and soil solution. The Na+ ions, in turn, cause the dispersion of smectites to translocate down the depth of soils. Thus the formation of PC and clay illuviation are two concurrent and contemporary pedogenic events that provide an appropriate soil environment to bring the pedogenic threshold during the dry climates of the Holocene until a further change in climate sets in [Pal et al. 2003a, 2009b].
Horizon
ESP
Base Saturation %
Saturation Extract Residual Sodium Carbonate (RSC) me/L
Zeolitic: Kheri soils: Subhumid moist: Typic Haplusterts 23 23.0 48.0 32 48.0 22 53.5 21 49.3 16 49.3 10 49.3 20 52.1
0.8 0.8 0.9 0.8 1.2 0.8 1.1
98 90 85 109 83 85 80
−0.06 1.31 1.55 1.99 1.33 0.22 0.65
Nonzeolitic: Paral soils: Semiarid dry: Sodic Haplusterts 17 4.0 54.4 5 56.5 2 47.8 3 51.8 1 52.5 1 43.3
1.4 4.1 8.1 14.2 16.7 21.0
84 79 92 97 91 100
0.07 0.94 5.89 6.95 NDb ND
Depth cm
Clay %
Fine Clay %
Ap Bw1 Bw2 Bss1 Bss2 Bss3 Bss4
0–14 14–32 32–61 61–82 82–112 112–133 133–156
51.1 53.7 46.3 53.6 46.6 44.6 47.7
27.3 32.2 31.0 28.7 33.7 27.0 34.8
Ap Bw1 Bss1 Bss2 Bss3 Bss4
0–9 9–35 35–69 69–105 105–132 132–150
55.3 58.9 56.9 62.6 61.8 56.3
22.6 30.7 29.5 35.6 37.6 37.6
sHC mm hr−1
sHC mm hr−1 weighted meana
CEC cmol(p+)kg−1
Formation and Management of Cracking Clay Soils (Vertisols)
TABLE 7.3 Hydraulic Properties of Vertisols as Influenced by ESP in the Presence and Absence of Ca-Zeolites
325
326
TABLE 7.3 (Continued) Hydraulic Properties of Vertisols as Influenced by ESP in the Presence and Absence of Ca-Zeolites
Horizon
Saturation Extract Residual Sodium Carbonate (RSC) me/L
3.6 2.5 1.5 1.6 1.6 1.9 4.2
98 108 105 110 106 109 112
−1.16 0.70 0.60 −1.58 1.11 0.05 1.04
Zeolitic: Vasmat soils: Semiarid moist (irrigated): Sodic Haplusterts 46.2 18 13.0 47.6 4.2 45.8 17 45.3 10.4 48.0 5 46.7 18.8 46.4 10 45.9 13.7 37.5 13 47.8 12.1 44.3 12 51.3 8.0
107 119 94 105 103 109
0.59 2.24 0.95 0.99 0.44 0.60
Depth cm
Clay %
Fine Clay %
Ap Bw1 Bw2 Bss1 Bss2 Bss3 Bss4
0–12 12–31 31–48 48–74 74–110 110–148 148–165
45.0 47.6 52.4 49.2 50.0 50.9 50.4
25.1 27.9 31.4 31.4 31.2 32.6 29.1
Ap Bw1 Bw2 Bss1 BC/Bss Bss2
0–20 20–42 42–68 68–102 102–131 131–150
64.0 66.9 66.3 66.4 59.1 70.9
sHC mm hr−1
CEC cmol(p+)kg−1
Zeolitic: Jhalipura soils: Semiarid dry: Typic Haplusterts 8 10.0 36.5 15 36.5 7 40.2 6 37.0 13 36.5 14 37.0 3 38.0
Source: Pal, D.K., et al. Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for land use planning. NRDMS, DST Project Report, Nagpur, India, 2003c. a Weighted mean of 0–100 cm.; b ND = not determined.
World Soil Resources and Food Security
ESP
Base Saturation %
sHC mm hr−1 weighted meana
327
Formation and Management of Cracking Clay Soils (Vertisols)
The dominant presence of poro/parallel/grano/reticulate-striated plasmic fabric indicates high shrink-swell activity of smectites in soils of HT and SHM climates [Pal et al. 2009b]. On the other hand, the dominant presence of stipple/mosaicspeckled plasma in soils of SHD, mosaic/crystallitic plasma in soils of SAM, mosaic/ stipple-speckled plasma in soils of SAD, and crystallitic plasmic in soils of AD climates suggests that shrink-swell magnitude is much less in the soils of drier climates compared to HT and SHM climates, as manifested in poor plasma separation. Thus, weak swelling is sufficient for the development of sphenoids and/or slickensides, but not adequate to cause a strong plasma separation [Pal et al. 2006, 2009b]. The sHC decreases rapidly with depth in all soils, but the decrease is sharper in both zeolitic and nonzeolitic soils of SAD and AD climates (Table 7.3), because of their subsoil sodicity [Pal et al. 2006, 2009b]. The decreased sHC restricts both vertical and lateral movement of water in the soils. During the very hot summer months, this would result in much less water in the subsoils of SAD and AD climates. This is evident from the deep cracks cutting through the Bss horizons (Figure 7.2). The lack of adequate soil water during the shrink-swell cycles restricts the swelling of smectite and results in weaker plasma separation. The subsoils of SAM, SAD, and AD remain under a smaller amount of water as compared to those of HT, SHM, and SHD climates [Pal et al. 2006, 2009b]. These polygenetic soils with very low water-supplying capacity to plants would lose further crop productivity in the event of more climatic adversities [Pal et al. 1989, 2001, 2009b]. Such situations shall remain a challenge for resource-poor farmers growing more than one crop in a year (Figure 7.3).
5
ESP 50
100
100
Arid
Depth (cm) 50
0
Subhumid Aridity Time
FIGURE 7.3 Projected view of the progressive development of soil sodicity from a wet to dry climate with time, a threat to resource-poor farmers.
328
World Soil Resources and Food Security
7.4 P EDOGENIC THRESHOLD AND SOIL AND CROP PRODUCTIVITY 7.4.1 Pedogenic Threshold in Dry Climates The pedogenic threshold in soils of dry climates during the Holocene has been realized as a natural soil degradation process induced by tectonic- and climatelinked events [Pal et al. 2003a, 2003b, 2009a, 2009b]. An example from benchmark Vertisols, with and without soil modifiers, representing a climosequence from SHM to AD climates (Table 7.4) indicates that dry climates during the late Holocene restricted further leaching and as a result, the formation of PC was favored [Pal et al. 2001]. The amount of PC in the first 1 m of the profile of soils of a representative climatic region indicates in general a progressive increase in the rate of formation of PC (from 0.39 to 2.12 mg per 100g soil per year) and ESP (from 1 to 28), and a decrease in sHC (from 33 to 2 mm hr−1) (Table 7.4).
7.4.2 Loss in Crop Productivity Due to Pedogenic Thresholds Vertisols have limitations that constrain their full potential to grow both rainyseason and winter crops (NBSS and LUP-ICRISAT 1991). This situation is observed at Nagpur, Amravati and the Akola districts of Maharashtra in central India. Vertisols of the western part of Amravati and the adjoining Akola district support either kharif/rabi (summer/winter) crops, whereas in the Nagpur district, both crops are grown with limited irrigation [Kadu et al. 2003]. The MAR in Akola, Amravati, and Nagpur is 877, 975, and 1127 mm, respectively. This indicates more aridity in Akola than in Amravati or Nagpur. Vertisols of these districts are deep-ploughed (30-cm soil depth) once every 2 to 3 years where a blade harrow is used each summer before the onset of monsoon season. The system is monoculture (cotton) with 4 months of fallow (February–May). Frequent blade harrowing is done after rains to produce the necessary tilth for sowing. Organic manures are added every 2 years. Improved cultivars are hand sown at a depth of about 5 cm on marker intersection points at appropriate times. The interrow spacing is 75–90 cm. Repeated intercultural operation is done to remove weeds, improve aeration, and to adsorb maximum moisture from receding rains, as well as to function as a dust mulch during the postrainy season. The first intercultural operation is carried out 20–25 days after sowing, which is followed by fertilizer application (N-P-K 18:18:10). Cotton yield (average of 5–6 years covering an area of about 200 ha in each village) was collected during the study by periodic contact with eight to ten farmers during 1991–2001. The variation in yield between years was about 40%. The maximum yield of cotton (1.8 t ha−1) obtained by the farmers following the typical management was taken as the optimum yield [FAO 1983] to evaluate the soils [Kadu et al. 2003]. Under similar soil management by farmers in 29 Vertisol (pedons) areas, and also under similar soil moisture and temperature regimes, cotton yields were better in the soils of Nagpur than those of Amravati and Akola (Table 7.5). The subsoils in Amravati and Akola are becoming sodic due to accelerated rates of formation and accumulation of PC.
329
Formation and Management of Cracking Clay Soils (Vertisols)
TABLE 7.4 Rate of Formation of PC and Concomitant Development of ESP in Vertisols of a Representative Climatic Region in India Soil Seriesa (District, State)
Nabibagh (Typic Haplusterts) (Bhopal, Madhya Pradesh)
Linga (Typic Haplusterts) (Nagpur, Maharashtra) Bhatumbra (Udic Haplusterts) (Bidar, Karnataka)
Jhalipura (Typic Haplusterts) (Kota, Rajasthan) Teligi (Sodic Haplusterts) (Bellary, Karnataka) Kovilpatti (Gypsic Haplusterts) (Thoothokudi, Tamil Nadu) Sollapuram (Sodic Haplusterts) (Anantapur, Andhra Pradesh) Paral (Sodic Haplusterts) (Akola, Maharashtra)
Sokhda (Sodic Calciusterts) (Rajkot, Gujarat)
Maximum ESP in 1-m Profile
sHC (mm hr−1)b
Subhumid Moist (MAR 1209 mm) 3.7 0.39
~1
20
Subhumid Dry (MAR 1011 mm) 7.8 0.76
1
23
4c
6c
3.6d
10d
CaCO3 (%)b
10.1
Rate of Formation of PCb
0.90
Semiarid Dry (MAR 842–583 mm) 5.5 0.57
9.6
0.94
17d
24d
7.9
1.02
1e
33e
17.5
1.32
18
2
10.4
1.48
14
4
28c
17c
Arid Dry (MAR–533 mm) 21.7 2.12
Source: Pal, D.K., et al., Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for better land use planning, NRDMS, DST Project Report, Nagpur, India, 2003c. Note: PC = pedogenic CaCO3; ESP = exchangeable sodium percentage; sHC = saturated hydraulic conductivity; MAR = mean annual rainfall. a Soil Survey Staff [2003], b weighted mean in the first 1 m of the profile, c irrigated, d Ca-zeolitic, e gypsic.
330
World Soil Resources and Food Security
TABLE 7.5 Range in Values of PC, ESP, sHC, and Yield of Cotton in Vertisols of Vidarbha, Maharashtra District, Maharashtra, Central India Nagpur (MAR 1011 mm) Amravati (MAR 975 mm) Akola (MAR 877 mm)
Soil Classification
PC (%)
Typic Haplusterts/ 3–6 Typic Calciusterts (a) Aridic Haplusterts 3–7 (b) Sodic Haplusterts 3–13 (a) Aridic Haplusterts 3–4 (b) Sodic Haplusterts 3–4
ESP
sHC (mm hr−1) Weighted Mean in the Profile (1 m)
Cotton Yield (tz ha−1) (seed + lint)
0.5–11
4–18
1.0–1.8
0.8–14 16–24 16–44 19–20
2–19 0.6–9.0 3–4 1–2
0.6–0.7 0.3–0.8 1.0 0.6
Source: Kadu, P.R., et al., Soil Use and Management, 19, 208–216, 2003. Note: PC = pedogenic CaCO3; ESP = exchangeable sodium percentage; sHC = saturated hydraulic conductivity.
This impairs their sHC and, hence, a significant positive correlation between cotton yield and sHC was observed [Kadu et al. 2003].
7.4.3 Loss in Soil Productivity Due to Irrigation Many productive cracking clay soils under rainfed conditions have been unrendered for agriculture through irrigation. Such a menacing situation is not only confined to irrigated (command) areas but also occurs in situations where river or well waters are used for irrigation [Nimkar et al. 1992; Pal et al. 2003c]. In 7 years of irrigation, nonzeolitic Vertisols at Chendkapur, Amravati district, Maharashtra have become more calcareous and ESP shows a four-fold increase as compared to nonirrigated soils. Likewise the ECe shows a two- to threefold increase. In addition, soils have become highly alkaline and the sHC has been impaired (Table 7.6). On the other hand, some zeolitic Vertisols (at Vasmat, Hingoli district, Maharashtra) in SAD climates are being irrigated through canals to produce sugarcane for the past 2 decades. These soils lack salt-efflorescence on the surface and are not waterlogged at present. This apparently suggests that these soils are not degraded due to better drainage (please see sHC in Tables 7.3 and 7.6). However, values of pH, ECe, CaCO3, and ESP of the same soil, with and without irrigation, indicate the development of sodicity in irrigated soils (Table 7.6). The presence of Ca-zeolites ensures a constant supply of soluble Ca 2+ ions that help in maintaining sHC > 10 mm hr −1 (Table 7.6). Natural endowment with modifiers in soils is not uncommon in other parts of the world. No ill effects of high ESP (>15) in crop production in the Vertisols of Gezira in Sudan or in Tanzania were observed [El Abedine et al. 1969; Robinson 1971; Ahmad 1996]; Vertisols of both the places showed high base saturation, mainly with Ca 2+ ions. It appears that these soils might contain soil modifiers like zeolites,
331
Formation and Management of Cracking Clay Soils (Vertisols)
TABLE 7.6 Comparative Soil Properties of Unirrigated and Irrigated Deep Black Soils (Haplusterts) of Maharashtra, India Depth (cm)
pH (1:2 water)
ECe (dS m−1)
CaCO3 (<2 mm) %
SAR
ESP
sHC (mm hr−1)
8.3 8.4 8.5 8.6 8.6
Unirrigated Chendkapur Soilsa 2.3 6.1 3.5 1.5 6.9 2.9 1.9 8.3 3.1 2.4 12.6 3.4 1.9 13.9 3.6
3.8 3.4 3.6 3.9 4.2
3.2 2.7 2.3 2.0 1.3
0–15 15–43 43–59 59–93 93–129
8.9 8.9 8.8 8.6 8.6
Irrigated Chendkapur Soilsa 5.7 10.4 18.2 6.5 10.9 20.7 5.2 11.8 18.7 3.4 12.3 11.6 3.2 12.7 7.9
17.8 18.2 15.3 11.6 7.0
3.1 3.2 3.2 0.9 0.9
0–18 18–45 45–77 77–108 108–142 142–166
8.4 8.2 8.2 8.7 9.2 9.2
0.94 2.50 2.64 1.85 0.67 0.32
Unirrigated Vasmat Soilsb 21.5 17.7 17.4 15.8 17.2 17.2
– – – – – –
4.0 4.6 5.5 6.0 5.5 3.9
26 34 35 33 13 12
0–20 20–42 42–68 68–102 102–131 131–150+
9.0 9.2 9.3 9.0 9.0 9.0
0.77 1.01 0.99 1.25 1.09 1.02
Irrigated Vasmat Soilsb 16.0 17.0 17.0 15.0 25.3 16.1
– – – – – –
4.2 10.4 18.8 13.7 12.1 8.0
18 17 5 10 13 12
0–17 17–44 44–67 67–100 100–130
Source: a Nimkar, A.M., et al., Agropedology, 2, 59–65, 1992; b Pal, D.K., et al., Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for better land use planning, NRDMS, DST Project Report, Nagpur, India, 2003c. Note: ECe = electrical conductivity of the saturation extract; SAR = sodium adsorption ratio; ESP = exchangeable sodium percentage; sHC = saturated hydraulic conductivity.
332
World Soil Resources and Food Security
as suggested by base saturation in excess of 100 [Bhattacharyya et al. 1999]. However, sustainability of crop productivity in dry climates, even in the presence of soil modifiers may not be possible because pedogenic thresholds in dry climates would make these soils more calcareous, sodic, and impermeable to water with time (Figure 7.3). Many Vertisols in higher MAR regions (>1400 mm) are cultivated to rice with well-water irrigation, like growing sugarcane in both nonacidic and near-neutral Vertisols in India [Pal et al. 2003c]. Despite the probable presence of Ca-zeolites, such Vertisols (Kheri, Table 7.3) have accumulated HCO3− and CO3− − ions in their soil solution, making residual sodium carbonate (RSC), ranging from 1 to 2 (Table 7.3). This practice caused the decline in wheat productivity and impaired sHC in the subsoils (sHC ≤ 20 mm hr−1) (Table 7.3). Thus, Vertisols of SHM to AD climates need to be cultivated to crops that are possible under rainfed agricultural management even though the soils have modifiers. However, the beneficial effect of soil modifiers can be realized in HT climates through sustained productivity of rice and sugarcane in both slightly acidic to acidic Vertisols of the Caribbean that are highly base-saturated (>100%), even in the presence of very high organic carbon content (>>1%) [Ahmad 1996].
7.5 M ANAGEMENT INTERVENTION IN VERTISOLS VIS- À-VIS ENHANCEMENT OF CROP PRODUCTIVITY The loss and gain of Ca2+ ions during the formation of PC and dissolution of soil modifiers have a relevance both in soil exchange and soil solution for crop productivity by improving the hydraulic property of soils [Pal et al. 2006] besides their (Ca2+ ions) role as environmental sensors [Nayyar 2003]. The cultivation of sugarcane and rice has been successful because of the continuous supply of Ca2+ ions by the soil modifiers, even in HT climates. Sustainability of such agricultural land use is likely to remain as a viable management intervention for years, until the Vertisols become devoid of soil modifiers forever. However, despite their role in improving sHC, the use of irrigation, either with canal or well water, cannot help sustain the good crop yield because of the development of high pH, ECe, CaCO3, and ESP in Vertisols in different climatic environments. The presence of CaCO3 (mainly the PC) in Vertisols has generally been considered of doubtful significance in replacing exchangeable Na+ ions by Ca2+ ions of CaCO3 at a pH around 8.0. However, it is generally affected by other factors such as the application of gypsum, followed by cropping. The beneficial effect of naturally endowed gypsum has been realized in the Vertisols of southern Peninsular India, even in SAD climates. Such gypsum-containing Vertisols have sHC > 30 mm hr−1 and ESP < 15 (Table 7.7), despite rapid formation of PC because of the much higher solubility of gypsum than Ca-zeolites [Pal et al. 2006]. Even after realizing the beneficial role of gypsum in the slightly to highly sodic soils of the Indo-Gangetic Plains (IGP) and Vertisols, in terms of better physical, chemical, and biological properties [Gupta and Abrol 1990; Rao and Ghai 1985], the use of gypsum as management intervention in Vertisols of the dry climates of western, central, and southern Peninsular India is not commonly practiced [Venkateswarlu 1984; Pal et al. 2009c], unlike in similar soils
333
Formation and Management of Cracking Clay Soils (Vertisols)
TABLE 7.7 Physical and Chemical Properties of Vertisols Endowed with Gypsum in Semiarid Dry Parts of Tamil Nadu, India Depth (cm) 0–6 6–20 20–41 41–74 74–104 104–128 128–140 140+
pH (1:2 Water)
ECe (dS m−1)
CaCO3 (<2 mm) %
ESP
sHC (mm hr−1)
8.0 8.0 8.0 8.0 7.9 7.9 7.4 7.5
0.2 0.3 0.5 0.4 0.2 0.6 2.7 –
5.4 4.3 5.3 7.9 12.5 12.8 15.6 17.4
0.5 0.9 0.6 0.9 1.1 1.4 1.8 0.3
19 22 44 30 37 34 32 48
Source: Pal, D.K., et al., Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for better land use planning, NRDMS, DST Project Report, Nagpur, India, 2003c. Note: ECe = electrical conductivity of the saturation extract; ESP = exchangeable sodium percentage; sHC = saturated hydraulic conductivity.
in Australia [McGarity et al. 1988]. In dry areas of Australia, the positive response to gypsum and gypsum deep tillage could be due to increased water storage in subsoils. This management intervention offers a cost-effective means of increasing crop productivity under rainfed agricultural systems [McGarity et al. 1988]. Under rain-fed conditions, the yield of deep rooted crops in Vertisols depends primarily on the amount of rain stored in the profile, and the extent to which this soil water is released during the crop growth [Kadu et al. 2003; Pal et al. 2006]. In the semiarid part of western and central India, rainfed cotton is grown under suboptimal conditions, with soil depth and moisture availability as the main limitations. Field experiments conducted in the Yavatmal district of Maharashtra (central India) [Venugopalan et al. 2004] on the comparison of soil properties of Vertisols of SAM climates under organic and nonorganic (conventional) cotton production systems indicate that the yield of cotton and component crops grown under the organic production system were higher than those of the nonorganic production system and, in general, the productivity was higher than the average productivity of the district. Even in a hot, semiarid climate, higher values of soil organic matter (SOC) (>0.6%) in the organic production system have been due to the sequestration of carbon (Table 7.8) as compared to conventional systems [Venugopalan et al. 2004]. The limits of SOC content of the typical soil association of smectitic and noncalcareous Mollisols–Alfisols–Vertisols of tropical India under various land uses indicate that the clay mineral type of soils could be one of the important factors influencing the build-up of SOC. Such agricultural management intervention can help the sequestration of SOC, even up to 1% [Bhattacharyya et al. 2005]. Due to improvement in SOC
334
World Soil Resources and Food Security
TABLE 7.8 Comparison of Chemical Properties of Surface Soils (0–20 cm) under Organic and Conventional Production Systems (Based on 55 Soil Samples) Conventional Farming Properties pH (1:2 water) OC (%) CaCO3 (%)
Organic Farming
Range
Mean
Range
Mean
7.7–8.4 0.20–0.80 2.4–12.2
8.0 0.54 6.2
7.1–8.1 0.30–1.70 1.1–12.5
7.7 0.76 5.3
Source: Venugopalan, M.V., et al., Effect of organic and conventional cotton production systems on soil properties: A case study in Yavatmal district, Maharashtra, International symposium on strategies for sustainable cotton production—A global vision, Vol. II, 118–121. 2004.
and the subsequent dissolution of CaCO3, the pH of soils under organic production systems remained below 8.1 (Table 7.8). A long-term heritage watershed experiment initiated in 1976 at the ICRISAT Centre, Patancheru, Andhra Pradesh, India under rainfed conditions to demonstrate how an improved system of catchment management (IM) in combination with an appropriate cropping system can sustain increased productivity and improve the soil quality of Vertisols [Wani et al. 2003, 2007], in comparison to the existing traditional farming (TM) system. The improved system followed soil and water conservation practices, where excess rainwater was removed in a controlled manner. The soil and water conservation practices consisted of improved, legume-based crop rotation and improved nutrient management. In the TM system, sorghum or chickpeas were grown in the postrainy season with organic fertilizers, and in the rainy-season, the field was maintained as a cultivated fallow. The updated results from this experiment (Figure 7.4) indicate that the average grain yield of the improved cropping system over 30 years was 5.1 t ha−1 yr−1, nearly a fivefold increase in the yield over the TM system with an average yield of about 1.1 t ha−1 yr−1. The annual gain in yield in the IM system was 82 kg ha−1 yr−1 as compared to 23 kg ha−1 yr−1 in the TM system (Figure 7.4). The IM system thus has a higher carrying capacity (21 persons versus 4.6 persons ha−1 of the TM system) [Wani et al. 2009]. The IM system shows increased rainwater use efficiency (65% versus 40%), reduced runoff (from 220 mm to 91 mm) and soil loss (from 6.64 t ha−1 to 1.6 t ha−1), along with increased crop productivity and carrying capacity of land [Wani et al. 2003]. All these benefits have, however, been possible in the improvement of the hydraulic properties of Vertisols under IM systems as compared to TM systems (Table 7.9). Vertisols under IM and TM system have comparable pH, clay, and fine clay content (weighted mean, WM, in the 0–100 cm), however, the sHC value (WM) of IM has increased by almost 2.5 times due to the reduction in ESP through the dissolution of CaCO3 (Table 7.9). The CaCO3 (WM) content of IM decreased from 6.2% under the TM system to 5.7% (WM). In the past 24 years, the rate of dissolution of CaCO3 is 21 mg yr−1 in the first 100 cm of the profile. Under the IM system, the inclusion of pigeonpea,
335
Formation and Management of Cracking Clay Soils (Vertisols)
which produces piscidic acid in root exudates that solubilize iron-bound phosphorous (Fe-P) [Ae et al. 1990] and the rootlets in soil through which rainwater passes, or other sources of CO2 could have caused the increase in solubility of PC and a slight increase in exchangeable Ca/Mg (Table 7.9). Increased SOC sequestration in soils of the tropics induced dissolution of native CaCO3 and its leaching [Bhattacharyya et al. 2001]. The further importance of inorganic carbon in sequestering carbon in soils of dry regions is highlighted by Sahrawat [2003]. The improvement in soil properties through the IM system is also reflected in the classification of Vertisols. The Vertisols under the IM system qualify as typic Haplusterts, after being originally classified as sodic Haplusterts in the TM system (Table 7.9). The contribution of the dissolution of CaCO3 to the improvement of soil quality of the Vertisols under the IM system validates the soil carbon transfer model [Bhattacharyya et al. 2004]. The IM system with better hydraulic properties has also helped the Vertisols to sequester more SOC since 1976. At present, soils under the IM system contain 0.53% SOC in the 0–100-cm soil depth, in comparison with soils under the TM system, which contain 0.42% (Table 7.9). The rate of addition of soil carbon for the past 24 years since 1977 has been around 5 mg yr−1 in the first 100 cm of the soil profile of the IM system (Table 7.9). A study was conducted by ICRISAT and its partners to determine the carbon status of Vertisols at 21 benchmark sites covering arid, semiarid, and moist HT locations in India to identify carbon sequestrating systems [ICRISAT 2004]. The study indicates that after 20 years, the Vertisols sequestered more organic carbon than ferruginous Alfisols. The legume-based systems (high management) sequestered more carbon than the cereals and the horticultural systems, whereas grasslands sequestered more carbon than the annual crops [Bhattacharyya et al. 2007a,c; Sahrawat et al. 2005; Ramesh et al. 2007]. 8
Observe potential yield
Yield (t/ha)
6
4
2
Rate of growth 82 kg ha–1 yr–1
Carrying capacity 21 persons ha–1
BW 1
Carrying capacity 4.6 persons ha–1 Rate of growth 23 kg ha–1 yr–1
0 1976
1979
BW 4C
1982
1985
1988
1991
1994
1997
2000
2003
2006
Year
FIGURE 7.4 Three year moving average of sorghum and pigeonpea grain yields under improved (IM) and traditional management (TM) in a deep Vertisol catchment at Patancheru, India. (Based on data from Wani, S.P. et al., International Journal of Environmental Studies 64, 719–727, 2007.)
336
Horizon
Fine Clay Clay (%), (%), Weighted Fine Weighted Depth Clay Mean in Clay Mean in cm % 0–100 cm % 0–100 cm
Organic CEC sHC mm Carbon CaCO3 (cmol(p+) Exchangeable hr−1, (%) (%), kg −1), Ca/Mg, ESP sHC Weighted pH Organic Weighted Weighted CEC Weighted Weighted Weighted mm Mean in H2O Carbon Mean in CaCO3 Mean in cmol(p+) Mean in Exchangeable Mean in Mean in hr −1 0–100 cm (1:2) % 0–100 cm % 0–100 cm kg−1 0–100 cm Ca/Mg 0–100 cm ESP 0–100 cm Kasireddipalli Soil (Sodic Haplusterts) under Traditional Management (TM)a
Ap
0–12
48.0
53.0
26.4
33.0
7.0
7.8
0.6
Bw1
12–30
51.4
29.7
6.0
4.0
7.8
0.4
0.42
6.0 6.2
6.2
48.7 52.1
52.2
3.2 2.8
2.2
2.0
Bss1
30–59
52.5
32.5
6.0
8.1
0.4
6.0
52.2
2.1
7.1
Bss2
59–101 55.6
36.4
2.0
8.3
0.4
6.4
53.5
1.8
13.0
4.0
Bss3
101– 130
59.4
30.8
2.0
8.3
0.4
6.5
57.8
3.1
8.0
BCk
130– 160
58.0
38.7
1.0
8.2
0.1
9.1
49.5
1.5
22.2
8.3
World Soil Resources and Food Security
TABLE 7.9 Modification of Physical and Chemical Properties of Vertisols through the Improved Management System at ICRISAT, Patancheru in the 24 Years since 1977
0–12
52.1
54.7
28.8
32.8
17.0
7.5
1.0
Bw1
12–31
51.5
28.1
16.0
11.0
7.8
0.6
0.53
4.5
4.2
5.7
54.3
50.4
56.0
2.4
2.9
2.4
2.0 2.0
Bss1
31–54
54.2
34.0
10.0
7.8
0.4
6.2
55.6
1.7
3.0
Bss2
54–84
57.3
40.0
9.0
8.2
0.4
5.1
56.4
1.9
7.0
Bss3
84–118 56.5
26.0
7.0
8.1
0.5
8.6
61.6
3.8
7.0
Bss4
118– 146
59.3
31.7
3.0
8.2
0.5
8.4
58.2
2.1
7.0
BC
146– 157
60.0
41.5
--
8.2
0.3
7.4
55.2
1.1
9.0
Source:
4.5
Pal, D.K., et al., Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for better land use planning, NRDMS, DST Project Report, Nagpur, India, 2003c; b Bhattacharyya, T., et al. Physical and chemical properties of selected benchmark spots for carbon sequestration studies in semi-arid tropics of India. Global Theme on Agro-ecosystems Report No. 35, 2007c, Andhra Pradesh, India, ICRISAT and ICAR. a
Formation and Management of Cracking Clay Soils (Vertisols)
Kasireddipalli Soil (Typic Haplusterts) under Improved Management (IM)b Ap
337
338
World Soil Resources and Food Security
Despite an overall benefit of IM systems in enhancing the crop productivity and improving the soil quality of Vertisols of semiarid tropics under rainfed conditions, its widespread adoption at the farmers’ level is still fraught with crop failure due to capricious rainfall patterns and socioeconomic constraints [Myers and Pathak 2001].
7.6 CONCLUDING REMARKS Vertisols are a relatively homogeneous soil group. They occur in a wide range of climatic conditions and exhibit remarkable variability in their properties, either in the presence or absence of soil modifiers (mainly Ca-bearing minerals like Ca-zeolites and gypsum). This review has created a window of updated knowledge that should assist the stakeholders of cracking clay soils (Vertisols and their intergrades) in better understanding the efficient use and management of their soils in the varied climatic environments of the world. Sustaining the productivity of rice and sugarcane in Vertisols endowed with soil modifiers is possible for a considerable period in HT climates. However, the pedogenic threshold (signifying the natural degradation process), in both zeolitic and nonzeolitic Vertisols of drier climates (SHM, SHD, SAM, SAD, and AD) is causing the decline in productivity of cereals, cash crops (cotton), and legumes. The rate of formation of CaCO3 and the concomitant development of subsoil sodicity in the Vertisols of India provide a realistic scenario as to how the dry climatic conditions pose a threat to agriculture (Figure 7.3), as it demands extra resources for raising crops (especially the winter crops) from resource-poor farmers [Pal et al. 2009a]. Research initiatives on the significance of PC and soil modifiers in the management of the Vertisols of dry climates at the NBSS and LUP (ICAR), Nagpur, India [Srivastava et al. 2002; Pal et al. 2006, 2009a, 2009b, 2009c] suggest that for sustained performance of crops in soils of dry climates, the replenishment of Ca2+ ions both in the soil solution and in the exchange complex appears to be a viable technological intervention. The solubility of PC can be enhanced by establishing crops through the IM system of ICRISAT, with inclusion of legumes and improved soil water and crop management options. The extra soluble Ca2+ ions would lower the equilibrium pH and ESP and make Vertisols more permeable to both air and water. The favorable soil water status would cause the enhancement of crop productivity. Soil modifiers may facilitate further improvement in water statuses in soils to release adequate amounts of water for crops. Vertisols of dry climates can thus show a natural resilience [Pal et al. 2009a]. Rainfed agriculture is predominant (80%) globally; however, current productivity levels are hovering around 1 to 1.5 t/ha−1 yr−1. Unlike the holistic approach taken for irrigated agriculture in Green Revolution areas in India, subsistence rainfed agriculture has so far been dealt with by compartments, such as soil conservation, water management, improved cultivars, and fertilizer application. The full potential of the technologies has not been realized, nor have these technologies been adapted on a sufficiently large scale to have a substantial impact [Wani et al. 2007]. In view of stagnating food grain production in the IGP areas, the maintenance of national
Formation and Management of Cracking Clay Soils (Vertisols)
339
buffer stock has become more dependent on the contributions by the few states of the northwestern part of the IGP that represent high crop productivity regions [Dhillon et al. 2010]. The total area of the IGP is 43.7 Mha, which produces 50% of the total food grain to feed 40% of the Indian population [Abrol and Gupta 1998; Pal et al. 2009d]. On the other hand, cracking clay soils (Vertisols and their intergrades) are less intensively cultivated as compared to the IGP areas [Bhattacharyya et al. 2007b], even though they occupy a nearly 66-Mha area [Bhattacharyya et al. 2009]. Therefore, areas dominated by cracking clay soils deserve immediate national attention so as to avoid the pitfalls encountered in the high productivity regions of the IGP [Bhattacharyya et al. 2007b; Dhillon et al. 2010]. Adaptation of the IM system may make Vertisols of dry climate more resilient and capable of producing more food grains required for the populous Indian subcontinent.
ACRONYMS AD arid hot ECe electrical conductivity of the saturation extract EMP exchangeable magnesium percentage ESP exchangeable sodium percentage HT humid tropical ICAR Indian Council of Agricultural Research IGP Indo-Gangetic Plains IM improved management MAR mean annual rainfall NBSS and LUP National Bureau of Soil Survey and Land Use Planning NPC nonpedogenic calcium carbonate PC pedogenic calcium carbonate SAD semiarid dry SAM semiarid moist sHC saturated hydraulic conductivity SHD subhumid dry SHM subhumid moist SOC soil organic carbon TM traditional management WDC water dispersible clay WM weighted mean
REFERENCES Abrol, I.P., and R.K. Gupta. 1998. Indo-Gangetic Plains—Issues of changing land use. LUCC Newsletter March, 1998, No. 3. Ae, N., Arihara, J., Okade, K., Yoshihara, T., and C. Johansen. 1990. Phosphorus uptake by pigeonpea and its role in cropping systems of the Indian subcontinent. Science 248:477–480. Ahmad, N. 1983. Vertisols. In Pedogenesis and soil taxonomy, Vol II. The soil orders, ed. L.P. Wilding, N.E. Smeck, and G.F. Hall, 91–123. Amsterdam: Elsevier.
340
World Soil Resources and Food Security
Ahmad, N. 1996. Occurrence and distribution of Vertisols. In Vertisols and technologies for their management, ed. N. Ahmad, and A. Mermut, 1–41, Amsterdam: Elsevier. Bhattacharyya, T., Chandran, P., Ray, S.K., et al. 2007a. Changes in levels of carbon in soils over years of two important food production zones of India. Current Science 93:1854–1863. Bhattacharyya, T., Chandran, P., Ray, S.K., et al. 2007c. Physical and chemical properties of selected benchmark spots for carbon sequestration studies in semi-arid tropics of India. Global Theme on Agro-ecosystems Report No. 35. Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), and New Delhi: Indian Council of Agricultural Research (ICAR). Bhattacharyya, T., Pal, D.K., and P. Srivastava. 1999. Role of zeolites in persistence of high altitude ferruginous Alfisols of the humid tropical Western Ghats, India. Geoderma 90:263–276. Bhattacharyya, T., Pal, D.K., and S.B. Deshpande. 1993. Genesis and transformation of minerals in the formation of red (Alfisols) and black (Inceptisols and Vertisols) soils on Deccan Basalt in the Western Ghats, India. Journal of Soil Science 44:159–171. Bhattacharyya, T., Pal, D.K., Chandran, P., and S.K. Ray. 2005. Land use, clay mineral type and organic carbon content in two Mollisols-Alfisols-Vertisols catenary sequences of tropical India. Clay Research 24:105–122. Bhattacharyya, T., Pal, D.K., Chandran, P., et. al. 2004. Managing soil carbon stocks in the Indo-Gangetic Plains, India. New Delhi: Rice–Wheat Consortium for the Indo-Gangetic Plains. Bhattacharyya, T., Pal, D.K., Chandran, P., Ray, S.K., Durge, S.L., and S.P. Wani. 2007b. Available K reserve of two major crop growing regions (alluvial and shrink-swell soils) in India. Indian Journal of Fertilizers 3:41–46, 49–52. Bhattacharyya, T., Sarkar, D., Sehgal, J., et al., 2009. Soil taxonomic database of India and the states (1:250,000 scale). NBSSLUP Publication 143. Nagpur: NBSS and LUP. Bhattacharyya, T., Pal, D.K., Velayutham, M., Chandran, P., and C. Mandal. 2001. Soil organic and inorganic carbon stocks in the management of black cotton soils of Maharashtra. Clay Research 20:21–29. Brunner, H. 1970. Pleistozane Klimaschwankungen im Bereich den Ostlichen MysorePlaleaus (SudIndien). Geologie 19:72–82. Coulombe, C.E., Wilding, L.P., and J.B. Dixon. 1996. Overview of Vertisols: Characteristics and impacts on society. In Advances in Agronomy, Vol. 57, ed. D.L. Sparks, 289–375. New York: Academic Press. Dhillon, B.S., Kataria, P., and P.K. Dhillon. 2010. National food security vis-à-vis sustainability of agriculture in high crop productivity regions. Current Science 98:33–36. Dudal, R. 1965. Dark clay soils of tropical and subtropical regions. FAO Agric. Dev. Paper 83, Rome: FAO. Dudal, R., and H. Eswaran. 1988. Distribution, properties and classification of Vertisols. In Vertisols: Their distribution, properties, classification and management, ed. L.P. Wilding and R. Puentes, 1–22. College Station, Texas: Texas A&M University Press. El-Swaify, S.A., Pathak, P., Rego, T.J., and S. Singh. 1985. Soil management for optimized productivity under rain-fed conditions in the semi-arid tropics. Advances in Soil Science 1:1–63. El Abedine, A.Z., Robinson, G.H., and V. Tyego. 1969. A study of certain physical properties of a Vertisol in a Gezira area, Republic of Sudan. Soil Science 108:358–366. Eswaran, H., and C. Sys. 1979. Argillic horizon in LAC soils formation and significance to classification. Pedologie (Ghent) 29:175–190. Eswaran, H., and E. van den Berg. 1992. Impact of building of atmospheric CO2 on length of growing season in the Indian sub-continent. Pedologie 42:289–296.
Formation and Management of Cracking Clay Soils (Vertisols)
341
FAO. 1983. Guidelines: Land evaluation for rainfed agriculture. FAO Soils Bulletin 52. Rome: FAO. Gupta, R.K., and I.P. Abrol. 1990. Salt-affected soils: Their reclamation and management for crop production. In Advances in soil science, Vol II, ed. B.A. Stewart, 223–288. Berlin: Springer-Verlaag. Harrison, W.H., and M.R. Sivan. 1912. A contribution to the knowledge of the black cotton soils of India. Pusa Memo. Dept. Agriculture India, 2, No. 5. ICRISAT. 2004. Identifying systems for carbon sequestration and increased productivity in semi-arid tropical environments (National Agricultural Technology Project—NATP) (Project Code: RNPS25). Project completion report submitted to the NATP Directorate, Santosh Nagar, Hyderabad, Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for Semi-Arid tropics. Jackson, M.L. 1964. Chemical composition of soils. In Chemistry of the soil, ed. F.E. Bear, 71–141. Calcutta: Oxford and IBM Publishing Co. Jodha, N.S. 1980. Some dimensions of traditional farming systems in semi-arid tropical India. In Proceedings of the International Workshop on Socio-Economic Constraints to Development of Semi-Arid Tropical Agriculture, 18–23 Feb. 1979, Hyderabad, India, 11–24, India: ICRISAT. Kadu, P.R., Vaidya, P.H., Balpande, S.S., Satyavathi, P.L.A., and D.K. Pal. 2003. Use of hydraulic conductivity to evaluate the suitability of Vertisols for deep-rooted crops in semi-arid parts of central India. Soil Use and Management 19:208–216. Leather, J.W. 1898. On the composium of Indian soils. Agriculture Ledger 4:81–164. McGarity, J.W., Mazloumi, H., and E.H. Hoult. 1988. The effect of soil amelioration on the yield of dryland crops on Vertisols in northern N.S.W. Australia. In Transactions international workshop swell-shrink soils, ed. L.R. Hirekerur, D.K. Pal, J.L. Sehgal, and S.B. Deshpande, 194–198. New Delhi: Oxford & IBH Publishing Co. Pvt. Ltd. McKee, G.S., and B. Hajek. 1973. Vertisols or cracking clay soils. In Soils of the southern states and Puerto Rico. Southern Cooperative Series Bulletin No. 174, 29–32, Washington, DC: USDA. Murthy, R.S., Bhattacharjee, J.C., Landey, R.J., and R.M. Pofali. 1982. Distribution, characteristics and classification of Vertisols. In Vertisols and rice soils of the tropics, Symposia paper II, 12th International Congress of Soil Science, 3–22. New Delhi: Indian Society of Soil Science. Myers, R.J.K., and P. Pathak. 2001. Indian Vertisols: ICRISAT’s research impact—past, present and future. ed. J.K. Syers, F.T. Penning de Vries, and P. Nyamudeza, 203–219. Wallingford: CAB International Publishing. Nayyar, H. 2003. Calcium as environmental sensor in plants. Current Science 84:893–902. NBSS and LUP. 2002. Soils of India, NBSS Pub. 94. Nagpur, India: NBSS and LUP. NBSS and LUP-ICRISAT. 1991. Suitability of Vertisols and associated soils for improving cropping systems in central India. Nagpur: NBSS and LUP, Patancheru: ICRISAT. Nimkar, A.M., Deshpande, S.B., and P.G. Babrekar. 1992. Evaluation of salinity problem in swell-shrink soils of a part of the Purna Valley, Maharashtra. Agropedology 2:59–65. Pal, D.K., and S.B. Deshpande. 1987. Characteristics and genesis of minerals in some benchmark Vertisols of India. Pedologie (Ghent) 37:259–275. Pal, D.K., Balpande, S.S., and P. Srivastava. 2001. Polygenetic Vertisols of the Purna Valley of central India. Catena 43:231–249. Pal, D.K., Dasog, G.S., and T. Bhattacharyya. 2009c. Pedogenetic processes in cracking clay soils (Vertisols) in tropical environments of India: A critique. Journal of the Indian Society of Soil Science 57:422–432. Pal, D.K., Srivastava, P., and T. Bhattacharyya. 2003a. Clay illuviation in calcareous soils of the semi-arid part of the Indo-Gangetic Plains, India. Geoderma 115:177–192.
342
World Soil Resources and Food Security
Pal, D.K., Bhattacharyya, T., Chandran, P., and S.K. Ray. 2009a. Tectonics-climate linked natural soil degradation and its impact in rainfed agriculture: Indian experience. In Rainfed agriculture: Unlocking the potential, ed. S.P. Wani, J. Rockstroem, and T. Oweis, 54–72. Oxfordshire: CAB International Publishing. Pal, D.K., Bhattacharyya, T., Ray, S.K., and S.R. Bhuse. 2003c. Developing a model on the formation and resilience of naturally degraded black soils of the Peninsular India as a decision support system for better land use planning. NRDMS, DST Project Report, Nagpur, India. Pal, D.K., Deshpande, S.B., Venugopal, K.R., and A.R. Kalbande. 1989. Formation of di- and trioctahedral smectite as evidence for paleoclimatic changes in southern and central Peninsular India. Geoderma 45:175–184. Pal, D.K., Srivastava, P., Durge, S.L., and T. Bhattacharyya. 2003b. Role of microtopography in the formation of sodic soils in the semi-arid part of the Indo-Gangetic Plains, India. Catena 51:3–31. Pal, D.K., Bhattacharyya, T., Srivastava, P., Chandran, P., and S.K. Ray. 2009d. Soils of the Indo-Gangetic Plains: Their historical perspective and management. Current Science 96:1193–1202. Pal, D.K., Dasog, G.S., Vadivelu, S., Ahuja, R.L., and T. Bhattacharyya. 2000. Secondary calcium carbonate in soils of arid and semi-arid regions of India. In Global climate change and pedogenic carbonates, ed. R. Lal, J.M. Kimble, H. Eswaran, and B.A. Stewart, 149–185. Boca Raton, FL: Lewis Publishers. Pal, D.K., Bhattacharyya, T., Ray, S.K., Chandran, P., Srivastava, P., Durge, S.L., and S.R. Bhuse. 2006. Significance of soil modifiers (Ca-zeolites and gypsum) in naturally degraded Vertisols of the Peninsular India in redefining the sodic soils. Geoderma 136:210–228. Pal, D.K., Bhattacharyya, T., Chandran, P., Ray, S.K., Satyavathi, P.L.A., Durge, S.L., Raja, P., and U.K. Maurya. 2009b. Vertisols (cracking clay soils) in a climosequence of Peninsular India: Evidence for Holocene climate changes. Quaternary International 209:6–21. Puentes, R., Harris, B.L., and C. Victoria. 1988. Management of Vertisols in temperate regions. In Vertisols: Their distribution, properties, classification and management, ed. L.P. Wilding and R. Puentes, 129–145, College Station, Texas: Texas A and M University. Rajaguru, S.N., and R. Korisetter. 1987. Quaternary geomorphic environment and culture succession in western India. Indian Journal of Earth Sciences 14:349–361. Ramesh, V., Wani, S.P., Rego, T.J., et al. 2007. Chemical characterisation of selected benchmark spots for C sequestration in the semi-arid tropics, India. Global Theme on Agroecosystems Report No. 32, Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) and New Delhi: Indian Council of Agricultural Research (ICAR). Rao, D.L.N., and S.K. Ghai. 1985. Urease and dehydrogenase activity of alkali and reclaimed soils. Australian Journal of Soil Research 23:661–665. Ritter, D.F. 1996. Is Quaternary geology ready for the future? Geomorphology 16:273–276. Robinson, G.H. 1971. Exchangeable sodium and yields of cotton on certain clay soils of Sudan. Journal of Soil Science 22:328–335. Rockstroem, J., Hatibu, N., Oweis, T., and S.P. Wanni. 2007. Managing water in rainfed agriculture. In Water for food, water for life: A comprehensive assessment of water management in agriculture, ed. D. Molden, 315–348. London, UK: Earthscan and Colombo, Srilanka: IWMI. Sahrawat, K.L. 2003. Importance of inorganic carbon in sequestering carbon in soils of the dry regions. Current Science 84:864–865. Sahrawat, K.L., Bhattacharyya, T., Wani, S.P., et al. 2005. Long term low land rice and arable cropping effects on carbon and nitrogen status of some semi-arid tropical soils. Current Science 89:2159–2163.
Formation and Management of Cracking Clay Soils (Vertisols)
343
Soil Survey Staff. 2003. Keys to soil taxonomy, 9th ed. Washington, DC: U.S. Department of Agriculture, Natural Resources Conservation Series. Soil Survey Staff. 2006. Keys to soil taxonomy. 10th ed. Washington, DC: U.S. Department of Agriculture, Natural Resources Conservation Services. Srivastava, P., Bhattacharyya, T., and D.K. Pal. 2002. Significance of the formation of calcium carbonate minerals in the pedogenesis and management of cracking clay soils (Vertisols) of India. Clays and Clay Minerals 50:111–126. Swindale, L.D. 1989. Approaches to agrotechnology transfer, particularly among Vertisols. In Management of Vertisols for improved agricultural production: Proceedings of an IBSRAM Inaugural workshop, 18–22 Feb. 1985, ICRISAT Centre, India. Patancheru, India: ICRISAT. Syers, J.K., Penningde Vries, F.T., and P. Nyamudeza (ed.). 2001. The sustainable management of Vertisols. Wallingford: CAB International Publishing 304pp. van der Merwe, A.J., DeVilliers, M.C., Buchmann, C., Beukes, D.J., and M.C. Walters. 2001. Vertisols management in South Africa, ed. J.K. Syers, F.T. Penning de Vries, and P. Nyamudeza, 85–100. Wallingford: CAB International Publishing. Venkateswarlu, J. 1984. Soil problems with Vertisols with particular reference to surface soil conditions and water relations. In ACIAR/IBSRAM Proceedings of the International workshop on soils, 12–16 September 1983, Townsville, 105–115. Venugopalan, M.V., Chandran, P., Pal, D.K., Challa, O., and S.L. Surge. 2004. Effect of organic and conventional cotton production systems on soil properties: A case study in Yavatmal district, Maharashtra. In International symposium on strategies for sustainable cotton production—A global vision, Vol II. Crop Production. 22–25 November, 2004. 118–121. Dharwad, Karnataka, India: University of Agricultural Sciences. Wani, S.P., Pathak, P., Jangawad, L.S., Eswaran, H., and P. Singh. 2003. Improved management of Vertisols in the semi-arid tropics for increased productivity and soil carbon sequestration. Soil Use and Management 19:217–222. Wani, S.P., Sahrawat, K.L., Sreedevi, T.K., Bhattacharyya, T., and Srinivas, Rao. 2007. Carbon sequestration in the semi-arid tropics for improving livelihoods. Int. J. Environ. Stud. 64:719–727. Wani, S.P., Rockstroem, J., and T. Oweis (ed.). 2009. Rain-fed agriculture: Unlocking the potential. Comprehensive Assessment of Water Management in Agriculture Series. Oxfordshire: CAB International Publishing. Yaalon, D.H. 1983. Climate, time and soil development. In Pedogenesis and soil taxonomy, I. Concepts and interaction, ed. L.P. Wilding, N.E. Smeck, and G.F. Hall, 233–251. Amsterdam: Elsevier.
of Nuclear and 8 Role Isotopic Techniques in Sustainable Land Management Achieving Food Security and Mitigating Impacts of Climate Change Long Nguyen, Felipe Zapata, Rattan Lal, and Gerd Dercon CONTENTS 8.1 Introduction...................................................................................................346 8.2 Basic Principles of Nuclear and Isotopic Techniques.................................... 347 8.2.1 Isotopes..............................................................................................348 8.2.2 Applications of Isotopic and Nuclear Techniques in SLM Research: Principles.......................................................................... 349 8.2.2.1 Radioactive Isotopes........................................................... 349 8.2.2.2 Stable Isotopes.................................................................... 350 8.2.3 Summary........................................................................................... 353 8.3 Toward SLM in Agroecosystems.................................................................. 357 8.3.1 Linkage between SLM, SOM, and Soil Quality............................... 357 8.3.2 Developing and Implementing the Soil-Water-Nutrient Management Approach...................................................................... 358 8.3.3 Use of NIT in Integrated Nutrient Management Studies.................. 359 8.3.4 Use of NIT in Water Management Studies in Agriculture................ 371 8.3.4.1 Introduction......................................................................... 371 8.3.4.2 Use of the SMNP in Water Use Efficiency Studies............ 372 8.3.4.3 Use of Stable Isotopes for Determining Water Used by Plants in Agroecosystems................................................... 373 8.3.4.4 Use of Carbon Isotope Discrimination in Plants as a Tool for Assessing Plant WUE........................................... 374 345
346
World Soil Resources and Food Security
8.4 Looking Ahead to Challenges Facing Agriculture........................................ 375 8.4.1 Contributing to Achieving Sustainable Land and Water Management while Mitigating Climate Change Impacts.................. 375 8.4.1.1 Combating Land Degradation and Improving Soil Quality................................................................................ 375 8.4.1.2 Development of an Integrated Approach to SoilWater- Plant Technologies in Agroecosystems.................... 376 8.5 Overview of SOC Cycling Studies in Agroecosystems................................ 382 8.5.1 Introduction....................................................................................... 382 8.5.2 Isotopic Techniques in C-Cycling Studies......................................... 382 8.5.2.1 Black Carbon...................................................................... 385 8.5.2.2 Biochar................................................................................ 386 8.6 Land Use, Management, and Soil C Sequestration in Agroecosystems........ 388 8.6.1 Introduction....................................................................................... 388 8.6.2 Assessment of Recommended Practices for SOC Management in Agroecosystems............................................................................. 389 8.7 Area-Wide (Watershed) Studies on Soil, Sediment, and SOC Redistribution and Identification of Sediment and Carbon Sinks in a Watershed...................................................................................................... 392 8.7.1 Introduction....................................................................................... 392 8.7.2 Use of Environmental Radionuclides in Soil Erosion Research in Agroecosystems............................................................................. 393 8.8 Mitigating GHG Emissions in Agroecosystems............................................ 394 8.8.1 Soils as Sinks and Sources of GHG.................................................. 394 8.8.2 Selected Applications of Isotope Techniques in GHG Emission Studies in Agroecosystems................................................................ 395 8.9 Conclusions.................................................................................................... 398 References............................................................................................................... 399
8.1 INTRODUCTION The present world population of 6.7 billion is expected to reach 8 billion by the year 2020. Most of the population increases will occur in developing countries, where the majority depend upon agriculture for their livelihoods. Against this background of projections of increased population growth and pressure on the worldwide availability of land and water resources, many developing countries will face major challenges to achieving sustainable food security considering their available per capita land area, the severe scarcity of fresh water resources, and particular infrastructure and socioeconomic conditions [Pinstrup-Andersen 1999; Lal 2000]. This scenario is further compounded by increased global land degradation, particularly increased risks of soil erosion and desertification in sub-Saharan Africa and South Asia [Scherr 1999]. Worldwide soil degradation is currently estimated at 1.9 billion hectares and is increasing at a rate of 5 to 7 million hectares each year [Lal 2006]. According to a recent study using data over a 20-year period, land degradation is increasing both in severity and extent in many parts of the world [FAO 2010a]. The consequences
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
347
of land degradation include productivity decline, food insecurity, damage to basic resources and ecosystems, and loss of biodiversity, all of which are intricately linked with long-term social, economic, and environmental impacts ultimately resulting in human migration and sociopolitical unrest [Doos 1994; Bruinsma 2003; UNEP 2010]. Besides these pressing issues, several environmental drivers that also need to be addressed include: (1) increasing risks and impacts of global warming and climatic variability, (2) rising energy demands, in particular renewable energy sources, (3) expanding urbanization and industrialization and related infrastructure development, and (4) deteriorating water and air quality. All of these issues are likely to have negative impacts and induce changes in agroecosystems, which will place increasing pressures on sustainable land and water resources to produce sufficient food, feed, fiber, and fuel for the ever-increasing world population [Lal 2000; Verchot and Cooper 2008]. Sustainable land management (SLM) will require the combined use of the following strategies to preserve land and water resources: (1) agricultural intensification on the best arable lands that are already being farmed to enhance food security with minimum environmental degradation, (2) rational utilization of marginal lands, and (3) combating land degradation and restoring degraded soils [Lal 2000]. A key element across all land types and an integral part of sustainable agriculture would be to enhance soil quality for environmental sustainability of agroecosystems [Karlen et al. 2001; Arshad and Martin 2002; Carter 2002]. In this context, there is a strong need for high quality, innovative research to develop—in a relatively short time— land-specific technologies that will address the most strategically important issues of SLM in the agroecosystems of the developing world. The purpose of this chapter is to highlight the role of isotopic and nuclear-based techniques in the development of integrated soil, water, nutrient, and plant management practices for SLM in agroecosystems. This review is made within the objectives, approaches and strategies, and main project activities of the Soil and Water Management and Crop Nutrition (SWMCN) subprogram of the Joint Food and Agriculture Organization (FAO) and International Atomic Energy Agency (IAEA) Division of Nuclear Techniques in Food and Agriculture [FAO/IAEA 2009]. It aims to provide information on the relevant nuclear and isotopic methods that can be used to address current and emerging issues related to SLM in agricultural research. As land and water in agroecosystems are dynamic components of terrestrial ecosystems, there are multitudes of applications of stable and radioactive isotopes as tracers of global biogeochemical and hydrological processes, their interactions, and main driving factors. However, these aspects are beyond the scope of this review and therefore not included here.
8.2 BASIC PRINCIPLES OF NUCLEAR AND ISOTOPIC TECHNIQUES This section contains a brief account of the basic principles of nuclear and isotopic (NTI) techniques to aid in understanding the use and application of these techniques in agroecosystems. For details, readers are referred to treatises, IAEA Web sites, and training manuals available on the subject [L’Annunziata 2003; IAEA 1990, 2001, 2002a, 2002b, 2003a, 2003b, 2008a, 2088b; US-EPA 2008; US-NNDC 2008].
348
World Soil Resources and Food Security
8.2.1 Isotopes Isotopes are defined as atoms of the same atomic number (Z) but different atomic weight (A). The atom is the basic structural unit of matter. The atom of a given element such as phosphorus (P) has a set number of positively-charged protons and a neutral neutron in the nucleus. The number of protons (Z) and neutrons (N) present in the nucleus is referred to as the mass number (also as the atomic mass or atomic weight, A = Z + N), while the number of protons (Z) is termed the atomic number. An atom of a given element may have two or more isotopes. For example, phosphorus (P) has three isotopes (1531P, 1532P, and 1533P), which have the same number of protons (Z = 15) but different atomic weights (A of 31, 32, and 33, respectively) and different numbers of neutrons (N, defined as the difference between A and Z, of 16, 17, and 18, respectively). Isotopes may exist in both stable and unstable (radioactive) forms, depending on the stability of the nucleus in an atom. For example, a sulfur atom consists of five isotopes (32S, 33S, 34S, 35S, and 36S); one of them (35S) is a beta emitter radioactive, while the four others (32S, 33S, 34S, and 36S) are stable. A radioactive isotope is an atom with an unstable nucleus that spontaneously emits radiation (alpha or beta particles and/or gamma electromagnetic rays). The nonstability occurs because the ratio of neutrons (N) over protons (Z) in the nucleus is outside the belt of stability (i.e., outside a particular number due to an excess of either protons or neutrons), which varies with each atom. In contrast, a stable isotope is an atom with a stable nucleus (i.e., the N:Z ratio in the nucleus of an atom is within the belt of stability) and hence it does not spontaneously emit any radiation. Stable isotopes exist in light and heavy forms with heavy isotopes (higher atomic weight than light isotopes) accounting for less than 1.5% (Table 8.1). Stables isotopes are measured by an elemental analyzer coupled to an isotope ratio mass spectrometry (IRMS) in which the sample is combusted into a gas, which is fed into the mass spectrometer, where the ratio of the stable isotopes of interest (e.g., 13C/12C, 2H/1H, 15N/14N, 18O/16O, 33S/32S) is���������������������������� determined. Recent developments in spectroscopic techniques such as wavelength-scanned cavity ring-down
TABLE 8.1 Average Abundances of Stable Isotopes (% Abundance in Brackets) of Major Elements Commonly Occurring in Agroecosystems Element
Heavy Isotope
Light Isotope
Carbon Hydrogen Nitrogen Oxygen
C (1.108%) H (0.0156%) 15N (0.366%) 18O (0.204%) 17O (0.037%) 33S (0.76%) 34S (4.22%) 36S (0.02%)
C (98.892%) H (99.984%) 14N (99.634%) 16O (99.759%)
Sulfur
13 2
12 1
S (95.02%)
32
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
349
spectroscopy (WS-CRDS) combined with a combustion module offer an advantage of a large analytical throughput of stable isotope determination [Crosson 2008]. In the case of radioisotopes, the radiation emitted can be tracked by means of specific radiation detection devices, for instance Geiger–Müller counters or scintillation counters for beta emitters, gamma spectrometers for gamma-emitting radionuclides, and alpha spectrometers for alpha-emitting radioisotopes. The international unit (SI) of activity decay is the Becquerel (Bq), which is equal to one disintegration per second (dps). The old unit commonly used was the Curie, which is equivalent to 3.7 × 1010 dps or 3.7 × 1010 Bq. Environmental radionuclides is a term commonly used to refer to those radionuclides (natural or manmade) that are widely distributed in the environment or landscape and, while occurring at very low levels, are readily measurable. In soil erosion and sedimentation investigations, work has focused on the use of a particular group of environmental radionuclides, namely fallout radionuclides such as cesium-137 (137Cs), excess lead-210 (210Pbex), and beryllium-7 (7Be), which are gamma-emitters. These fallout radionuclides, which are deposited on the soil surface by dry deposition and rainfall, are ������������������������������������������������������������������� adsorbed to fine soil particles and their distribution in soil profiles and across agricultural landscapes resulting from the movement of soil particles can be used to document soil erosion and deposition rates and patterns [Zapata and Nguyen 2009].
8.2.2 A pplications of Isotopic and Nuclear Techniques in SLM Research: Principles Both stable and radioactive isotopes are used in quantifying nutrient and water pools and fluxes in the soil-plant system. Specific chemical sources and pollutants can be also traced in the system using stable isotopic tracers. Stable isotopes do not pose safety hazards and can be used in a wide range of conditions (laboratory, glasshouse, and field conditions) at natural abundance or enriched levels. In contrast, the use of radioactive isotopes is normally restricted to laboratory and glasshouse conditions because of stringent safety measures and protection procedures relating to storage and transport, source handling, sampling, analyses, and waste disposal required to prevent the harmful effects of the radiation from radioactive isotopes on environments and living organisms. Strict compliance to international and national regulations is required. 8.2.2.1 Radioactive Isotopes Radioactive isotopes are used as tracers to investigate the kinetics of applied nutrients such as phosphorus (P) or sulfur (S) in the soil-plant system and also to assess the fertilizer recovery of labeled fertilizers by the crop as affected by soil type and fertilizer management practices (e.g., timing, method, source, etc.). In these studies, radioactive isotopes (or radioisotopes) of the same element (nutrient) for instance 32P for investigating P or 35S for S are added to the soil and the specific radioactivity (SR) of radioactive isotopes in soils and plants are determined. The SR is used to describe the amount of radioactivity per unit of material added (e.g., 32P or 35S added to soils as labeled fertilizer products or plant materials) as well as the amount of radiotracer
350
World Soil Resources and Food Security
per unit of the element being traced. For instance, when using 32P in a tracer experiment, the SR of P in the soil or plant material is Bq 32P per 1 g of soil or 1 g of plant dry matter and the SR of P in the soil P or in the plant P is Bq 32P mg−1 P. The usefulness of a radioisotope as a tracer depends on the following properties: i) its halflife (the time required for the radioactivity of a radioisotope to decay to half of its initial activity); ii) its decay mode or type of nuclear transformation and radiation emitted; and iii) its decay energy (amount of energy released in a particular nuclear decay). The basic energy unit is the electron volt. Energies are ranging typically from several kilo electron volts (Kev) to several mega electron volts (MeV). The halflife of a radioisotope must be suitable (sufficiently long) in relation to the duration of the experiment. The mode and energy of decay will determine how the radioisotope will be measured. 8.2.2.2 Stable Isotopes Stable isotopes offer an advantage over the conventional, nonisotopic techniques because they can be used to address a range of important issues such as:
1. The extent of nutrient cycling in the soil-plant systems and the fate of fertilizers in agricultural landscapes 2. The sources of pollutants from agricultural landscapes 3. The sources of water used by plants and crop water use efficiency 4. The extent of atmospheric nitrogen capture (biological nitrogen fixation) by crops for their growth and its contribution to the following crops 5. The decomposition and turnover rates of carbon and nitrogen from crop residues 6. The extent of nitrous oxide and carbon dioxide emissions from soil organic matter (SOM) and recently added organic manure, crop residues, or wastewater.
In the soil-plant system, both heavy and light isotopes of the same element take part in physical-chemical-biological processes, but because of different atomic weights, they react at different rates. For example, plants preferentially take up carbon dioxide containing the lighter carbon isotope (12C) in photosynthesis, but the degree of carbon isotopic discrimination against the heavier 13C depends on the type of plants and water availability. Thus, the 13C/12C ratios in plant tissues provide a measure of crop water use efficiency. Similarly, physical processes such as soil evaporation can discriminate oxygen (O) in soil water against heavy 18O isotope because the lighter 16O is more readily evaporated. 8.2.2.2.1 Natural Abundance Isotopic Tracer Studies Natural abundance studies rely on the natural differences in the ways that heavy and light isotopes are fractionated by physical, chemical, and microbiological processes that lead to either enrichment or deletion of the heavy isotopes. These enrichment– deletion processes can be useful to track changes of an element in the environment. In these natural abundance studies, isotopic data are expressed as the ratio (R) of the heavy to light isotopes in the sample (R Sample) compared to the same ratio in
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
351
an international standard (R Standard), using the delta (δ) notation. Since the differences in the absolute isotopic ratios (R) between the sample and standard are very small, they are expressed as parts per thousand or per mil (‰) deviation from the standard. For example, for δ values of nitrogen (N) and oxygen (O) in samples (e.g., soils or plants):
δ15N sample = [(15N/14N sample) / (15N/14N standard) − 1] × 1000
δ18O sample = [(18O/16O sample) / (18O/16O standard) – 1] × 1000
For N, the international standard is AIR (Atmospheric air; Table 8.2) with an accepted absolute 15N/14N ratio of 0.003676 (Table 8.2). Samples with ratios of 15N/14N greater than 0.003676 have positive delta values and those with ratios of 15N/14N lower than 0.003676 have negative delta values. 8.2.2.2.2 Isotopically Enriched Studies Heavy isotopes of essential elements in the soil-plant systems such as C, H, N, and O do not exist in abundance in nature (natural abundances of <1.5%; Table 8.1). Thus isotopically enriched tracers are often added to soils to track the labeled elements with better resolution, instead of relying on changes in the natural abundance over a period of investigation. For example, by adding 15N as ammonium (NH +4 ) or nitrate (NO3−) and monitoring both 15N and 14N in the soil and plants, it is possible to investigate N use efficiency by plants as influenced by types of fertilizers (NH +4 or NO3−) and to determine the rate of N transformations in soils including nitrification, immobilization, mineralization, leaching beyond the main plant rooting zone (normally <1–2 m depth), and denitrification losses as nitrous oxide and dinitrogen gases.
TABLE 8.2 Absolute Isotope Ratios (R) of International Reference Standards for the Five Elements Commonly Found in Agroecosystems—Both Hydrogen and Oxygen Elements Have More than One International Standard Element Carbon Hydrogen
Ratio Measured (R)
R of Standards
C/ C H/1H
0.0112372 0.00015575
H/1H
0.000089089
N/14N O/16O
0.003676 0.0020052
O/16O O/16O
0.0020672 0.0018939
S/32S
0.045005
13
12
2
2
Nitrogen Oxygen
15 18
18 18
Sulfur
34
Standards Vienna Pee Dee Belemnite (VPDB) Vienna Standard Mean Ocean Water (VSMOW) Standard Light Antarctic Precipitation (SLAP) Air (Atmospheric air) Vienna Standard Mean Ocean Water (VSMOW) Vienna Pee Dee Belemnite (VPDB) Standard Light Antarctic Precipitation (SLAP) Vienna Canyon Diablito Troilite (VCDT)
352
World Soil Resources and Food Security
Isotopic data obtained from isotopically enriched studies are usually reported as atom percent (At%) or atom percent excess (At% excess). The At% provides the absolute number of atoms of a given isotope present in 100 atoms of total element; for examples, At% for 15N- and 13C-enriched studies are notated as follows:
At%15N = (15N / 14N + 15N) (100 At%)
At%13C = (13C / 12C + 13C + 14C) (100 At%)
In calculating At% for 13C, 14C is usually treated as negligible and only the sum of 12C and 13C is taken to be total C. The At% excess is calculated as the difference between the experimental value and the background natural abundance expressed as % (Table 8.1) This At% excess unit specifies the level of isotopic abundance above a given background, which is normally the natural abundance. 8.2.2.2.3 Nuclear Magnetic Resonance and Neutron/Gamma Probes Nuclear magnetic resonance (NMR) is another technique used to characterize inorganic and organic compounds in biological materials and to investigate the mechanisms of the chemical reactions by detecting not only isotopic differences but also giving an indication of the position of the atom. Nuclei with an odd mass or odd atomic number have nuclear spin (in a similar fashion to the spin of electrons). This includes 1H and 13C (but not 12C). The spins of nuclei are sufficiently different that NMR experiments can be sensitive for only one particular isotope of one particular element. The NMR behavior of 1H and 13C nuclei has been exploited by organic chemists since they provide valuable information that can be used to deduce the structure of organic compounds. The neutron/gamma probes contain a mixture of americium and beryllium isotopes that are used as a source of neutrons and an isotope of caesium as a source of gamma rays. The working principle of the soil moisture neutron probe (SMNP) is based on the interaction of emitted neutrons and soil water content [Greacen 1981; IAEA 2003a]. The neutron source emits fast neutrons that interact with soil particles and water that surrounds the probe. Through collisions with soil particles and water, fast neutrons (high energy > 2 MeV) lose energy (a process called moderation or thermalization) and become slow or thermal neutrons (low energy < 0.025 eV). Because hydrogen is the target nucleus that most efficiently reduces neutron energy, hydrogen is said to be a good neutron moderator. Thus, because of its hydrogen content, water is also a good neutron moderator. With the exception of the SOM, which may gradually change with time, soil materials containing hydrogen remain constant and are taken into account during calibration of the SMNP. When the neutron probe is lowered into a soil profile through the access tube, a stable spherical cloud of slow neutrons develops quickly around the source, with a diameter of about 30 cm. The drier the soil, the greater is the diameter of the cloud. The number of slow neutrons per unit volume at each point of the cloud remains constant and is proportional to the water content of the soil within the cloud. Since the slow neutron detector (3He chamber) in SMNP is positioned inside the cloud volume, the count rate per minute
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
353
(cpm) or per second (cps) is proportional to the soil water content (θ) of the same volume. The instrument is calibrated with soil samples of known θ measured by gravimetric method. Dual type probes allow the simultaneous measurement of both soil bulk density and soil water content. There are both depth and surface probes. The processes involved with the surface probes to measure the average water content in the surface layer (0–15 cm) and the bulk density of various layers from 2.5 to 30 cm thickness are described briefly. Soil water content measurements for surface probes are identical to those described before for the SMNP except that the fast neutron source and slow neutron detector are fixed to the base of the shield, precluding measurements at various soil depths. With respect to soil bulk density measurements, surface probes rely on both backscattering and attenuation of gamma radiation while depth probes rely only on backscattering of gamma radiation. In the backscattering process, the gamma ray detector measures the photons that return to the soil surface after interacting (backscattering) with atoms from the soil particles while in the attenuation process, measurements are made at a desired depth, the gamma ray detector counts both the number of photons that cross the soil sample of thickness X located between the gamma ray source and the detector, and the number of backscattered gamma rays. As the soil may contain water, the dry soil bulk density can be obtained from the wet soil bulk density. The gamma source of a surface neutron/gamma probe has two modes of operation. In the first mode, the gamma source is not lowered into the soil and can occupy two positions, 1) BS (backscattering), a little above the soil surface, and 2) AC (asphaltconcrete) at the soil surface. Measurements in both positions are made by backscattering only, and the bulk density measurement is made on the surface layer. In the second mode of operation, the gamma source is lowered into the soil to the desired depth (from 5 to 30 cm in 2.5 cm steps) and the bulk density is measured by both gamma ray backscattering and attenuation processes. For both modes of operation, the average soil water content of the soil surface (0–15 cm) is measured by neutron moderation using the neutron source located at the base of the probe–soil interface (placed on the soil surface).
8.2.3 Summary Isotopic and nuclear-based techniques comprise stable and radioactive isotopes (natural abundance and artificial labeling), and the use of radiation sources such as neutron and gamma density probes. Isotopes provide unique and quantitative data on nutrient and water pools and fluxes in the soil-plant system. Specific chemical sources and pollutants can also be traced in the system. SMNPs are used for monitoring soil water content changes and constructing field water balance useful for irrigation scheduling. Gamma density probes are used to measure changes in soil bulk density. For details on the principles of isotopic and nuclear techniques and their application in soil, water, and plant nutrient studies in agroecosystems refer to the IAEA Training Manuals [IAEA 1990, 2001]. Nuclear-based techniques are a complement not a substitute to nonnuclear techniques. They are applied in the context of agricultural research under field and
354
World Soil Resources and Food Security
TABLE 8.3 Nuclear and Isotopic Techniques Commonly Used in Soil-Water-Nutrient Research Studies and Their Advantages and Disadvantages Soil moisture neutron (radiation source) probe (SMNP) Advantages • Fast, economical, and non-destructive field monitoring of soil water content, particularly in salt-affected soils • Permits characterization of soil hydraulic properties Limitations • Safety and security issues (sealed radiation source/shielding) • National radiation protection authority/infrastructure monitoring services/import permit/ licensing/personal dosimetry/control Gamma (radiation) density probe Advantages • Fast, economical, and nondestructive field monitoring of soil bulk density compared to hand auger/cylinder method • Allows identification of changes in soil compaction and characterization of soil physical conditions Limitations: Same as for SMNP Nuclear magnetic resonance (NMR) Advantages: • Allows use of both solution and solid state of 31P NMR techniques for characterization of soil P forms • Solid state 13C NMR and 31P NMR can be used to study characterization and transformations of soil organic P and C Limitations • Lack of resolution of solid state 31P NMR compared to the solution 31P NMR • Presence of paramagnetic species (Fe, Mn) limits quantitative potential of solid state 31P NMR • Skilled staff and specialized facilities are required Natural abundance variations in deuterium (2H) and 18O (stable isotopes) in water, soil, and plant materials Advantages • Allows sources of water transpired by plants to be identified • Provides isotopic signatures for the sources of hydraulic-lifted water in trees Limitations: • Skilled staff and specialized facilities, in particular an expensive IRMS (isotope ratio mass spectrometer), which is costly to maintain
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
355
TABLE 8.3 (Continued) Nuclear and Isotopic Techniques Commonly Used in Soil-Water-Nutrient Research Studies and Their Advantages and Disadvantages Natural abundance variations in 13C (stable isotope) in plant materials Advantages • 13 C isotope discrimination technique can be used to evaluate crop genotypes with superior water use efficiency and tolerance to drought/salinity Limitations: • Same as for 2H and 18O P- and 33P (radioactive isotopes)-labeled materials Advantages 32
• Direct method to measure fertilizer phosphorus (P) use efficiency by crops • Indirect method to measure P supply from sources that cannot be labeled with 32P or 33P (e.g., phosphate rocks and organic wastes) Limitations • Short halflife of 32P and 33P for crop uptake • Only applicable to laboratory and greenhouse experiments because of specific safety, storage, and handling requirements for radioactive isotopes S (stable)- or 35S (radioactive)-labeled materials Advantages 34
• Sufficiently long halflife of radioactive 35S (87 days) for crop uptake and residue decomposition studies • Direct method to estimate of fertilizer S uptake by crops (34S or 35S-enriched sources) Limitations • Precise analysis (34S) requires skilled staff and an expensive IRMS, which is costly to maintain • Time consuming 35S analyses and high cost associated with 34S-enriched materials Cl and 22Na isotopic tracers Advantages: 36
• Able to investigate salt movement and plant uptake (36Cl ) in biosaline agriculture • Able to investigate soil adsorption and plant uptake of sodium (22Na) in biosaline agriculture Limitations • Same as indicated for phosphorus radioisotopes (continued)
356
World Soil Resources and Food Security
TABLE 8.3 (Continued) Nuclear and Isotopic Techniques Commonly Used in Soil-Water-Nutrient Research Studies and Their Advantages and Disadvantages C (radioactive) and 13C (stable) natural abundance or 13C-labeled materials Advantages 14
• Estimates decomposition rates of labeled organic residues (residual 13C or labeled CO2 respired) • Estimates mean residence time of SOM using 14C dating • Estimates turnover time of soil OM (change in soil delta 13C) when C change from C3 to C4 vegetation or vice versa • Identifies sources (C3 or C4) of OM breakdown on the basis of 13C/12C isotopic signatures • Identifies diffusion of respired gases in soils in relation to OM breakdown and source (18O and 13C) Limitations • • • • •
National radiation protection authority/ infrastructure services/import/permit/dosimetry/control Safety/security issues (open source), training and licensing (14C) Disposal of 14C-labeled materials, including soil contaminated with 14C Prohibitive cost and sophistication of 14C-dating Precise analysis (13C natural abundance) requires skilled staff and specialized facilities, in particular an expensive IRMS (isotope ratio mass spectrometer), which is costly to maintain
Environmental radionuclides (137Cs , 210Pbex, 7Be) Advantages • Fewer resources required compared with the use of erosion plots • Spatially distributed data of erosion/sedimentation can be readily coupled to GIS and spatial geostatistical tools • Data are used for the assessment of the effectiveness of soil conservation technologies Limitations • Low 137C inventories (especially in southern hemisphere soils) require extended counting times (low sample throughput) • Limited availability of gamma-counting equipment in developing countries • Variable and uncertain inputs of 210Pbex and 7Be in some locations and difficulties in measuring 210Pb activities ex N (stable isotope) natural abundance or labeled materials Advantages 15
• Provides a direct method to measure uptake and recovery from 15N-labeled materials by crops • Enables direct and indirect estimates of biological N fixation and the transfer of biologicallyfixed N from legume to nonlegume crops • Allows assessment of soil N process rates (nitrification, mineralization, immobilization, nitrate reduction and denitrification)
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
357
TABLE 8.3 (Continued) Nuclear and Isotopic Techniques Commonly Used in Soil-Water-Nutrient Research Studies and Their Advantages and Disadvantages Limitations • Precise analysis of 15N-enriched samples requires trained staff and expensive analyzer (emission spectrometer with modern electronics) • Precise analysis of natural abundance samples (delta 15N) requires skilled staff and an expensive IRMS (isotope ratio mass spectrometer), which is costly to maintain • Relatively high cost of 15N-enriched materials Note: The techniques outlined above are commonly used in SWMCN studies. However, the list is not exhaustive. For detailed information on the principles and applications of these techniques, the reader is referred to http://www.iaea.org/Publications/index.html and http://www-naweb.iaea.org/ nafa/swmn/index.html.
greenhouse conditions. They offer comparative advantages over conventional techniques, but they demand skilled and trained personnel and adequate laboratory facilities, in particular measurement equipment and techniques or, alternatively, financial resources for analytical services. In the case of radioactive isotopes, strict compliance with safety regulations and radiation protection procedures is required. Nuclear-based techniques like any other techniques have advantages and limitations. Sometimes several techniques (nuclear and non-nuclear) are available to achieve the required information, while in other situations, the nuclear technique is the only available tool. It is therefore the task of a research team leader to assess the usefulness and effectiveness of the nuclear-based techniques to meet specific research objectives using available resources. The value of the information obtained should be properly evaluated. The above table (Table 8.3) summarizes the advantages and limitations of the most commonly used nuclear-based techniques in soil science research projects.
8.3 TOWARD SLM IN AGROECOSYSTEMS 8.3.1 Linkage between SLM, SOM, and Soil Quality The soil quality concept has evolved with time but in its simplest term it is the capacity of a soil to function productively and sustainably. It is considered that the basic soil functions and their balance demand the integration of three following major components: (i) sustained biological productivity; (ii) environmental quality; and (iii) support to biodiversity (plant and animal health) [USDA-NRCS 2009; Karlen et al. 2001]. Within agroecosystems, the productive capacity of a cropping system depends on the complex interactions between the basic soil properties and ecosystem components such as climatic conditions (rainfall and temperatures), cropping systems (crop rotation, crop species, and cultivar), and agronomic management practices
358
World Soil Resources and Food Security
(e.g., tillage, chemical fertilizers and irrigation water inputs, animal manures, and crop residues disposal). Similarly, the levels of SOM and its turnover rates (SOM dynamics) are intricately linked to soil physical, chemical, and biological properties and are important in sustaining the function of soil quality in terms of productive capacity and environmental sustainability. As soil quality concepts are commonly used to evaluate SLM in agroecosystems, SOM plays an important role in assessing sustainable land management [Carter 2002].
8.3.2 Developing and Implementing the Soil-WaterNutrient Management Approach In the 1990s, in response to the action plan of Agenda 21 to arrest soil degradation, research studies were initiated, adopting an integrated approach to soil, water, and nutrient management to enhance the basic soil function of productive capacity while ensuring environmental quality for sustainable agricultural production in agroecosystems. Furthermore, since the World Summit on Sustainable Development held in 2000 and the establishment of the Millennium Development Goals (MDGs), there has been a firm commitment and increased emphasis on the role of soil and water resources in SLM and environmental sustainability [UNMDG 2009]. Intensification of agricultural production on prime agricultural land demands more refined management of external inputs of water and nutrients and, thus, an increased need for both nuclear and nonnuclear methods to develop better water and nutrient management practices in both rainfed and irrigated agricultural systems. The development of an integrated nutrient management package (involving not only manufactured fertilizers but also natural sources of nutrients such as rock phosphates, biological N fixation, animal and green manures, etc.) along with the recycling of crop residues has resulted in a greater demand for the use of 15N, 32P, and sulfur-35 (35S) isotopes as tracers to develop efficient agronomic practices tailored to the specific cropping systems and local conditions. Significant advances have also been made during the past decade in the development and application of natural variations in the abundance of stable isotopes (deuterium, 13C, 15N, 18O, and 34S) to assess the dynamics of nutrients and water in the soil-plant system. These developments have been possible due to advances in automated systems for stable isotope ratio measurements in soil, plant, water, and gas samples [Chalk et al. 2002]. The underlying philosophy in developing and implementing projects for the SWMCN Subprogram of the Joint FAO/IAEA Program in Food and Agriculture was to adopt an integrated approach to soil, water, and nutrient management addressing issues of major concern and relevance in soil quality and sustainable crop production in the developing world. Table 8.4 provides an overview of the main issues addressed, strategies adopted, including the use of nuclear and isotopic techniques (NIT) in the medium-term networked research projects called coordinated research projects (CRPs) implemented by the SWMCN Subprogram [IAEA 2009a, 2009b]. The strategy adopted in these studies was to work within regional rather than global CRPs, targeting cropping systems or agroecological zones with the highest agricultural potential (e.g., maize/food legumes in the savannahs of Africa and South America; the rice-wheat system in South East Asia). The importance of the
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
359
ecosystem approach and the use of inputs of both organic and inorganic nutrient sources for the maintenance of soil fertility have been recognized in the agroforestry CRP [Chalk et al. 2002; Nguyen and Zapata 2006; FAO/IAEA 2009; Nguyen et al. 2010]. According to the principle of knowledge-based management, it is necessary to understand and quantify nutrient and water pools and fluxes in a given system before sustainable management practices can be formulated. This table (Table 8.4) illustrates only major applications of nuclear and isotopic techniques and this list is not exhaustive. There is a wide range of isotopic (stable and radioactive) and nuclear techniques that can be potentially used to study specific issues related to soil quality and health, plant nutrition, crop physiology, and crop water productivity and water use efficiency. A major challenge was to measure soil erosion and associated sedimentation (soil redistribution) in the landscape and to control soil and land degradation in agroecosystems at the watershed scale. Environmental radionuclides of 7Be, 137Cs, and 210Pb have been used to obtain information on short-term (<30 days), medium-term (~40 years), and long-term (~100 years) average soil redistribution rates and patterns in the landscape, respectively, and hence area-wide (watershed) erosion and sedimentation through the implementation of two CRPs on the topics [Zapata 2002a]. In the past, these techniques have been mainly applied in other disciplines such as hydrology (sedimentology/siltation), and earth sciences (geomorphology, soil geography), but, with time, their potential use for measuring and controlling erosion and sedimentation processes in highly dynamic agroecosystems have been widely recognized. Initial work focused on the development of refined and standardized protocols for the worldwide application of the 137Cs technique in agricultural research under various environmental conditions [Zapata 2002a, 2003]. Human and institutional capacities in developing countries were established and strengthened through technical cooperation projects (TCPs) of the IAEA Technical Cooperation Program [IAEA 2009c]. Opportunities were found for interdisciplinary research between plant breeders and soil and water management specialists in the area of tolerance to abiotic stress factors (aluminium toxicity, N and P deficiency, drought, salinity, etc.). Isotopic techniques can be used in physiological studies such as root growth and distribution, transpiration efficiency, nutrient acquisition, and utilization. They have demonstrated great potential to assist in the identification of suitable germplasm with tolerance to particular abiotic stress factors (drought and salinity, and N- and P-deficient soils) and are becoming more widely applied in plant breeding programs. The techniques are also used for evaluating improved genotypes under varying abiotic stress conditions and identifying integrated soil-plant management practices for enhancing crop production under these conditions [FAO/IAEA 2009].
8.3.3 Use of NIT in Integrated Nutrient Management Studies In most developing countries, inherent low soil fertility, N and P deficiencies in particular, are the main factors limiting crop production. Extensive tracts of land in Asia, Africa, and Latin America contain infertile tropical and subtropical soils. These areas generate low crop yields and are prone to land degradation as a result of
CRP Management of nutrients and water in rainfed, arid, and semiarid areas for increasing crop production (D1.20.06;1997–2002)
Strategies • Water conservation measures • Crop rotations (cereals /legumes) using genotypes adapted to water-limited environments • Drought-tolerant crop cultivars • Integrated soil fertility management (ISFM) • Characterization of AF systems • Tree or crop management • Resource (nutrients and water) use efficiency • Integrated soil fertility management (ISFM) • Al-tolerant cultivars • P efficient cultivars • Liming • Direct application of rock phosphates • Minimum or no tillage (direct seeding); mulch-based cropping (DMC) • Crop rotations, including legumes and ley pastures • Recycling crop residues • Improved fertilizer N management • Organic matter additions
Nuclear and Isotopic Techniques • SMNP • 15N direct (labeled fertilizers) and indirect (biological N fixation, BNF) • 32P isotope dilution or exchange • 13C isotope discrimination (cultivars)
• • • • • • • • •
SMNP (soil water status and use efficiency) Foliar labeling (13C, 15N) including stem injection δ13C shift in soil (C stocks and dynamics) D, 18O (hydraulic lift) 15N isotope dilution (BNF) 32P isotope dilution or exchange SMNP (water status and use efficiency) Gamma probes (bulk density) Foliar labeling (15N) for below-ground N stored (BGN) by legumes • 32P isotope dilution and exchange kinetics • 15N direct (fertilizers) and indirect (BNF) • δ13C shift in SOM (dynamics of soil organic C)
World Soil Resources and Food Security
Use of nuclear techniques for developing integrated nutrient and water management practices for agroforestry systems (AF) (D1.20.07; 1998–2005) Development of management practices for sustainable crop production systems on tropical acid soils through the use of nuclear and related techniques (D1.50.06; 1999–2005)
Issues • Inefficient capture and storage of erratic rainfall • Drought or dry spells • Low crop water use efficiency • Low soil fertility (OM, N, P status) • Shifting cultivation and reduced fallows • A range of AF systems • Low soil fertility • Water–nutrient interactions (competition or synergy) • Soil acidity (Al toxicity) • Soil N and P deficiencies • Low SOM • Low soil buffering capacity
360
TABLE 8.4 Issues, Strategies, and Use of Nuclear and Isotopic Techniques in the Implemented FAO/IAEA Networks of CRPs to Support Sustainable Intensification of Crop Production Systems
Assess the effectiveness of soil conservation techniques for sustainable watershed management using fallout radionuclides (D1.50.08; 2002–2007)
• Soil degradation and erosion, particularly in intensively cultivated land • Changes in land use or management
Integrated soil, water, and nutrient management in conservation agriculture (D1.50.09; 2005–2009)
• Cropland degradation • Soil losses (SOM and nutrient losses) • Deterioration of soil quality
Managing irrigation water to enhance crop productivity under water-limiting conditions: A role for isotopic techniques (D1.20.09; 2009–2012)
• Water scarcity • Water use efficiency • Low crop water productivity
• Innovative crop establishment and planting systems • Improved fertilizer N management practices • Irrigation water use efficiency and water savings • Crop residue management • Crop rotations (legumes) • Rice cultivar adaptation • Measuring rates and patterns (spatial and temporal) of soil erosion and associated sedimentation • Measuring changes in soil quality • Assessing soil conservation technologies to reduce soil losses and sediment production • Changes in cropping systems • Zero or minimum tillage • Recycling crop residues and organic wastes • Use of legume green manures and cover crops • Testing of recommended management practices (RMP) • Water-saving irrigation technologies • Improved irrigation water management practices • Reducing soil evaporation • Changing cropping systems and management practices
• 15N direct method (fertilizer N recovery, balance in soil-crop rotations) • SMNP (water status and use efficiency)
• Combined use of fallout radionuclides 137Cs, 210Pb , 7Be ex • 15N direct method (fertilizer N recovery, balance in soil-crop rotations) • SMNP (water status and use efficiency)
• 15N direct (labeled fertilizers, N dynamics in crop rotations) • 15N / 13C-labeled residues (nutrient recycling) • 15N isotope dilution technique (BNF inputs) • SMNP (water balance) • Gamma probe (bulk density) • δ13C shift in SOM pools (dynamics of soil organic C pools) • Natural abundance 2H and 18O variations in soil, plants, and water • Plant variations in 13C isotope discrimination • SMNP (water status and balance) • 15N direct (labeled fertilizers) technique
361
• Traditional degradative flooded systems • Low soil fertility • High N losses (rice) • Inefficient water usage • Poor physical conditions of soil
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
Integrated soil, water, and nutrient management for sustainable rice-wheat cropping systems in Asia (D1.50.07; 2001–2006)
362
World Soil Resources and Food Security
deforestation, overgrazing, and inadequate farming practices. Therefore, substantial fertilizer inputs are required to improve soil fertility for achieving optimum plant growth and adequate food and fiber production. However, limited availability and the high costs of manufactured and imported NPK fertilizers (containing nitrogen, phosphorus, and potassium) have resulted in little or no replacement of nutrients removed in the off-take farm products, leading to a degradation process called nutrient mining. According to Sanchez et al. [1997], annual nutrient losses in Africa are equivalent to 7.9 million tons of NPK fertilizers, which are 6 times the annual fertilizer consumption. Lal [2001a] reported that the depletion of the SOC pool occurs at a rate of 2%–12% per year and the cumulative loss can be as high as 50%–70% of the original pool over a cultivation period of 10 years. All these soil degradation processes ultimately lead to low crop production, food insecurity, and extreme rural poverty. This trend can be reversed by the development and implementation of integrated soil fertility (also called integrated nutrient) management, which involves the judicious use of all nutrient sources available in a farm, village, or region. These include locally available nutrient sources (crop residues, animal manures, organic wastes suitable for nutrient recycling, biological nitrogen fixation, biofertilizers, etc.) supplemented by manufactured fertilizers. Over the past 30 years, additional nutrients applied as fertilizers have been responsible for about 55% of the yield increases in developing countries [FAO 1998]. However, the utilization of these technologies requires the assessment of the nutrient supply from the locally available materials applied as nutrient sources, their tailoring to specific cropping systems, and the provision of site-specific recommendations for their application [FAO 1998; Chalk et al. 2002]. Nitrogen is the main factor limiting agricultural productivity worldwide. It is reported that adequate food production, in particular cereals for present and future populations, will not be achieved without external inputs of N fertilizers [FAO 2008]. Thus, management practices of fertilizer N inputs should be efficient to optimize crop production while minimize adverse effects (e.g., leaching losses beyond the plant rooting zone or greenhouse gaseous losses as nitrous oxide from fertilizers) on the environment. Fertilizer use efficiency is an important factor that needs to be taken into consideration in agricultural production systems as inefficient use of fertilizer inputs represents not only an environmental hazard but also a substantial economic loss [Keerthisinghe et al. 2003]. To achieve high use efficiency of the applied fertilizer N, management practices such as timing, placement, sources, etc. tailored to specific cropping systems and local conditions can be evaluated using 15N-labeled fertilizers. 15N-enriched techniques have been extensively used by ordinary (direct method) and reverse (indirect method) dilution [Hauck and Bremner 1976; Van Cleemput et al. 2008]. In order to get a thorough insight of issues related to fertilizer use efficiency, these studies may involve sequential experiments to: i) determine the recovery of the applied fertilizer N by the crop and related management practices; ii) ascertain the fate of the applied fertilizer N in the soil-plant system and obtain an estimate of unaccounted losses; and iii) measure direct losses by gas production processes (ammonia volatilization, denitrification) and/or leaching to reduce negative impacts on the environment [IAEA 1980, 1984]. Also, interactions of fertilizer N with other nutrients,
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
363
sources (i.e., organic sources), amendments, and management practices such as water availability, plant genotype, irrigation and tillage system, etc. can be assessed [Van Cleemput et al. 2008]. In view of the essential role that soil and fertilizer N play in maintaining and increasing crop yields in modern intensive agriculture, the stable isotope 15N has been widely used in several projects of the FAO/IAEA Program as a tracer to quantitatively determine amounts and movement in plants and soil of the N derived from applied fertilizers. Initial research projects conducted in the 1960s and 1970s focused mainly on the application of stable isotopic nitrogen-15 (15N) and radioactivephosphorus (32P)-enriched materials for measuring fertilizer nutrient (N or P) recovery and assessing the efficiency of N or P fertilizer management practices (timing, methods, sources, etc.) for major food grain crops [FAO 1980]. This information was essential for achieving the high yield potential of the cereal crop varieties developed by the Green Revolution. As there were raising concerns about the potential hazard of high fertilizer N residues in soils as potential pollutants of water bodies, research projects in the 1980s aimed at determining the fate of the applied fertilizer N in the soil-plant system and monitoring its recovery over a crop rotation in a number of environments worldwide [IAEA 1980, 1984; Zapata and Hera 1996]. Although it is well recognized that the application of mineral fertilizers plays an important role in intensification of crop production, lack of affordable and adequate supplies of fertilizers in substantial parts of the tropics and subtropics remains the major constraint to crop production. This is not due to a lack of knowledge of the importance of fertilizers in soil fertility, plant nutrition, and crop production but due to socioeconomical factors such as high prices resulting from weakly-developed markets, lack of domestic production capacity, poor infrastructure, and inefficient marketing and financial systems [Pinstrup-Andersen 1999]. For example, annual use of nutrients in Africa averages only about 10 kg of NPK ha−1 [Henao and Baanante 2001]. Under these circumstances, it is important to investigate cropping systems and nutrient management practices to optimize the integrated use of all nutrient sources (e.g., fertilizers, organic manures, and waste materials) suitable for recycling nutrients and biofertilizers for the maintenance of soil fertility and crop productivity. In the context of integrated nutrient management, the use of alternative N sources such as organic residues and biological nitrogen fixation (BNF) should be increased and optimized. Nitrogen-fixing legumes (grown as intercrops or in rotation, cover and green manure crops, pasture leys, mixed swards, multipurpose trees) are included in farming systems to add nitrogen inputs and also to provide other benefits such as fuel wood and feed for animals. Any legume improvement program aiming at improving productivity and protein production needs suitable methodologies to measure BNF under various environmental and management conditions to maximize the inputs of nitrogen fixation in agroecosystems. Some 30 years ago, in connection with the oil crisis of 1974 and associated increases in fertilizer prices, projects on BNF were initiated with a CRP on 15N methodology for measuring BNF in grain legume crops. Other projects followed, with quantification of BNF in forage, pasture, and tree legumes and in Azolla, in a range of environments. The use of the 15N isotope dilution technique will enable the accurate determination of BNF by the
364
World Soil Resources and Food Security
studied legume. Extensive research has been conducted to measure BNF in herbaceous (grain legumes), forage, and woody legumes at both 15N-enriched and natural abundance levels [Jensen et al. 2008]. Studies to improve or optimize the BNF potential in farming systems have been conducted by investigating the legume genotypic differences; effectiveness and competitiveness of the Rhizobium strains and improved agronomic management practices [Hardarson and Atkins 2003; Jensen et al. 2008]. Recent international meetings have identified the need for more research on specific aspects of BNF oriented to the development and introduction of viable and cost-effective technologies to farming systems considering the needs and constraints of small-scale resource-poor farmers in developing countries [Hardarson and Broughton 2003]. More research is needed to obtain quantitative information on the below-ground contributions of C and N of legumes to their respective balances in agroecosystems and their effects on nitrogen economies of the cropping system. Poth et al. [1986], using a 15N dilution technique, showed that about 53% to 71% of the nitrogen fixed by pigeon pea over a period of 225–252 days (equivalent to 150 to 180 kg N ha−1) was recovered in the soil after removal of the coarse roots, indicating that the below-ground contribution can be substantial. Chalk [1998] reviewed the dynamics of biologically fixed N in legume-cereal rotations and highlighted the need for wider use of 15N-based methodologies to estimate additions of legume N to the soil and its effects on subsequent crops so that more accurate N balances can be made. Moreover, quantitative information is also needed to study the fate of residual N on the N nutrition of subsequent crops under different soil-crop management practices. 15N foliar labeling techniques can be used in particular for perennial legumes (forages and tree species) to better quantify root N yields and study root N turnover over time and its uptake by subsequent or neighboring crops [Russell and Fillery 1996; McNeill et al. 1998; McNeill 1999]. Nitrogen dynamics in different agroecosystems have been studied for several decades, however increasing concerns over potential environmental hazards and ecological impacts have led to the development of research areas related to mechanistic studies of soil N transformations and measurement methods to control and mitigate N losses [Follett et al. 1991; Freibauer et al. 2001]. 15N techniques, either enriched or natural abundance, can be used in nitrogen cycling and dynamics studies by determining pools, processes, and flow rates of nitrogen and their influencing factors at various physical scales in agroecosystems [Van Cleemput et al. 1996; Mosier et al. 2004; IAEA 2005, 2006, 2008a]. A recent IAEA publication entitled Nitrogen Management in Agricultural Systems provides a comprehensive coverage of key topics dealing with the utilization of all sources of N in farming systems, in particular the application of 15N tracer technologies in research to improve the overall N use efficiency in agricultural systems while increasing crop yields in a sustainable manner [IAEA 2008a]. Applications of the 15N tracer techniques in agroecosystems can be found in works by the IAEA [2005, 2006]. The extremely variable composition of the organic residues and the site-specific environmental and management conditions have a great influence on the decomposition and release of nutrients, which often do not match the crop demands. The
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
365
accurate assessment of the nutrient supply and selection of appropriate management practices to increase the efficiency of recycling of nutrients in agroecosystems call for the use 15N, 32P, 35S, and 13C as tracers to identify efficient management practices tailored to specific cropping systems and local conditions. Direct and indirect (ordinary and reverse dilution) isotopic methods can be used in the laboratory, greenhouse, and field conditions [IAEA 2003b; Hood-Novotny et al. 2008]. Increasing research attention has been paid to the characterization of organic residues using standardized methods and the establishment of databases on quality parameters of important locally available products and to better predict their nutrient supply with time in order to match crop demands and their effect on soil properties [Cadisch and Giller 1997; Palm 2001; Vanlauwe et al. 2002; FAO 2003; ORD 2004]. Scope and limitations in the management of organic matter in tropical soils [Martius et al. 2001] and effects of organo-mineral combinations (mixtures of organic materials with mineral fertilizers) on soils and crops and their fate in cropping systems have been assessed; however, it is evident that more studies are needed [Vanlauwe et al. 2001]. In view of the great variability in organic composition, the detailed characterization of organic amendments like composts, manures, and biosolids provides essential information for evaluating their agronomic value (or nutrient supply) with time, assessing the impacts of dietary modification in animal production and mitigating the adverse environmental impacts on surface water eutrophication. In this context, the simple characterization of P in manures into labile and nonlabile forms based on their solubility in chemical extraction methods does not provide information on the structure and chemical form of the various P compounds found in the manures [Leytem et al. 2004, 2008]. Both liquid state and solid state NMR can be used to study the speciation of C, N, and P and characterize the specific P forms present in foods, animal feed [O’Neill et al. 1980; Kemme et al. 1999], digesta, and organic amendments [Frossard et al. 1994a; Cade-Menum and Preston 1996; Turner 2004; Turner and Leytem 2004; Toor et al. 2005]. 13C cross polarization NMR spectroscopy with magic angle spinning has been employed for studies on litter acid insoluble SOM factions [Wilson et al. 1983]. 31P-NMR has been found to be a reliable technique for phytate determination and it is recommended that the entire range of P compounds (both inorganic and organic) of a sample needs to be identified in one simple extraction procedure [Leytem et al. 2008]. Moreover, the use of x-ray absorption near edge structure (XANES) to characterize P pools in manures is reported to have several advantages over conventional chemical extraction methods [Peak et al. 2002; Toor et al. 2005; Sato et al. 2005]. Several authors have made combined use of these techniques in their studies, for instance Frossard et al. [2002] utilized solid state 31P NMR sequential extraction and isotopic exchange techniques to investigate forms and exchangeability of inorganic phosphate in composted solid organic wastes. More recently, Ajiboye et al. [2007] employed the sequential chemical extraction with solution 31P NMR and XANES in their studies to provide a detailed molecular speciation of P in various organic amendments. Urbanization of modern societies is leading to increased production of human, animal, and industrial wastes that require disposal, creating environmental problems
366
World Soil Resources and Food Security
mainly in developed industrialized countries but also in some regions of the developing world. This safe disposal would imply the assessment of both: i) the agronomic value or their nutrient supply to crops for the purpose of recycling them in the soilplant system, and ii) the potential hazard of some harmful elements such as heavy metals present in the applied organic wastes, which accumulate in the soil and can be released to the environment. Isotopic tracer techniques such as the isotope dilution and isotope exchange kinetics technique used in soil nutrient studies have several advantages over conventional methods [Sheppard 1962; Fardeau and Jappe 1976; Fardeau et al. 1979, 1985; Frossard et al. 1992, 1994a]. The main advantage of the isotopic methods is that they do not modify the chemical equilibria of the element studied in the soil and make it possible to measure the size of the various compartments including the labile pool, which is considered plant-available. Depending on the method of measurement, this plant-available or bioavailable pool is referred to as the E (or exchangeable when measured using chemical methods in the laboratory) value or L (labile measured using plants in the greenhouse) value. These methods were initially developed for studies of gross transformation rates of nutrients (N, P, and S) in the soil and plant nutrient availability [Kirkham and Bartholomew 1954; Barraclough 1995; Di et al. 1997, 2000] and further extended to study the bioavailability of heavy metals (e.g., Zn, Cd) of sewage sludge (biosolids) applied to land. Both methods, the E and L values, have been used either alone or combined in the same study for phosphorus [Frossard et al. 1994a]. Increases in extractable soil P and plant P uptake have been reported from the applications of organic amendments by several authors. However the relative agronomic efficiency (compared to water soluble P fertilizer) of various organic materials at comparable P application rates were very variable, ranging from 10% to 264% [Pommel 1982; Sikora et al. 1982; Bezzola et al. 1994; Frossard et al. 1996]. These large differences were attributed to differences in the organic materials, soil, and climatic conditions, plant species and methodologies used (Frossard et al. 2002; Sinaj et al. 2002). Using the radioisotopes 109Cd and 65Zn, E values have been determined for Cd [Nakhone and Young 1993] and for Cd and Zn [Young et al. 2000]. L values were determined for Cd using soybean grown in soils with a history of sludge application [Lloyd et al. 1981]. Hamon et al. [1997, 1998] determined the L values of Cd and Zn with different plant species in Australian soils. Several workers have measured and compared the E and L values of heavy metals in sludge-treated soils in the same study. For example, Tiller et al. [1972] found different E and L values for Zn, while Echevarria et al. [1997] found similar values for Ni. Smolders et al. [1999] found that the L values for Cd of 10 Belgian soils were from 1.05 to 1.4 times that of the corresponding E values. Stacey et al. [2001] investigating the effects of the aging of biosolids on the availability of metals found that the E values were consistently lower than the corresponding L values measured using wheat plants. Their data showed that the composition of the biosolids, even in aged (5 year old) sewage sludge, controlled the release of metals to the plant-available pool. Stable isotopes have also been used to determine the E and L values of heavy metals from sewage sludge-amended soils [Gäbler et al. 1999; Ahnstrom and Parker
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
367
2001; Gray et al. 2003]. For example, Ayoub et al. [2003] used 114Cd and 67Zn and the isotopic dilution method for determining their E and L values in a nonamended and a sewage-sludge-amended soil. The isotopic exchange kinetics (IEK) method has been developed and tested to describe the transfer of radioactive 32PO4 (phosphate) ions from the soil solution to the soil solid phase [Fardeau et al. 1985, Fardeau 1993]. The IEK method has been applied successfully to study the dynamics and availability of inorganic nutrients P, K, S, and Zn [Frossard and Sinaj 1997], the K release and fixation processes [Poss et al. 1991], the Zn exchangeability in soils [Sinaj et al. 1999], the phyto-availability of Ni in Ni-polluted soils [Echevarria et al. 1998], and Cd availability in four alkaline soils of wide variation in total Cd concentration [Gérard et al. 2001]. Gray et al. [2004] studied the optimal experimental conditions to determine Cd in soils using the IEK method. Pilot and industrial scale installations for wastewater purification and sludge treatment have demonstrated that radiation-processing-based technologies can help to mitigate environmental degradation. The IAEA publication Radiation Processing: Environmental Applications reports on current uses of these radiation-based technologies for the treatment of wastewater and sludge. The information is provided to facilitate the preparation of guidelines in feasibility studies for further implementation of radiation processing technologies [IAEA 2007]. A networked research project involving several countries was conducted from 1995 to 2000 to develop management practices for the efficient use of irradiated sewage sludge for application to agricultural lands under a wide range of environments [IAEA 1997]. Sewage sludge from different origins was treated with a moderate dose (2 to 25 kGy) of gamma irradiation and both the pathogen elimination and agronomic value of the irradiated sewage sludge were evaluated using 15N and 32P isotopic techniques in a number of environments [IAEA 2002b]. Municipal wastewater treatment processes can produce slightly 15N-enriched sewage sludge (biosolids). The degree of 15N enrichment in biosolids varies with wastewater treatment and their characteristics [Wang et al. 2005]. When wastes with a high δ15N are applied to an ecosystem where the soil background N has a significantly lower δ15N, the fate of the waste-derived N can be tracked using δ15N values. The 15N natural abundance has been also used to trace the fate of applied sewage sludge (biosolids) to forestland in New Zealand [Wang et al. 2005]. In another study, treated sewage effluent was applied to a pine plantation forest and the δ15N of the plant tissues were used to estimate the proportion derived from the following sources: effluent N, N stored in the tree and soil N [Wang et al. 2005]. A further extension of the application of the 15N natural abundance was made to construct an N balance of a catchment in Rotorua, New Zealand [Tozer et al. 2005]. In Malaysia, the 15N reverse dilution technique was used to assess the N supply from palm oil mill effluent (POME) applied to young oil palm plants [Mohd Hashim and Zaharah 1994]. In grassland ecosystems, ruminant livestock production research has mainly focused on the chemical composition of cattle dung in terms of NPK availability. Studies have been made by feeding animals with 15N-enriched plant materials to produce 15N-labeled manure or 15N-spiking cattle dung pats for research on the fate
368
World Soil Resources and Food Security
of this manure in the soil-plant system [Catchpoole and Blair 1990; Shepherd et al. 2000]. Similarly, studies using the 13C natural abundance have been made to unravel the fate of the dung-derived C into the soil, in particular the C sequestration potential from dung in temperate grasslands. In this case either C3- or C4-labeled dung can be produced using a monitored feeding diet to stalled cattle [Amelung et al. 1999; Bol et al. 2000, 2004]. Most acid soils of the subtropical and tropical regions show a widespread phosphorus deficiency and high P sorption capacity, thus becoming the main constraint to agricultural productivity. In spite of their inherent low soil P fertility, P depletion due to a continuous cultivation without replenishment is severely affecting agricultural production in sub-Saharan Africa. According to Smaling et al. [1997], approximately 75 kg P ha−1 has been lost over the past 30 years from 200 million ha of cultivated land in 37 African countries. This is equivalent to an annual loss of 0.5 million ton P for the African region. While additional N inputs can be obtained from sources such as BNF, crop residues, and other organic sources, external P inputs need to be applied in order to improve the soil P status and ensure normal plant growth and adequate yields. The application of manufactured water soluble phosphatic (WSP) fertilizers such as superphosphate is usually recommended to correct P deficiencies. However, most developing countries import WSP fertilizers, which are often in limited supply and represent a major outlay for resource-poor farmers. In addition, intensification of agricultural production in these regions necessitates the addition of P inputs, not only to increase crop production but also to improve soil P status in order to avoid further soil degradation. Therefore, it is imperative to explore the sound utilization of alternative P inputs. In this context, the direct application of indigenous phosphate rocks (PRs) is particularly attractive under some conditions to develop an effective and economic phosphate management program [Rajan and Chien 2003; Zapata and Roy 2004]. In the framework of a networked research project on the use of nuclear and related techniques for evaluating the agronomic effectiveness of phosphatic fertilizers, rock phosphates in particular, investigations on several key issues related to soil and fertilizer P management for crop production systems in tropical agroecosystems were conducted from 1993 to 1998 [IAEA 2002b]. The performance of several chemical extraction methods for determining plant-available soil P was evaluated in mostly Ultisols, Inceptisols, and Oxisols using the refined 32P IEK as a reference method [Fardeau et al. 1985; Fardeau 1993, 1996; IAEA 2002b]. The 32P IEK provides a good description of soil P status, i.e., intensity (Cp), quantity (E1), and capacity (Q) factors. It was found to be very valuable to assess P dynamics in soil with or without the addition of P fertilizers (WSP and PR). It was concluded that there is no single soil chemical P test that can be universally used to estimate available P in soils amended with PR and WSP fertilizers. The Pi strip test works well in determining available soil P in soils amended with either PR or WSP fertilizer. A modification of the method was made using 0.02 M KCl instead of 0.01 M CaCl2·2H2 0 for use in soils treated with PR and superphosphate [Habib et al. 1998]. The soil P fixation capacity can be also determined from the 32P IEK measurements [Frossard et al. 1993]. The evaluation was made after about one month of incubation. Soil/P fertilizer can be used to predict the agronomic effectiveness of
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
369
P fertilizers [Morel and Fardeau 1991]. Furthermore, the method can be used to identify various kinetic pools of soil P and their changes with time [Fardeau 1993; Fardeau et al. 1995]. Investigations were conducted to develop standardized protocols to characterize PR sources and evaluate their relative agronomic effectiveness and, where necessary, to find ways and means of enhancing their effectiveness [IAEA 2002b]. 32P and 33P isotopic techniques have been used extensively in both laboratory and glasshouse experiments to measure P uptake and utilization from the applied P fertilizers, in particular PR-based products [Zapata 1990; Zapata and Axmann 1995; Zaharah and Zapata 2003]. Adequate soil-plant-fertilizer management practices that can be put in place to enhance the efficient use of soil P and added external P inputs, in particular the application of PR to build up the soil P capital in tropical acid soils, were developed [Zapata 1995, 2002b; IAEA 2000b, 2002a, 2002b; Zapata and Roy 2004]. The generated information has been compiled in a PR database with chemical and agronomic entries and a DSS for the use of phosphate rock (PR-DSS) was developed in a joint venture with IFDC [Sale and Heng 2004; Smallberger et al. 2006]. A dedicated Web site on direct application of phosphate rock (DAPR) has been also developed for wide dissemination of the results to land managers and policy makers (http://www-iswam.iaea.org/dapr/srv/en/about). Plants differ greatly in their ability to grow on low P soils because they have developed specific physico-chemical mechanisms and processes to utilize P compounds in these low P fertility soils, often with the intervention of microorganisms present in the soil-root interface called the rhizosphere [Hinsinger 1998; Hocking 2001]. A FAO/IAEA coordinated research project [FAO/IAEA 2009] is operational to investigate the main processes occurring in the rhizosphere that determine P acquisition by plants. 32P isotopic techniques will be used to unravel the main factors influencing these fluxes in the rhizosphere and the extent of P availability to selected genotypes of food crops. Phosphorus deficiency is also a major constraint for crop and livestock production in temperate grasslands, where the low concentration of solution P available for plant uptake limits the continuous production of animal products and crops. Applications of external sources of P are necessary to enhance and sustain crop and pasture production [Donald 1964; Nguyen and Goh 1992; Bertrand et al. 2006]. Phosphate fertilization is of great economic importance to pastoral farming in New Zealand and Australia. In these countries, legume-based pastures rely on a regular supply of adequate available P so that the rhizobium in the legume root nodules is able to fix atmospheric N [Lewis and Clarke 1975; Nguyen et al. 1989]. The direct application of reactive phosphate rocks (RPRs) has been shown to be as effective as soluble P fertilizers for well-improved pastures where soil conditions are favorable for RPR dissolution [Rajan 1987; Rajan et al. 1996; Bolan et al. 1990; Sale et al. 1997]. In intensively cultivated areas, the continuous use of external P inputs can lead to its accumulation in the topsoil and increased pollution risks and eutrophication of water bodies, if such accumulated P finds its way to streams and rivers. In these situations, 32P isotopic techniques can be used to unravel the main factors that
370
World Soil Resources and Food Security
influence these fluxes so that appropriate soil-plant-fertilizer management practices can be put in place to enhance the efficient use of soil P and added external P inputs, thus minimizing the adverse effects of surplus P and potential losses to the environment. Agricultural intensification using high yielding varieties of crop species with high nutrient export, reduced S atmospheric deposition inputs from industrial sources, and the sustained use of high grade, low, or free S-containing fertilizers has led over time to the development of S deficiency in many parts of the world, which were previously S sufficient areas [Blair and Lefroy 1987; Messick and Fan 1999]. Several studies have used both 35S (radioactive) and 34S (stable) as isotopic tracers to follow the pathways of sulfur in soil-plant or soil-plant-animal systems and to construct sulfur budgets in grazing systems [Till 1981; Chen et al. 1999]. The direct method involving the use of 35S-labeled materials has been applied in studies to determine: 1) the S uptake and recovery of 35S-labeled fertilizer and its fate in flooded rice systems [Samosir et al. 1993]; and 2) the availability of subsoil sulfate using 35S-labeled gypsum to crops [Bole and Pittman 1984]. Sometimes primary effects and interactions involving C, N, P, and S should be investigated in these systems. In these cases, multilabeled plant materials should be produced and used [Konboon et al. 2000; Basilio-Sanchez et al. 2000] to study their nutrient release. The 35S reverse dilution technique has been employed to determine i) the ability of plants to acquire S from the atmospheric SO2 [Hoeft et al. 1972]; ii) the release of S from elemental sources and sulfate sources and its uptake by plants [Shedley et al. 1979]; iii) the sources of S taken up by ryegrass and measured by chemical extractants [Chinoim et al. 1997]; iv) the time course of S uptake from the sulfur-coated urea by crops [Yasmin 2003]. The 34S natural abundance has been used extensively to identify sources and to trace the fate of S in the environment. This approach requires that the S isotopic signatures of the different sources are known to be able to apportion the contribution of S from the sources [Krouse 1977; Mayer et al. 1995; Alewell et al. 1999; Novak et al. 2001]. This technique has been applied to study long-term changes in sulfur deposition in the Broadbalk experiment, UK [Zhao et al. 2003]. In a recent study, ion exchange membranes have been successfully used to collect soil water sulfate for investigating its isotope (sulfur and oxygen isotope fractionation) composition [Kwon et al. 2008]. An IAEA publication entitled Guidelines for the Use of Isotopes of Sulfur in Soil-Plant Studies provides a good overview of the sulfur cycle and basic principles and practical procedures for the use of the S isotopic techniques, including selected practical applications of the sulfur isotopes [IAEA 2003b]. The increasing deficiency of micronutrients is becoming a matter of concern due to their impacts on agricultural production, in particular in intensively cultivated systems in southeast Asia. Among the micronutrients, zinc (Zn) deficiency is the most acute followed by boron (Bo). According to Sillanpää [1990], Zn deficiency is the most commonly occurring micronutrient deficiency problem limiting crop growth in many parts of the world. In India, application of Zn resulted in spectacular yield increases in wheat-growing areas [Takkar et al. 1989]. Isotopic-aided studies using 65Zn have also been conducted in southeast Asia to improve the Zn nutrition of flooded rice [IAEA 1983].
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
371
8.3.4 Use of NIT in Water Management Studies in Agriculture 8.3.4.1 Introduction The four major factors driving the increasing water demand are population growth, industrial development, the expansion of irrigated agriculture, and the impacts of climate change and climate variability. Water is a major input for agricultural production under both rainfed and irrigated conditions. Agriculture is the predominant user (75%–80%) of the available freshwater resources in many parts of the world. About one-third of the world’s population lives in countries suffering from moderate to high water stress. By 2020, water use is expected to increase by 40% and 17% more water will be required to produce the food requirements of the growing population [Rosengrant and Ximing 2001]. Competition among different sectors for scarce water resources and increasing public concern about water quality for human, animal, and industrial consumption and recreational activities, have focused more attention on water management in agriculture. As water resources shrink and competition from other sectors grows, agriculture faces a dual challenge: to produce more food with less water and to prevent the deterioration of water quality through contamination with runoff, soil erosion, and sedimentation, and associated nutrients and agrochemicals [Meizen-Dick and Rosengrant 2001]. Water use efficiency (WUE) is a concept that can be defined in many ways. For farmers and land managers, WUE is the yield of harvested crop product achieved from the water available to the crop through rainfall, irrigation, and the contribution of soil water storage. With regard to WUE assessment as a tool for improving water management in agriculture, it is essential to recognize: 1) spatial and temporal scales are critical, and 2) WUE depends on both inherent and dynamic properties, processes, and interactions occurring in the soil-plant system. Improving water use efficiency in agriculture will require an array of strategies going from plant mechanistic studies to increase crop water productivity to a reduction in water losses from the plant rooting zone at the field level to ensure adequate storage of moisture (and nutrients) required for optimizing crop production at both the farm and watershed levels. Increasing WUE is of a paramount importance, particularly in arid and semiarid areas. Under rainfed conditions, soil water can be lost from the soil surface through evaporation or from plant leaves through plant transpiration. It can also be lost through runoff and deep infiltration. Water losses associated with soil evaporation and plant transpiration are often referred to as evapotranspiration. When irrigation is available, water losses also include the mismanagement of irrigation water from its source to the crop roots. Usually, more than 50% of irrigation water is lost for the crop to use at the field-scale level. However, at the watershed level, it might be less due to the possible recoveries from the subsoil and groundwater. The offsite losses of water can result from either inappropriate land management practices to capture a substantial part of the rainfall within an agricultural landscape and retain it in the plant rooting zone, or excessive use of irrigation water. Such losses not only lead to water wastage but also potential hazards of soil salinity and water pollution resulting from the transport of nitrate, phosphate, sediments, and agrochemicals to receiving water bodies.
372
World Soil Resources and Food Security
The role of nuclear and isotopic techniques as tools for improving water use efficiency in agroecosystems will be described below. 8.3.4.2 Use of the SMNP in Water Use Efficiency Studies The SMNP is a portable nuclear equipment for periodically measuring soil water content at different depths through access tubes installed permanently in the soil profile. The SMNP provides a rapid, reliable, cost-effective, and non-destructive technique and is still considered as the reference equipment for field monitoring of soil water content [Greacen 1981]. Data generated from this monitoring are used for calculating the soil water balance and estimating the soil water used by plants in function of soil depth and time. Thus, the SMNP provides an evaluation of evapotranspiration (ET), i.e., the bulk amount of soil water loss associated with soil evaporation and transpiration out of plant leaves. The SMNP has been widely utilized in agricultural research studies on water use efficiency [IAEA 2000a, 2001, 2003a, 2008b]. Although other soil water monitoring methods have been developed, the SMNP is still relevant in areas where soil conditions such as salinity can affect the performance of other equipments [IAEA 2000a, 2008b]. The applications of the SMNP in agriculture range from characterization of soil physical properties to improving water use efficiency in rainfed and irrigated cropping systems. The soil hydraulic conductivity is measured easily in situ through the so-call internal drainage method [Watson 1966; Van Bavel et al. 1968]. Water use efficiency is measured at the field level using a network of SMNP access tubes. Very often, sets of tensiometers or porous ceramic cups are associated with the access tubes for monitoring hydraulic heads [Richards et al. 1956] and for the calculation of water fluxes and, for example, N fertilizer leaching [Moutonnet and Fardeau 1997]. The Joint FAO/IAEA Program has implemented several coordinated research and TCPs with the objective of optimizing the use of agricultural water (rainfed soil moisture in arid and semiarid areas, supplemental irrigation, irrigation scheduling, fertigation), and the results were published in the IAEA TECDOC series [IAEA 1996, 1998, 2000b, 2002a, 2005]. The judicious application of N fertilizer increased the WUE by crops as measured by SMNP due to improved groundcover and hence reduced evaporation from the soil. In most of the cases, substantial savings of 20% to 35% of the irrigation water were achieved. Up to 50% improvement of WUE could be achieved by changing the management practices according to the pattern of rainfall during the growing season in Jordan. In Syria, most of the water was saved under drip irrigation during early vegetative development stages of cotton. The SMNP is also used in studies to evaluate WUE under different irrigation technologies like drip irrigation, subsurface drip irrigation, deficit irrigation, regulated deficit irrigation, alternate root-zone drying, and alternate furrow irrigation, that aim to minimize soil evaporation and increasing water availability for plant use. It can be employed to assess the quantity of water supplied by surface irrigation and the uniformity of its distribution. For example, in a FAO/IAEA CRP, crop yield responses to imposed deficit irrigation were evaluated for a wide range of field crops (IAEA 1996). A series of yield response factors showed a wide range of variation from 0.08 (no impact on yield) to 1.75 (severe impact on yield). These results were
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
373
published in a book [Kirda et al. 1999] and later on in an FAO/IAEA Water Report series [FAO/IAEA 2002]. 8.3.4.3 U se of Stable Isotopes for Determining Water Used by Plants in Agroecosystems Stable isotopes of water, 2H, and 18O at the natural abundance level have been extensively used in plant-water relations research to investigate physiological and hydrological processes from plant to ecosystem scale. Thus, stable isotopes of water are routinely measured for tracking sources of water in vegetation. In agriculture, the crop water requirements are evaluated as bulk evapotranspiration, which includes both soil evaporation and plant transpiration. Due to the natural fractionation among isotopes of hydrogen and oxygen, soil water at increasing depths in the soil profile and plant materials have their own fingerprint. Therefore, it is possible to partition evapotranspiration into its two components in order to obtain information to minimize soil evaporation, the non-productive loss of water. Both hydrogen (H) and oxygen (O), which are element constituents of water can exist as light and heavy isotopes. The light isotopes (1H and 16O) evaporate more readily than the heavy isotopes (2H and 18O). Thus the stable isotopic ratios of H (2H/1H) and O (18O/16O), which are often expressed (see Section 8.2.2) as the enrichment of 2H (δ 2H) and 18O (δ 18O) in soil water. Water vapor within the plant canopy and plant leaves can provide an estimate of soil evaporation (E) and plant transpiration (T). The following studies illustrate the application of these techniques:
1. Studies on the daily changes in E following surface irrigation of olive trees in Morocco showed that the δ2H values estimated for olive transpiration and soil evaporation provided very distinct isotopic signals and E ranged from 0% prior to irrigation to 31% of total ET after surface irrigation [Williams et al. 2004]. 2. Partition of transpiration from overstory trees from that of understory grasses in the SW of the United States showed that during the postrainy period, the total ecosystem ET was 3.5 mm/day, of which 70% was from tree transpiration, 15% was from the grass layer, and 15% came from soil evaporation E [Yepez et al. 2003]. 3. The role of tree size and hydraulic lift on water use by trees and forests was investigated using isotopic, energy balance, and transpiration analyses [Dawson 1996]. 4. Several studies caution about isotope fractionation in water during uptake by roots. Evidence has been provided for hydrogen fractionation during water uptake by woody halophytic and xerophytic species. In these cases, δ 18O values of xylem water of the same species that fractionated hydrogen isotopes in water did accurately reflect the δ 18O of the soil water. Standardized procedures should be used in sampling and sample preparation for isotopic analyses [Dawson and Ehleringer 1993; Ellsworth and Williams 2007].
Present day woodlands often have hydrological characteristics very different from those of the original grasslands in the Rio Grande Plains of North America.
374
World Soil Resources and Food Security
Compared to grassland plants, tree and shrub species are rooted more deeply. Differentiation in isotopic signatures confirms that grassland species acquire soil water from the upper 0.5 m of the soil profile. In contrast, tree and shrubs utilize soil water from the deeper 4 m of the profile [Boutton et al. 1999]. Similar studies have been recently made to assess hydraulic lift and water sources of native trees growing in parklands of the dry savannahs of the West African Sahel to develop improved agroforestry practices [Bayala et al. 2008]. The stable isotopes methodology is being further explored in a networked research project to test the efficacy of innovative irrigation techniques with the objective to increase irrigation water use efficiency [Hillel and Vlek 2005]. 8.3.4.4 U se of Carbon Isotope Discrimination in Plants as a Tool for Assessing Plant WUE Water use efficiency (WUE) is often considered as the mass of carbon dioxide fixed by the plant in exchange for the mass of water transpired. During photosynthesis, carbon dioxide is subjected to an isotopic fractionation, with lighter 12C more readily taken up by plants than 13C. The kinetics of diffusion of carbon dioxide and water through the leaf stomata openings depends on the size of their aperture. Under drought stress, the plant closes its stomata, a process which impedes both exchanges. A side effect is the discrimination of the heavy isotopes of C, O, and H; therefore, water scarcity results in a plantdepleted content in 13C, for example, which is measured by the ratio of 13C/12C (i.e., δ13C) in its dry matter. A cultivar that is resistant to water scarcity should display less depletion in 13C compared to a susceptible cultivar. This carbon isotope discrimination (CID) may be used as a tool for assessing crop WUE [Condon et al. 2002, 2004]. Soil scientists as well as plant breeders have used this method strategically for screening and evaluating WUE of crops such as barley [Hubick and Farquhar 1989], bread wheat [Rebetzke et al. 2002], peanut [Wright et al. 1994], and upland rice [Dingkuhn et al. 1991]. In addition, CID measured in different plant parts at harvest can be used as an indicator of how water availability varied during the cropping season [Dercon et al. 2006]. The IAEA has implemented two CRPs involving the use and application of the CID technique under water-limited conditions:
1. Nutrient and water management practices for increasing crop production in rainfed arid and semiarid areas (D1.20.06). In this project, sixteen teams of scientists from different countries developed and pilot tested nutrient and water management practices tailored to local conditions of the rainfed agroecosystems. Also the CID technique was employed as a diagnostic tool for predicting WUE and yield. Promising results relating Δ13C and wheat grain yields were found [IAEA 2005]. 2. Selection for greater agronomic water-use efficiency in wheat and rice using carbon isotope discrimination (D1.20.08). This network includes twelve research teams from developing and developed countries working on selections of wheat genotypes that are tolerant to drought and irrigated rice strains tolerant to salinity. Results obtained showed that grain CID in wheat (Triticum L.) can be used as a selection criterion for wheat yield
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
375
under a wide range of environmental conditions [FAO/IAEA 2009]. For rice (Oryza sativa L.), the CID values from flag leaves can potentially be used to select rice for salinity tolerance, since flag leaf CID correlated well with grain CID, and both positively correlated with grain yield.
8.4 LOOKING AHEAD TO CHALLENGES FACING AGRICULTURE 8.4.1 Contributing to Achieving Sustainable Land and Water Management while Mitigating Climate Change Impacts 8.4.1.1 Combating Land Degradation and Improving Soil Quality In a recent situation analysis of the food and agricultural sector in the developing world, it was noted that among several issues of global significance, the status of land degradation has continued to worsen in many parts of the world despite some bright spots (see Section 8.1). The impacts from human-driven global warming resulting from climate change have emerged as the main driver exacerbating pressures on the natural resources. This increased worldwide degradation of extensive tracts of land causes a decline in productivity, disrupts vital ecosystem functions, negatively impacts biodiversity, and increases vulnerability to climate change [Nguyen 2007; Nguyen et al. 2010]. Recognizing that environmental degradation, poverty, and food security are strongly intertwined, one of the key elements of strategic planning documents of UN organizations such as the FAO Strategic Framework (2000–2015) and the IAEA Midterm Strategy (2006–2011) is to support environmental sustainability [FAO/IAEA 2009; IAEA, 2009a, 2009b]. The overall aim is to develop and implement project activities to help countries and regions achieve sustainable use and management of natural resources and mitigate climate change impacts, thus contributing to meeting the sustainable development targets set by the World Summit for Sustainable Development and the UN Millennium Development Goals of poverty reduction, food security, and environmental sustainability [UNMDG 2009]. In view of this situation, the underlying philosophy in developing strategies for the SWMCN Subprogram was to identify and address global issues of major concern and relevance to sustainable land and water management for food and agriculture in the developing world [FAO/IAEA 2009]. Soil quality, in terms of productivity and sustainability, is governed to a great extent by SOM, of which soil organic carbon (SOC) is the building block through its effect on many soil properties and processes [Janzen et al. 1997; Arshad and Martin 2002]. Therefore, SOM, in particular the amount of SOC, is commonly used as an indicator of soil quality in mineral soils in agroecosystems [Herrick 2000; Carter 2002; Smith 2003]. All ecosystem services provided by soil are affected by the quality and quantity of SOM and its dynamics [Lal 2004]. In most cultivated soils of the world, SOM has been depleted and needs to be restored through implementing an integrated approach to soil-water-plant technologies in agroecosystems. Because of the importance of C sequestration in soils, soil quality should be evaluated in terms of its SOC content [Lal and Kimble 1997]. The assessment of the carbon ecosystem budgets in terms of soil sequestration and CO2 emission is needed with respect to
376
World Soil Resources and Food Security
recommended land and water management practices for food security and environmental sustainability. 8.4.1.2 D evelopment of an Integrated Approach to SoilWater-Plant Technologies in Agroecosystems Reducing the impacts from global warming necessitates an action plan addressing the urgent global issues of carbon sequestration and biofuel production and represents a key requirement for sustainable management of land and water resources [Lal 2004]. In a systems approach to activities relating to soil C sequestration, climate, and human interactions, a common link among these issues is the biogeochemical cycling of C and its coupling with those of N, P, S, and water. Interactive processes affecting coupled cycling of these elements are strongly influenced by the increasing global energy demand. But present knowledge of global cycling of elements, particularly the global C cycle remains limited by uncertainties in quantitative aspects of SOC storage and dynamics [Nguyen 2007]. Nuclear and isotopic techniques can play a major role in addressing these issues of global significance. The diagram in Figure 8.1 shows the role of isotopic techniques in addressing important issues of food and biomass production such as C
Im
pr
o
g vin
cy
cli
ng
&
Nuclear & isotopic technologies in soil & plant analyses
15N, 32P
Pu
rifi
cat
ion
Nutrient/ elemental cycling
Gl
ob
al
Fo o
d&
En
erg
yS
ecu
12C, 13C, 14C
n ni tio tra ems s e qu st se o s y on o-ec b r r Ca Ag
Social, ethnic & policy issues
rit
y
Carbon sequestration & climate change mitigation Economic issues
Re
14C, 15N
2H, 18O
Water resource, quality & availability
W ate r
W
UE
&
E NU
Species selection
Identifying niches for species
Tree species with deep root system & recalitrant compound
Food biomass production
The human dimensions
FIGURE 8.1 The role of isotopic techniques in addressing global issues of carbon sequestration to advance food security and mitigate climate change. (From Nguyen, M. L., Integrated soil-water-plant solutions for biomass production and environmental performance as influenced by climate change, Report of the FAO/IAEA Consultants Meeting, Vienna, Austria, November 12–14, 2007. With permission.)
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
377
sequestration in agroecosystems, bioenergy production on degraded or marginal soils, and declining availability and quality of water resources [Nguyen 2007]. In this carbon-climate-human system, there are also strong interactions between changing climate and species selection and adaptation and human dimensions such as related economic, social, cultural, ethnical, and political factors (dotted circles in Figure 8.1). The design and implementation of successful land and water management programs on soil C sequestration and biofuel production aimed at advancing food security and mitigating climate change requires an integrated approach to soil-water-plant technologies in agroecosystems to gather the following information:
1. Quantitative data on C fluxes and residence time and budgets under different soil and crop management practices in site-specific environments. 2. Information on incremental changes in the C pool at different temporal and spatial scales. 3. Reliable information on improved efficiency in land and water management practices for mitigating climate change and enhancing food and bioenergy production.
The focus of future project activities will, therefore, be on improving management of SOM by enhancing the net rate of sequestration of carbon in soils and terrestrial ecosystems in agroecosystems to increase biomass production in land biota (plants) and soils, enhance soil quality, and mitigate the impacts from climate change (win/win strategy options) [Lal and Spruce 1999]. Improving SOC sequestration will be achieved through integrated strategies for sequestering atmospheric carbon in biomass and soil and reducing emissions from greenhouse gases emissions. Therefore, soil quality enhancement would require improvement of SOC through: 1) defining appropriate land use; 2) selection and adoption of effective recommended management practices for enhancing SOC; and 3) gathering science-based knowledge on SOC management. An ecoregional approach for the diagnosis of degradation status of land and water resources in the tropics and subtropics will be followed. Thus, soil quality enhancing technologies will be developed and pilot-tested within well-defined land classes (arable, marginal, and degraded lands) in predominant agroecological zones using an integrated approach to soil, water, and nutrient management through enhancing resource use efficiency and improving soil carbon sequestration in agroecosystems. The overall goal will be to contribute to breaking the vicious cycle of declining soil quality, depleting SOC stock, and lowering yields, and to combating hunger and poverty in the developing world [Lal 2004]. In defining research activities of the SWMCN Subprogram on integrated approaches to soil-water-plant technologies for biomass production and environmental performance, a number of research topics were identified, of which four were considered relevant to the IAEA and FAO program of work [Nguyen 2007; Nguyen et al. 2010]. The tentative priority titles and main elements of the strategy are given below: • Carbon sequestration in agroecosystems for enhancing food production and mitigating climate change.
378
World Soil Resources and Food Security
• Ecosystem carbon budgeting and environmental sustainability for bioenergy production. • Restoration of degraded and desertified soils for mitigating climate change. • Selecting germplasm from farming systems under extreme climate conditions for sustainable food and biomass production. The primary focus will be to develop integrated approaches for assessing impacts of food and bioenergy production on C sequestration and carbon budgets in agroecosystems, and to address mitigation of the global warming problem through the use of isotopic and nuclear techniques. The research activities will be designed to underpin the selection and implementation of targeted sustainable land-use and management practices aimed at conserving the natural resources base (land and water) and enhancing soil quality and biodiversity. Nuclear and isotopic techniques will be an essential component of the project activities. Stable isotopic—natural abundance and isotopic-labeled SOM using bulk and compound-specific 13C isotope analysis (CSIA)—tracers and fallout radionuclides (137Cs, 7Be, and 210Pbex) will be used in combination with conventional measurement techniques to gather quantitative data of SOM storage processes and dynamics in relation to land use, climate, and gas composition of the atmosphere. This goal is essential to be able to achieve sustainable management of management of land and water resources [Nguyen 2007]. The combined use of isotopic techniques with thermochemical techniques such as mid-infrared spectroscopy have the potential to assess the net rate of carbon sequestration of recommended management practices (RMPs) in terms of resource (fertilizer and water) and energy use efficiency for biomass production (net primary productivity). As there are no blueprints, networked research needs to be conducted to obtain information on the relative efficacy of the interventions designed to enhance soil carbon sequestration through the integrated approach to soil-water-plant technologies for selecting the most effective and best locally adapted RMPs for agricultural, marginal, and degraded lands. This information is essential for socioeconomic studies and climate change policy formulation [Lal and Spruce 1999]. Nuclear-based techniques can also play an essential role in the research and development of suitable soil C sequestration technologies in agricultural lands to provide data backed by scientific evidence. Because of the need to stabilize the stored C in the soil, key mechanistic studies would be conducted to get a better understanding of processes and driving factors that control the dynamics (transformations/turnover) of specific compounds of SOC, as well as the functions of soil biota (biological component of soil quality), thus demanding the refined quantification of C, nutrient, and water pools as well as fluxes in a given agroecosystem. The measurement of natural variations in the abundance of stable isotopes (deuterium, 13C, 15N, 18O, 34S) in components of the agroecosystem (SOM, standing biomass, ground and surface water, atmospheric gases) can provide unique information on such pools and fluxes. Significant advances have been made in the development and application of isotopic techniques in the fields of biology, ecology, biogeochemistry, and environmental sciences. These developments have been possible due to novel isotope techniques such as compound-specific isotope analysis, advances in instrumentation, analytical techniques, electronic data acquisition, and automated systems for isotope ratio
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
379
measurements in soil, plant, water, and gas samples [Crosson 2008]. However, these advanced techniques need to be further refined and protocols harmonized for worldwide application in agricultural research. The advent of synchrotron facilities that allow the exploitation of particular qualities of synchrotron radiation as a research tool and the continuing development of synchrotron-based techniques (such as x-ray absorption, florescence and tomography) to improve spatial resolution and sensitivity [Lombi and Susini 2009] offer exciting opportunities to unravel processes and factors influencing soil-water-nutrient-plant-rhizosphere interactions. Because of the urgent need for making better use of extensive tracts of marginal lands and restoration of degraded lands in the tropics and subtropics, bioenergy plantations should be established in these lands with the overall objective of improving the environment, protecting soil and water resources, and meeting energy needs without exacerbating the problem of global warming. A major challenge would be to select, evaluate, and develop plant genotypes for bioenergy plantations that can accumulate large amount of biomass under water-limited soil conditions or where soils are affected by salinity or acidity or are depleted of nutrients. It is very important to determine whether biofuels are good or bad for the planet through assessment of impacts of biofuel production on water use efficiency and C and N cycles in ecosystems [Nguyen 2007]. There is also a great scope for interdisciplinary research between plant breeders and soil and water management specialists in the area of crop resilience to climatic change and resulting abiotic stresses (drought, flooding, salinity, variable nutrient/ water availability, etc.). There is insufficient information on the ecology, physiology, and genetics of almost every species, which is necessary to predict the evolutionary responses of different species to climate change, i.e., extinction versus adaptation to climate change, and to develop genotypes that can be matched with environment. Soil typematching plus appropriate external inputs could be the key to adaptation to climate change. Isotope techniques can be used in plant breeding/crop improvement programs to examine the response/adaptation of plant genotypes to climate variations and extreme events. The use of the integrated approach to soil, water, and plant management requires the use of a suite of nuclear-based and conventional (non-nuclear) techniques, which means multidisciplinary and often interinstitutional teams, demanding networking, close coordination, information technology (IT) exchange, and overall increases in the cost and complexity of the research efforts (Table 8.5). The new integrated approach also calls for the creation of databases and application of models to integrate large and complex sets of data obtained under a range of agroecological conditions. The experimental data can be also used to validate and refine existing models as well as to develop DSSs to make better informed decisions on the matching of technologies with environments in specific regions and farming and catchment systems. Because of the nature of the sustainability issues, long-term field experiments with treatments of continuous natures are needed to develop and pilot test adequate technologies over time. In view of the large spatial variability over short distances across the landscape, innovative area-wide (catchment/watershed) approaches and related geoinformation systems (GIS) and techniques (geostatistical tools) are required to assess land and water connectivity issues and also issues between neighboring natural ecosystems and agroecosystems [FAO/IAEA 2009].
380
World Soil Resources and Food Security
TABLE 8.5 Summary of Soil C Cycling in Agroecosystems: Research and Development Needs Knowledge Gaps SOC measurement and monitoring
Developing a DSS
Issues/Topics • Measurement of GHG emissions in the landscape • Spatial variability of SOC vertically (in depth) within the soil profile and over short distances in the landscape • Below-ground C contribution Information on databases and modeling
Ecosystem carbon budgeting for bioenergy production under changing climates Processes and factors involved in SOC sequestration (storage and stability) in agricultural watersheds
Modeling carbon, energy, water, and nutrient dynamics under bioenergy plantations Carbon redistribution in watershed and carbon losses (net erosion) from farming systems
Processes and factors involved in SOC sequestration (storage and stability) in agroecosystems
Assessment of the relative value of specific RMP
Activities • Networked research • Standardized methodological framework • Harmonized protocols • Faster and better techniques/ tools • Coordination and IT exchange
• Pilot testing, coordination, and IT exchange • Creation of database and development of DSS • Field validation DSS • Adoption and policy formulation • Research and development • Field validation • Model calibration • Soil redistribution (rates and patterns) and its relationship and linkage to SOC redis tribution in the landscape and within the watershed • Use and calibration of spatially distributed models • Assessment of carbon losses by other processes such as leaching (DOC) and mineralization • Characterization of SOM (SOC) forms • Identification of specific compounds (active forms) as soil quality indicators for SOC sequestration • Implement SOC changes under conservation agriculture • Establish relationship to total SOC (detect changes over short period of time) baseline and thresholds • Impacts on fluxes, GHGs, and relationship to other elements
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
381
TABLE 8.5 (Continued) Summary of Soil C Cycling in Agroecosystems: Research and Development Needs Knowledge Gaps
Issues/Topics
Extrapolation of results from RMP
Simulation modeling climate change mitigation measures
Processes and factors involved in SOC storage and stability in vegetation and landscapes
Biomass burning
Processes and factors involved in SOC storage and stability in landscapes
Changes in land use and management, particularly impacts from deforestation and afforestation
Mechanistic studies
Better understanding of complex environmental and management influencing factors in a range of agroecosystems
Refined C-cycling studies
Mechanistic studies Soil and microbial ecology Ecosystem functioning
Importance of SIC in desert ecosystems Studies relating SOC to soil quality
Fires (wild and prescribed)
• Knowledge of the microbial community for SOC management • Understanding underground interactions Quantification and stability of SIC pool in semiarid and arid areas Better understanding of SOC sequestration and its relationship to soil quality
Activities • • • •
Research and development Field validation Model calibration Characterization of SOM, especially SOC forms • Identification of specific compounds and its turnover • Assess the role of black carbon (biochar) under changing climate • Study SOC losses and accumulation at various scales • Improve assessment of below-ground SOC contribution by deep-rooted plants and perennials • Improve characterization of SOC • Identification of old and new carbon • Measure SOC turnover (pools, fluxes, and residence times) processes • Measure GHG emissions at various scales • Assess structure and functions of main microbial groups related to SOC cycling • Manage soil biota (biological component of soil quality) for improved soil health Measure SIC (secondary carbonates) formation and assessing stability of SIC forms • Establish relationship between C pools and soil quality • Relate soil quality changes to biomass and agronomic production
382
World Soil Resources and Food Security
8.5 OVERVIEW OF SOC CYCLING STUDIES IN AGROECOSYSTEMS 8.5.1 Introduction In examining C cycling in agricultural systems, there are some important processes affecting the enhancement or depletion of the SOC pool and its stabilization that deserve further investigation. The change in total SOC over time is usually considered as an indicator of soil quality for agricultural lands. However, there is a need to develop a general harmonized framework for measuring and monitoring these changes over time to derive comparative SOC data, as well as to facilitate interpreting and reporting the results. Guidelines, current limitations, and research needed to improve soil C measurements are available [Lal et al. 2001; Smith 2003]. Recently mid-infrared (MIR) and near-infrared diffuse reflectance has been tested to measure soil C inventories on landscapes. MIR appears more promising but it shows limitations in the presence of carbonates [McCarty et al. 2002]. Table 8.5 presents an overview of the main research issues (needs and gaps) related to soil C cycling in agroecosystems (terrestrial ecosystems) [Smith 2003; Nguyen 2007].
8.5.2 Isotopic Techniques in C-Cycling Studies C isotope techniques can be used to measure C-cycling processes and identify the main factors that influence the pools, fluxes, and balances in agroecosystems. This information can be also employed to assess the effectiveness of judicious land use and the RMPs for enhancing soil C sequestration in agroecosystems. However, in order to optimize the use of C isotopes in these studies, it is important to understand the main factors affecting soil C sequestration and how the processes involved relate to the global C cycle, since the C pools and flows are interlinked, and ultimately influence the adaptation of agriculture to climate change through improved soil health and resistance to degradation, as well as through reducing the soil C emissions as greenhouse gases to the atmosphere [IAEA 2010]. The SWMCN Subprogram of the Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture—in its efforts to enhance productivity and sustainability of agroecosystems through an integrated approach to soil, water, and nutrient management using nuclear and isotopic-based techniques—has produced a technical document entitled Guidelines on the Use of Carbon Isotopes in Investigating Soil Carbon Sequestration in Agroecosystems (in press). This volume of guidelines is organized into five chapters. The first two chapters provide insights on soil C sequestration. The next two chapters deal with assessment methods of soil C sequestration and the decomposition of organic residues using C isotopic techniques (Chapter 3) and, in particular, C isotope labeling procedures and safety issues when using radioactive C (Chapter 4). The final chapter contains a suite of selected case studies on the application of C isotopic techniques to demonstrate to scientists from developing countries how these tracer technologies can be used in soil C management research to improve soil quality in agricultural systems, while increasing their productivity and sustainability, thus contributing to the arrest of land degradation, enhancing food security, and mitigating climate change [IAEA 2010]. The approaches and implications for the mass balance measurements of C loss to assess soil C sequestration over a period of time are examined in the introduction
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
383
to Chapter 3. Five different carbon (stable 13C or radioactive 14C) isotope methods are presented, i.e., tracer-labeling techniques, 13C natural abundance techniques, 14C (radiocarbon) dating techniques, bomb 14C (carbon) techniques, and SOM models. They can all be employed to determine the SOM decomposition or turnover rate, a key element to assessing soil C sequestration over a period of time. Detailed descriptions of these techniques, including the advantages and limitations and key references are available in the guidelines. The 13C natural abundance techniques have been widely used to measure SOM turnover in agroecosystems. However, care should be taken in these soil organic C allocation studies based on C3 and C4 isotopic signatures of organic matter derived from C3 and C4 plants because it has been found that C4-derived SOC decomposes faster than its C3 counterpart in mixed soils [Wynn and Bird 2007]. C4 plants such as maize, sugar cane, and saltbush absorb more 13C than C3 plants such as barley, potatoes, rice, soybean, tomatoes, and wheat. The δ13C (i.e., 13C/12C) in C3 plants ranges from −22 to −30 parts per thousand, while δ13C in C4 plants ranges from −8 to −11 parts per thousand. Tracer-labeling techniques in these studies have also been applied using either the stable 13C [Stewart and Metherell 1999; Hood et al. 2004] or the radioactive 14C [Saggar et al. 1996; Saggar and Searle 1995; Saggar and Hedley 2001]. Other studies involve the use of radiocarbon dating [Hedges 1987] and the bomb 14C to estimate turnover rates of SOM [Balesdent 1987; Scharpenseel et al. 1989; Levin and Hesshaimer 2000; Baisden et al. 2002]. A recent review highlights the potential role of radiocarbon for establishing soil carbon balance over long timescales and larger spatial scales (regional studies). The information obtained can be used to test long-term models of soil carbon dynamics in ecosystems [Trumbore 2009]. The mechanistic studies are focused on C-cycling process dynamics and driving factors to gather essential knowledge for better understanding and management of SOC sequestration in agricultural lands. Table 8.6 contains reviews and key applications of C isotopes in investigations related to SOC sequestration in agroecosystems. As these studies are preferentially conducted in situ in agroecosystems, a majority of them rely on stable C isotopes (natural abundance and labeling techniques, see Section 8.2.2). The development of innovative techniques and related mass spectrometric methods for CSIA after the 1990s was a milestone in environmental research [Hayes et al. 1990; Lichfouse 2000]. Currently these techniques have been developed further for investigating transformations and turnover of specific C compounds in soils and ecosystems [Glaser 2005]. Stable isotope probing (SIP) is a very powerful technique in microbial ecological research. It utilizes DNA or RNA labeled with 13C and enables researchers to study the substrate use and identify in situ active populations in the ecosystem. Recent developments have enabled the combined use of stable C isotope techniques with biomarkers and SIP, thus increasing the potential of unravelling the role of biodiversity on soil C cycling and improving the understanding of soil-vegetation relations in global C modeling [Staddon 2004; Amelung et al. 2008]. Recent reports highlight the need for better understanding of C budgets and the relationship between C pools and soil quality in a range of environments. As the biogeochemical cycle of C and its coupling with those of N, P, S, and water is highly complex, and there is a need to integrate novel C- and nutrient-tracking techniques for
384
World Soil Resources and Food Security
TABLE 8.6 List of Selected Applications of Carbon Isotopes in Investigations Related to SOC and Soil Quality in Agroecosystems Issue/Topic Impacts of land-use changes (C3 and C4 plants) on SOC sequestration in agricultural soils in northern Italy Impacts of conventional or zero tillage on changes in soil nitrogen and carbon stocks in southern Brazil The influence of the SOC changes with time of cultivation (long- and medium-term turnover rates) in temperate Europe Relative contribution of vegetation types (C3 trees and C4 crops) to soil carbon reserves in West Africa savannah Validating the ROTHC SOM model using the changes of SOC and their 13C signatures in a long-term alley cropping field trial at Ibadan, southwestern Nigeria SOM pools and 13C signatures in particle size fractions of a long-term agricultural field experiment receiving organic amendments Studies using 13C-enriched plant material on the entry of new carbon into SOC pools and its turnover Characterization of SOC forms by 13C NMR and 15N NMR spectroscopy Analytical approaches for characterizing SOM, including CSIA Lignin turnover studies using CSIA of 13C natural abundance and artificial labeled biomass Sediment tracking studies in estuarine areas using CSIA techniques Source partitioning using stable isotopes Use of carbon isotopes in functional soil ecology Stable isotopes and biomarkers in microbial ecology Linkage of microbial populations to specific geochemical processes by 13C labeling of biomarkers Stable isotope probing in microbial ecology: Coupling molecular (DNA and RNA) and compound-specific stable isotope analysis Methods for compound-specific stable C isotope analysis in soil carbon studies, CSIA Current applications of CSIA techniques in ecosystem research studies of biomarkers (molecular markers) and soil constituents of environmental interest such as black carbon and polycyclic aromatic carbons
References Del Galdo et al. 2003
Sisti et al. 2004 Balesdent et al. 1987, 1988, 1990; Balesdent and Mariotti 1996 Traore et al. 2004; Bayala et al. 2006
Diels et al. 2004
Gerzabek et al. 2001
Van Kessel et al. 2000; Leavitt et al. 2001; Van Groenigen et al. 2002; Wiesenberg et al. 2008 Novotny et al. 2006; Smernik and Baldock 2005a, 2005b Kogel-Knabner 2000 Rasse et al. 2006; Heim and Schmidt 2007 Gibbs 2008 Phillips and Gregg 2003 Staddon 2004 Boschker and Middelburg 2002 Boschker et al. 1998 Radajewski et al. 2000; Staddon 2004
Glaser 2005 Glaser 2005; Zech and Glaser 2008
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
385
TABLE 8.6 (Continued) List of Selected Applications of Carbon Isotopes in Investigations Related to SOC and Soil Quality in Agroecosystems Issue/Topic Combined use of biomarkers and compound-specific stable isotope analysis for assessing the fate, mean residence times (MRT) in particular, of individual SOM compounds Intramolecular, compound-specific, and bulk carbon isotope patterns in C3 and C4 plants: a review and synthesis Radiocarbon and soil C-cycling studies
References Amelung et al. 2008
Hobbie and Werner 2004
Hedges 1987; Levin and Hesshaimer 2000; Trumbore 2009
quantifying C pools, fluxes, and residence time in agroecosystems. This comprehensive information will provide the basis for assessing the potential to mitigate climate change through C sequestration in agroecosystems and to provide guidelines for sustainable management of land and water resources [Nguyen 2007]. In such studies of C fluxes in predominant agroecosystems, two main approaches have been proposed for further investigation: 1) the use of stable isotope (natural abundance and CSIA) techniques to assess the net rate of C sequestration and nutrient/water dynamics in agroecosystems, and 2) the use of fallout radionuclides to assess rates of soil redistribution at different temporal and spatial scales and determine eroded C mobilization, transfer, and storage within watersheds. The combination of both approaches and their integration with non-nuclear/isotopic techniques and modeling will provide the basis for establishing the comprehensive C budgets needed for climate change mitigation and sustainable land and water management [Nguyen 2007]. Several research groups have focused their attention on black carbon and biochar studies to elucidate its quality and its role in the C cycle in agroecosystems. 8.5.2.1 Black Carbon Black Carbon (BC) is a form of carbon produced by incomplete combustion of fossil fuels, biofuel, and biomass. Thus, BC exists as a continuum from partly charred material to highly graphitized soot particles [Kuhlbusch 1998]. It is reported that >80% of BC produced ends up in the soil, where it can influence the extent of C storage and plant nutrient availability in soils. In order to understand the potential C storage in soils, it is important to know mechanisms involved in the degradation and stabilization of BC. The chemical and thermal stability (and hence longevity) of BC is either due to chemical recalcitrance that results from its aromatic structure or physical protection due to its surface functionality and binding with minerals and other organic compounds [Brodowski et al. 2005, 2007; Lehman et al. 2005, 2006; Forbes et al. 2006]. Many analytical methods are available involving the removal of the non-BC components from the sample by thermal or chemical means or a combination of both.
386
World Soil Resources and Food Security
The remaining C, operationally defined as BC, is quantified via mass balance, elemental composition, or by exploiting benzene carboxylic acids as molecular markers or applying 13C MAS NMR (magic angle spinning nuclear magnetic resonance) spectroscopy [Schmidt et al. 2001; Hammes et al. 2007, 2008; Schmidt 2007]. 8.5.2.2 Biochar Biochar is a charcoal (fine-grained, biomass-derived BC) produced by the thermal treatment (heating) of natural organic materials (e.g., crop waste, wood chip, municipal waste, organic wastes, biofuel crops, or manure) in an oxygen-limited environment. Due to its aromatic structure, biochar is chemically and biologically more stable compared with the carbon from which it was made (i.e., uncharred biomass), making biochar difficult to break down [Baldock and Smernik 2002]. This means that in some cases it can remain stable in soil for hundreds to thousands of years. The production of biochar via pyrolysis also yields bioenergy in the form of synthesis gas (or ‘syngas’) such as carbon monoxide, carbon dioxide, hydrogen, methane, and higher hydrocarbons, which in turn can be used to produce heat and power. As a soil amendment, biochar provides the following benefits [Lehmann et al. 2003; CSIRO 2010; UNCCD 2008]:
1. Improves fertilizer use efficiency because of its high affinity (adsorption) for nitrate and phosphate. This in turn results in a reduction of phosphorus (P) in runoff into surface waters and nitrate leaching losses to groundwater. 2. Enhances soil physical and biological properties and soil water-holding capacity. 3. Reduces soil N2O and CH4 emissions.
The benefits of biochar as a soil amendment depend on the quality of the biochar and the quality of the soil [Steiner et al. 2007; Biochar Fund 2008]. The quality of the biochar is greatly affected by the pyrolysis process (temperature and solids residence time) and the type of natural organic materials used in biochar production. With regard to the quality of the soil, highly degraded and nutrient-poor soils usually show increased crop production after biochar application, whereas fertile and healthy soils do not always yield a positive change. In addition to its soil amendment benefits, biochar can help combat climate change via C sequestration. Producing biochar and bioenergy via pyrolysis is a C-negative (reducing carbon emission from biomass) process because a substantial portion of the C that has been captured by photosynthesis is retained by biochar systems, instead of being returned to the atmosphere during combustion. The organic materials being burnt are naturally part of the photosynthesis cycle, so taking the C out of the cycle and locking it in biochar and biogases means that there is a net decrease of C in the atmosphere. Thus biochar sequestration is C-negative, serving as a net withdrawal of atmospheric carbon dioxide, which is then stored in highly recalcitrant soil C stocks [Lehmann 2007; Gaunt and Lehmann 2008]. The environmental implications of the use of biochar as amendment are reviewed by Woolf [2008]. A wide range of isotopes, nuclear and conventional analytical techniques, can be used in the characterization of BC properties as follows:
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
387
• Compound-specific stable-isotope (δ13C) analysis by GC-IRMS [Glaser 2005]. 13C solid-state NMR [Novotny et al. 2006] and 15N NMR spectroscopy [Smernik and Baldock 2005a, 2005b] for detailed characterization of BC-rich samples. • High-resolution online determination of δ13C in soil (thin sections) and plant (tree rings) samples, combining laser ablation (LA), combustion (C), gas chromatography (GC), and IRMS (LA-C-GC-IRMS). This technique can be used to understand the distribution of newly formed organic C in the soil (13C-labeling experiments). • Carbon isotope geochemistry and nanomorphology (cell structure level) of soil BC by NanoSIMS [Schmidt et al. 2002]. • Nuclear-based analytical techniques can be used for the characterization of BC. They include x-ray diffraction (XRD), scanning electron microscopy energy dispersive spectroscopy (SEM/EDX), electron microprobe analysis (EPMA), Raman spectroscopy (microscopy), Fourier transform infrared (FTIR) spectroscopy in combination with thermal analysis, and near-edge x-ray absorption fine structure (NEXAFS) spectroscopy for mapping nanoscale distribution of organic and BC in soil [Lehmann et al. 2005]. With regard to the studies on C dynamics and turnover in soil and sediments and the processes involved, several techniques are available: • Stable C isotope tracer techniques (natural abundance and enriched) to study soil C dynamics over short-term periods and small (point and soil profile) scales [Glaser 2005]. • Radiogenic carbon (C-14) isotope techniques for BC dating and to establish regional soil C balances over long timescales and larger spatial scales. The information can be used to test long-term models of soil C dynamics in ecosystems [Trumbore 2009]. • Stable isotope probe techniques can assist in the assessment of microbial functions in the context of BC dynamics studies [Radajewsky et al. 2000]. For the biochar investigations, in addition to the techniques mentioned above for BC research, there is a great potential to utilize a suite of isotopic techniques to study the ameliorating effects of biochar application to tropical lands. One recent development is the use of high time resolution instrumentation based on WS-CRDS, which can be deployed in remote unattended locations for long-term continuous in situ isotopic carbon dioxide (CO2) measurements, enabling the observation of diurnal and seasonal trends in the biosphere–atmosphere CO2 exchange processes occurring in ecosystems. This information helps validate C transport and diffusion models of terrestrial ecosystems and is key to understanding dynamics influencing global atmospheric C budgets [Van Pelt et al. 2008]. Tunable diode laser absorption spectroscopy (TDLAS) is another technique used for the gas-phase stable isotope analysis in studies of ecosystem–atmosphere CO2 exchange [Bowling et al. 2003; Schaeffer et al. 2008]. In combination with micrometeorological techniques, it has been applied for measuring field-scale isotopic CO2 fluxes [Griffis et al. 2004].
388
World Soil Resources and Food Security
Frequency modulated spectroscopy (FMS), also based on TDLAS is an ultrasensitive means of detecting isotopes of CO2 and is being applied for tracking CO2 geological sequestration seepage at or below ambient CO2 concentrations [Fessenden and Clegg 2008]. The following sections provide an insight into relevant topics and issues for soil C sequestration studies using nuclear and isotopic techniques. Some studies may contain intertwined and crosscutting issues; for instance, the impacts from soil erosion on net SOC sequestration are poorly understood. Soil erosion affects not only the size and quality of the SOC pool and fluxes, but also influences GHG emissions at different scales in agroecosystems. While the measurement of soil erosion will provide practical information to select the most effective RMP to control soil C losses and deposition, the mechanistic studies will be mainly devoted to gaining knowledge and better understanding of the processes and drivers involved in the C dynamics as related to removal, transport, and deposition, thus providing the knowledge base for better management of soil organic C in agricultural lands.
8.6 L AND USE, MANAGEMENT, AND SOIL C SEQUESTRATION IN AGROECOSYSTEMS 8.6.1 Introduction World soils constitute the third largest global pool comprising 1550 Gt (i.e., Gt = giga ton; 1Gt = 109 tons and 1 ton = 1000 kg) of SOC and 950 Gt of soil inorganic C to 1-m depth. This soil C pool of 2500 Gt is 3.3 times the biotic pool, and 4.5 times the atmospheric pool and plays an important role in the global C cycle. Changes (gains and losses) in soil C in terrestrial ecosystems result from the balance between inputs from the net primary productivity (NPP) and the organic matter (OM) decomposition rate and additional losses from erosion, combustion, and other processes [UNESCO-SCOPE 2006]. Each of these mechanisms has human and natural drivers (i.e., deforestation, changes in energy consumption, changes in hydrology, desertification). Land-use changes as influenced by climate–human interactions can result in significant pulses of soil C. Hydrological changes may also have a pronounced impact on the C fluxes from peat lands and wetlands [Gruber et al. 2004]. Another important strategy among the technological options to mitigate global warming resulting from climate change is the management of world soils to sequester atmospheric CO2 through a dual approach of enhancing C storage in the soil improving soil quality and mitigating GHG emissions from agricultural lands. Changes in management practices of agricultural production systems such as biomass and yield increases due to reduction in tillage, fertilization and irrigation, mulching, incorporation (no burning) of crop residues, use of cover crops, and expansion of areas under perennial crops can all contribute to C sinks [Lal 2001a, 2001b, 2007; FAO 2010b, 2010c). Therefore, sound management of the SOC is the key to the productivity and sustainability of agroecosystems. In this context, it is of great importance to develop strategies and adopt practices to enhance soil C sequestration in agricultural lands.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
389
8.6.2 A ssessment of Recommended Practices for SOC Management in Agroecosystems Appropriate and strategic management of SOC to enhance soil quality can be achieved through the adoption of the so-called recommended management practices (RMP), which are a set of judicious practices based on fairly universal good agricultural principles (Table 8.7). They have been formulated in generic terms according to the degree of soil degradation [Lal 2002, 2004]. However, there are no blueprints and technologies specific for land degradation class and agroecosystem or farming system need to be assessed and pilot tested in terms of biomass production, SOC changes, and resource and energy use efficiency. Moreover, the impact of recommended land and water management practices on carbon sequestration, fluxes of GHGs, and other elements needs to be defined and better understood before cost-effective and efficient land and water management practices can be envisaged [Nguyen 2007]. Some guidelines suggested for such an assessment are as follows:
1. The first step is to assess the status of human induced soil degradation according to the information provided in the National Soil Degradation Maps (area, location, extent, and severity degree) of the GLASOD study [FAO 2010a]. 2. Select the farming or watershed system with similar soil type and degree of degradation severity. According to the GLASOD study, there are six classes of degradation degree and reduction in productive capacity. Five are set in order of increasing severity from none, light, moderate, severe, and very severe. One extra category is assigned as non-classified due to the lack of data. It should be noted that this classification is only an approximation and there is a need for a better definition of degree of severity. Additional information on the GLASOD assessment methodology and soil degradation types and causes is available [FAO 2010a, 2010f]. 3. For the purpose of the selection and application of the RMPs, it is suggested to consider the classes none and light as roughly equivalent to prime agricultural or arable croplands; the moderate class may be equivalent to marginal lands, and severe and very severe belong to the category of degraded lands. 4. In order to quantify and select changes in SOC and/or the soil quality index in agroecosystems, the steps or procedures proposed by Arshad and Martin [2002] can be followed.
Integrated studies involving the use of NIT will also be included to gather the following information: • Quantitative data on C fluxes��������������������������������������������� and residence time�������������������������� ������������������������� and budgets ������������� under different soil and crop management practices in site-specific environments • Information on incremental changes in the C pool at different temporal and spatial scales
390
World Soil Resources and Food Security
TABLE 8.7 Recommended Management Practices (RMP) to Enhance SOC and Develop or Improve Soil Quality Land Prime agricultural lands
Marginal lands
Management
Practices
Effects/Impacts
Maintaining and improving soil quality through the adoption of best RMP of conservation agriculture for sustainable intensification of agricultural production
1. Conservation and zero tillage with mulching; direct planting mulch-based cropping (DMC) systems
Conservation of natural resources soil and water; erosion control; favorable soil moisture and temperature
2. Improved cropping systems (crop rotations including best adapted crop varieties, ley pastures) and appropriate agroforestry systems 3. Integrated soil fertility management (chemical fertilizers and crop residues, biological nitrogen fixation, below-ground nutrient contribution) 4. Soil cover using cover and green manure crops
Better growth; high biomass; better use efficiency of water and nutrients; deeper root systems
Taking land out of production and improving soil quality through ecological land management to enhance resource use efficiency and increase NPP while protecting the environment
5. Water management irrigation and water-saving technologies; water conservation 1. Conservation reserve and water reserve programs 2. Restorative land use: • Natural vegetation regrowth • Afforestation • Planting vegetation such as grasses, shrubs, and multipurpose fast growing trees tolerant to relevant abiotic stresses • Improved agroforestry parklands and agrosilvopastoral systems
Improved soil fertility levels and nutrient balance and cycling; higher biological and agronomic production Erosion control; nutrient cycling; forage; high biomass Improved soil water storage and availability to plants; crop productivity Conservation and increased use efficiency of natural resources (soil and water); erosion control; nutrient cycling; biomass production; enhancing biological activity; forage and biofuel production
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
391
TABLE 8.7 (Continued) Recommended Management Practices (RMP) to Enhance SOC and Develop or Improve Soil Quality Land
Management
Practices
Effects/Impacts
• Conservation buffer zones and plant species with deep rooting systems • Land application of organic wastes (biosolids and others) Degraded lands
Land reclamation and restoration for developing soil quality through ecological land management to enhance resource use efficiency and increase NPP while protecting the environment
1. Soil reclamation (combined use of amendments and adapted plant genotypes) 2. Biological rehabilitation (land use and vegetation change) 3. Effective soil conservation measures 4. Wetlands restoration and riparian zones 5. Land disposal of biosolids and filling 6. Forming soil and land
Conservation and increased use efficiency of natural resources (soil and water); erosion control; nutrient cycling; biomass production; enhancing biological activity; forage/biofuel production
• Reliable information on efficiency of improved land and water management practices for mitigating climate change and enhancing food and bioenergy production. Conservation agriculture (CA) technologies are RMPs mainly recommended for sustainable agriculture in prime agricultural lands. The basic elements are: 1) minimum soil disturbance through no till or conservation tillage; 2) permanent soil cover due to crop residue recycling and mulching and the use of cover crops; and 3) diversified farming systems including crop rotations of annual crops and plant associations of perennial crops (agroforestry and agrosilvopastoral systems) [FAO 2010d]. A comparative review of conservation tillage in tropical and temperate environments has been made by Lal [1989]. Currently, the extent of no till adoption worldwide covers some 95.5 millions ha [Hobbs et al. 2008]. This combination of practices, when applied on a continuous basis over a period of time, provides a means of enhancing soil C sequestration on the best agricultural lands and thus, sustaining and increasing SOM levels and crop production and improving environmental quality [Lal et al. 1998; Lal 2007]. They are also proposed as an adaptive and mitigation strategy to combat climate change [PACA 2009]. According to the FAO [2010e], better farming practices through CA, among many benefits, could help to bury about 10% of the atmospheric C from emissions caused by human activity over the next 25 years.
392
World Soil Resources and Food Security
Research is, therefore, needed to identify and pilot test CA-specific packages of integrated technologies and practices tailored to targeted agroecosystems and local conditions. Furthermore, this information would be essential for a comprehensive assessment of socioeconomic and environmental benefits required to develop appropriate policies to facilitate their adoption by farmers. While initial CA investigations have focused on improving soil fertility and increasing agronomic crop yields, much research is needed to assess the CA benefits in terms of resource (water and fertilizer inputs and energy) use efficiency, as well as their environmental impacts (soil carbon sequestration and ecosystems services). In this context, NITs have an important role to play in gathering such relevant information, as mentioned above. Currently a CRP on integrated soil, water, and nutrient management for conservation agriculture using NIT is being implemented to investigate the individual and interactive effects of the main CA practices on carbon and nutrient (N and P) dynamics, water balance and its use efficiency, and soil erosion losses [FAO/IAEA 2009].
8.7 A REA-WIDE (WATERSHED) STUDIES ON SOIL, SEDIMENT, AND SOC REDISTRIBUTION AND IDENTIFICATION OF SEDIMENT AND CARBON SINKS IN A WATERSHED 8.7.1 Introduction Soil erosion (removal) and associated deposition (accumulation) can cause redistribution (mobilization and transport) in a differential manner of both soil particles and SOC along the landscape, and ultimately may result in C losses from the watershed as emissions of CO2 and CH4 and/or deposition/burial in sediment sinks and neighboring aquatic systems. According to estimates of Lal [1995], the average total transport or movement of C displaced by the soil erosion is 1.59 Gt yr−1 ranging from a low of 0.80 Gt yr−1 to a high of 2.40 Gt yr−1. It is also assumed that the sediment delivery ratio may be as low as 10%, thus on average 0.16 Gt yr−1 may be transported out from tropical watersheds. Erosion control in tropical lands has a potential for soil C sequestration of 10 to 25 Gt over a 50-year period [Lal 1995]. Terrestrial sedimentation and land-ocean transfer of suspended sediment are important components of the global C budget. Current estimates of the particulate organic C flux associated with land-ocean suspended sediment transfer are 0.2 to 0.3 Gt yr−1 [Stallard 1998]. Some authors [Bajracharya et al. 1998; Lal 1995] have reported that, while deposition of eroded soil does not necessarily lead to the direct accumulation of SOC, it is likely to increase the overall soil C sequestration by leading to an accumulation of organic material with greater potential to be stabilized as SOC in the aggregates. Thus, the presence in the watershed of depositional and non-eroded sites may increase the potential for SOC accumulation, probably through soil aggregation. However, erosion would lead to a gradual depletion of the SOC through aggregate destruction and continuous removal of soil particles. Studies on SOC redistribution highlight that it is important not only to assess C storage but also its biodegradability, i.e., permanence or stability [Lal 2004]. Walling et al. [2006a], investigating carbon accumulation associated with overbank deposition in the floodplains of six rivers in southern England, observed
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
393
appreciable spatial variability in SOC content of overbank sediment deposits, both between rivers (values ranging between 2.17% and 5.07%) and between individual sites along a floodplain. Estimated mean values of soil C sequestration rate (upper 24 cm) obtained for the floodplains of the individual rivers ranged from 69.2 and 114.g m−2 yr−1, thus confirming their significant capacity as C sinks. These studies indicate that there is a need for a comprehensive assessment of the magnitude of sediment and SOC redistribution in erosional and depositional sites and their accumulation and isotopic signatures in soil aggregates and/or specific particle size fractions. Whenever possible, a watershed sediment budget should be established. Controlling erosion-caused soil degradation, in particular soil quality decline and associated off-site detrimental impacts in the watershed/catchment and connected ecosystems is a major challenge for productivity and sustainability of agroecosystems [OECD 2003, 2004].
8.7.2 Use of Environmental Radionuclides in Soil Erosion Research in Agroecosystems Substantial progress has been made using fallout, also called environmental radionuclides (e.g., 137Cs and 210Pb), to estimate soil redistribution at the watershed scale and area-wide erosion and sedimentation through the implementation of two CRPs on the topics [Zapata 2002a, 2003, 2004]. A significant number of research groups distributed worldwide across developed and developing countries now have the capacities to conduct such investigations and a wealth of information has been produced over the past 10 years [Ritchie and Ritchie 2009]. The information on soil redistribution rates and patterns obtained using environmental radionuclides have been applied to derive improved understanding of soil loss and gains related to other parameters such as soil quality and soil organic redistribution in the landscape (Table 8.8). In a recently terminated CRP the combined use of the 137Cs technique and other environmental radionuclides has been applied to measure soil redistribution at several spatial and temporal scales in a number of environments [Mabit et al. 2008b] and identify practical and cost-effective soil conservation measures at both farm and catchment scales [Zapata and Li 2007]. Two main approaches to C fluxes have been identified for further investigation: 1) the use of stable isotope techniques to assess the net rate of C sequestration and nutrient dynamics in agroecosystems, and 2) the use of FRNs to assess soil and C redistribution (mobilization, transfer, and storage) within watersheds over different timescales. The combination of both approaches and their integration with non-nuclear/isotopic monitoring techniques will provide the basis for establishing comprehensive C budgets needed for sustainable management of land and water resources [Nguyen 2007]. Current concerns about agricultural land-use impacts on the environment have highlighted the important role of sediment (and associated nutrients and chemicals) in degrading water quality and in a range of other environmental problems [Horowitz and Walling 2005]. Sediment transport is the key to understanding the movement and fate of many nutrients (e.g., nitrogen and phosphorus) and contaminants [McDowell and Sharpley 2003] and organic carbon and mobilization within the watershed [Lal 2007]. A networked research project on precision conservation
394
World Soil Resources and Food Security
TABLE 8.8 Selected Applications of Fallout Radionuclides in Soil Erosion and Sedimentation Studies Related to SOC and Soil Quality Issue/Topic The relationship between soil quality parameters and soil erosion as inferred from 137Cs redistribution The relationship between 137Cs-derived soil redistribution and SOC distribution patterns
The influence of tillage in soil redistribution
Tracing sources of suspended sediment in river basins The contribution of sediment and sediment-P from erosion processes and land-use types in a mixed land-use catchment Identification of spatial patterns of soil erosion for use in precision conservation studies Impacts of severe bush fires on redistribution of soil and organic material Sediment budget studies using fallout radionuclides Validation of spatially distributed soil erosion and sediment yield models
References Pennock 1998, 2000, 2003; Bacchi et al. 2000, 2003; Li et al. 2000; Fulajtar 2000, 2003; Theocharopoulos et al. 2000 Pennock and Frick 2001; McCarty and Ritchie 2002; Ritchie and McCarty 2003; Venteris et al. 2004; Li et al. 2006, 2007a, 2007b; Mabit et al. 2008a; Mabit and Bernard 2009 Lobb et al. 1995; Govers et al. 1996; Li and Lindstrom 2001; Li et al. 2007a, 2007b; Schuller et al. 2003, 2004, 2007; Zhang et al. 2003; Zheng et al. 2007 Sogon et al. 1999; Wallbrink et al. 1999; Walling 2005 Wallbrink et al. 2003
Schumacher et al. 2005 Wallbrink et al. 2005 Blake et al. 2002; Walling et al. 1998; 2006b Walling et al. 2003
has recently begun studies to identify and target critical source areas of soil loss and sediment production in catchments and thus reduce sediment-associated pollution risks [Gibbs 2008]. There is also potential for identifying sink and sources of organic C in the watershed. In view of the number of issues to be tackled and its complexity within the watershed, a suite of isotopic (environmental radionuclides and compound-specific stable isotope analysis) and non-nuclear techniques will be used to obtain the information required for the use and calibration of selected models [Nguyen 2008; Zapata and Nguyen 2009].
8.8 MITIGATING GHG EMISSIONS IN AGROECOSYSTEMS 8.8.1 Soils as Sinks and Sources of GHG It is now widely accepted that the rising anthropogenic emissions of GHGs in the atmosphere are contributing to the global warming potential and climate change, becoming one of the most pressing environmental problems and a key issue for global sustainable development. GHGs differ greatly in their radiative forcing effect.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
395
The major GHGs are carbon dioxide (CO2), nitrous oxide (N2O), and methane (CH4). Over a 100-year time horizon, N2O is estimated to have approximately 310 times the effect on climate as CO2 and CH4, roughly 21 times the effect of CO2. When comparing the impact of GHG emissions, the emissions of different gases are translated into CO2 equivalents. On the basis of CO2 equivalents, CO2 emissions contribute about 82%, CH4 emission about 10%, and N2O about 7% of the total global warming potential of the GHGs emissions in the OECD countries. The share of agriculture in total GHGs emissions from the OECD countries in CO2 equivalents is less than 8%, while for CH4 and N2O, agriculture emissions contribute about 40% and 60%, respectively [OECD 2002]. Currently there is available information on country emission inventories following guidelines of the Intergovernmental Panel on Climate Change (IPCC) but a considerable range of uncertainty (variable measurement protocols and reporting sources) on the data exists [IPCC 2001]. Monitoring the role of agriculture as a source and sink for GHGs is a matter of concern to both farmers and policy makers. It is generally accepted that agriculture is an important contributor of emissions of the three most important GHGs: CO2, N2O, and CH4. Carbon dioxide emissions are primarily due to land-use changes, in particular deforestation, intensive land tillage, or fossil fuel combustion from agricultural operations. N2O emissions occur mainly as a result of transformations of mineral nitrogen in agricultural soils under crop cultivation using manure and synthetic fertilizers, while CH4 emissions are generally related to livestock and rice production [IAEA 1992; OECD 2002]. Soils are an important source of the three major GHGs, but the emission rates are strongly dependent on their use and management. If appropriate management practices are used in agricultural soils they can also act as a sink, contributing to reduce the GHGs emissions to the atmosphere and mitigate the impacts of climate change. Thus, in order to assess the net rate of C sequestration in agroecosystems, it is necessary to measure both the GHGs emitted and the changes in soil C storage over time [Lal 2001a, 2001b, 2004; Nguyen 2007]. These aspects are key topics of environmental research. Extensive reviews, books, and reports of meetings/workshops on GHG emissions from agroecosystems at the laboratory, field, country, regional, and continental scales are available. They report on the progress made in the measurement approaches and techniques for measuring GHG emissions, levels produced, and main influencing factors and environmental conditions.
8.8.2 Selected Applications of Isotope Techniques in GHG Emission Studies in Agroecosystems Stable isotope techniques can be used in either basic research to investigate the processes of GHG production in soils and gain a better understanding of the main factors influencing the GHG production and consumption processes, or in applied research to assess the value or effectiveness of selected management practices on the mitigation of GHG emissions in agroecosystems. All this knowledge is essential not only to policymakers for formulating adequate local regulations and policies to meet international commitments to reduce atmospheric concentrations of GHGs, but also
396
World Soil Resources and Food Security
TABLE 8.9 Selected Applications of Isotopic Techniques in Greenhouse Gas (GHG) Studies Isotopic Techniques
GHG N2O
Natural abundance (15N and 18O isotopic signatures)
N natural abundance and enrichment methods
15
N enrichment methods
15
N and 18O enrichment
15
CO2
Natural abundance (13C and 18O)
Experimental Work Soil microbial processes affecting nitrous oxide (N2O) emissions and factors influencing nitrification and denitrification in tropical Amazon rain forest soils Factors affecting nitrous oxide (N2O) emissions from extensively managed grassland in Bavaria, Germany Factors affecting soil nitrogen dynamics and N2O emissions in European grasslands Extent of N2O and N2 emissions from soil and those produced by nitrifying and denitrifying bacteria Study of N2O topsoil fluxes and the dynamics of N2O formation, consumption, and emission Soil N2O emissions and global N2O budgets Five 15N approaches used for estimating denitrification in soils: 1) variations in 15N natural abundance; 2) mass balances using 15N-labeled fertilizers; 3) use of 15NO in 3 isotope dilution and modeling of N cycling; 4) addition of 15NO3 and measuring 15N2 gas emissions; and 5) 15N2 gas isotope dilution studies 15N-based methods used for estimating denitrification include isotope fractionation, isotope dilution, 15N mass balances, and direct measurement of 15N-labeled gases upon addition of 15N-nitrate- and 15N-ammoniumlabeled salts. The addition of 15N-labeled salts to achieve high levels of enrichment increase the availability of N and may result in overestimation of denitrification. Use of 18O-labeled water and 15N-labeled ammonium nitrate to distinguish respective N2O contributions from i) nitrification, ii) nitrifyer denitrification, and iii) denitrification 15N-labeled nitrate to estimate denitrification in sediments 13C signatures of respired CO to investigate 2 SOC dynamics in grassland under elevated CO2
References Perez et al. 2003
Tilsner et al. 2003
Wrage et al. 2004 Webster and Hopkins 2004 Van Groningen et al. 2005 Kim and Craig 1993 Myrold 1990
Groffman et al. 2006
Wrage et al. 2005
Nielsen 1992 Pendall and King 2007
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
397
TABLE 8.9 (Continued) Selected Applications of Isotopic Techniques in Greenhouse Gas (GHG) Studies Isotopic Techniques
GHG CO2
Natural abundance (13C and 18O)
C natural abundance and 13C enrichment
13
C enrichment
13
CH4
C natural abundance
13
Enriched products (13C–CH4 and 14NH 15NO , 4 3 15NH 15NO ) 4 3
Experimental Work C signatures in air, plant, and soil samples to assess SOC storage and dynamics in relation to rising air CO2 concentrations 13C signatures in air, plant, and soil samples to assess the relative contributions of root and soil heterotrophic respiration to total soil respiration in situ Interannual variations 18O signatures in atmospheric CO2 from assimilation and respiration 18O signatures in soil water and soil CO to 2 partition evapotranspiration (ET) into its components under normal and elevated CO2 conditions in a semiarid short grass steppe 18O signatures in soil water and soil CO to study 2 isotopic equilibration of 18O between carbon dioxide and water in soils, which is a major component of the 18O isotopic balance of atmospheric CO2 13C natural abundance and 13C enrichment to study impacts of climate change on ecosystem C cycling with a focus on below-ground soil-plant responses (SOM, roots, and rhizosphere) to elevated CO2 and temperatures 13C labeling of the elevated CO to evaluate 2 changes in root biomass and C/N ratios over 5 years, and to quantify the input rate of new C or rhizodeposition in ambient and elevated CO2 treatments 13C signatures from DOC isolates found in water, peat, and agricultural drains to examine relationships with formation of trihalomethane (THM) To study the effects of elevated CO2 concentrations on the emissions of N2O and CH4 (oxidation/emission), 14NH415NO3 and 15NH 15NO enable the determination of 4 3 respective contributions of nitrification and denitrification to N2O emissions 13
References Leavitt et al. 1994 Pendall and King 2007 Andrews et al. 1999
Cuntz et al. 2003
Ferretti et al. 2003
Tans 2002
Pendall et al. 2004a
Pendall et al. 2004b
Bergamaschi et al. 1999
Baggs and Blum 2004
398
World Soil Resources and Food Security
to farmers for adoption of appropriate management practices to enhance adaptation to climate change and variability in agricultural systems. In earlier studies (2 decades ago), using stable isotope techniques, scientists were confronted with several analytical limitations, in particular a lack of adequate instrumentation to detect sizeable differences in isotopic composition (isotopic signatures) of the reactants and products or very low emission rates of the studied gas with regard to the background (for instance N2O related to the yield of N2 and the large background of atmospheric N2). For these reasons, these techniques were not commonly employed in these studies [IAEA 1992]. Currently with development of advanced instrumentation and modern electronic data acquisition and processing systems, most of these technical limitations have been overcome, but their use remains limited to groups specialized in these studies. Table 8.9 displays selected investigations and reviews to illustrate how stable isotopic (13C, 2H, 18O, 15N) techniques are employed to generate knowledge at the process level, i.e., mechanisms, emission rates, and controlling factors at several spatial and temporal scales. This isotopic information in agricultural and natural landscapes [termed isoscapes; West et al. 2010] can be further used to derive effective management strategies to mitigate GHG emissions and enhance nutrient and water use efficiencies.
8.9 CONCLUSIONS NITs are valuable tools to develop integrated soil-water-plant technologies supporting the enhancement of soil quality for productivity (biological and agronomic production) and sustainability (sustainable use and management of land and water resources), thus contributing to achieving the MDG of food security and environmental sustainability. In this context, they have the potential to address basic scientific issues and topics for better understanding of biogeochemical processes, as well as to assess the relative value and effectiveness of novel management technologies designed to remove constraints and limitations to soil quality for productivity and/or sustainability in agroecosystems. Moreover, research is needed to provide data backed by scientific evidence and to inform land-care practitioners and climate change policymakers. NITs, like other techniques, have advantages and limitations. They offer comparative advantages of high specificity, accuracy, and sensitivity over conventional non-nuclear techniques. Because of these advantages, they generate quantitative data, providing direct answers to the questions posed and thus, saving time and efforts. However, a set of preconditions is required to take advantage of the chosen NITs in terms of efficiency and cost-effectiveness, in particular the availability of skilled and trained human resources and accessibility to adequate analytical facilities. The overall assessment of these conditions for using NITs should be made by the research leader or project officer during the planning stage. NITs are not used in isolation and they must be properly inserted in national and/or regional research project plans. Their utilization must be properly framed in the research program plan from planning to generation and interpretation of data until the dissemination of results to beneficiaries and endusers, including policy- and decision-makers. Some apparent limitations can be overcome while working in networked research with
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
399
multidisciplinary and interinstitutional teams. In complex investigations at large area-wide scales conducted in agroecosystems, a suite of NITs is often employed in combination with conventional techniques, including molecular tools in biological research. Also the large set of data generated is commonly used for the creation of databases and calibration or validation as well as further development or refinement of models. Because most research is conducted at the ecosystem level, a range of stable isotope techniques is widely employed in various disciplines such as biology, chemistry, ecology, hydrology, and others related to earth sciences and in general environmental research. Because of their uniqueness and specific comparative advantages, some radiation sources such as SMNP and radioisotopes like 32P and 14C are employed under controlled conditions to address some specific research issues. In these cases, strict compliance with international and national safety regulations and radiation procedures in storage, handling, transport, and disposal is required. This overview highlights the applications of a range of promising NITs to address relevant integrated soil-water-plant research activities supporting the enhancement of soil quality for productivity (biological/agronomic production) and sustainability (sustainable use and management of land and water resources). The information provided will help researchers working in scientific research to focus on global warming and adapting issues related to sustainable land and water management and food security and environmental sustainability. Also, it will serve agricultural scientists working on enhancing productivity and ensuring food security to shift their attention to assess the impacts of agricultural activities on issues of environmental sustainability. A common key challenge for the future is to exploit the use of these techniques for mitigating the impacts and adapting to global warming from climate change. Networking, coordination, and information and communication technology exchanges will be important tools for further development and wide application of these techniques, especially in developing countries.
REFERENCES Ahnstrom, Z.A.S., and D.R. Parker. 2001. Cadmium reactivity in metal-contaminated soils using a couple stable isotope dilution-sequential extraction procedure. Environ. Sci. Technol. 35:121–126. Ajiboye, B., O.O. Akinremi, Hu Yongfen, and D.N. Flatten. 2007. Phosphorus speciation of sequential extracts of organic amendments using nuclear magnetic resonance and XANES spectroscopies. J. Environ. Qual. 36:1563–1576. Alewell, C., M.J. Mitchell, G.E. Likens, and H.R. Krouse. 1999. Sources of stream sulphate at the Hubbard Brook Exp. Forest: Long term analysis using stable isotopes. Biogeochem. 44:281–299. Amelung, W., R. Bol, and C. Friedrich. 1999. Natural 13C abundance: A tool to trace the incorporation of dung-derived carbon into soil particle size fractions. Rapid Comm. Mass Sp. 14:1291–1294. Amelung, W., S. Brodosky, A. Sandhage-Hofmann, and R. Bol. 2008. Combining biomarker with stable isotope analyses for assessing the transformation and turnover of soil organic matter. Adv. Agron. 100:155–250.
400
World Soil Resources and Food Security
Andrews, J., K.G. Harrison, R. Matamala, and W.H. Schlesinger. 1999. Separation of root respiration from total soil respiration using carbon-13 labelling during free-air CO2 enrichment (FACE). Soil Sci. Soc. Amer. J. 63:1429–1435. Arshad, M.A., and S. Martin. 2002. Identifying critical limits for soil quality indicators in agro-ecosystems. Agric. Ecosyst. Environ. 88:153–160. Ayoub, A.S., B.A. McGaw, C.A. Shand, and A.J. Midwood. 2003. Phytoavailability of Cd and Zn in soil estimated by stable isotope exchange and chemical extraction. Plant Soil 252:291–300. Bacchi, O.O.S., K. Reichardt, G. Sparovekand, and S.B.L. Ranieri. 2000. Soil erosion evaluation in a small watershed in Brazil through 137Cs fallout redistribution analysis and conventional methods. Acta Geol. Hispanica 35(3–4):251–259. Bacchi, O.O.S., K. Reichardt, and G. Sparovek. 2003. Sediment spatial distribution evaluated by three methods and its relation to some soil properties. Soil Till. Res. 69:117–126. Baggs E., and H. Blum. 2004. CH4 oxidation and emission of CH4 and N2O from Lolium perenne swards under elevated atmospheric CO2. Soil Biol. Biochem. 36:713–723. Baisden, W.T., R. Amundson, A.C. Cook, and D.L. Brenner. 2002. The turnover and storage of C and N in five density fractions from California annual grassland surface soil. Global Biogeochem. Cycles. 16:1117. Bajracharya, R.M., R. Lal, and J.M. Kimble. 1998. Soil organic carbon distribution in aggregates and primary particle fractions as influenced by erosion phases and landscape position. In Soil processes and the carbon cycle. Advances in Soil Science, eds. R. Lal, J.M. Kimble, R.F. Follett, and B.A. Stewart, 353–367. Boca Raton, FL: CRC Press. Baldock, J.A., and R.J. Smernik. 2002. Chemical composition and bioavailability of thermally, altered Pinus resinosa (red pine) wood. Organic Geochem. 33:1093–1109. Balesdent, J. 1987. The turnover of soil organic fractions estimated by radiocarbon dating. Sci. Total Environ. 62:405–408. Balesdent, J., A. Mariotti, and B. Guillet. 1987. Natural 13C abundance as a tracer for studies of soil organic matter dynamics. Soil Biol. Biochem. 19:25–30. Balesdent, J., G.H. Wagner, and A. Mariotti. 1988. Soil organic matter turnover in long-term field experiments as revealed by carbon 13 natural abundance. Soil Sci. Soc. Am. J. 52:118–124. Balesdent, J., A. Mariotti, and D. Boisgontier. 1990. Effect of tillage on soil organic carbon mineralization estimated from 13C abundance in maize fields. J. Soil Sci. 41:587–596. Balesdent, J., and A. Mariotti. 1996. Measurement of soil organic matter turnover using 13C natural abundance. In Mass spectrometry of soils, eds. T.W. Boutton and S. Yamasaki, 83–111. New York: Marcel Dekker. Barraclough, D. 1995. 15N dilution techniques to study soil nitrogen transformations and plant uptake. Fert. Res. 42:185–192. Basilio-Sanchez, P., G.J. Blair, A.R. Till, and M. Faint. 2000. Production of labelled plant materials to trace the fate of residue-carbon, nitrogen, and sulphur. In Proceedings of the International Symposium on Nuclear Techniques in Integrated Plant Nutrient, Water and Soil Management. Advances in Soil Science, ed. IAEA. October 16–20, 2000. Vienna, Austria. Bayala, J., J. Balesdent, C. Marol, F. Zapata, Z. Teklehaimanot, S.J. Ouedraogo, 2006. Relative contribution of trees and crops to soil organic carbon content in a parkland system in Burkina Faso using variations in natural 13C abundance. Nutr. Cycling Agroecosyst. 76:193–201. Bayala, J., L.K. Heng, M. van Noordwijk, S.J. Ouedraogo. 2008. Hydraulic redistribution study in two native tree species of agroforestry parklands of West African dry savanna, Acta Oecol. 34:370–378. Bergamaschi, B.A., M.S. Fram, R. Fujii, G.R. Aiken, C. Kendall, and S.R. Silva. 1999. The carbon isotopic composition of trihalomethanes formed from chemically distinct dissolved organic carbon isolates from the Sacramento-San Joaquin River Delta, California, USA. In
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
401
US geological survey toxic substances hydrology program—USGS water resources investigations report 99-40118B, eds. D.W. Morganwalp and H.T. Buxton, 99–107. http://water .usgs.gov/nrp/proj.bib/kendall.html. Berling, D.J., C.N. Hewitt, J.A. Pyle, and J.A. Raven. 2007. Critical issues in trace gas biogeochemistry and global change. Philos. T. Roy. Soc. A. 365:1629–1642. Bertrand, I., M.J. McLaughlin, R.E. Holloway, R.D. Armstrong, and T. McBeath. 2006. Changes in P bioavailability induced by liquid and powder sources of P, N and Zn fertilizers in alkaline soils. Nutr. Cycling Agroecosyst. 74:27–40. Bezzola, L.C., S.C. Lopez, and N.O. Barbaro. 1994. Effectiveness of different phosphatic fertilizers measured using labelled superphosphate and phosphorus taken up by plants. Fert. Res. 39:31–37. Biocahr Fund. 2008. Balance carbon and restore soil fertility. http://www.biochar.org/ joomla/. Blair, G.J., and R.D.B. Lefroy. 1987. Sulphur cycling in tropical soils and the agronomic impact of increasing use of S free fertilizers, increased crop production and burning of crop residues. In Proceeding of the International Symposium on Fertilizer Sulphur Requirements and Sources in Developing Countries of Asia and the Pacific, 12–17. FADINAP, FAO, The Sulphur Institute, ACIAR. Bangkok, Thailand. Blake, W.H., D.E. Walling, and Q. He. 2002. Using cosmogenic beryllium-7 as a tracer in sediment budget investigations. Geografiska Annaler: Series A, Physical Geography 82A:89–102. Bol, R., W. Amelung, and C. Friedrich. 2004. Role of aggregate surface and core fraction in the sequestration of carbon from dung in a temperate grassland soil. Eur. J. Soil Sci. 55:71–77. Bol, R., W. Amelung, C. Friedrich, and N.J. Ostle. 2000. Tracing dung-derived carbon in temperate grassland using 13C natural abundance measurements. Soil Biol. Biochem. 32:1337–1344. Bolan, N.S., R.E. White, and M.J. Hedley. 1990. A review of the use of phosphate rocks as fertilizers for direct application in Australia and New Zealand. Aust. J. Exp. Agric. 30:297–313. Bole, J.B., and U.J. Pittmann. 1984. Availability of soil sulphates to barley and rapeseed. Can. J. Soil Sci. 64:301–312. Boschker, H.T.S, S.C. Nold, P. Wellsbury, D. Bos, W. de Graaf, R. Pel, R.J. Parkes, and T.E. Cappenberg. 1998. Direct linking of microbial populations to specific biogeochemical processes by 13C labelling of biomarkers. Nature 392:801–805. Boschker, H.T.S., and J.J. Midelburg. 2002. Stable isotopes and biomarkers in microbial ecology. FEMS Microbiol. Ecol. 40:85–95. Boutton, T. W., S.R. Archer, and A.J. Midwood. 1999. Stable isotopes in ecosystem science: structure, function and dynamics of a subtropical savannah. Rapid Commun. Mass Spectrom. 13:1263–1277. Bowling, D.R., S.D. Sargent, B.D. Tanner, and J.R. Ehleringer. 2003. Tunable diode laser absorption spectrometer for stable isotope studies of ecosystem-atmosphere CO2 exchange. Agric. For. Meteorol. 118:1–19. Brodowski S., W. Amelung, L. Haumaier, and W. Zech. 2007. Black carbon contribution to stable humus in German arable soils. Geoderma 139:220–228. Brodowski, S., W. Amelung, L. Haumaier, C. Abetz, and W. Zech. 2005. Morphological and chemical properties of black carbon in physical soil fractions as revealed by scanning electron microscopy and energy-dispersive x-ray spectroscopy. Geoderma 128: 116–129. Bruinsma, J. 2003. World agriculture: Towards 2015/2030. An FAO perspective. London: Earthscan Publications. Cade-Menun, B.J., and C.M. Preston. 1996. A comparison of soil extraction procedures for 31P NMR spectroscopy. Soil Sci. 161:770–785.
402
World Soil Resources and Food Security
Cadisch, G., and K.E. Giller. 1997. Driven by nature: Plant litter quality and decomposition. Wallingford, UK: CAB International. Carter, M.R. 2002. Soil quality for sustainable land management: Organic matter and aggregation interactions that maintain soil functions. Agron. J. 94:38–47. Catchpoole, D.W., and G.J. Blair. 1990. Forage tree legumes. III. Release of nitrogen from leaf, faeces, and urine derived from leucaena and gliricidia leaf. Aust. J. Agric. Res. 41:539–547. Chalk, P.M. 1998. Dynamics of biologically fixed N in legume-cereal rotations: A review. Aust. J. Agric. Res. 49:303–316. Chalk, P.M., F. Zapata, and G. Keerthisinghe. 2002. Towards integrated soil, water and nutrient management in cropping systems: The role of nuclear techniques. In Transactions of the 17th World Congress of Soil Science, Symposium 59, Paper No. 2164. August 14–20, 2002, Bangkok, Thailand. Chen, W., G.J. Blair, J. Scott, and R. Lefroy, 1999. Nitrogen and sulphur dynamics of contrasting grazed pastures. Aust. J. Agric. Res. 50:1381–1392. Chinoim, N., R.D.B. Lefory, and G.J. Blair. 1997. The effect of crop duration and soil type on the ability of soil sulphur tests to predict plant response to sulphur. Aust. J. Agric. Res. 35:1131–1141. Condon, A.G., R.A Richards, G.J. Rebetzke, and G.D. Farquhar. 2002. Improving intrinsic water-use efficiency and crop yield. Crop Sci. 42:122–131. Condon, A.G., R.A Richards, G.J. Rebetzke, and G.D. Farquhar. 2004. Breeding for high water-use efficiency. J. Exp. Bot. 55:2447–2460. Crosson, E.R. 2008. WS-CRDS: Precision trace gas analysis and simplified stable isotope measurements. Am. Lab. November/December, 2008:37–41. CSIRO 2010. URL: http://www.csiro.au/clw. Cuntz, M., P. Peylin, P. Ciais, G. Hoffmann, R.J. Francey, P.P. Tans, C.E. Allison, and J.W. White. 2003. Interannual variations of 18O in atmospheric CO2-Data analysis and global modeling. Geophys. Res. Abs. 5:12768. Dawson, T.E. 1996. Determining water use by trees and forests from isotopic, energy balance and transpiration analyses—the roles of tree size and hydraulic lift. Tree Physiol. 16:263–272. Dawson, T.E., and J.R. Ehleringer. 1993. Isotopic enrichment of water in the woody tissues: Implications for plant water source, water uptake, and other studies which use the stable composition of cellulose. Geochim Cosmochim Ac. 57:3487–3492. Del Galdo, I., J. Six, A. Peressotti, and M. F. Cotufro. 2003. Assessing the impact of land-use change on soil C sequestration in agricultural soils by means of organic matter fractionation and stable C isotopes. Global Change Biol. 9:1204–1213. Dercon, G., E. Clymans, J. Diels, R. Merckx, and J. Deckers. 2006. Differential 13C isotopic discrimination in maize at varying water stress and at low to high nitrogen availability. Plant Soil 282:313–326. Di, H.J., K.C. Cameron, and R.G. McLaren. 2000. Isotopic dilution methods to determine the gross transformation rates of nitrogen, phosphorus, and sulfur in soil: A review of the theory, methodologies, and limitations. Aust. J. Soil Res. 38:213–230. Di, H.J., L.M. Condron, and E. Frossard. 1997. Isotope techniques to study phosphorus cycling in agricultural and forest soils: A review. Biol. Fert. Soil. 27:1–12. Diels, J., B. Vanlauwe, M.K. Van Der Meersch, N. Sanginga, and R. Merck. 2004. Long term soil organic carbon dynamics in a subhumid tropical climate: 13C data in mixed C3/C4 cropping and modeling with ROTHC. Soil Biol. Biochem. 36:1739–1750. Dingkuhn, M., G.D. Farquhar, S.K. De Datta, and T.C. O’Toole. 1991. Discrimination of 13C among upland rices having different water use efficiencies. Aust. J. Agric. Res. 42:1123–1131. Donald, C.M. 1964. Phosphorus in Australian agriculture. J. Aust. Inst. Agric. Sci. 6:75–105. Doos, B.R. 1994. Environmental degradation, global food production, and risk for large-scale migrations. Ambio 23:124–130.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
403
Echevarria, G., J.L. Morel, J.C. Fardeau, and E. Leclere-Cessac. 1998. Assessment of photoavailability of nickel in soils. J. Environ. Qual. 27:1064–1070. Echevarria, G., S. Klein, J.C. Fardeau, and J.L. Morel. 1997. Mesure de la fraction assimilable des éléments en traces du sol par la méthode des cinétiques d’échange isotopique: Cas du nickel. C. R. Seances 324 (serie IIa):221–227. Elliott, E.M., C. Kendall, S.D. Wankel, D.A. Burns, E.W. Boyer, K. Harlin, D.J. Bain, and T.J. Butler. 2007. Nitrogen isotopes as indicators of NOx source contributions to atmospheric nitrate deposition across the Midwestern and Northeastern United States. Environ. Sci. Technol. 41:7661–7667. Ellsworth, P.Z., and D.G. Williams. 2007. Hydrogen isotope fractionation during water uptake by woody xerophytes. Plant Soil 291:93–107. Food and Agriculture Organization of the United Nations (FAO). 1980. Maximizing the efficiency of fertilizer use by grain crops. FAO fertilizer and plant nutrition bulletin No 3. Rome: FAO. FAO. 1998. Guide to efficient plant nutrition management, 18. Rome: FAO. FAO. 2003. Database on commercially available organic fertilizers and water-retaining products. Rome: FAO/AGL. http://www.fao.org/ag/agl/agll/orgfert. FAO. 2008. Current world fertilizer trends and outlook to 2011/12, 57. Rome: FAO. ftp://ftp .fao.org/agl/agll/docs/cwfto11.pdf. FAO. 2010a. UNEP/ISRIC Global assessment of soil degradation (GLASOD). National soil degradation maps. Rome: FAO. http://www.fao.org/landandwater/agll/glasod/glasod maps.jsp; http://www.isric.org/UK/About+ISRIC/Projects/Current+Projects/GLADA.htm. FAO. 2010b. World Agriculture: Towards 2015/2030. Executive summary. Rome: FAO. http:// www.fao.org/docrep/004/y3557e/y3557e00.htm. FAO. 2010c. Soil carbon sequestration. Rome: FAO. http://www.fao.org/nr/land/sustainableland-management/soil-carbon-sequestration/en/. FAO. 2010d. Conservation agriculture. Rome: FAO. http://www.fao.org/ag/ca/. FAO. 2010e. Intensifying crop production with conservation agriculture. Rome: FAO. http:// www.fao.org/ag/ags/AGSE/agse_e/general/Contents.htm. FAO. 2010f. TERRASTAT. Land resource potential and constraints at country and regional level. Rome: FAO. http://www.fao.org/AG/aGL/agll/terrastat/wsr.asp. FAO/International Atomic Energy Agency (IAEA). 2002. Deficit irrigation practices. Water report 22, 102. Rome: FAO. ftp://ftp.fao.org/agl/aglw/docs/wr22e.pdf. FAO/IAEA. 2009. The Joint FAO/IAEA programme. Subprogramme soil and water management and crop nutrition. Rome: FAO. http://www-naweb.iaea.org/nafa/swmn/index.html. Fardeau������������������������������������������������������������������������������������ , J.C. 1993. Le Phosphore disponible des sols: Sa représentation par un modèle fonctionnel à plusieurs compartiments. Agron. J. 13:317–331. Fardeau, J.C. 1996. Dynamics of phosphate in soils: An isotopic outlook. Fert. Res. 45:91–100. Fardeau���������������������������������������������������������������������������������� , J.C., and J. Jappe. 1976. Nouvelle méthode de determination du phosphore assimilable par les plantes: Extrapolation des cinétiques de dilution isotopique. C. R. Seances Paris 282:1137–1140. Fardeau, J.C., C. Morel, and J. Jappé. 1985. Cinétique d’échange des ions phosphate dans les systémes sol-solution. Vérification expérimentale de l’équation théorique. C. R. Seances Paris 300:1039–1042. Fardeau, J.C., G. Guiraud, and C. Marol. 1995. Bioavailability soil P as a key to sustainable agriculture. Functional model determined by isotopic tracers. In Proceedings of the International Symposium on Nuclear and Related Techniques in Soil-Plant Studies on Sustainable Agriculture and Environmental Preservation, ed. IAEA, 131–144. October 17–21, 1994, Vienna, Austria. Fardeau, J.C., J.M. Hétier, and J. Jappé. 1979. Potassium assimilable du sol: Identification du compartiment des ions isotopiquement diluables. Cr. Soc. Biol. Paris 288:1039–1042.
404
World Soil Resources and Food Security
Ferretti, D.F., E. Pendall, J.A. Morgan, J.A. Nelson, D. LeCain, and A.R. Mosier. 2003. Partitioning ET fluxes from a Colorado grassland using stable isotopes: Seasonal variations and ecosystem implications of elevated atmospheric CO2. Plant Soil 254:291–303. Fessenden, J.E., and S.M. Clegg. 2008. Carbon dioxide stable isotope detection of geological seepage. Eos Trans. AGU 89 (53), abstract #U41C-0023. American Geophysical Union, Fall Meeting 2008. December 2008, San Francisco, USA. Follett, R.F., D.R. Keeney, and R.M. Cruze (eds.). 1991. Managing nitrogen for groundwater quality and farm profitability. Madison, WI: SSSA. Forbes, M.S., R.J. Raison, and J.O. Skjemstad. 2006. Formation, transformation and transport of black carbon (charcoal) in terrestrial and aquatic ecosystems. Sci. Total Environ. 370:190–206. Freibauer, A., A. Mosier, J. Clemens, U. Dammgen, M. Esala, F. Hartung, A. Jensen, M. Kaltschmitt, C. Kroeze, S. Petersen, R. Phillips, L. Valli, R. Vandre, and M. Worgetter (eds.). 2001. Biogenic emissions of greenhouse gases caused by arable and animal agriculture. Nutr. Cycling Agroecosyst. Special Issue 60:(1–3):1–326. Frossard, E., J.C. Fardeau, M. Brossard, and J.L. Morel, 1994a. Soil isotopically exchangeable phosphorus: A comparison between E and L values. Soil Sci. Soc. Am. J. 58:846–851. Frossard, E., J.C. Fardeau, M. Ognalaga, and J.L. Morel. 1992. Influences of agricultural practices, soil properties and parent material on the phosphate buffering capacity of cultivated soils developed in temperate climates. Eur. J. Agron. 1:45–50. Frossard, E., C. Feller, H. Tiessen, J.W.B. Stewart, J.C. Fardeau, and J.L. Morel. 1993. Can an isotopic method allow for the determination of the phosphate-fixing capacity of soils? Commun. Soil Sci. Plant Anal. 24:367–377. Frossard, E., and S. Sinaj. 1997. The isotope exchange kinetic technique: A method to describe the availability of inorganic nutrients. Applications to K, P, S and Zn. Isot. Environ. Health Stud. 33:61–77. Frossard, E., S. Sinaj, L.-M. Zhang, and J.L.Morel. 1996. The fate of sludge phosphorus in soil-plant systems. Soil Sci. Soc. Am. J. 60:1248–1253. Frossard, E., P. Skrabal, S. Sinaj, F. Bangerter, O. Traore. 2002. Forms and exchangeability of inorganic phosphate in composted solid organic wastes. Nutr. Cycling. Agroecosyst. 62:103–113. Frossard, E., P. Tekley, and J.Y. Grimal. 1994b. Characterization of phosphate species in urban sludges by high-resolution solid-state 31P NMR. Eur. J. Soil Sci. 45:403–408. Fulajtar, E. 2000. Assessment of soil erosion through the use of 137Cs at Jaslovske Bohunice, Western Slovakia. Acta Geol. Hisp. 35(3–4):291–300. Fulajtar, E. 2003. Assessment of soil erosion on arable land using 137Cs measurements; A case study from Jaslovske Bohunice, Slovakia. Soil Till. Res. 69:139–152. Gäbler, H.-E., A. Bahr, B. Mieke. 1999. Determination of the interchangeable heavy-metal fraction in soils by isotope dilution mass spectrometry. Fresenius J. Anal. Chem. 365:409–414. Gaunt, J.L., and J.A. Lehmann. 2008. Energy balance and emission associated with biochar sequestration and pyrolysis bioenergy production. Environ. Sci. Technol. 42: 4152–4158. Gérard, E., G. Echevarria, C. Morel, T. Sterckeman, and J.L. Morel. 2001. Isotopic exchange kinetic method for assessing cadmium availability in soils. In Trace elements in soil- bioavailability, flux and transfer, eds. I.K. Islander and M.B. Kirk, 127–143. Boca Raton, FL: Lewis Publishers. Gerzabek, M.H., G. Habenbauer, and H. Kicman. 2001. Soil organic matter pools and 13C natural abundances in particle size fractions of a long term agricultural field experiment receiving organic amendments. Soil Sci. Soc. Am. J. 65:352–358.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
405
Gibbs, M. 2008. Identifying source soils in contemporary estuarine sediments: A new compound specific isotope method. Estuaries and Coasts 31:344–359. Glaser, B. 2005. Compound-specific stable isotope (δ 13C) analysis in soil science. J. Plant Nutr. Soil Sci. 168:633–648. Govers, G., T.A. Quine, P.J.J. Desmet, and D.E. Walling. 1996. The relative contribution of soil tillage and overland flow erosion to soil redistribution in agricultural land. Earth Surf. Proc. Land. 21:929–946. Gray, C.W., R.G. McLaren, and J. Shiowatana. 2003. The determination of labile cadmium in some biosolids-amended soils by isotope dilution mass spectrometry. Aust. J. Soil Res. 41:589–597. Gray, C.W., R.G. McLaren, D. Günther, and S. Sinaj. 2004. An assessment of cadmium availability in cadmium-contaminated soils using isotope exchange kinetics. Soil Sci. Soc. Am. J. 68:1210–1217. Greacen, E.L. (ed.). 1981. Soil water assessment by the neutron method. Adelaide, Australia: CSIRO. Griffis, T.J., J.M. Baker, S.D. Sargent, B.D. Tanner, and J. Zhang. 2004. Measuring field-scale isotopic CO2 fluxes with tunable diode laser absorption spectroscopy and micrometeorological techniques. Agric. For. Meteorol. 124:15–29. Groffman, P.M., M.A. Altabet, J.K. Bohlke, K. Butterbach-Bhal, M.B. David, M.K. Firestone, A.E. Giblin, T.M. Kana, L.P. Nielsen, and M.A. Boytek. 2006. Methods for measuring denitrification: Diverse approaches to a difficult problem. Ecol. Appl. 16:2091–2122. Gruber, N., P. Friedlingstein, C.B. Field, R. Valentini, M. Heimann, J.E. Richey, P. RomeroLankao, E.D. Schulze, and C.-T. A. Chen. 2004. The vulnerability of the global carbon cycle in the 21st century: An assessment of carbon-human interactions. In The global carbon cycle, eds. C.B. Field and M.R. Raupach, Scope 62, 45–76. Washington, DC: Island Press. Habib L., S.H. Chien, R.G. Menon, and G. Carmona. 1998. Modified iron oxide-impregnated paper strip test for soil treated with phosphate fertilizers. Soil Sci. Soc. Am. J. 62:972–976. Hammes K. et al. 2007. Comparison of black carbon quantification methods to measure firederived (black/elemental) carbon in soils and sediments using reference materials from soil, water, sediment and the atmosphere. Global Biogeochem. Cycles 21:GB3016, doi:10.1029/2006GB002914. Hammes, K., M.S. Torn, A.G. Lapenas, and M.W.I. Schmidt. 2008. Centennial black carbon turnover observed in a Russian steppe soil. Biogeosciences 5:1339–1350. Hamon, R., J. Wundke, M. McLaughlin, and R. Naidu. 1997. Availability of zinc and cadmium to different plant species. Aust. J. Soil Res. 35:1267–1277. Hamon, R.E., M.J. McLaughlin, R. Naidu, and R. Correll. 1998. Long-term changes in cadmium bioavailability in soil. Environ. Sci. Technol. 32:3699–3703. Hardarson, G., and C. Atkins 2003. Optimising biological nitrogen fixation by legumes in farming systems. Plant Soil 252:41–54. Hardarson, G., and W.J. Broughton (eds.). 2003. Maximising the use of biological nitrogen fixation in agriculture. Plant Soil Special Issue 252:1. Hauck, R.D., and J.M. Bremner. 1976. Use of tracers for soil and fertilizer nitrogen research. Adv. Agron. 28:219–266. Hayes, J.M., K.H. Freeman, B.N. Po, and C.H. Hoham. 1990. Compound specific isotope analysis. A novel tool for reconstruction of ancient biochemical processes. Adv. Organ. Geochem 16:1115–1128. Hedges, R.E.M. 1987. Radiocarbon dating by accelerator mass spectrometry—some recent results and applications. Philos. Trans. R. Soc. Lond. A 323:57–73. Heim, A., and W.I. Schmidt. 2007. Lignin turnover in arable soil and grassland analysed with two different labeling approaches. Eur. J. Soil Sci. 58:599–608.
406
World Soil Resources and Food Security
Henao, J., and C. Baanante. 2001. Nutrient depletion in the agricultural soils of Africa. 2020 Vision, Brief 62, Washington, DC: International Food Policy Research Institute. Herrick, J.E. 2000. Soil quality: An indicator of sustainable land management? Appl. Soil Ecol. 15:75–83. Hillel, D., and P. Vlek. 2005. The sustainability of irrigation. Adv. Agron. 87:55–84. Hinsinger, P. 1998. How do plant roots acquire mineral nutrients? Adv. Agron. 64:225–265. Hobbie, E.A., and R.A. Werner. 2004. Intramolecular, compound-specific, and bulk carbon isotope patterns in C3 and C4 plants: A review and analysis. New Phytol. 161:371–385. Hobbs, P., K. Sayre, and R. Gupta. 2008. The role of conservation agriculture in sustainable agriculture. Philos. Trans. R. Soc. B. 363:543–555. Hocking, P.J. 2001. Organic acids exuded from roots in phosphorus uptake and aluminium tolerance of plants in acid soils. Adv. Agron. 74:63–93. Hoeft, R.G., D.R. Keeney, and L.M. Walsh. 1972. Nitrogen and sulphur precipitation and sulfur dioxide in the atmosphere in Wisconsin. J. Environ. Qual. 1:203–208. Hood, R.C., M. Khan, A. Haque, M. Khadir, J.P. Bonetto, R. Syamsul, L. Mayr, and M. Heiling. 2004. Carbon sequestration and estimated carbon credit values as measured using 13C labelling and analysis by means of an optical breath test analyser. Commun. Soil Sci. Plant. Anal. 379:242–246. Hood-Novotny, R., C. van Kessel, and B. Vanlauwe. 2008. Use of tracer technology for the management of organic sources. In Guidelines on nitrogen management in agricultural soils, ed. IAEA, IAEA-TCS-29CD, 181–234, Vienna, Austria. Horowitz, A.J., and D.E Walling (eds.). 2005. Sediment budgets 1 and 2. IAHS Publications 291 and 292. Wallingford, UK:IAHS Press. Hubick, K., and G. Farquhar. 1989. Carbon isotope discrimination and the ratio of carbon gains to water lost in barley cultivars. Plant Cell Environ. 12:795–804. IAEA. 1980. Soil nitrogen as fertilizer or pollutant, Panel Proceedings. Series STI/PUB/535. Vienna, Austria: IAEA. IAEA. 1983. Zn fertilization of flooded rice. IAEA-TECDOC-245. Vienna, Austria: IAEA. IAEA. 1984. Soil and fertilizer nitrogen. Technical Report Series No 244. Vienna, Austria: IAEA. IAEA. 1990. Use of nuclear techniques in studies of soil-plant relationships. IAEA training course series No.2. Vienna, Austria: IAEA. IAEA. 1992. Manual on measurement of methane and nitrous oxide from agriculture. IAEATECDOC 674. Vienna, Austria: IAEA. IAEA. 1996. Nuclear techniques to assess irrigation schedules for field crops. IAEATECDOC-888. Vienna, Austria: IAEA. IAEA. 1997. Sewage sludge and wastewater for use in agriculture. IAEA-TECDOC-971. Vienna, Austria: IAEA. IAEA. 1998. Management of nutrients and water in rainfed arid and semi-arid areas. IAEATECDOC-1026. Vienna, Austria: IAEA. IAEA. 2000a. Comparison of soil water measurement using the neutron scattering, time domain reflectometry and capacitance methods. IAEA-TECDOC-1137. Vienna, Austria: IAEA. IAEA. 2000b. Optimizing nitrogen fertilizer application to irrigated wheat. IAEA-TECDOC1164. Vienna, Austria: IAEA. IAEA. 2001. Use of isotope and radiation methods in soil and water management and crop nutrition. IAEA training course series No 14. Vienna, Austria: IAEA. IAEA. 2002a. Water balance and fertigation for crop improvement in West Asia. IAEATECDOC 1266. Vienna, Austria: IAEA. IAEA. 2002b. Assessment of soil phosphorus status and management of phosphatic fertilizers to optimize crop production. IAEA-TECDOC-1272 (also available on CD ROM). Vienna, Austria: IAEA.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
407
IAEA. 2003a. Neutron and gamma probes: Their use in agronomy, 2nd ed. IAEA training course series No 16. Vienna, Austria: IAEA. IAEA. 2003b. Guidelines for the use of isotopes of sulphur in soil-plant studies. IAEA training course series No 20. Vienna, Austria: IAEA. IAEA. 2005. Nutrient and water management practices for increasing crop production in rainfed arid/semiarid areas. IAEA-TECDOC-1468. Vienna, Austria: IAEA. IAEA. 2006. Management practices for improving sustainable crop production in tropical acid soils. IAEA Proceedings Series, STI/PUB/1285. Vienna, Austria: IAEA. IAEA. 2007. Radiation processing: Environmental applications. IAEA-RPEA. Vienna, Austria: IAEA. IAEA. 2008a. Guidelines on nitrogen management in agricultural soils. IAEA-TCS-29CD. Vienna, Austria: IAEA. IAEA. 2008b. Field estimation of soil water content: A practical guide to methods, instrumentation and sensors. IAEA training course series No 30. Vienna. Austria: IAEA. IAEA. 2009a. Home/work. http://www.iaea.org/OurWork/index.html. IAEA. 2009b. Co-ordinated research activities. NACA welcome. http://www-crp.iaea.org/ html/welcome.html. IAEA. 2009c. Technical co-operation programme. http://tc.iaea.org/tcweb/default.asp. IAEA. 2010. Guidelines on the use of carbon isotopes in investigating soil carbon sequestration in agro-ecosystems. IAEA training course series. Vienna, Austria: IAES. Forthcoming. Intergovernmental Panel on Climate Change (IPCC). 2001. Climate change 2001: The scientific basis. Cambridge Univ. Press, Cambridge, UK. Janzen, H.H., C.A. Campbell, B.H. Ellert, and E. Bremer. 1997. Soil organic matter dynamics and their relationship to soil quality. In Soil quality for crop production and ecosystem health. Developments in soil science, Vol. 25, eds. E.G Gregorich and M.R. Carter, 399–430. Amsterdam: Elsevier. Jensen, E.S., G. Hardarson, M.B. Peoples. 2008. Use of tracer technology in biological nitrogen fixation research. In Guidelines on nitrogen management in agricultural soils, IAEA-TCS-29CD, ed. IAEA, 127–180. Vienna, Austria: IAEA. Karlen, D.L., S.S. Andrews, and J.W. Doran. 2001. Soil quality: Current concepts and applications. Adv. Agron. 74:1–40 Keerthisinghe G., F. Zapata, and P.M. Chalk. 2003. Plant nutrition: Challenges and tasks ahead. In Proceedings of the IFA-FAO Agricultural Conference on Global Food Security and the Role of Sustainable Fertilization. March 26–28, 2003, Rome, Italy. http://www .fertilizer.org/ifa/publicat/PDF/2003_rome_keerthisinghe.pdf and slides; http://www .fertilizer.org/ifa/publicat/PDF/2003_rome_keerthisinghe_slides.pdf. Kemme, P.A., A. Lommen, L.H. DeJonge, J.D. Van der Klis, A.W. Jongbloed, Z. Mroz, and A.C. Beynen. 1999. Quantification of inositol phosphates using 31P nuclear magnetic resonance spectroscopy in animal nutrition. J. Agric. Food Chem. 47:5116–5121. Kim, K.R., and H. Craig. 1993. Nitrogen-15 and oxygen-18 characteristics of nitrous oxide: A global perspective. Science 262:1855–1857. Kirda, C., P. Moutonnet, C. Hera, and D.R. Nielsen (eds.). 1999. Crop yield response to deficit irrigation. Dordrecht, the Netherlands: Kluwer Academic Publishers. Kirkham, D., and W.V. Bartholomew. 1954. Equations for following nutrient transformations in soil, utilizing tracer data. Soil Sci. Soc. Am. Proceedings 18:33–34. Kogel-Knabner, I. 2000. Analytical approaches for characterizing soil organic matter. Org. Geochem. 31:609–625. Konboon, Y., G.J. Blair, R. Lefroy, and A. Whitbread. 2000. Tracing the nitrogen, sulphur and carbon released from plant residues in a soil/plant system. Aust. J. Agric. Res. 38:699–710. Krouse, H.R. 1977. Sulphur isotope abundance elucidates uptake of atmospheric sulphur emissions by vegetation. Nature 265:45–46.
408
World Soil Resources and Food Security
Kuhlbusch, T.A.J. 1998. Black carbon and the carbon cycle. Science 280: 1903–1904. Kwon, J.-S., B. Mayer, S.-T. Yun, and M. Nightingale. 2008. The use of ion exchange membranes for isotope analyses on soil water sulfate: Laboratory experiments. J. Environ. Qual. 37:501–508. L’ Annunziata, M.F. (ed.). 2003. Handbook of radioactivity analysis, 2nd ed. San Diego, CA: Academic Press Ltd. Lal, R. 1989. Conservation tillage for sustainable agriculture. Adv. Agron. 42:85–197. Lal, R. 1995. Global soil erosion by water and carbon dynamics. In Soils and global change, eds. R. Lal, J.M. Kimble, E. Levine, and B.A. Stewart, 131–142. Boca Raton, FL: CRC Press. Lal, R. 2000. Soil management in the developing countries. Soil Sci. 165:57–72. Lal, R. (ed.). 2001a. Soil carbon sequestration and the greenhouse effect. SSSA Spec. Publ. 57. Madison, WI: SSSA. Lal, R. 2001b. World cropland soils as a source or sink for atmospheric carbon. Adv. Agron. 71:145–191. Lal, R. 2002. The potential of soils of the tropics to sequester carbon and mitigate the greenhouse effect. Adv. Agron. 76:1–30. Lal, R. 2004. Soil carbon sequestration impacts on global climate change and food security. Science 304:1623–1627. Lal, R. 2006. Encyclopaedia of soil science, 2nd ed. Boca Raton, FL: CRC Press. Lal, R. 2007. Farming carbon. Soil Till. Res. 96:1–5. Lal, R., and J.M. Kimble. 1997. Conservation tillage for carbon sequestration. Nutr. Cycling Agroecosyst. 49:243–253. Lal, R., J.M. Kimble, R.F. Follett, and B.A. Stewart (eds.). 1998. Soil processes and the carbon cycle. Advances in soil science. Boca Raton, FL: CRC Press. Lal, R., and J.P. Spruce. 1999. The potential of world cropland soils to sequester C and mitigate the greenhouse effect. Environ. Sci. Pol. 2:177–185. Lal, R., J.M. Kimble, R.F. Follett, and B.A. Stewart. 2001. Assessment methods for soil carbon. Boca Raton, FL: CRC Press. Leavitt, S.W., E. Pendall, E.A. Paul, T. Brooks, B.A. Kinball, P.J. Pinter Jr., H.B. Johnson, A. Matthias, G.W. Wall, and R.L. LaMorte. 2001. Stable-carbon isotopes and soil organic carbon in wheat under CO2 enrichment. New Phytol. 150:305–314. Leavitt, S.W., E.A. Paul, B.A. Kimball, G.R. Hendrey, J.R. Mauney, R. Rauschkolb, H. Rogers, K.F. Levin, J. Nagy, P.J. Pinter Jr, and H.B. Johnson. 1994. Carbon isotope dynamics of free-air CO2 enriched cotton and soils. Agric. For. Meteorol. 70:87–101. Lehmann, J. 2007. A handful of carbon. Nature 447:143–144. Lehmann, J., J.P. da Silva Jr, S. Christoph, T. Nehls, W.W. Zech, and B. Glaser. 2003. Nutrient availability and leaching in an archaeological Anthrosol and a Ferralsol of the Central Amazon basin: Fertilizer, manure and charcoal amendments. Plant Soil 249:343–357. Lehmann, J.A., B. Liang, D. Solomon, M. Lerotic, F. Luizao, J. Kinyangi, T. Schafer, S. Wirick, and C. Jacobsen. 2005. Near-edge x-ray absorption fine structure (NEXAFS) for mapping nano-scale distribution of organic forms in soil: Application to black carbon particles. Global Biogeochem. Cycles 19:GB1013, doi: 10.1029/2004GB002435. Lehmann, J., J. Gaunt, and M.. Rondon. 2006. Bio-char sequestration in terrestrial ecosystems—A review. Mitigation and Adaptation Strategies for Global Change 11:403–427. Levin, I., and V. Hesshaimer. 2000. Radiocarbon—A unique tracer of global carbon cycle dynamics. Radiocarbon. 42:69–80. Lewis, D.C., and A.L. Clarke. 1975. Superphosphate for acid South East sands. J. Agric. South Aust. 78:74–77. Leytem, A.B., B.L. Turner, and P.A. Thacker. 2004. Phosphorus composition of manure from swine fed low-phytate grains: Evidence from hydrolysis in the animal. J. Environ. Qual. 33:2380–2383.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
409
Leytem, A.B., K. Kwanyuen, P.W. Plusmtead, R.O. Maguire, and J. Brake. 2008. Evaluation of phosphorus characterization in broiler ileal digesta, manure and litter samples: 31PNMR vs HPLC. J. Environ. Qual. 37:494–500. Li, S., D.A. Lobb, M.J. Lindstrom, and A. Farenhorst. 2007a. Tillage and water erosion on different landscapes in the northern North American Great Plains evaluated using the 137Cs technique and soil erosion models. Catena 70:493–505. Li, Y., and M.J. Lindstrom. 2001. Evaluating soil quality-soil redistribution relationships on terraces and steep hillslope. Soil Sci. Soc. Am. J. 65:1500–1508. Li, Y., M.J. Lindstrom, J. Zhang, and J. Yang. 2000. Spatial variability of soil erosion and soil quality on hillslopes in the Chinese Loess Plateau. Acta Geol. Hisp. 35:261–270. Li, Y., Q.W. Zhang, D.C. Reicosky, L.Y. Bai, M.J. Lindstrom, and L. Li. 2006. Using 137Cs and 210Pb for quantifying soil organic carbon redistribution affected by intensive tillage on ex steep slopes. Soil Till.Res. 86:176–184. Li, Y., Q.W. Zhang, D.C. Reicosky, M.J. Lindstrom, L.Y. Bai, and L. Li. 2007b. Changes in soil organic carbon induced by tillage and water erosion on a steep cultivated hillslope in the Chinese Loess Plateau from 1898–1954 and 1954–1998. J. Geophys. Res. 112:G01021. Lichfouse, E. 2000. Compound specific analysis. Application to archaology, biomedical sciences, biosynthesis, environment, extraterrestrial chemistry, food science, forensic science, humic substances, microbiology, organic geochemistry, soil science, and sport. Rapid Commun. Mass Sp. 14:1337–1344. Lloyd, C.A., R.L. Chaney, and S.B. Hornick. 1981. Labile cadmium in soils of long-term sludge utilization farms. In Agronomy Abstracts, 29. Madison, WI: Agronomy Society of America. Lobb, D.A., R.G. Kachanoski, and M.H. Miller. 1995. Tillage translocation and tillage erosion on shoulder slopes landscape positions measured using 137Cs as a tracer. Can. J. Soil Sci. 75:211–218. Lombi, E., and J. Susini. 2009. Synchrotron-based techniques for plant and soil science: Opportunities, challenge and future perspectives. Plant Soil 320:1–35. Mabit, L., and C. Bernard. 2009. Spatial distribution and content of soil organic matter in an agricultural field in Eastern Canada, as estimated from geostatistical tools. Earth Surf. Proc. Landf. 35:278–283. Mabit, L., C. Bernard, M. Makhlouf, and M.R. Laverdière. 2008a. Spatial variability of erosion and soil organic matter content estimated from 137Cs measurements and geostatistics. Geoderma 145:245–251. Mabit, L., M. Benmansour, and D.E. Walling. 2008b. Comparative advantages and limitations of fallout radionuclides (137Cs, 210Pb, and 7Be) to assess soil erosion and sedimentation. J. Environ. Radioact. 99:1799–1807. Magesan, G.N., H. Wang. 2003. Application of municipal and industrial residuals in New Zealand forests: An overview. Aust. J. Soil Res. 41:557–569. Martius, C., H. Thiessen, and P.L.G. Vlek. 2001. Managing organic matter in tropical soils: Scope and limitations. Nutr. Cycling Agroecosyst. Special Issue 61(1–2):1–213. Mayer, B., K.H. Feger, A. Giesemann, and H.J. Jager. 1995. Interpretation of sulfur cycling in two catchments in the Black Forest (Germany) using stable sulfur and oxygen-isotope data. Biogeochem. 30:31–58. McNeill, A. 1999. Enriched stable isotopes techniques to study soil organic matter accumulation and decomposition in agricultural systems. In Application of stable isotope techniques to study plant physiology, plant water uptake and nutrient cycling in terrestrial ecosystems, ed. M. Unkovich, 105–121. Western Australia: Center for Legumes in Mediterranean Agriculture. McNeill, A.M., C. Zhu, and I.R.P. Fillery. 1998. A new approach to quantifying the N benefit from pasture legumes to succeeding wheat. Aust. J. Agric. Res. 49:427–436.
410
World Soil Resources and Food Security
McCarty, G.W., J.B. Reeves III, V.B. Reeves, R.F. Follett, and J.M. Kimble. 2002. Mid-infrared and near-infrared diffuse reflectance spectroscopy for soil carbon measurement. Soil Sci. Soc. Am. J. 66:640–646. McCarty, G.W., and J.C. Ritchie. 2002. Impact of soil movement on carbon sequestration in agricultural ecosystems. Environ. Pollut. 116(3):423–430. McDowell, R.W., and A.N. Sharpley. 2003. Phosphorus uptake and release from stream sediments. J. Environ. Qual. 32:937–948. McLain, J.E.T., and D.A. Martens. ���������������������������������������������������������� 2005. Nitrogen oxide flux from soil amino acid mineralization. Soil Biol. Biochem. 37:289–299. Meizen-Dick, R.S., M.W. Rosengrant. 2001. Overview. 2020 Vision, Focus 9 (Overcoming water scarcity and quality constraints), Brief 1-14. Washington, DC: International Food Policy Research Institute. Messick, D.L., and M.X. Fan. 1999. The role of sulphur in oil crop production. Paper presented at the IFA Regional Conference for Asia and the Pacific. November 14–17, 1999. Kuala Lumpur, Malaysia. Mohd Hashim, T., and A.R. Zaharah. 1994. Nitrogen contribution by oil palm mill effluent to young oil palm (Elaeis guineensis Jacq) as measured by 15N isotope dilution technique. Pertanika. J. Trop. Agric. Food Sci. 17(2):81–87. Morel, C., and J.C. Fardeau. 1994. Phosphorus availability of fertilizers: A predictive laboratory method for its evaluation. Fert. Res. 28:1–9. Mosier, A.R., K. Syers, and J.R. Freeney (ed.). 2004. Agriculture and the nitrogen cycle: Assessing the impacts of fertilizer use on food production and the environment. Washington, DC: Island Press. Moutonnet, P., and J.C. Fardeau. 1997. Inorganic nitrogen in soil solution collected with tensionic samplers. Soil Sci. Soc. Am. J. 61:822–825. Myrold, D.D. 1990. Measuring denitrification in soils using 15N tracer techniques. In Denitrification in soil and sediment, eds. N.P. Resvbech and J. Sorensen, 181–198. New York: Plenum Press. Nakhone, L.N., and S.D. Young. 1993. The significance of (radio-) labile cadmium pools in soil. Environ. Pollut. 82:73–77. Nguyen, M.L. 2007. Integrated soil-water-plant solutions for biomass production and environmental performance as influenced by climate change. Report of the FAO/IAEA Consultants Meeting. November 12–14, 2007, Vienna, Austria. Unpublished. Nguyen, M.L. 2008. Integrated approaches for the assessment of land use impacts on soil loss/ sediment production and related environmental problems. Report of the Consultants Meeting. November 5–7, 2007, Vienna, Austria. http://www-naweb.iaea.org/nafa/ swmn/index.htm. Nguyen, M.L., F. Zapata, and G. Dercon. ������������������������������������������������� 2010. Zero-tolerance on land degradation for sustainable intensification of agricultural production. In Land degradation and desertification: Assessment, mitigation and remediation, eds. P. Zdruli, M. Pagliai, S. Kapur, and F. Faz Cano, 37–47. Heidelberg, London, New York: Springer Dordrecht. Nguyen, M.L., and F. Zapata. 2006. Use of nuclear techniques in addressing soil-water-nutrient issues for sustainable agricultural production. In 18th World Congress of Soil Science, CD-ROM Abstracts-Session No.1-1, July 2006, Philadelphia, PA. Nguyen, M.L., D.S. Rickard, and S.D. McBride. 1989. Pasture production and changes in phosphorus and sulphur status in irrigated pastures receiving long-term applications of superphosphate fertilizer. New Zeal. J. Agric. Res. 32:245–262. Nguyen, M.L., and K.M. Goh. 1992. Nutrient cycling and losses based on a mass balance model in grazed pastures receiving long-term superphosphate applications in New Zealand. Phosphorus. J. Agric. Sci. 119:89–106. Nielsen, L.P. 1992. Denitrification in sediment determined from nitrogen isotope pairing. FEMS Microbiol. Ecol. 86:357–362.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
411
Novak, M., I. Jackova, and E. Prechova. 2001 Temporal trends in the isotope signature of airborne sulphur in Central Europe. Environ. Sci. Technol. 35:255–260. Novotny, E.H., M.H.B. Hayes, E.R. de Azevedo, and T.J. Bonagamba. 2006. Characterization of black carbon-rich samples by 13C solid state nuclear magnetic resonance. Natur wissenschaften 93:447–450. O’Neill, I.K., M. Sargent, and M.L. Trible. 1980. Determination of phytate in foods by phosphorus-31 fourier transform nuclear magnetic resonance spectrometry. Anal. Chem. 52:1288–1291. Organization for Economic Co-operation and Development (OECD). 2002. Agricultural practices that reduce greenhouse gas emissions: Overview and results of survey instrument. Joint working party of the Environment Policy Committee and the Committee for Agriculture. COM/ENV/EPOC/AGR/CA(1998)/149/FINAL. Paris: OECD. OECD. 2003. Agriculture and land conservation: Developing indicators for policy analysis. In Proceedings of the Japan/OECD Expert Meeting, May 2003, Kyoto, Japan. OECD. 2004. Agricultural impacts on soil erosion and soil biodiversity: Developing indicators for policy analysis. In Proceeding of the OECD Expert Meeting, March 2003, Rome, Italy. Oeschger H. 2000. Perspectives on global change science: Isotopes in the earth system, past and present. Quatern. Sci. Rev. 19:37–44. Organic Resources Database (ORD). 2004. ftp://iserver.ciat.cgiar.org/webciat/ ORD. Professional Alliance for Conservation Agriculture (PACA). 2009. Conservation agriculture as an adaptive and mitigation strategy to combat climate change. PACA Newsletter, Issue 8. http://www.conservagri.org. Palm, C.A. 2001. Organic inputs for soil fertility management in tropical agroecosystems: Application of an organic resource database. Agric. Ecosyst. Environ. 83:27–42. Peak, D., J.T. Sims, and D.L. Sparks. 2002. Solid-state speciation of natural and alum-amended poultry litter using XANES-spectroscopy. Environ. Sci. Technol. 36:4253–4261. Pendall, E., S. Bridgman, P.J. Hanson, B. Hungate, D.W. Kicklighter, D.W. Johnson, B.E. Law, Y. Luo, J.P. Megonigal, M. Olsrud, M.G. Ryan, S. Wan. 2004a. Below-ground process responses to elevated CO2 and temperature: A discussion of observations, measurement methods, and models. New Phytol. 162:311–322. Pendall, E., and J.Y. King. 2007. Soil organic dynamics in grasslands under elevated CO2: Insights from long term incubations and stable isotopes. Soil Biol. Biochem. 39:2628–2639. Pendall, E., A.R. Mosier, and J.A. Morgan. 2004b. Rhizodeposition stimulated by elevated CO2 in a semi-arid grassland. New Phytol. 162:447–458. Pennock, D.J. 1998. New perspectives in the soil erosion-soil quality relationship. In Use of 137Cs in the study of soil erosion and sedimentation, IAEA TECDOC-1028, ed. IAEA,13–25. Vienna, Austria: IAEA. Pennock, D.J. 2000. Suitability of 137Cs redistribution as an indicator of soil quality. Acta Geol. Hisp. 35(3–4):213–217. Pennock, D.J. 2003. Terrain attributes, landscape segmentation, and soil redistribution. Soil Till. Res. 69:15–26. Pennock, D.J., and A.H. Frick. 2001. The role of field studies in landscape-scale applications of process models: An example of soil redistribution and soil organic carbon modeling using CENTURY. Soil Till. Res. 58:183–191. Perez, T., D. Garcia, S. Trumbore, S. Tyler, P. de Camargo, M. Moreira, M. Piccolo, S. Park, K. Boering, and C. Cerri. 2003. Characterization of the N2O isotopic composition (15N, 18O and N O Isotopomers) emitted from incubated Amazon forest soils. Geophys. Res. 2 Abstr. 5:1475. Phillips, D.L., and J.W. Greg. 2003. Source partitioning using stable isotopes: Coping with too many sources. Oecol. 136:261–269.
412
World Soil Resources and Food Security
Pinstrup-Andersen, P. 1999. Towards ecologically sustainable world food production. UNEP Industry and Environment 22:10–18. Pommel, B. 1982. Aptitude de plusieurs dechets urbains a fournir du phosphore aux cultures. Agron. 2:851–857. Poss, R., J.C. Fardeau, and H. Saragoni. 1991. Potassium release and fixation in Ferralsols (Oxisols) from Southern Togo. J. Soil Sci. 42:649–660. Poth, M., J.S. La Favre, and D.D. Focht. 1986. Quantification by direct 15N dilution of fixed N2 incorporation into soil by Cajanus cajan (pigeon pea). Soil Biol. Biochem. 18:125–127. Radajewski, S., P. Ineson, N.R. Parekh, and J.C. Murrell. 2000. Stable-isotope probing as a tool in microbial ecology. Nature 403:646–649. Rajan, S.S.S. 1987. Phosphate rock and phosphate rock/sulphur granules as phosphate fertilizers and their dissolution in soil. Fert. Res. 11:43–60. Rajan, S.S.S., and S.H. Chien (eds.). 2003. Direct application of phosphate rock and related technology: Latest developments and practical experiences. Proceedings of an International Meeting. July 16–20, 2001, Kuala Lumpur. Special Publications IFDCSP-37. Muscle Shoals, AL: International Center for Soil Fertility. Rajan, S.S.S., J.H. Watkinson, and A.G. Sinclair. 1996. Phosphate rocks for direct application to soils. Adv. Agron. 57:77–159. Rasse, D.P., M.-F. Dignac, H. Bahri, C. Rumpel, A. Mariotti, and C. Chenu. 2006. Lignin turnover in agricultural soils: From plant residues to soil-protected fractions. Eur. J. Soil Sci. 57:530–538. Rebetzke, G.J., A.G. Condon, R.A. Richards, and G.D. Farquhar. 2002. Selection for reduced carbon isotope discrimination increases aerial biomass and grain yield of rainfed bread wheat. Crop Sci. 42:739–745. Richards, L.A., W.R. Garner, and G. Ogata. 1956. Physical processes determining water loss from soil. Soil Sci. Soc. Am. Proc. 20:310–314. Ritchie, J.C., G.W. McCarty. 2003. 137 Cesium and soil carbon in a small agricultural watershed. Soil Till. Res. 69:45–51. Ritchie, J.C., and C.A. Ritchie. 2009. Bibliography of publications of 137Cs studies related to erosion and sediment deposition. http://www.ars.usda.gov/Main/docs.htm?docid=16037. Rosengrant, M.W., C. Ximing. 2001. Water for food production. 2020 Vision, Focus 9 (Overcoming water scarcity and quality constraints), Brief 2-14. Washington, DC: International Food Policy Research Institute. Russell, C.A., and I.R.P. Filler. 1996. In situ 15N labelling of lupin below-ground biomass. Aust. J. Agric. Res. 47:1035–1046. Saggar, S., and C.B. Hedley. 2001. Estimating seasonal and annual carbon inputs, and root decomposition rates in a temperate pasture following field 14C pulse-labelling. Plant Soil 236:91–103. Saggar, S., A. Parshotam, G.P. Sparling, C.W. Feltham, and P.B.S. Hart. 1996. 14C–labeled ryegrass turnover and residence times in soils varying in clay content and mineralogy. Soil Biol. Biochem. 28:1677–1686. Saggar, S., and P.L. Searle. 1995. A simple chamber technique for in situ labelling of pasture sward with 14C. Commun. Soil Sci. Plant Anal. 26:1547–1563. Sale, P.W.G., and L.K. Heng. 2004. Decision-support systems for the use of phosphate rocks. In Use of phosphate rocks for sustainable agriculture. FAO fertilizer and plant nutrition bulletin No. 13, eds. F. Zapata and R.N. Roy, 69–75. Rome: FAO. Sale, P.W.G., P.G. Simpson, C.A. Anderson, and L.L. Muir (eds.). 1997. The role of reactive phosphate rock fertilizers for pastures in Australia. Aust. J. Exp. Agric. 37:845–1023. Samosir, S.S.R., G.J. Blair, and R.D.B. Lefroy. 1993. Effects of placement of elemental S and sulfate on the growth of two rice varieties under flooded conditions. Aust. J. Agric. Res. 44:1775–1788.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
413
Sanchez, P.A., R.J. Buresh, and R.R.B. Leaky. 1997. Trees, soils and food security. Philos. Trans R. Soc. Lond. B. Biol. Sci. 352:949–961. Sato, S., D. Solomon, C. Hyland, Q.M. Kettering, and J. Lehman. 2005. Phosphorus speciation in manure and manure-amended soils using XANES spectroscopy. Environ. Sci. Technol. 39:7485–7491. Schaeffer, S.M., J.B. Miller, B.H. Vaughn, J.W. White, and D.R. Bowling. 2008. Long-term field performance of a tunable diode laser absorption spectrometer for analysis of carbon isotopes of CO2 in forest air. Atmos. Chem. Phys. Disc. 8:531–9568. Scharpenseel, H.W., P. Becker-Heldmann, H.U. Neue, and K. Tsutsuki. 1989. Bomb-carbon, 14C dating and 13C measurements as tracers of organic matter dynamics as well as morphogenetic and turbation processes. Sci. Total Environ. 81/82:99–110. Scherr, S.J. 1999. Soil degradation: A threat to developing country food security by 2020? Food, agriculture, and the environment. Discussion Paper 27. Washington, DC: International Food Policy Research Institute. Schmidt, M.W.I. 2007. Interdisciplinary intercomparison of black carbon analysis in soil and sediment. Eos. 88:344. Schmidt, M.W.I., and A.G. Noack. 2007. Black carbon in soils and sediments: Analysis, distri bution, implications and current challenges. Global Biogeochem. Cycles 14(3):777–793. Schmidt, M.W.I., J.O. Skjemstad, C.I. Czimczik, B. Glaser, K.M. Prentice, Y. Gelinas, and T.A.J. Kuhlbusch. 2001. Comparative analysis of black carbon in soils. Global Biogeochem. Cycles 15:163–167. Schmidt, M.W.I., J.O. Skjemstad, and C. Jager. 2002. Carbon isotope geochemistry and nanomorphology of soil black carbon: Black Chernozenic soils of central Europe originate from ancient burning biomass. Global Biogeochem. Cycles 16:1123, Doi 10.1029/2002GB001939. Schuller, P., A. Ellies, A. Castillo, and I. Salazar. 2003. Use of the 137Cs technique to estimate tillage- and water-induced soil redistribution rates on agricultural land under different use and management in central-south Chile. Soil Till. Res. 69:45–51. Schuller, P., D.E. Walling, A. Sepulveda, R.E. Trumper, J.L. Rouanet, I. Pino, and A. Castillo. 2004. Use of Cs-137 measurements to estimate changes in soil erosion rates associated with changes in soil management practices on cultivated land. Appl. Radiat. Isot. 60:759–766. Schuller, P., D.E. Walling, A. Sepulveda, A. Castillo, and I. Pino. 2007. Changes in soil erosion associated with the shift from conventional tillage to a no-tillage system, documented using Cs-137 measurements. Soil Till. Res. 94:183–192. Schumacher, J.A., T.C. Kaspar, J.C. Ritchie, T.E. Schumacher, D.L. Karlen, E.R. Venteris, G.W. McCarty, T.S. Colvin, D.B. Jaynes, M.J. Lindstrom, and T.E. Fenton. 2005. Identifying spatial patterns of erosion for use in precision conservation. J. Soil Water Conserv. 60:355–361. Shedley, C.D., A.R. Till, and G.J. Blair. 1979. A radiotracer technique for studying the nutrient release from different fertilizer materials and its uptake by plants. Comm. Soil Sci. Plant Anal. 10:737–745. Shepherd, M.J., J.M.A. Anderson, R. Bol, and D.K. Allen. 2000. Incorporation of N from spiked cattle dung pats into soil under two moorland plant communities. Rapid Comm. Mass Sp. 14:1361. Sheppard, C.W. 1962. Basic principles of the tracer method. New York: John Wiley and Sons. Sikora, L.J., C.F. Tester, J.M. Taylor, and J.F. Parr. 1982. Phosphorus uptake by fescue from soils amended with sewage sludge compost. Agron. J. 74:27–33. Sillanpää, M. 1990. Micronutrient assessment at the country level: An international study. FAO Soil Bulletin No 63. Rome: FAO. Sinaj, S., F. Machler, and E. Frossard. 1999. Assessment of isotopically exchangeable zinc in polluted and non-polluted soils. Soil Sci. Soc. Am. J. 63:1618–1625.
414
World Soil Resources and Food Security
Sinaj, S., O. Traore, and E. Frossard. 2002. Effects of compost and soil properties on the availability of compost phosphate for white clover (Trifolium repens L.). Nutr. Cycling Agroecosyst. 62:89–102. Sisti, C.P.J., H. dos Santos, R. Kohhann, B.J.R. Alves, S. Urquiaga, and R.M. Boddey. 2004. Change in carbon and nitrogen stocks in soil under 13 years of conventional or zero tillage in southern Brazil. Soil Till. Res. 76:39–58. Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. 1997. Soil fertility in Africa is at stake. In Replenishing soil fertility in Africa, An SSSA Special Publication, eds. R.J. Buresh, P.A. Sanchez, and F. Calhoun, 47–61. Madison, WI: SSSA. Smallberger, S.A., U. Singh, S.H. Chien, J. Henao, and P.W. Wilkens. 2006. Development and validation of a phosphate rock decision support system. Agron. J. 98: 471–483. Smernik, R.J., and J.A. Baldock. 2005a. Does solid-state 15N NMR spectroscopy detect all soil organic nitrogen? Biogeochem. 75:507–528. Smernik, R.J., and J.A. Baldock. 2005b. Solid-state 15N NMR analysis of highly 15N-enriched plant materials. Plant Soil 275:271–283. Smith, C.A.S. (ed.). 2003. Soil organic carbon and agriculture: Developing indicators for policy analysis. Proceedings of an expert meeting. Ottawa, Canada. Ottawa: Agriculture and Agri-Food Canada; Paris: Organization for Economic Co-operation and Development. Smolders, E., K. Brans, A. Foldi, and R. Merckx. 1999. Cadmium fixation in soils measured by isotopic dilution. Soil Sci. Soc. Am. J. 63:78–85. Sogon, S., M.J. Penven, P. Bonte, and T. Muxart. 1999 Estimation of sediment yield and soil loss using suspended sediment load and 137Cs measurements on agricultural land, Brie Plateau, France. Hydrobiol. 410:251–261. Stacey, S., G. Merrington, and M.J. McLaughlin. 2001. The effect of aging biosolids on the availability of cadmium and zinc. Eur. J. Soil Sci. 52:313–321. Staddon, Ph. L. 2004. Carbon isotopes in functional soil ecology. Trends Ecol. Evol. 19: 148–154. Stallard, R.F. 1998. Terrestrial sedimentation and the carbon cycle: Coupling weathering and erosion to carbon burial. Global Biogeochem. Cycles 12:231–257. Steiner, C., W.G. Teixeira, J. Lehmann, T. Nehls, J.L. Vasconcelos de Macêdo, W.E.H. Blum, and W. Zech. 2007. Long term effects of manure, charcoal and mineral fertilization on crop production and fertility on a highly weathered Central Amazonian upland soil. Plant Soil 291:275–290. Stewart, D.P.C., and A.K. Metherell. 1999. Carbon 13C uptake and allocation in pasture plants following pulse labeling. Plant Soil 210:61–73. Takkar, P.N., I.M. Chibba, and S.K. Mehta. 1989. Twenty years of co-ordinated research of micronutrients in soil and plants. 1967–1987. Bhopal: Indian Institute of Soil Science. Tans, P. 2002. Oxygen isotopic equilibrium between carbon dioxide and water in soils. Tellus. B. Chem. Phys. Meteorol. 50(92):163–178. Theocharopoulos, S.P., H. Florou, P. Kritidis, D. Belis, F. Tsoloucha, M. Christou, P. Kouloumbis, and T. Nikolaou. 2000. Use of the 137Cs isotopic technique in soil erosion studies in Central Greece. Acta Geol. Hisp. 35:301–309. Till, A.R. 1981. The study of sulphur recycling processes in a grazing system using S. In Animal World Science, B1, Grazing Animals, 33–54. Amsterdam: Elsevier. Tiller, K.G., J.L. Honeysett, and M.P.C. de Vries. 1972. Soil Zn and its uptake by plants: 1. Isotopic exchange equilibria and the application of tracer techniques. Aust. J. Soil Res. 10:151–164. Tilsner, J., N. Wrage, J. Lauf, and G. Gebauer. 2003. Emission of gaseous nitrogen oxides from an extensively managed grassland in NE Bavaria, Germany. Biogeochem. 63: 249–267. Toor, G.S., B.J. Cade-Menun, and J.T. Sims. 2005. Establishing a linkage between phosphorus forms in dairy diets, faeces and manures. J. Environ. Qual. 34:1380–1391.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
415
Tozer, W., K. Wilkins, H. Wang, M. van den Heuvel, T. Charleson, and W.B. Silvester. 2005. Using the 15N to derive a budget for effluent-derived nitrogen applied to forest. Isotopes Environ. Health Stud. 41:13–30. Traore, K., F. Ganry, R. Oliver, and J. Gigou. 2004. Litter production and soil fertility in a catena in Southern Mali. Arid Res. Manag. 18:359–368. Trumbore, S. 2009. Radiocarbon and soil carbon dynamics. Ann. Rev. Earth Planet. Sci. 37:47–66. Turner, B.L. 2004. Optimizing phosphorus characterization in animal manures by solution 31P NMR spectrosocopy. J. Environ. Qual. 33:757–766. Turner, B.L., and A.B. Leytem. 2004. Phosphorus compounds in sequential extracts of animal manures: Chemical speciation and a novel fractionation procedure. Environ. Sci. Technol. 38:6101–6108. The United Nations Convention to Combat Desertification (UNCCD). 2008. Use of biochar (charcoal) to replenish soil carbon pools, restore soil fertility and sequester CO2. Fourth Session of the Ad Hoc Working Group on Long-term Cooperative Action under the Convention (AWG-LCA 4). December 1–10, 2008, Poznan, Poland. UNEP 2010. The United Nations System-wide Earthwatch. http://earthwatch.unep.ch/emerg ingissues/desertification/landdegradation.php. UNESCO-SCOPE. 2006. The global carbon cycle. UNESCO-SCOPE Policy briefs, October 2006, No.2. Paris: UNESCO-SCOPE. United Nations Millennium Development Goals (UNMDG). 2009. http://www.un.org/ millenniumgoals/. U.S. Department of Agriculture-National Resource Conservation Service (USDA-NRCS). 2009. Soil quality. http://soils.usda.gov/sqi. U.S. Environmental Protection Agency (EPA). 2008. Radiation protection. http://www.epa .gov/radiation/radionuclides/. U.S. National Nuclear Data Center (US-NNDC). 2008. Interactive chart of nuclides. Brookhaven National Laboratory. http:// www.nndc.bnl.gov. Van Bavel, C.H.M., G.B. Stirck, and K.J. Brust. 1968. Hydraulic properties of a clay loam soil and the field measurement of water uptake by roots: 1. Interpretation of water content and pressure profiles. Soil Sci. Soc. Am. Proc. 32:310–317. Van Cleemput, O., G. Hofman, and A. Vermoesen (eds.). 1996. Progress in nitrogen cycling. Proceedings of the 8th Nitrogen Workshop. September 5–8, 1994, Ghent, Belgium. Dordrecht, the Netherlands: Kluwer. Van Cleemput, O., F. Zapata, and B. Vanlauwe. 2008. Use of tracer technology in mineral fertilizer management. In Guidelines on nitrogen management in agricultural soils. IAEATCS-29CD, ed. IAEA, 19–126. Vienna, Austria: IAEA. Van Groenigen, J.W., K.B. Zwart, D. Harris, and C. van Kessel. 2005. Vertical gradients of δ15N and δ18O in soil atmospheric N20-temporal dynamics in a sandy soil. Rapid Commun. Mass Sp. 19:1289–1295. Van Groeningen, K.-J., D. Harris, W.R. Horwath, U.A. Hartwig, and C. Van Kessel. 2002. Linking sequestration of C and N in aggregates in a pasture soil following 8 years of elevated CO2. Glob. Change Biol. 11:1094–1108. Van Kessel, C., J. Nitschelm, W.R. Horwath, D. Harris, F. Walley, A. Luscher, and U.A. Hartwig. 2000. Carbon-13 inputs and turnover in a pasture soil exposed to long term elevated atmospheric CO2. Glob. Change Biol. 6:123–135. Van Pelt, A.D., T.A. Rahn, D. Risk, N. McDowell, H. Powers, N.R. Nickerson, K. Bible, C. Rella, and E. Wahl. 2008. Continuous, in situ isotopic carbon dioxide measurements of ecosystem exchange processes and soil respiration using a WS-CRDS Analyzer. American Geophysical Union, Fall Meeting 2008, abstract #B14A-06. Vanlauwe, B., C.A. Palm, H.K. Murwira, and R. Merckx. 2002. Organic resource management in sub-Saharan Africa: Validation of a residue quality-driven decision support system. Agron. 22:839–846.
416
World Soil Resources and Food Security
Vanlauwe, B., J. Wendt, and J. Diels. 2001. Combined application of organic matter and fertilizer. In Sustaining soil fertility in West Africa. SSSA Special Publication No 58, 247– 280. Madison, WI: SSSA. Venteris, E.R., G.W McCarty, J.C. Ritchie, and T.J. Gish. 2004. Soil organic carbon and soil redistribution: Influence of site history and site variables. Soil Sci. 169(11):787–795. Verchot, L.V., and P. Cooper. 2008. International agricultural research and climate change: A focus on tropical systems. Agric. Ecosyst. Environ. 126:1–3. Wallbrink, P.J., W. Blake, S. Doerr, R. Shakesby, G. Humphreys, and P. English. 2005. Using tracer based sediment budgets to assess redistribution of soil and organic material after severe bush fires. In Sediment budgets. IAHS Publication 292, eds. A.J. Horowitz and D.E. Walling, 223–230 Wallingford, UK: IAHS Press. Wallbrink, P.J., C.E. Martin, and C.J. Wilson. 2003. Quantifying the delivery of sediment sediment-P and fertilizer-P from forested, cultivated and pasture areas at the land use and catchment scale using fallout radionuclides and geochemistry. Soil Till. Res. 69:53–68. Wallbrink, P.J., A.S. Murray, and J.M. Olley. 1999. Relating suspended sediment to its original soil depth using fallout radionuclides. Soil Sci. Am. J. 63:369–378. Walling, D.E. 2005. Tracing suspended sediment sources in catchments and river systems. Sci. Total Environ. 344:159–184. Walling, D.E., A.L. Collins, P.A. Jones, G.J.L. Leeks, and G. Old. 2006b. Establishing finegrained sediment budgets for the Pang and Lambourn LOCAR study catchments. J. Hydrol. 330:126–141. Walling, D.E., D. Fang, A.P. Nicholas, R.J. Sweet. 2006a. River floodplains as carbon sinks. In Sediment dynamics and the hydromorphology of fluvial systems. IAHS Publication 306, 460–470. Wallingford, UK: IAHS Press. Walling, D.E., Q. He, and P.J. Whelan. 2003. Using 137Cs measurements to validate the application of the AGNPS and ANSWERS erosion and sediment yield models in two small Devon catchments. Soil Till. Res. 69:27–44. Walling, D.E., P.N. Owens, and G.J.L. Leeks. 1998. The role of channel and floodplain storage in the suspended sediment budget of the River Use, Yorkshire, UK. Geomorphology 22:225–242. Wang, H., G.N. Magesan, P.W. Clinton, and J.M. Lavery. 2005. Using natural 15N abundances to trace the fate of waste-derived nitrogen in forest ecosystems: New Zealand case studies. Isot. Environ. Health Stud. S. 41:31–38. Watson, K.K. 1966. An instantaneous profile method for determining the hydraulic conductivity of unsaturated porous materials. Water Resour. Res. 2:709–715. Webster, E.A., and D.W. Hopkins. 2004. Nitrogen and oxygen isotope ratios of nitrous oxide emitted from soil and produced by nitrifying and denitrifying bacteria. Bio. Fert. Soils 22(4):326–330. West, J.B., G.J. Bowen, T.E. Dawson, and K.P. Tu (eds.). 2010. Understanding movement, pattern, and processes on earth through isotope mapping. New York: Springer. Wiesenberg, G.L.B., J. Scharzbauer, M.W.I. Schmidt, and L. Schwark. 2008. Plant and soil lipid modification under elevated atmospheric CO2 conditions: II. Stable carbon isotopic values (δ13C) and turnover. Organic Geochem. 39:103–117. Williams, D.G., W. Cable, K. Hultine, J.C.B. Hoedjes, E.A. Yepez, V. Simonneaux, S. Er-Raki, G. Boulet, H.A.R. De Bruin, A. Chehbouni, O.K. Martogensis, and F. Timouk. 2004. Evapotranspiration components determined by stable isotopes, sap flow and eddy covariance techniques. Agric. For. Meteorol. 125:241–258. Wilson, M.A., S. Heng, K.M. Goh, R.G. Pugmire, and D.M. Grant. 1983. Studies of litter and acid insoluble soil organic matter fractions using 13C-cross polarisation nuclear magnetic resonance spectroscopy with ‘magic angle’ spinning. J. Soil Sci. 34:83–97.
Role of Nuclear and Isotopic Techniques in Sustainable Land Management
417
Woolf, D. 2008. Biochar as a Soil Amendment—A review of the Environmental Implications, January 2008. Swansea University, United Kingdom. World Meteorological Organization (WMO). Global Atmosphere Watch (GAW). 2007. 14th WMO/IAEA Meeting of experts on carbon dioxide, other greenhouse gases and related tracers measurement techniques. GAW Report No. 186. Geneva, Switzerland. http:// www.wmo.int/pages/prog/arep/gaw/gaw-reports.html. Wrage, N., J. Lauf, A. Del Prado, M. Pinto, S. Pietrzak, S. Yamuki, O. Oenema, and G. Gebauer. 2004. Distinguishing sources of N2O in European grasslands by stable isotope analysis. Rapid Comm. Mass Sp. 18:1201–1207. Wrage, N., J.W. Van Groenigen, O. Oenema, and E.M. Baggs. 2005. A novel dual isotope labeling method for distinguishing between soil sources of N2O. Rapid Comm. Mass Sp. 19:3298–3306. Wright, G., R.C.N. Rao, and G.D. Farquhar. 1994. Water-use efficiency and carbon isotope discrimination in peanut under water deficit conditions. Crop Sci. 34:92–97. Wynn, J.G., and M.I. Bird. 2007. C4 derived soil organic carbon decomposes faster than its C3 counterpart in mixed soils. Glob. Change Biol. 13:2206–2217. Yasmin, N. 2003. Availability of fertilizer S and suitable time of application for rice, mustard and cotton. In Guidelines for the use of isotopes of sulphur in soil-plant studies. IAEA TCS-20, ed. IAEA, 97–98. Vienna, Austria: IAEA. Yepez, E.A., D.G. Williams, R.L. Scott, and G. Lin. 2003. Partitioning overstory and understory evapotranspiration in a semiarid savanna woodland from the isotopic composition of water vapour. Agric. For. Meteorol. 119:53–68. Young, S.D., A. Tye, A. Carstensen, L. Resende, and N. Crout. 2000. Methods for determining labile Cd and Zn in soil. Eur. J. Soil Sci. 51: 129–136. Zaharah, A.R., and F. Zapata. 2003. The use of P-32 isotope techniques to study soil P dynamics and to evaluate the agronomic effectiveness of phosphate fertilizers. In Direct application of phosphate rock and related technology: Latest developments and practical experiences. Proceedings of an International Meeting, July 16–20, 2001, Kuala Lumpur. Special Publications IFDC-SP-37, eds. S.S.S. Rajan and S.H. Chien, 225–235. Muscle Shoals, AL: IFDC. Zapata, F. 1990. Isotope techniques in soil fertility and plant nutrition studies. In Use of nuclear techniques in studies of soil-plant relationships. IAEA training course series No. 2, ed. IAEA, 61–128, Vienna, Austria: IAEA. Zapata, F. (ed.). 1995. Evaluation of the agronomic effectiveness of phosphate fertilizers through the use of nuclear and related techniques. Special issue. Fert. Res. 41:167–242. Zapata, F., and H. Axmann. 1995. P-32 isotopic techniques for evaluating the agronomic effectiveness of rock phosphate materials. Fert. Res. 41:189–195. Zapata, F., and C. Hera. 1996. FAO/IAEA International networking in soil/plant nitrogen research. In Nitrogen economy in tropical soils, ed. N. Ahmad, 415–425. Dordrecht, the Netherlands: Kluwer. Zapata, F. (ed.). 2002a. Handbook for the assessment of soil erosion and sedimentation using environmental radionuclides. Dordrecht, the Netherlands: Kluwer. Zapata, F. (ed.). 2002b. Utilisation of phosphate rocks to improve soil P status for sustainable crop production in acid soils. Nutr. Cycling Agroecosyst. Special issue 63:1–98. Zapata, F. (ed.). 2003. Field application of the Cs-137 technique in soil erosion and sedimentation. Soil Till. Res. Special Issue 69:1–153. Zapata, F. 2004. Soil erosion and sedimentation studies using fallout radionuclides: Recent developments and future applications. In Agricultural impacts on soil erosion and soil biodiversity. Developing indicators for policy analyses, ed. R. Francaviglia, 59–74. Proceedings of the OECD Expert Meeting, March 2003, Rome, Italy. Zapata, F., and R.N. Roy (eds.). 2004. Use of phosphate rocks for sustainable agriculture. FAO fertilizer and plant nutrition bulletin No 13. Rome: FAO.
418
World Soil Resources and Food Security
Zapata, F., and Y. Li. 2007. Report of the fourth and final research co-ordination meeting of the FAO/IAEA Co-ordinated Research Project to Assess the Effectiveness of Soil Conservation Technologies for Sustainable Watershed Management using Fallout Radionuclides. IAEA-311-D1-RC 888. Vienna, Austria: IAEA. Zapata, F., and M.L. Nguyen. 2009. Soil erosion and sedimentation studies using environmental radionuclides. In Environmental radionuclides: Tracers and timers of terrestrial processes, 295–322, Vol. 16 of Radioactivity in the environment. Vienna, Austria: IAEA. Zech, M., and B. Glaser. 2008. Improved compound-specific analysis of n-alkanes for application in palaeoenvironmental studies. Rapid Comm. Mass Sp. 22:135–142. Zhang, X.B., Y. Zhang, A.B. Wen, and M. Feng. 2003. Assessment of soil erosion losses on cultivated land by using the 137Cs technique in the Upper Yangtze River Basin of China. Soil Till. Res. 69:99–105. Zhao, F.J., J.S. Knights, Z.Y. Hu, and S.P. McGrath. 2003. Stable sulphur isotope ratio indicates long term changes in sulphur deposition in the Broadbalk Experiment since 1845. J. Environ. Qual. 32:33–39. Zheng, J.-J., X. He, D.E. Walling, X. Zhang, D. Flanagan, and Y.Q. Qi. 2007. Assessing soil erosion rates on manually-tilled hillslopes in the Sichuan hilly basin using 137Cs and 210Pb measurements. Pedosphere 17:273–283. ex
Paradigm to 9 New Unlock the Potential of Rainfed Agriculture in the Semiarid Tropics Suhas P. Wani, Johan Rockstrom, B. Venkateswarlu, and A. K. Singh CONTENTS 9.1 Challenges for Humankind in the Twenty-First Century.............................. 420 9.1.1 High and Increasing Risk with Climate Change............................... 420 9.2 Current Status of Rainfed Agriculture.......................................................... 423 9.2.1 Crop Yields in Rainfed Areas........................................................... 424 9.2.2 Rainfed Agriculture—A Large Untapped Potential.......................... 426 9.2.3 Constraints in Rainfed Agriculture Areas......................................... 427 9.3 Strategies for Harnessing the Potential of Rainfed Agriculture.................... 428 9.3.1 Understand the Issue of Water Scarcity............................................. 428 9.3.2 Water Alone Cannot Achieve Food Security.................................... 431 9.3.2.1 Soil Health: An Important Driver for Enhancing Water Use Efficiency..................................................................... 431 9.3.3 Water Resources Management.......................................................... 436 9.3.4 Shifting Nonproductive Evaporation to Productive Transpiration.... 437 9.4 New Paradigm Is a Must for Unlocking the Potential of Rainfed Agriculture..................................................................................................... 437 9.4.1 New Paradigm for Water Management in Rainfed Agriculture........ 437 9.4.2 Holistic Watershed Approach through Integrated Genetic and Natural Resource Management (IGNRM)........................................ 439 9.4.3 Watershed Development—A Growth Engine for Development of Rainfed Areas................................................................................ 441 9.4.4 Common Features of the Watershed Development Model................ 441 9.4.5 Recent Additions to the Watershed Model........................................ 442 9.4.6 Learning from Meta-Analysis of Watersheds in India...................... 443 9.4.7 Results of Meta-Analysis Regression................................................446 9.4.8 Business Model..................................................................................448 9.4.9 Promoting Collective Action in the Community...............................448 9.4.9.1 Convergence and Collective Action....................................448 419
420
World Soil Resources and Food Security
9.4.9.2 Consortium for Technical Backstopping............................449 9.4.9.3 Discard Artificial Divide between Irrigated and Rainfed Agriculture............................................................449 9.4.9.4 Pilot-Scale Model Community Watershed—A Site of Learning..............................................................................449 9.4.10 Multiple Benefits and Impacts........................................................... 450 9.4.10.1 Reducing Rural Poverty...................................................... 450 9.4.10.2 Increasing Crop Productivity.............................................. 452 9.4.10.3 Improving Water Availability............................................. 452 9.4.10.4 Supplemental Irrigation...................................................... 454 9.4.10.5 Sustaining Development and Protecting the Environment........................................................................ 455 9.4.10.6 Conserving Biodiversity..................................................... 461 9.4.11 Scaling-Up......................................................................................... 461 9.4.12 Unlocking the Potential of Rainfed Agriculture—A Beginning Is Made in India................................................................................. 463 Acknowledgments...................................................................................................464 References...............................................................................................................464
9.1 C HALLENGES FOR HUMANKIND IN THE TWENTY-FIRST CENTURY The twenty-first century has presented complex and multiple challenges for humankind and the main challenge is to achieve food security for all with scarce water resources and increasing land degradation. The world is facing a severe water scarcity that is already complicating national and global efforts to achieve food security in several parts of the world. Agriculture is the world’s second largest consumer of water after forestry. The second important factor controlling world food production is soil health, which is severely affected due to land degradation. The growing human population is reducing the per capita availability of land as well as water differently in different parts of the world. Growing water scarcity and land degradation are emerging as the major biophysical factors affecting food security in the world.
9.1.1 High and Increasing Risk with Climate Change Rainfed agriculture in the semiarid tropics is a fragile and risk-prone ecosystem due to high spatial and temporal variability of rainfall. Rainfall is concentrated in a short rainy season (approximately 3 to 5 months), with few intensive rainfall events that are unreliable in temporal distribution that is manifested by high deviations from the mean rainfall (coefficients of variation of rainfall are as high as 40% in semiarid regions) [Wani et al. 2004]. In fact, even if water is not always the key limiting factor for yield increase, rainfall is the only truly random production factor in the agricultural system. This is manifested through high rainfall variability causing recurrent flooding, droughts, and dry spells. Established but incomplete evidence suggests that the high risk for water-related yield loss makes farmers risk averse, which in turn determines farmers’ perceptions on investments in other production factors (such as labor and improved seed
New Paradigm to Unlock the Potential of Rainfed Agriculture
421
and fertilizers). Smallholder farmers are usually aware of the effects of shortage and/or variability of soil moisture on the variety, quantity, and quality of produce, leading to a very narrow range of options for commercialization. This, together with the fluctuations in yields, makes it hard for resource-poor men and women in semiarid areas to respond effectively to opportunities made possible by emerging markets, trade, and globalization. Therefore, temporal and spatial variability of climate, especially rainfall, is a major constraint to yield improvements, competitiveness, and commercialization of rainfed crops, tree crops, and livestock systems in most of the tropics. Management options should therefore start by focusing on reducing rainfall-induced risks. Evidence is emerging that climate change is making the variability more intense with increased frequency of extreme events such as drought, floods, and hurricanes [IPCC 2007]. A recent study assessing rainfed cereal potential under different climate change scenarios, with varying total rainfall amounts, concluded that it is difficult to estimate the exact degree of regional impact. But most scenarios resulted in losses of rainfed production potential in the most vulnerable developing countries. In these countries, the loss of production area was estimated at 10%–20%, with an approximate potential of 1–3 billion people affected by 2080 [IIASA 2002]. In particular, sub-Saharan Africa is estimated to lose 12% of the cultivation potential. This loss is mostly projected in the Sudan–Sahelian zone, which is already subject to high climatic variability and adverse crop conditions. Because of the risk associated with climate variability, smallholder farmers are generally and rationally keen to start reducing risk of crop failure due to dry spells and drought before they consider investments in soil fertility, improved crop varieties, and other yieldenhancing inputs [Hilhost and Muchena 2000]. Global warming and associated impacts of climate change will have adverse impacts on water availability and food production and, here again, the developing tropical region countries are likely to be affected more by impacts of climate change [IPCC 2007]. Poverty and food security are very closely related and are affected by the growing water scarcity and land degradation, which are also affected by growing population pressure on the limited land and water resources essential for food production. An adequate human diet requires about 4000 liters of water per day to produce, which is over 90% of the daily human water requirement. The increasing water scarcity resulting from population growth, rising incomes, and climate change limits the amount of water available for food production and threatens food security in many countries. To meet the food demand of the world’s growing population and rising incomes with the current production options, we will need an additional 1600 km3 water per year just to achieve the UN Millennium Development Goal of halving hunger by 2015 [SEI 2005], and another 4500 km3/yr with current water productivity levels in agriculture to feed the world in 2050 (Figure 9.1) [Falkenmark et al. 2009; Rockström et al. 2009]. This is more than twice the current consumptive water use in irrigation, which is already contributing to depleting several large rivers before they reach the ocean. It is becoming increasingly difficult on social, economic, and environmental grounds to supply more water to farmers. There is a correlation between poverty, hunger, and water stress [Falkenmark 1986]. The UN Millennium Development Project has identified the “hot spot” countries in the world suffering from the largest prevalence of malnourishment. These countries coincide closely with the countries in the world hosted in semiarid and dry subhumid hydroclimates in the world (Figure 9.2),
422
World Soil Resources and Food Security
Undernourished
as % of total popn. 2001 <5 5–20 20–35 >35 Koeppen Climate Zones
Savanna/Steppe
FIGURE 9.1 The prevalence of undernourished in developing countries (as percentage of population 2001/2002), together with the distribution of semiarid and dry subhumid hydroclimates in the world, i.e., savannah and steppe agroecosystems. These regions are dominated by sedentary farming subject to the world’s highest rainfall variability and occurrence of dry spells and droughts. (From SEI, Sustainable Pathways to Attain the Millennium Development Goals—Assessing the Role of Water, Energy and Sanitation. Document prepared for the UN World Summit, Sept. 14, 2005, New York, SEI, Stockholm, Sweden, http://sei-international .org/mediamanager/documents/Publications/Water-sanitation/sustainable_pathways_mdg .pdf, 2005. With permission.)
Savanna/Steppe
Dry tropical/temperate Wet tropical Dry tropical (temperate winter)
Other
Desert Other
FIGURE 9.2 The zone with savanna type hydroclimate – the zone with large hunger eradication challenges but also huge potential for additional food production. (From SEI, Sustainable Pathways to Attain the Millennium Development Goals—Assessing the Role of Water, Energy and Sanitation. Document prepared for the UN World Summit, Sept. 14, 2005, New York, SEI, Stockholm, Sweden, http://sei-international.org/mediamanager/documents/ Publications/Water-sanitation/sustainable_pathways_mdg.pdf, 2005. With permission.)
423
New Paradigm to Unlock the Potential of Rainfed Agriculture
i.e., savannahs and steppe ecosystems, where rainfed agriculture is the dominating source of food, and where water constitutes a key limiting factor to crop growth [SEI 2005]. Of the 850 million undernourished people in the world, essentially all live in poor developing countries, which are predominantly located in tropical regions [UNStat 2005]. Drought (water scarcity) and land degradation are interlinked in a cause-and-effect relationship, and the two combined are the main causes of poverty in farm households. This unholy nexus between drought (water scarcity), poverty, and land degradation has to be broken to meet the Millennium Development Goal of halving the number of food insecure poor by 2015 [Wani et al. 2006].
9.2 CURRENT STATUS OF RAINFED AGRICULTURE The importance of rainfed agriculture varies regionally, but it produces most food for poor communities in developing countries. In sub-Saharan Africa, more than 95% of the farmed land is rainfed, while the corresponding figure for Latin America is almost 90% and is about 60% for South Asia, 65% for East Asia, and 75% for Near East and North Africa [FAOStat 2010] (Table 9.1). Most countries in the world TABLE 9.1 Global and Continent-Wide Rainfed Area and Percentage of Total Arable Land Continent Regions World Africa Northern Africa Sub-Saharan Africa Americas Northern America Central America and Caribbean Southern America Asia Middle East Central Asia Southern and Eastern Asia Europe Western and Central Europe Eastern Europe Oceania Australia and New Zealand Other Pacific Islands
Total Arable Land (million ha)
Rainfed Area (million ha)
% of Rainfed Area
1551.0 247.0 28.0 218.0 391.0 253.5 15.0 126.0 574.0 64.0 40.0 502.0 295.0 125.0 169.0 46.5 46.0 0.57
1250.0 234.0 21.5 211.0 342.0 218.0 13.5 114.0 362.0 41.0 25.5 328.0 272.0 107.5 164.0 42.5 42.0 0.56
80.6 94.5 77.1 96.7 87.5 86 87.7 90.8 63.1 63.4 63.5 65.4 92.3 85.8 97.1 91.4 91.3 99.3
Source: FAO, AQUASTAT: FAO’s Information System on Water and Agriculture, http://www.fao.org/nr/ aquastat, 2010; FAO, FAOStat, http://faostat.fao.org/site/567/DesktopDefauly.aspx?pageID= 567#ancor, 2010.
424
World Soil Resources and Food Security
depend primarily on rainfed agriculture for their grain food. Despite large strides made in improving productivity and environmental conditions in many developing countries, a great number of poor families in Africa and Asia—where rainfed agriculture is the main agricultural activity—still face poverty, hunger, food insecurity, and malnutrition. These problems are exacerbated by adverse biophysical growing conditions and the poor socioeconomic infrastructure in many areas in the arid and semiarid tropics (SAT) and the subhumid regions. Even with growing urbanization, globalization, and better governance in Africa and Asia, hunger, poverty, and vulnerability of livelihoods to natural and other disasters will continue to be greatest in the rural tropical areas. These challenges are complicated by climatic variability, the risk of climate change, population growth, health pandemics (AIDS, malaria), a degrading natural resource base, poor infrastructure, and changing patterns of demand and production [Ryan and Spencer 2001; Walker 2010]. The majority of poor in developing countries live in rural areas; their livelihoods depend on agriculture and overexploitation of the natural resource base, pushing them downward into a spiral of poverty. The importance of rainfed sources of food weighs disproportionately on women, given that approximately 70% of the world’s poor are women [WHO 2000]. Agriculture plays a key role in economic development [World Bank 2005] and poverty reduction [Irz and Roe 2000], with evidence indicating that every 1% increase in agricultural yields translates to a 0.6% to 1.2% decrease in the percentage of absolute poor [Thirtle et al. 2002]. On an average for sub-Saharan Africa, agriculture accounts for 35% of the gross domestic product (GDP) and employs 70% of the population [World Bank 2000], while more than 95% of the agricultural area is rainfed [FAOStat 2010]. Agriculture will continue to be the backbone of economies in Africa and South Asia for the foreseeable future. As most of the poor are farmers and landless laborers [Sanchez et al. 2005], strategies for reducing poverty, hunger, and malnutrition should be driven primarily by the needs of the rural poor, and should aim to build and diversify their livelihood sources. Substantial gains in land, water, and labor productivity, as well as better management of natural resources, are essential to reverse the downward spiral of poverty and environmental degradation, apart from the problems of equity, poverty, and sustainability and, hence, the need for greater investment in SAT areas [World Bank 2005; Rockström et al. 2007; Wani et al. 2008a, 2009].
9.2.1 Crop Yields in Rainfed Areas Over the past 40 years, agricultural land use has expanded by 20%–25%, which has contributed approximately 30% to the overall grain production growth during the period [FAO 2002]. The remaining yield outputs originated from intensification through yield increases per unit land area. However, the regional variation is large, as is the difference between irrigated and rainfed agriculture. In developing countries, rainfed grain yields are on an average 1.5 t/ha, compared to 3.1 t/ha for irrigated yields [Rosegrant et al. 2002], and the increase in production from rainfed agriculture has mainly originated from land expansion. Trends are clearly different for different regions. With 99% rainfed production of main cereals such as maize, millet, and sorghum, the cultivated cereal area in sub-
425
New Paradigm to Unlock the Potential of Rainfed Agriculture
Saharan Africa has doubled since 1960, while the yield per unit land has nearly been stagnant for these staple crops [FAOStat 2010]. In South Asia, there has been a major shift away from more drought-tolerant, low-yielding crops such as sorghum and millet, while wheat and maize has approximately doubled in area since 1961 [FAOStat 2010]. During the same period, the yield per unit land for maize and wheat has more than doubled (Figure 9.3). For predominantly rainfed systems, maize crops per unit land have nearly tripled and wheat more than doubled during the same time period. (a)
5 SSA
Maize yield (t/ha)
4
Lat Am. S.Asia
3
2
1
0 1960 (b)
6
Wheat yield (t/ha)
5 4
1970
1980
Year
1990
2000
2010
SSA Lat Am. S.Asia Europe N.America
3 2 1 0 1960
1965
1970
1975
1980
1985 Year
1990
1995
2000
2005
2010
FIGURE 9.3 Grain yield of predominantly rainfed maize and wheat for different regions during 1961–2010. (From FAO, FAOStat, http://faostat.fao.org/site/567/DesktopDefauly .aspx?pageID=567#ancor, 2010. With permission.)
426
World Soil Resources and Food Security
Rainfed maize yield differs substantially between regions (Figure 9.3a). In Latin America (including the Caribbean) it exceeds 3 t/ha, while in South Asia it is around 2 t/ha, and in sub-Saharan Africa it only just exceeds 1 t/ha. This can be compared with maize yields in the United States or Southern Europe, which normally amount to approximately 7–10 t/ha (most maize in these regions is irrigated). The average regional yield per unit land for wheat in Latin America (including the Caribbean) and South Asia is similar to the average yield output of 2.5–2.7 t/ha in North America (Figure 9.3b). In comparison, wheat yield in Western Europe is approximately twice as large (5 t/ha), while in sub-Saharan Africa it remains below 2 t/ha. In view of the historic regional difference in development of yields, a significant potential appears to exist for raised yields in rainfed agriculture, particularly in sub-Saharan Africa and South Asia.
9.2.2 Rainfed Agriculture—A Large Untapped Potential In several regions of the world, rainfed agriculture generates yields among the world’s highest. These are predominantly temperate regions, with relatively reliable rainfall and inherently productive soils. Even in tropical regions, particularly in the subhumid and humid zones, agricultural yields in commercial rainfed agriculture exceed 5–6 t/ha [Rockström and Falkenmark 2000; Wani et al. 2003a, 2003b] (Figure 9.4). At the same time, the dry subhumid and semiarid regions have experienced the lowest yields and the weakest yield improvements per unit land. Here, yields oscillate between 0.5 to 2 t/ha, with an average of 1 t/ha in sub-Saharan Africa and 1–1.5 t/ha in South Asia, Central Asia, and West Asia and North Africa (CWANA) for rainfed agriculture [Rockström and Falkenmark 2000; Wani et al. 2003a, 2003b]. 8
Yield (t ha–1)
6
4
BW1
Carrying capacity 21 persons ha–1
Rate of growth 82 kg ha–1 y–1
Carrying capacity 4.6 persons ha–1
2 BW4C
Rate of growth 23 kg ha–1 y–1
0 1976
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
2009
Year
FIGURE 9.4 Three-year moving average of crop yields in improved and traditional management systems during 1976–2009 at ICRISAT, India.
427
New Paradigm to Unlock the Potential of Rainfed Agriculture 5000
Max. rainfed potential Mean rainfed potential District mean
4000 3000
3000
2000
2000
1000 0
Max. rainfed potential Mean rainfed potential District mean
4000
Yield (kg ha–1)
Yield (kg ha–1)
5000
1000 Soybean Groundnut Pigeonpea Chickpea
0
Kharif sorghum Rabi sorghum
Legumes
Pearl millet
Sorghum and Pearl millet
7000 Max. rainfed potential Mean rainfed potential District mean
Yield (kg ha–1)
6000 5000 4000 3000 2000 1000 0
Rice
Cotton
Mustard
Wheat
Other crops
FIGURE 9.5 Rainfed potential yields and yield gaps of crops in India. (From Singh, P., et al., Yield Gap Analysis: Modeling of Achievable Yields at Farm Level in Rain-Fed Agriculture: Unlocking the Potential, CAB International, Wallingford, UK, 81–123, 2009. With permission.)
Yield gap analyses carried out for comprehensive assessment (CA) for major rainfed crops in semiarid regions in Asia and Africa and rainfed wheat in West Africa and North Africa (WANA), revealed large yield gaps, with farmers’ yields being a factor 2 to 4 times lower than achievable yields for major rainfed crops (Figures 9.5 and 9.6). Detailed yield gap analysis for major rainfed crops in different parts of the world are discussed [Singh et al. 2009; Fisher et al. 2009]. In countries in Eastern and Southern Africa, the yield gap is very large (Figure 9.6). Similarly, in many countries in West Asia, farmers’ yields are less than 30% of achievable yields, while in some Asian countries, the figure is closer to 50%. Historic trends present a growing yield gap between farmers’ practices and farming systems that benefit from management advances [Wani et al. 2003b, 2009a].
9.2.3 Constraints in Rainfed Agriculture Areas An insight into the inventories of natural resources in rainfed regions shows a grim picture of water scarcity, fragile ecosystems, drought, and land degradation due to soil erosion by wind and water, low rainwater use efficiency (35%–45%), high population pressure, poverty, low investments in water use efficiency (WUE) measures, poor infrastructure, and inappropriate policies [Wani et al. 2003b, 2003c, 2009; Rockström et al. 2007]. These rainfed areas are also prone to severe land degradation.
428
World Soil Resources and Food Security 100
Observed yield gap (%)
90 80 70 60 50 40 30 20
Yemen
Syria
Pakistan
Jordan
Morocco
Iraq
Iran
India
Vietnam
Thailand
Botswana
Zimbabwe
Niger
Burkina Faso
Uganda
Ethiopia
Kenya
Zambia
0
Tanzania
10
FIGURE 9.6 Examples of observed yield gap (for major grains) between farmers’ yields and achievable yields (100% denotes achievable yield level and columns contain the actual observed yield levels). (From Rockström, J., et al., In Water for Food, Water for Life: A Comprehensive Assessment of Water Management in Agriculture, Earthscan and IWMI, London and Colombo, Sri Lanka, 315–348, 2007. With permission.)
Reduction in the producing capacity of land due to wind and water erosion of soil, loss of soil humus, depletion of soil nutrients, secondary salinization, diminution and deterioration of vegetation cover, as well as loss of biodiversity, is referred to as land degradation. The root cause of land degradation is inappropriate land use and management practices. Land degradation represents a diminished ability of ecosystems or landscapes to support the functions or services required for sustaining livelihoods. A global assessment of the extent and form of land degradation showed that 57% of the total area of drylands occurring in two major Asian countries, namely China (178.9 million ha) and India (108.6 million ha), are degraded [UNEP 1997].
9.3 S TRATEGIES FOR HARNESSING THE POTENTIAL OF RAINFED AGRICULTURE 9.3.1 Understand the Issue of Water Scarcity Water scarcity is a relative concept. Using the conventional approach and assessing the amount of renewable surface and groundwater per capita (i.e., so-called blue water), suggests that water stress is increasing in a number of countries and that regions are moving into increasing water-stressed conditions. Although the global amount of fresh water has not changed, the amount available per person is much less than it was in 1950, with a significant difference between countries and regions. Water is not equally scarce in all parts of the world. As Figure 9.7a illustrates, Southeast Asia and the Middle East North Africa (MENA) regions are the worst affected in terms of blue water scarcity. However, this picture may be misleading
New Paradigm to Unlock the Potential of Rainfed Agriculture
429
(a)
Blue water year 2000 (m3 cap–1 y–1) 1 – 500 500 – 1000 1000 – 2000 2000 – 5000 5000 – 10000 10000 – 50000 50000 – 100000 > 100000
(b)
Savanna/steppe
Dry tropical/temperate Wet tropical Dry tropical (temperate winter)
Others
Desert Other
FIGURE 9.7 (a) Renewable liquid freshwater (blue water – hatched in dark) stress per capita using LPJ dynamic modeling year 2000. (From Rockström, J., et al., Water Resources Research, 45, W00A12, 2009.) (b) Renewable rainfall (green and blue water – hatched in semi dark and dark) stress per capita using LPJ dynamic modeling year 2000. (From Rockström, J., et al., Water Resources Research, 45, W00A12, 2009.)
because the average amount of water per capita in each pixel could obscure large differences in actual access to a reliable water source. In addition, these water quantities only include blue water. The full resource of rainfall, and notably green water, i.e., soil moisture used in rainfed cropping and natural vegetation, is not included. In a recent assessment that included both green and blue water resources, the level of water scarcity changed significantly for many countries (Figure 9.7b). Among the regions that are conventionally (blue) water scarce, but still have sufficient green and blue water to meet the water demand for food production, are large parts of sub-Saharan Africa, India, and China. If green water (on current agricultural land) for food production is included, per capita water availability in countries such as Uganda, Ethiopia, Eritrea, Morocco, and Algeria more than doubles or triples. Moreover, low ratios of transpiration to evapotranspiration (T/ET) in countries
430
World Soil Resources and Food Security
Water productivity (m3/t grain)
such as Bangladesh, Pakistan, India, and China indicate high potential for increasing water productivity through vapor shift [Rockström et al. 2009]. Absolute water stress is found most notably in arid and semiarid regions with high population densities such as parts of India, China, and the MENA region. The MENA region is increasingly unable to produce the food required locally due to increasing water stress from a combination of population increase, economic development, and climate change, and will have to rely more and more on food (and virtual water) imports. For the greater part of the world, the global assessment of green and blue water suggests that water stress is primarily a blue water issue and large opportunities are still possible in the management of rainfed areas, i.e., the green water resources in the landscape [Rockström et al. 2009]. The current global population that has blue water stress is estimated to be 3.17 billion, expected to reach 6.5 billion in 2050. If both green and blue water are considered, the number currently experiencing absolute water stress is a fraction of this (0.27 billion), and will only marginally exceed today’s blue water stressed in 2050. Given the increasing pressures on water resources and the increasing demands for food and fiber, the world must succeed in producing more food with less water. Hence, it is essential to increase water productivity in both humid and arid regions. Some describe the goal as increasing the “crop per drop” or the “dollars per drop” produced in agriculture. Regardless of the metric, it is essential to increase the productivity of water and other inputs in agriculture. Success will generate greater agricultural output, while also enabling greater use of water in other sectors and in efforts to enhance the environment. Water productivity can vary with household income, as farmers’ yields vary as a result of local input and management styles. In a household level study of 300 farmers in eight sub-Saharan countries, the more wealthy farmers had generally higher yield levels [Holmen 2004], and subsequently better water productivity (Figure 9.8). The differences were significant between the wealthier classes and poorest classes. 3000 2500 2000 1500 1000 500 0
Very poor
Below average wealth
Average wealth
Above average
Very wealthy
FIGURE 9.8 Water productivity for maize yields and income levels for smallholder farming systems in sub-Sahara Africa. (Based on Holmen, H., Currents, 34, 12–16, 2004.)
New Paradigm to Unlock the Potential of Rainfed Agriculture
431
More than 1000 m3 additional water was required per ton of maize grain produced by the poorest farmers compared to the wealthiest farmers. Data suggest that yield improvements for the purpose of poverty alleviation can also significantly improve water productivity, especially in current low-yielding rainfed (green water) agriculture in sub-Saharan Africa and parts of South Asia. Improved water use efficiency and productivity can improve food security. A sectoral approach to managing water is a cause of low water use efficiency.
9.3.2 Water Alone Cannot Achieve Food Security 9.3.2.1 S oil Health: An Important Driver for Enhancing Water Use Efficiency Soil health is severely affected due to land degradation and is in need of urgent attention. Often, soil fertility is the limiting factor to increased yields in rainfed agriculture [Stoorvogel and Smaling 1990; Rego et al. 2005]. Soil degradation through nutrient depletion and loss of organic matter causes serious yield decline, closely related to water determinants as it affects water availability for crops due to poor rainfall infiltration and plant water uptake due to weak roots. Nutrient mining is a serious problem in smallholder rainfed agriculture. In sub-Saharan Africa, soil nutrient mining is particularly severe. It is estimated that approximately 85% of African farmland in 2002–2004 experienced a loss of more than 30 kg/ha of nutrients per year [IFDC 2006]. The International Crops Research Institute for the Semi-Arid Tropics’ (ICRISAT) on-farm diagnostic work in different community watersheds in different states of India as well as in Southern China, North Vietnam, and Northeast Thailand showed severe mining of soils for essential plant nutrients. Exhaustive analysis in selected states in India showed that 80%–100% of farmers’ fields are deficient not only in total nitrogen but also micronutrients like zinc, boron, and secondary nutrients such as sulfur beyond the critical limits (Table 9.2a and b) [Rego et al. 2007; Sahrawat et al. 2007]. A substantial increase in crop yields was experienced after micronutrient amendments, and a further increase of 70% to 120% occurred when both micronutrients and adequate nitrogen and phosphorus were applied to a number of rainfed crops (maize, sorghum, mung bean, pigeonpea, chickpea, castor, and groundnut) in farmers’ fields [Rego et al. 2005; Sahrawat et al. 2007; Srinivasarao et al. 2010]. Evidence from on-farm participatory trials in different rainfed areas in India clearly indicated that investments in soil fertility improvement directly improved water management, resulting in increased rainwater productivity. Rainwater productivity (i.e., for grain yield per mm of rainfall) was significantly increased in the example above as a result of micronutrient amendment. The rainwater productivity for grain production was increased by 70%–100% for maize, groundnut, mungbean, castor, and sorghum by adding boron, zinc, and sulfur [Rego et al. 2005]. In terms of net economic returns, rainwater productivity was substantially higher by 1.50 to 1.75 times. Similarly, rainwater productivity was increased significantly when integrated land, nutrient, and water management options were adopted as well as use of improved cultivars in semiarid regions of India [Wani et al. 2003; Sreedevi and Wani 2009]. Gains in rainwater use efficiency with improved land, nutrient, and water management options were far higher in low rainfall years (Figure 9.9).
432
TABLE 9.2a Percent of Farmers’ Fields Deficient in Available Nutrients in Various States (Districts within States) of India State Andhra Pradesh
Chattisgarh Gujarat Jharkhand Karnataka
Rajasthan Tamilnadu Total
Adilabad, Ananthapuram, Kadapa, Khammam, Kurnool, Mahabubnagar, Medak, Nalgonda, Prakasam, Rangareddy, Warangal Kanker Junagadh Gumla, Kharsawan Bengaluru, Rural, Bijapur, Chamrajanagar, Chikballapur, Chitradurga, Dharwad, Haveri, Kolar, Raichur, Tumkur Kollam, Pathanamthitta, Thiruvananthapuram Badwani, Dewas, Guna, Indore, Jhabua, Mandla, Raisen, Rajagarah, Sagar, Sehore, Shajapur, Vidisha Alwar, Banswara, Bhilwara, Bundi, Dungarpur, Jhalwar, Sawai Madhopur, Tonk, Udaipur Kanchipuram, Karur, Salem, Tirunelveli, Vellore
OC a %
Av P ppm
Av K ppm
Av S ppm
Av B ppm
Av Zn ppm
3650
76 b
38
12
79
85
69
40 82 115 17,712
– 12 42 70
63 60 65 46
10 10 50 21
90 46 77 84
95 100 97 67
50 85 71 55
28 341
11 22
21 74
7 1
96 74
100 79
18 66
421
38
45
15
71
56
46
119 22,508
57 69
51 45
24 19
71 83
89 70
61 58
Source: Rego, T.J., et al., Journal of Plant Nutrition, 30, 1569–1583, 2007; Sahrawat, K.L., et al., Current Science, 93(10), 1–6, 2007; Wani, S.P., et al., In Conservation Farming: Enhancing Productivity and Profitability of Rain-Fed Areas, Soil Conservation Society of India, New Delhi, 163–178, 2008; and unpublished data sets of ICRISAT. a OC = organic carbon; AvP = available phosphorus; AvK = available potash; AvS = available sulfur; AvB = available boron, AVZn = available zinc. b = Per cent of farmers fields deficient, i.e., below critical limit for a particular nutrient. * = Extensive soil sampling undertaken to interpolate analysis at district level using GIS.
World Soil Resources and Food Security
Kerala Madhya Pradesh
District
No. of Farmers
State Andhra Pradesh Mean Range Chattisgarh Mean Range Gujarat Mean Range Jharkhand Mean Range Karnataka
Mean Range Kerala Mean Range
District Adilabad, Ananthapuram, Kadapa, Khammam, Kurnool, Mahabubnagar, Medak, Nalgonda, Prakasam, Rangareddy, Warangal
Kanker
Junagadh
Gumla, Kharsawan
Bengaluru, Rural, Bijapur, Chamrajanagar, Chikballapur, Chitradurga, Dharwad, Haveri, Kolar, Raichur, Tumkur
Kollam, Pathanamthitta, Thiruvananthapuram
No. of Farmers
OC a %
Av P ppm
Av K ppm
Av S ppm
Av B ppm
Av Zn ppm
0.41 0.08–3.00
9.1 0.0–247.7
129 0–1,263
9.6 0.0–801.0
0.34 0.02–4.58
0.81 0.08–35.60
– –
6.99 0.0–63.6
128.9 4.1–11.66
6.53 1.4–34.6
0.25 0.1–0.78
0.91 0.4–3.07
0.77 0.21–1.90
6.9 0.4–42.0
104 30–635
16.0 1.1–150.4
0.22 0.06–0.49
0.44 0.18–2.45
0.53 0.19–1.13
5.3 0.0–72.4
63 8–247
7.8 1.3–50.0
0.17 0.06–0.80
0.68 0.24–2.90
0.43 0.01–3.60
12.3 0.0–480.0
133 4–3750
13.2 0.1–4647.4
0.57 0.02–26.24
0.97 0.06–235.00
1.04 0.36–2.57
22.0 1.2–137.0
101 33–313
3.4 1.0–11.0
0.31 0.18–0.48
1.88 0.56–7.20 (continued)
3650
40
82
115
17712
New Paradigm to Unlock the Potential of Rainfed Agriculture
TABLE 9.2b Mean and Range Values of Nutrient Content in Soil Samples in Various States (Districts within States) of India
28
433
434
TABLE 9.2b (Continued) Mean and Range Values of Nutrient Content in Soil Samples in Various States (Districts within States) of India State Madhya Pradesh Mean Range Rajasthan
Badwani, Dewas, Guna, Indore, Jhabua, Mandla, Raisen, Rajagarah, Sagar, Sehore, Shajapur, Vidisha
Alwar, Banswara, Bhilwara, Bundi, Dungarpur, Jhalwar, Sawai Madhopur, Tonk, Udaipur
Kanchipuram, Karur, Salem, Tirunelveli, Vellore
OC a %
Av P ppm
Av K ppm
Av S ppm
Av B ppm
Av Zn ppm
0.65 0.28–2.19
5.0 0.1–68.0
190 46–716
9.6 1.8–134.4
0.43 0.06–2.20
0.72 0.10–3.82
0.72 0.09–2.37
8.1 0.2–44.0
116 14–1,358
10.6 1.9–274.0
0.60 0.08–2.46
1.27 0.06–28.60
0.51 0.14–1.37
9.2 0.2–67.2
122 13–690
11.3 1.0–93.6
0.34 0.06–2.18
0.78 0.18–5.12
0.44 0.01–3.60
11.5 0.0–480.0
133 0–3750
12.4 0.0–4647.4
0.53 0.02–26.24
0.94 0.06–235.00
341
421
119
22,508
Source: Rego, T.J., et al., Journal of Plant Nutrition, 30, 1569–1583, 2007; Sahrawat, K.L., et al., Current Science, 93(10), 1–6, 2007; Wani, S.P., et al., In Conservation Farming: Enhancing Productivity and Profitability of Rain-Fed Areas, Soil Conservation Society of India, New Delhi, 163–178, 2008; and unpublished data sets of ICRISAT. a OC = Organic Carbon; AvP = Available Phosphorus, AvK = Available Potash, AvS = Available Sulfur, AvB = Available Boron, AVZn= Available Zinc. * = Extensive soil sampling undertaken to interpolate analysis at district level using GIS.
World Soil Resources and Food Security
Mean Range Tamilnadu Mean Range Total Mean Range
District
No. of Farmers
435
New Paradigm to Unlock the Potential of Rainfed Agriculture
Rainfall Improved system (BW1)
14
Traditional system (BW4C)
11
Rainfall (mm)
Rainfall use efficiency (kg/ha/mm of rainfall)
16
8 5 2000 3 0
1000 1976
1979
1982
1985
1988
1991
1994
1997
2000
2003
2006
2009
0
Year
FIGURE 9.9 Increased rainwater use efficiency in low rainfall years.
In addition, soil organic matter—an important driving force for supporting biological activity in soil—is very much in short supply, particularly in tropical countries. In addition to its importance for sustainable crop production, low soil organic matter in tropical soils is a major factor contributing to their poor productivity [Lee and Wani 1989; Syers et al. 1996; Katyal and Rattan 2003], the accelerated decomposition of soil organic carbon (SOC) due to agriculture and release of carbon (C) in the atmosphere also contributes to global warming [IPCC 1990; Lorenz and Lal 2005]. Management practices that augment soil organic matter and maintain it at a threshold level are needed. Sequestration of C in soil has attracted the attention of researchers and policymakers alike as an important mitigation strategy for minimizing impacts of climate change [Lal 2004; Velayutham et al. 2000; ICRISAT 2005; Bhattacharya et al. 2009; Srinivasarao et al. 2009]. Agricultural soils are among the earth’s largest terrestrial reservoirs of C and hold potential for expanded C sequestration [Lal 2004]. Improved agricultural management practices in the tropics such as intercropping with legumes, application of balanced plant nutrients, suitable land and water management, and use of stress-tolerant high-yielding cultivars improved SOC content and also increased crop productivity [Wani et al. 1995, 2003a, 2005, 2007; Lee and Wani 1989; ICRISAT 2005; Srinivasarao et al. 2009]. Farm bunds and degraded common lands in the villages could be productively used for growing nitrogen-fixing shrubs and trees to generate nitrogen-rich loppings. For example, growing Gliricidia sepium at a close spacing of 75 cm on farm bunds could provide 28–30 kg nitrogen per ha in addition to valuable organic matter. Also, large quantities of farm residues and other organic wastes could be converted into valuable sources of plant nutrients and organic matter through vermicomposting [Nagavallama et al. 2005]. Vermicompost is a good source of plant nutrients along with organic C addition
436
World Soil Resources and Food Security
TABLE 9.3 Nutrient Composition of Vermicompost and Garden Compost Nutrient Element Organic carbon Nitrogen Phosphorus Potassium Calcium Magnesium Sodium Zinc Copper Iron Manganese
Vermicompost (%) 9.8–13.4 0.51–1.61 0.19–1.02 0.15–0.73 1.18–7.61 0.093–0.568 0.058–0.158 0.0042–0.110 0.0026–0.0048 0.2050–1.3313 0.0105–0.2038
Garden Compost (%) 12.2 0.8 0.35 0.48 2.27 0.57 <0.01 0.0012 0.0017 1.1690 0.0414
Source: Nagavallemma, K.P., et al., Vermicomposting: Recycling Wastes into Valuable Organic Fertilizer, Global Theme on Agroecosystems Report No. 8, ICRISAT, Patancheru, Andhra Pradesh, India, 2004.
to the soil (Table 9.3). Through collective action, women self-help groups in the watersheds are earning additional income with vermicomposting [Wani et al. 2008; Sreedevi et al. 2007; Sreedevi and Wani 2009] and also contributing to enhancement of agricultural productivity and disposal of agricultural wastes thru environmentfriendly processes.
9.3.3 Water Resources Management For enhancing rainwater use efficiency in rainfed agriculture, the management of water alone cannot result in enhanced water productivity as the crop yields in these areas are limited by factors additional to water limitation. ICRISAT’s experience in rainfed areas has clearly demonstrated that more than water quantity per se, management of water is the limitation in the SAT regions [Wani et al. 2005]. An analysis in Malawi indicates that over the past three decades, only a fraction of the years that have been politically proclaimed as drought years were actually years subject to meteorological droughts (i.e., years where rainfall totals fall under minimum water needs to produce food at all) [Mwale 2003]. As indicated by Agarwal [2000], India would not have to suffer from droughts if local water balances were managed properly. Even during drought years, watershed development efforts of improving rainfall management have benefited Indian farmers [Shiferaw et al. 2006]. Evidence from water balance analyses on farmers’ fields around the world shows that only a small fraction, less than 30% of rainfall, is used as productive green water flow (plant transpiration) supporting plant growth [Rockström 2003]. In arid areas, typically as little as 10% of the rainfall is consumed as productive green water
New Paradigm to Unlock the Potential of Rainfed Agriculture
437
flow (transpiration), 90% flows as nonproductive evaporation flow, i.e., no or very limited blue water generation [Oweis and Hachum 2001]. In temperate arid regions, such as WANA, a large portion of the rainfall is generally consumed in the farmers’ fields as productive green water flow (45%–55%) that results in higher yield levels (3–4 t/ha as compared to 1–2 t/ha) and 25%–35% of the rainfall flows as nonproductive green water flow. The remaining 15%–20% generate blue water flow. These indicate a large scope of opportunity. Low agricultural yields in rainfed agriculture, often blamed as rainfall deficits, are in fact caused by other factors than rainfall. Still, what is possible to produce on-farm will not always be produced by resourcepoor small-scale farmers. The farmers’ reality is influenced by other constraints such as labor shortage, insecure land ownership, capital constraints, and limitation in human capacities.
9.3.4 Shifting Nonproductive Evaporation to Productive Transpiration Rainwater use efficiency in agricultural systems in arid and SAT is 35% to 50%. This suggests a scope for improvement of green water productivity, as it entails shifting nonproductive evaporation to productive transpiration, with no downstream water trade-off. This vapor shift (or transfer) through improved management options is a particular opportunity in arid, semiarid, and dry subhumid regions [Rockström et al. 2007]. Field measurements of rainfed grain yields and actual green water flows indicate that by doubling yields from 1 to 2 t/ha in semiarid tropical agroecosystems, green water productivity may improve from approximately 3500 m3/t to less than 2000 m3/t. This is a result of the dynamic nature of water productivity improvements when moving from very low yields to higher yields. At low yields, crop water uptake is low and evaporative losses are high, as the leaf area coverage of the soil is low. This results in high losses of rainwater as evaporation from soil. When yield levels increase, shading of soil improves.
9.4 N EW PARADIGM IS A MUST FOR UNLOCKING THE POTENTIAL OF RAINFED AGRICULTURE 9.4.1 New Paradigm for Water Management in Rainfed Agriculture Business as usual in managing rainfed agriculture as subsistence agriculture with low resource use efficiency cannot sustain the economic growth and needed food security. There is an urgent need to develop a new paradigm for upgrading rainfed agriculture. The conventional sectoral approach to water management produced low water use efficiencies resulting in increased demand for water to produce food. We need to have a holistic approach based on converging all the necessary aspects of natural resource conservation—their efficient use, production functions, and income enhancement avenues through value chain and enabling policies and much needed investments in rainfed areas [Wani et al. 2003b, 2009; Rockström et al. 2007]. Policy on water resource management for agriculture remains focused on irrigation, and the framework for integrated water resource management (IWRM) at
438
World Soil Resources and Food Security
catchment and basin scales are primarily concentrated on allocation and management of blue water (irrigation water) in rivers, groundwater, and lakes. The evidence from the CA indicated that water for agriculture is larger than irrigation, and there is an urgent need for a widening of the policy scope to include explicit strategies for water (green and blue) management in rainfed agriculture, including grazing and forest systems. Effective integration is a must to have a focus on the investment options of water management across the continuum (range) from rainfed to irrigated agriculture. This is the time to abandon the obsolete sectoral divide between irrigated and rainfed agriculture, which would place water resource management and planning more centrally in the policy domain of agriculture at large and not, as today, as a part of water resource policy [Molden et al. 2007]. Furthermore, the current focus on water resource planning at the river basin scale is not appropriate for water management in rainfed agriculture, which overwhelmingly occurs on farms of <5 ha at the scale of small catchments, below the river basin scale. Therefore, the focus should be on managing water at the catchment scale (or small tributary scale of a river basin), and opening for much needed investments in water resource management and also in rainfed agriculture [Rockström et al. 2007]. Evidence collected during the CA of water for food and water for life revealed that business as usual in global agriculture would not be able to meet the goal of food security and poverty reduction. If the situation continues, it will lead to crises in many parts of the world [Molden et al. 2007]. However, the world’s available land and water resources can satisfy future demands by taking the following steps: • Upgrading rainfed agriculture by investing more to enhance agricultural productivity (rainfed scenario) • Discarding the artificial divide between rainfed and irrigated agriculture and adopting an IWRM approach for enhancing resource efficiency and agricultural productivity • Investing in expanding irrigation where the scope exists and improving the efficiency of the existing irrigation systems (irrigation scenario) • Conducting agricultural trade within and between countries (trade scenario) • Reducing gross food demand by influencing diets and reducing postharvest losses, including industrial and household waste To upgrade rainfed agriculture in the developing countries community, a participatory and integrated watershed management approach is recommended and is found effective through a number of islands of success in Asia and Africa [Wani et al. 2002, 2003; Rockström et al. 2007; Wani et al. 2008]. In the rainfed areas of the tropics, water scarcity and growing land degradation cannot be tackled thru farmlevel interventions alone, and community-based management of natural resources for enhancing productivity and improving rural livelihoods are urgently needed [Wani et al. 2003, 2009; Rockström et al. 2007]. We need to have a holistic approach based on converging all the necessary aspects of natural resource conservation, their efficient use, production functions, and income enhancement avenues through value
New Paradigm to Unlock the Potential of Rainfed Agriculture
439
chain and enabling policies and much needed investments in rainfed areas. A major research and development challenge in upgrading rainfed agriculture is to establish convergence among different stakeholders and scientific disciplines by coming out of disciplinary compartments and translating available blueprints into operational plans and implementing them [Wani et al. 2003, 2006, 2009a; Rockström et al. 2007]. We know what to do but we face the challenge of how to do it. The community-based management of natural resources calls for new approaches (technical, institutional, and social) that are knowledge-intensive and need strong capacity-building measures for all the stakeholders, including policymakers, re searchers, development agents, and farmers. The small and marginal farmers are deprived of the new knowledge and materials produced by the researchers. There is a disconnect between the farmers and the researchers as the extension systems in most developing countries are not functioning at the desired level. There is an urgent need to bring in change in the way we are addressing the issues of rainfed agriculture to achieve food security and alleviate poverty to meet the MDGs.
9.4.2 Holistic Watershed Approach through Integrated Genetic and Natural Resource Management (IGNRM) Traditionally, crop improvement and natural resource management (NRM) were seen as distinct but complementary disciplines. ICRISAT has deliberately blurred these boundaries to create the new paradigm of IGNRM [Twomlow et al. 2006] to solve farming problems. Improved varieties and improved resource management are two sides of the same coin. The systems approach looks at various components of the rural economy—traditional food grains, new potential cash crops, livestock and fodder production, as well as socioeconomic factors such as alternative sources of employment and income. Crucially the IGNRM approach is participatory, with farmers closely involved in technology development, testing, and dissemination. ICRISAT’s studies in Africa and Asia have identified several key constraints to more widespread technology adoption [Ryan and Spencer 2001]. Other institutes have independently reached similar conclusions for other agroecosystems. So there is general agreement on the key challenges before us: • Lack of a market-oriented smallholder production system where research is market-led, demand-driven, and follows the commodity chain approach to address limiting constraints along the value chain. • Poor research-extension–farmer linkages, which limit transfer and adoption of technology. • Need for policies and strategies on soil, water, and biodiversity to offset the high rate of natural resource degradation. • Need to focus research on soil fertility improvement, soil and water management, development of irrigation, promotion of integrated livestock– tree–crop systems, and development of drought mitigation strategies. • Need to strengthen capacities of institutions and farmers’ organizations to support input and output marketing and agricultural production systems. • Poor information flow and lack of communication on rural development issues.
440
World Soil Resources and Food Security
• Need to integrate a gender perspective in agricultural research and training as seen in ICRISAT’s work on community watershed, VASAT, and villagelevel studies. In much of agricultural research, the multidisciplinary team approach has often run into difficulties in achieving impact because of the perceived disciplinary hierarchy [Shambhu et al. 2005]. The IGNRM approach in The Community Watershed Consortium pursues integration of the knowledge and products of the various
TABLE 9.4 Effect of Climate Variability on Pearl Millet Crop Performance and IGNRM Options in Mali Climate Parameters Late onset of rains
Early drought
Midseason drought
Terminal drought
Excessive rainfall Increased temperature
Unpredictability of drought stress Increased CO2 levels
Increased occurrence of dust storms at onset of rains Increased dust in the atmosphere
Effects on Crops and Natural Resources Shorter rainy season, risk that long-cycle crops will run out of growing time Difficult crop establishment and need for partial or total resowing Poor seed setting and panicle development, fewer productive tillers, reduced grain yield per panicle/plant Poor grain filling, fewer productive tillers Downy mildew and other pests, nutrient leaching Poor crop establishment (dessication of seedlings), increased transpiration, faster growth See above Faster plant growth through increased photosynthesis, higher transpiration Seedlings buried and damaged by sand particles Lower radiation, reduced photosynthesis
IGNRM Options Early-maturing varieties, exploitation of photoperiodism, P fertilizer at planting P fertilizer at planting, water harvesting and runoff control, delay sowing (but poor growth due to N flush), exploit seedling heat and drought tolerance Use of pearl millet variability: differing cycles, high tillering cultivars, optimal root traits, etc., water harvesting and runoff control Early-maturing varieties, optimal root traits, fertilizer at planting, water harvesting and runoff control Resistant varieties, pesticides, N fertilizer at tillering Heat tolerance traits, crop residue management, P fertilizer at planting (to increase plant vigor), large number of seedlings per planting hill Phenotypic variability, genetically diverse cultivars Promote positive effect of higher levels through better soil fertility management Increase number of seedlings per planting hill, mulching, ridging (primary tillage) Increase nutrient inputs (i.e., K)
New Paradigm to Unlock the Potential of Rainfed Agriculture
441
research disciplines into useful extension messages for development workers that can sustain increased yields for a range of climatic and edaphic conditions (Table 9.4).
9.4.3 Watershed Development—A Growth Engine for Development of R ainfed Areas In several countries, central and state governments and development investors have emphasized management of rainfed agriculture under various programs. Important efforts, for example, have been made under the watershed development programs in India, Thailand, Vietnam, and China in Asia, and Ethiopia and Rawanda in Africa. Originally, these programs in India were implemented until 2008 by different ministries such as the Ministry of Agriculture, the Ministry of Rural Development, and the Ministry of Environment and Forestry, causing difficulties for integrated watershed management. Recently, steps were taken to unify the program according to the common watershed guidelines developed by the Government of India in 2008 [GoI 2008].
9.4.4 Common Features of the Watershed Development Model Government agencies, development thinkers, donors, researchers, and NGOs have gradually learned one from another, though some are ahead of the field and others are deficient in some aspect or other, principally in people participation or in the science. But generally, the better models of today have some or all of the following features in common: • Participation of villagers as individuals, as groups, or as a whole, increasing their confidence, enabling their empowerment, and their ability to plan for the future and for self determination • Capturing the power of group action in the village, between villages, and from federations, e.g., capturing economies of scale by collective marketing • The construction of basic infrastructure with contributions in cash or labor from the community • Better farming techniques, notably the improved management of soil, water, diversifying the farming system and integrating the joint management of communal areas and forest • The involvement of the landless, often in providing services • Arrangements for the provision of basic services and infrastructure • The establishment of village institutions and links with the outside world • Improved relationships between men and women • Employment and income generation by enterprise generation in predominantly, but not exclusively, agriculture-related activities And sometimes: • The fusion of research and development (R&D) by capturing the extraordinary power of participatory technology development, including variety selection with direct links to germplasm collections
442
World Soil Resources and Food Security
• Complete avoidance of corruption so that trust is engendered and all the benefits pass to the community • Involvement with enforced migration
9.4.5 Recent Additions to the Watershed Model More recently, the following features have been added: • The pragmatic use of scientific knowledge as the entry point rather than money, leading to tangible economic benefits from low-cost interventions that generate rapid and substantial returns at an acceptable low level of risk. Among these are novel interventions focusing on soil health, participatory evaluation of improved cultivars, integrated pest management (IPM), use of micronutrients, and soil conservation and water table recharge structures. • A broad-based approach to income generation involving private sector links associated with scientific advances and markets: e.g., in the remediation of micronutrients deficiencies; in the marketing of medicinal and aromatic plants; with premium payments paid by industrial processors for aflatoxinfree maize and groundnut; with high sugar sorghum, soybean, and selected crops sold to industry for processing; and with the production for sale of commercial seed, hybrid varieties, and biopesticides. • Using new science methodologies to improve performance: remote sensing for monitoring and feedback to farmers; yield gap analysis; and rapid assessment of the fertility status of the watershed. • Building productive partnerships and alliances in a consortium for research and technical backstopping with the members brought together from the planning stage. • A desire to create resilience in the watershed and its community to climate change and to events occurring after program intervention. Where best applied, the model has led to profound farming system changes, improved food self sufficiency, expanded employment and commerce, and enhanced incomes. Where indifferently executed, the approach has led to no better than ad hoc development schemes as we shall see. There is indeed something here analogous to the yield gap exhibited between research stations and farmers’ yields. Much of the difference can be captured by implementing agencies catching up with best practices. The more recent linking of natural resource science with the private sector, markets, and with peoples’ broader livelihoods in consultation with them, is transforming the dynamics and success rate of development efforts. The watershed approach is a paradigm that works in all rainfed circumstances, has delivered important benefits and impacts, and needs to be implemented on a large scale. But watershed impact covers a spectrum from no better than ad hoc development schemes to impressive improvements of the natural resource endowment and agricultural production and a transformation of the socioeconomy. To consolidate and build upon the foundation already laid and universally gain the impact that is possible requires governments doing some difficult things, most
New Paradigm to Unlock the Potential of Rainfed Agriculture
443
noticeably introducing a new mind set or different form of approach that accepts that: • Watershed development is not just a means to increase production or to conserve soil and water, but an opportunity for the fully integrated and sustained development of human and natural resources. • The approach is valid across various rainfall regimes over vast tracts of tropical rainfed areas and can contribute in large measure to the simultaneous achievement of government’s production, environmental, and social goals. • Sustainability and better social impact and equity are very important issues with pro-poor interventions, not as a spin-off or afterthought, but planned and integral to the whole. • There are vast opportunities to reduce costs and increase output by improving the appropriateness and reach of technology. There is obvious value in converging different schemes in the interest of impact and sustainability, rather than a spread of activity: this is particularly important in the case of water and of schemes aimed to reach the poor.
9.4.6 Learning from Meta-Analysis of Watersheds in India The descriptive summary of multiple benefits derived from 636 watersheds, as indicated in numerous studies, is shown in Table 9.5. It is obvious that watershed programs are silently revolutionizing the rainfed areas with a mean benefit/cost ratio of 2.0 with the benefits ranging from 0.82 to 7.30. It indicates that, on average, even in fragile and high-risk rainfed environments, watershed programs were able to generate benefits that were more than double their costs. In many of the watersheds, benefits were even higher. About 18% of watersheds generated benefit/cost ratios above 3, which is fairly modest (Figure 9.10). However, 68% of watersheds performed below average (B/C ratio of 2.0) and indicated a large scope to enhance the impact of watershed projects in the country. Merely 0.6% of watersheds failed commensurate with the cost of the project [Joshi et al. 2008]. The mean internal rate of return of 27.43% was significantly high and comparable with any successful government programs (Table 9.5). The internal rates of return in 41% of watersheds were in the range of 20% to 30%, whereas about 27% of watersheds yielded an internal rate of return (IRR) of 30% to 50% (Figure 9.11). The watersheds with IRRs below 10% were only 1.9%. Another important purpose of the watershed programs was to generate employment opportunities to address the equity concerns of landless laborers and marginal and small farmers. The results of meta-analysis indicated that watershed programs have generated significant and substantial employment opportunities in the watershed areas. The mean additional annual employment generation in the watershed area on various activities and operations was about 154 person days/ha/yr (Table 9.5). It was as high as 900 person days/ha/yr in those watersheds that included multiple activities. Generating employment opportunities for the rural poor means raising
444
TABLE 9.5 Summary of Benefits from the Sample Watersheds Using Meta-Analysis Particulars
No. of Studies
Median
Minimum
311
2.01
1.70
1.70
0.82
7.30
35.09
IRR
%
162
27.43
25.90
25.00
2.03
102.70
21.75
Equity
Employment
99
154.53
286.67
56.50
0.05
900.00
8.13
Sustainability
Increase in irrigated area
Person days/ ha/yr %
93
51.55
34.00
63.43
1.28
204
10.94
Increase in cropping intensity
%
339
35.51
5.00
21.00
3.00
283.00
14.96
Runoff reduced
%
83
45.72
43.30
42.53
0.38
96.00
9.36
Soil loss saved
t/ha/yr
72
1.12
0.91
0.99
0.11
2.05
47.21
World Soil Resources and Food Security
Mode
Ratio
Source: Joshi, P.K., et al., In Global Theme on Agroecosystems, Report No. 46, ICRISAT, Patancheru, Andhra Pradesh, India, 2008.
Maximum
t Value
Mean
B/C ratio
Efficiency
Unit
445
Watersheds (%)
New Paradigm to Unlock the Potential of Rainfed Agriculture
80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0
Watershed (%) 67.5
13.2
12.2
0.6 <1
1 to 2
2 to 3 3 to 4 Benefitcost ratio
2.6
3.9
4 to 5
>5
FIGURE 9.10 Distribution (%) of watersheds according to benefit/cost ratio (BCR). (From Joshi, P.K., et al., In Global Theme on Agroecosystems, Report No. 46, ICRISAT, Patancheru, Andhra Pradesh, India, 2008. With permission.)
Watersheds (%)
their purchasing power and, in turn, alleviating rural poverty and income disparities. This has an important implication in that the watershed investment may be characterized as a poverty alleviation program in the fragile ecosystem areas. The important objective of the watershed programs is to improve the livelihood of poor rural households, who encounter disproportionate uncertainties in rainfed agriculture due to precarious environments, acute degradation of soil, and water scarcity. The estimates show that watershed programs were quite effective in addressing the problems of land degradation due to soil erosion and loss of water due to excessive runoff. Soil loss of about 1.12 t/ha/yr was saved due to interventions in the watershed framework. Conserving soil means raising farm productivity and transferring good soils to the next generation. It was noted that, on average, about 38 ha m additional water storage capacity was created in a watershed of 500 ha as a result of the watershed program. Augmenting water storage capacity contributed in (i) reducing the rate of runoff, and (ii) increasing groundwater recharge. On average, runoff loss was 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0
41.4 30.2
8.6
11.1
1.9 <10
10 to 20 20 to 30 30 to 40 40 to 50 Internal rate of return (%)
6.8
>50
FIGURE 9.11 Distribution (%) of watersheds according to IRR. (From Joshi, P.K., et al., In Global Theme on Agroecosystems, Report No. 46, ICRISAT, Patancheru, Andhra Pradesh, India, 2008. With permission.)
446
World Soil Resources and Food Security
reduced by 46% because of various watershed interventions, and the groundwater table was also augmented by 3.6 m in the watershed areas. These improvements have direct impacts in expanding the irrigated area, increasing cropping intensity, and diversifying systems with high-value crops. On an average, the irrigated area increased by about 52%, while the cropping intensity increased by 35.5%. In some cases, the irrigated area increased up to 204%, while the cropping intensity increased by 283%. Such an impressive increase in the cropping intensity was not realized in many surface irrigated areas in the country. These benefits confirm that the watershed programs perform as a viable strategy to overcome several externalities arising due to soil and water degradation [Joshi et al. 2008]. The above evidence suggests that watershed programs, which have been specifically launched in rainfed areas with the sole objective of improving the livelihood of poor rural households in a sustainable manner, have paid rich dividends and were successful in raising income levels, generating employment opportunities, and augmenting natural resources in the rainfed areas. These benefits have far-reaching implications for rural masses in the rainfed environment [Wani et al. 2008].
9.4.7 Results of Meta-Analysis Regression The results of meta-analysis regression further showed that the benefits vary depending upon the location, size, type, rainfall, implementing agency, and people’s participation, among other factors. The coefficient of multiple determination (R2) shows the variables included in the model and explains the more than 56% variation in the benefit:cost ratio. The positive value of intercept also indicates a positive impact of watershed programs on augmentation of income. A number of factors determine the economic efficiencies of watershed programs. Geographical location, rainfall pattern, focus of watershed program, implementing agency, status of target population, and people’s participation are some of the critical factors that play a deterministic role in the performance and efficiency of watersheds [Joshi et al. 2008]. Consideration of the time gap between implementation and evaluation of the program is also important. However, the effect of the time gap between implementation and evaluation could not be captured, as the variable was statistically nonsignificant. However, a positive sign of the variable indicates a larger benefit associated with intervention with time and suggests that performance of the watershed program should not be judged immediately after the implementation [Wani et al. 2008a]. Macrowatersheds (>1200 ha) achieved better impact than micros of 500 ha. Development needs to be undertaken in clusters of at least four to six microwatersheds together (2000–3000 ha) and the common guidelines has addressed this [GoI 2008], adopting sizes of 1000 to 5000 ha by selecting microwatersheds in a cluster. Macrounits offer economies of scale, more technical options, and greater hydrological efficiency and would ease collaboration between agencies and their interface with the community. Between 700 mm and 1100 mm of rainfall, there is good technology available. Above and below these amounts, the appropriateness and range of current technologies is not good enough and needs to be researched in concert with watershed communities to enhance the impact of watershed programs in all rainfed regions.
New Paradigm to Unlock the Potential of Rainfed Agriculture
447
Use of new scientific tools such as crop simulation and water balance models, GIS, remote sensing and information and communication technology (ICT), participatory research and development (PR&D), and collective action for planning, implementation, monitoring, and evaluation is needed to manage natural resources more efficiently and sustainably in the watersheds. The drivers of success are tangible economic benefits to large number of people, empowerment through knowledge, equal partnership, trust and shared vision, good local leadership, transparency and social vigilance in financial dealings, equity through low-cost structures, a predisposition to work collectively, activities targeted at the poor and women, increased drinking water availability, and income-generating activities for women. The current allocations are insufficient to treat a complete watershed or to adopt the livelihoods approach. Higher investments are a must to make watersheds engines of growth. The new IWMP of the Government of India has increased investments from Rs.6000 (US $133) to Rs.12000 (US $266) per ha in plains and Rs.15000 (US $333) in hilly areas [GoI 2008] and has adopted a livelihood approach to ensure tangible economic benefits to people in a watershed. There is opportunity to reduce costs: more cost-effective water structures; economies of scale from using the macrowatershed as the development unit; convergence of action to avoid duplication; getting things right the first time to avoid repeat expenditures; and avoiding the adverse costs of environmental deterioration. The cost/benefit ratio would be much improved by more efficient use of technology to increase productivity, by bringing wasteland into productive use, and by a total accounting of socioeconomic and environmental benefits. Interventions are needed to benefit women and vulnerable groups and help them to develop social capital and increased sustainability. National and state planning for and selection of watersheds might best be based on a matrix of the potentials for impact on production, poverty, the environment, and community involvement. Moving forward requires that a lack of capacity to effectively implement programs is addressed. Implementing agencies need to expand and broaden their capacities and skills and reach and communities need to strengthen their institutions and their skills. This will require a longer implementation period of 7 to 8 years with more time spent in preparation and in postintervention support. Additional funds and more flexibility in using budgets and the engagement of specialist service providers will also be required. The new common guidelines [GoI 2008] have addressed these recommendations and project duration is increased up to 7 years and 5% of the total budget is earmarked for capacity-building using quality service providers. One of the weakest aspects lies in the generation and dissemination of technology. A big improvement is needed in making appropriate technology and information accessible to the watershed community. The remedy lies in devising technology for the drier and wetter parts of the rainfed area and in more PR&D in forming consortia and employing agencies to provide specialist technical backstopping. There is a crucial need to improve monitoring and evaluation (M&E) and the feedback of the information obtained to constantly improve performance. Only a few key indicators need to be monitored in all watersheds. At one or two representative watersheds in each district, a broad range technical and socioeconomic parameters
448
World Soil Resources and Food Security
should be measured to provide a scientific benchmark and a better economic valuation of impact than is currently possible.
9.4.8 Business Model Watersheds should be seen and developed as a business model. This calls for a shift in approach from subsidized activities to knowledge-based entry points and from subsistence to marketable surplus ensuring tangible economic benefits for the population of the watershed at large. This is being done with productivity enhancement, diversification to high-value enterprises, income-generating activities, market links, public–private partnerships, microentrepreneurship, and broad-based community involvement. Strengths of rainfed areas using available water resources efficiently through involvement of private entrepreneurs and value addition can be harnessed by linking small and marginal farmers to markets through public–private partnership business model for watershed management [Wani et al. 2008, 2008a].
9.4.9 Promoting Collective Action in the Community Collective action occurs when the benefits from lower transaction costs of doing business are higher than the additional costs involved (sacrifices made) in complying with collective rules. When the anticipated benefits of cooperation are lower than the expected costs, households are unlikely to engage in collective efforts. For example, this may be the case for very marginal farmers who produce very small quantities such that the benefits per unit of transaction are small and do not warrant additional costs from cooperation. Previous studies have shown that these costs and benefits are likely to differ from one household to another depending on location, volume of production, endowment of assets, education, managerial skills, etc. [Kerr et al. 2002]. This shows that the benefits of collective action are likely to be unequally distributed and it may not be useful to some households unless some interventions are designed to enhance their participation. 9.4.9.1 Convergence and Collective Action Convergence of actors and their actions at the watershed level is needed to harness the synergies to unlock the potential of rainfed agriculture and maximize the benefits through efficient and sustainable use of natural resources. The integrated watershed management approach is science-based and knowledge-driven; it demands synthesis of knowledge from different sectors and translation into messages that small and marginal farmers could understand and adopt. The convergence approach to benefit small and marginal farmers through increased productivity per unit of resource is recommended as a large benefit of watershed programs has been missed due to a compartmentalized approach [Wani et al. 2003, 2003b, 2008]. New institutional mechanisms are also needed at district, state, and national levels to converge various watershed programs implemented by several ministries and development agencies to enhance the impact and efficiency by overcoming duplicity and confusion. For example, in 2005, the National Commission on Farmers in India recommended a holistic, integrated watershed management approach, with a focus on rainwater harvesting and improving soil health for sustainable development
New Paradigm to Unlock the Potential of Rainfed Agriculture
449
of drought-prone rainfed areas [GoI 2005]. Recently, the Government of India has established National Rain-fed Areas Authority (NRAA) with the mandate to converge various programs for integrated development of rainfed agriculture in the country. The common watershed guidelines issued by the Government of India have also emphasized the need for convergence and collective action [GoI 2008]. Thus, it has become increasingly clear that water management for rainfed agriculture requires a landscape perspective, and involves cross-scale interactions from farm household scale to watershed/catchment scale. 9.4.9.2 Consortium for Technical Backstopping Enhancing partnerships and institutional innovations through a consortium approach was a major impetus for harnessing community watershed’s potential to reduce household poverty [Wani et al. 2003b]. Complex issues were effectively addressed by a efforts of ICRISAT in collaboration with key partners, namely National Agricultural Research Systems (NARSs), nongovernmental organizations (NGOs), government organizations, agricultural universities, community-based organizations (CBOs), and other private interest groups with farm households as the key decision-makers. Self-help groups (SHGs) like village seedbanks were established not just to provide timely and quality seeds, but also to provide technical support and build the capacity of members, including women, for management, conservation, and livelihood-development activities. Incorporating knowledge-based entry points in the approach led to the facilitation of rapport and, at the same time, enabled the community to make rational decisions for their own development [Wani et al. 2008b]. As demonstrated by ICRISAT, the strongest merit of the consortium approach is in the area of capacity-building, where farm households are not the sole beneficiaries. Researchers, development workers, and students of various disciplines are also trained, and policymakers from the NARSs are sensitized to the entire gamut of community watershed activities. Private–public partnership has provided the means for increased investments, not only for enhancing productivity but also for building institutions as engines for people-led NRM [Wani et al. 2008b]. 9.4.9.3 Discard Artificial Divide between Irrigated and Rainfed Agriculture There is an urgent need to have sustainable water-use policies to ensure sustainable development and adopt an IWRM approach in the watersheds by discarding the artificial divide between rainfed and irrigated agriculture. In the absence of suitable policies and mechanisms for sustainable use of groundwater resources, benefits of watershed programs can be undone—quickly and easily—by overexploitation of the augmented water resources [Sreedevi et al. 2006]. Cultivation of water inefficient crops like rice and sugarcane needs to be controlled through suitable incentive mechanisms for rainfed irrigated crops and policies must be evolved to stop cultivation of high water requiring crops [Wani et al. 2008a]. 9.4.9.4 Pilot-Scale Model Community Watershed—A Site of Learning Based on detailed studies and synthesis of the results, impacts, shortcomings, lessons learned from many watershed programs, and on-farm experiences gained, the ICRISAT-led consortium developed an innovative farmers’ participatory consortium model for integrated watershed management [Wani et al. 2002, 2003b,
450
World Soil Resources and Food Security
2003c]. ICRISAT-led watershed espouses the IGNRM approach, where activities are implemented at the landscape level at benchmark sites representing the different agroecoregions of the SAT. The entire process revolves around the four E’s (empowerment, equity, efficiency, and environment), which are addressed by adopting specific strategies prescribed by the four C’s (consortium, convergence, collective action, and capacity-building). The consortium strategy brings together institutions from the scientific, nongovernment, government, and farmers’ groups for knowledge management. Convergence allows integration and negotiation of ideas among actors. Cooperation enjoins all stakeholders to harness the power of collective actions. Capacity-building engages in empowerment for sustainability [Wani et al. 2003b]. The important components of the new model, which are distinct from the earlier ones, are the following: • Collective action by farmers and participation from the beginning through cooperative and collegiate modes in place of contractual modes—a PR&D approach. • Principle of “users pay” adopted from the beginning; no free rides in the program. • Demand-driven approach and no supply-driven technologies; users have to pay in cash/kind. • Integrated water resource management and holistic system approach through convergence for improving livelihoods vs. a traditional compartmental approach. • A consortium of institutions for technical backstopping. • Knowledge-based entry point to build rapport with community and enhance participation of farmers and landless people through empowerment. • Tangible economic benefits to individuals through on-farm interventions enhancing efficiency of conserved soil and water resources and targeted income-generating activities for women and vulnerable groups through allied sector activities and rehabilitation of wastelands for improved livelihoods and environmental protection. • Low-cost and environment-friendly soil and water conservation measures throughout the toposequence for more equitable benefits to large numbers of farmers.
9.4.10 Multiple Benefits and Impacts Through the use of new tools [i.e., remote sensing, geographical information systems (GIS), and simulation modelling) along with an understanding of the entire food production–utilization system (i.e., food quality and market) and genuine involvement of stakeholders, the ICRISAT-led consortium approach effected remarkable impacts on SAT resource-poor farm households. 9.4.10.1 Reducing Rural Poverty Reduction of rural poverty in the watershed communities is evident in the transformation of their economies. The ICRISAT model ensured improved productivity with
451
New Paradigm to Unlock the Potential of Rainfed Agriculture
the adoption of cost-efficient water harvesting structures (WHSs) as an entry point for improving livelihoods. Crop intensification and diversification with high-value crops is an example of allowing households to achieve production of basic staples and surplus for modest incomes. The model has provisions for improving the capacity of farm households through training and networking and for alleviating poverty, improved livelihood, and enhanced participation, especially of the most vulnerable groups like women and the landless. Building on social capital made a huge difference in addressing the rural poverty of watershed communities. This is evident in the case of Adarsha Watershed, Kothapally, in Andhra Pradesh, India. Today, it is a prosperous village on the path to long-term sustainability and has become a beacon for science-led rural development. In 2001, the average family income from agriculture, livestock, and nonfarming sources was US$945, compared with the neighboring nonwatershed village income of US$613 (Figure 9.12). The villagers proudly professed: “We did not face any difficulty for water even during the drought year of 2002. When surrounding villages had no drinking water, our wells had sufficient water.” To date, the village prides itself in households owning 6 tractors, 8 lorries, 7 cars, and 30 auto-rickshaws. People from surrounding villages come to Kothapally for on-farm employment. With more training in livelihood and enterprise development, migration is bound to cease. Additional income of Rs.2000 to Rs.5000 per ha was obtained by the dryland farmers through in situ rainwater conservation, farm ponds, and suitable cropping systems [Venkateswarlu et al. 2008; Rao et al. 2010]. Crop–livestock integration is another facet harnessed for poverty reduction. The Lucheba watershed in Guizhou province of southern China has transformed its economy through modest injection of capital-allied contributions of labor and finance to create basic infrastructures like access to roads and drinking water supply. With
Watershed
15%
48%
Crops Livestock Nonfarm
28.9
2002
37%
Nonwatershed 12% 13%
75%
20.2
36%
10%
54%
42.5
2001
Watershed
44%
Nonwatershed 0
4
8
7%
12
49%
27.6
16 20 24 28 32 Actual values (Rs 1000)
36
40
44
FIGURE 9.12 Effect of integrated watershed management on flow of household net income. (From Shiferaw, B., et al., Journal of SAT Agricultural Research, 2, 1, 2006. With permission.)
452
World Soil Resources and Food Security
technical support from the consortium, the farming system was intensified from rice and rape seed to tending livestock (raising pigs) and growing horticultural crops (fruit trees like Ziziphus; vegetables like beans, peas, chillies, and sweet potatoes) and groundnuts. In forage production, wild buckwheat was specifically important as an alley crop, as it was good forage grass for pigs [Sreedevi and Wani 2009]. This cropping technology was also effective in controlling erosion and increasing farm income in sloping lands. This holds true in many watersheds of India, where the improvement in fodder production has intensified livestock activities like breed improvement (by artificial insemination and natural means) and livestock center/ health camp establishment [Wani et al. 2006]. In Tad Fa and Wang Chai watersheds in Thailand, there was a 45% increase in farm income within 3 years. Farmers earned an average net income of US$1195 per cropping season. A complete turnaround in the livelihood system of farm households was inevitable in ICRISAT-led watersheds. In the Kothapally, Adarsha watershed, milk production increased substantially over the past 6 years, resulting in a marketable surplus of 800 L day−1. At present, Reliance Industries, a supermarket chain, has established a milk procurement center at Kothapally and farmers get good value for milk based on its fat content. 9.4.10.2 Increasing Crop Productivity Increasing crop productivity is a common objective of all the watershed programs, and enhanced crop productivity is achieved after the implementation of soil and water conservation practices, along with appropriate crop and nutrient management. For example, the implementation of improved crop management technology in the benchmark watersheds of Andhra Pradesh increased the maize yields 2.5 times (Table 9.6) and sorghum yields threefold [Wani et al. 2006]. Overall, in the 65 community watersheds (each measuring approximately 500 ha), implementing best-bet practices resulted in significant yield advantages, varying with crops from 63%–197% (Table 9.7). The crop responses varied with location as well as with crops, the increases ranged in sorghum from 35%–270%, in maize from 30%–174%, in pearl millet from 72%–242%, in groundnut from 28%–179%, in sole pigeonpea from 97%–204%, and in intercropped pigeonpea from 40%–110% [Sreedevi and Wani 2009]. In Thanh Ha watershed of Vietnam, yields of soybean, groundnut, and mung bean increased by threefold to fourfold (2.8–3.5 t/ha) as compared with baseline yields (0.5 to 1.0 t/ha), reducing the yield gap between potential farmers’ yields. A reduction in nitrogen fertilizer (90–120 kg urea per ha) by 38%, increased maize yield by 18%. In Tad Fa watershed of northeastern Thailand, maize yield increased by 27%–34% with improved crop management [Wani et al. 2006]. 9.4.10.3 Improving Water Availability Improving water availability in the watersheds was attributed to efficient management of rainwater and in situ conservation, establishment of WHS, and improved groundwater levels. Even after the rainy season, the water level in wells nearer to WHS sustained good groundwater yields. In the various watersheds of India like Lalatora (in Madhya Prdesh), treated area registered a groundwater level rise of 7.3 m. At Bundi, Rajasthan, the average rise was 5.7 m and the irrigated area increased from 207 ha to 343 ha. In Kothapally watershed in Andhra Pradesh, the groundwater
Yield (kg ha–1) Before 1998
1999– 2000
2000– 2001
2001– 2002
2002– 2003
Sole maize Maize/pigeonpea intercrop system Sorghum/pigeonpea intercrop system Sole sorghum
– –
3250 5260
3760 6480
3300 5600
3480 5650
–
5010
6520
5830
–
5780
–
4360
4590
3570
2960
2740
Sole maize Sorghum/pigeonpea intercrop system Hybrid cotton BT cotton Mean CV% SE
1500 1980
1700 2330
1600 2170
1600 2750
1800 3190
– –
2295 – 3477 11.9 415
7050 – 4970 31.4 1559
6600 – 3833 10.7 410
6490 – 4018 8.0 323
Cropping Systems
2003– 2004
2004– 2005
2005– 2006
2006– 2007
2007– 2008
2008– 2009
Mean
CV%
SE
3920 6390
3630 6170
4680 6120
4810 6680
3820 5960
17.8 16.7
80 116
4790
5290
5310
–
–
5500
13.4
154
3020
2860
2500
–
–
3330
23.9
141
Farmers practice 2040 1950 3310 3000
2250 3360
2150 3120
– –
– –
1890 2900
17.2 19.2
53 110
– 6210 4584 10.8 495
– 5590 4320 12.2 525
– 7310 6268 16.7 1049
– 9380 7396 16.2 1201
5880 7120
37.0 26.1
511 315
Improved systems 3920 3420 6290 4990
6950 – 4814 14.5 698
– – 3651 20.3 742
New Paradigm to Unlock the Potential of Rainfed Agriculture
TABLE 9.6 Average Crop Yields (kg ha−1) with Equivalent of Maize Crop with Different Cropping Systems at Adarsha Watershed, Kothapally, 1999–2008
Note: The farmers practice sorghum/pigeonpea intercrop system; improved pigeonpea variety ICPL 87119 was grown along with local sorghum variety (Pacha Jonna) from 2001 onward. The old variety, which was highly susceptible for fusarium wilt, was discontinued.
453
454
World Soil Resources and Food Security
TABLE 9.7 Crop Yields as Influenced by Best-Bet Options in Districts of Andhra Pradesh under APRLP–ICRISAT Watersheds and Karnataka under Sujala–ICRISAT Project Grain Yield (t ha−1) Crop
Farmers’ Practice
Improved Practice
% Increase over Farmers’ Practice
Andhra Pradesh Groundnut Greengram Pearl millet Maize Sorghum
0.95 0.98 0.88 3.10 1.13
Groundnut Finger millet Sunflower Maize Soybean
1.00 1.15 0.76 3.45 1.35
1.52 1.67 2.13 5.24 2.30
63 70 142 69 104
1.97 1.93 2.26 5.87 2.47
97 68 197 70 82
Karnataka
Source: Sreedevi, T.K., and Wani, S.P. In Rain-Fed Agriculture: Unlocking the Potential, Comprehensive Assessment of Water Management in Agriculture Series, CAB International, Wallingford, UK, 222–257, 2009.
level rise was 4.2 m in open wells (Figure 9.13). The various WHSs resulted in an additional groundwater recharge per year of approximately 428,000 m3, on average, in a watershed. With this improvement in groundwater availability, the supply of clean drinking water was guaranteed. In Lucheba watershed in China, a drinking water project, which constitutes a water storage tank and pipelines to farm households, was a joint effort of the community and the watershed project. This solved the drinking water problem for 62 households and more than 300 livestock. Previously, every farmer’s household spent 2–3 hours per day fetching drinking water. This was the main motivation for the excellent participation by farmers in the project. On the other hand, in Thanh Ha watershed in Vietnam, collective pumping of well water established an efficient water distribution system and enabled the farmers’ group to earn more income by growing watermelon with reduced drudgery as women previously had to carry water a long distance on their heads to irrigate watermelon crops [Wani et al. 2006]. 9.4.10.4 Supplemental Irrigation Supplemental irrigation can play a very important role in reducing the risk of crop failures and in optimizing productivity in the SAT. In these regions, there is a good potential for delivering excess rainwater to storage structures or groundwater because, even under improved systems, there is a loss of 12%–30% of the rainfall as
455
New Paradigm to Unlock the Potential of Rainfed Agriculture
0
2004
2005
2006
6
1000
500
12
Rainfall (mm)
Water level in well (m)
Bundi watershed, Rajasthan 2003
2002
0
18 Rainfall
Near check dam
Away from check dam
Estimated additional groundwater recharge was 6,75,0000 m3 per year.
0
2001
2002
2003
2004
2005
2006
1500
6
1000
12
500
18
0 Rainfall
Near check dam
Rainfall (mm)
Water level in well (m)
Adrasha watershed, Andhra Pradesh 2000
Away from check dam
FIGURE 9.13 The impact of watershed interventions on groundwater levels at two benchmark sites in India. (Note: Estimated additional groundwater recharge due to watershed interventions is 675,000 m3/yr in Bundi watershed and 427,800 m3/yr in Adarsha Watershed.)
runoff. Striking results were recorded from supplemental irrigation on crop yields in ICRISAT benchmark watersheds in Madhya Pradesh. On-farm studies made during the 2000–2003 postrainy seasons showed that chickpea yields (1.25 t/ha) increased by 127% over the control yields (0.55 t/ha), and groundnut pod yields (1.3 t/ha) increased by 59% over the control yields (0.82 t/ha) by application of two supplemental irrigations of 40 mm. Similar yield responses in mung bean and chickpea crops were obtained from supplemental irrigation at the ICRISAT center in Patancheru [Pathak et al. 2009]. 9.4.10.5 Sustaining Development and Protecting the Environment Sustaining development and protecting the environment are the two-pronged achievements of the watersheds. The effectiveness of improved watershed technologies was evident in reducing runoff volume, peak runoff rate and soil loss, and improving groundwater recharge. This is particularly significant in Tad Fa watershed where interventions such as contour cultivation at midslopes, vegetative bunds planted with Vetiver, fruit trees grown on steep slopes, and relay cropping with rice bean reduced seasonal runoff to less than half (194 mm) and soil loss to less than one-seventh (4.21 t/ha) of the conventional system (473 mm runoff and soil loss of 31.2 t/ha). This
456
World Soil Resources and Food Security
holds true with the peak runoff rate, where the reduction is approximately one-third (Table 9.8). A large number of fields (80%–100%) in the SAT were found severely deficient in zinc, boron, and sulfur, as well as nitrogen and phosphorus (Tables 9.2a and b). Amendment of soils with the deficient micronutrients and secondary nutrients increased crop yields by 30% to 70%, resulting in an overall increase in water and nutrient use efficiency (Table 9.9) [Rego et al. 2007]. The Introduction of IPM in cotton and pigeonpea substantially reduced the number of chemical insecticidal sprays in Kothapally, India during the season and thus reduced the pollution of water bodies with harmful chemicals. Introduction of IPM and improved cropping systems decreased the use of pesticides to 66 per ha, saving US$44 [Ranga Rao et al. 2007]. Crop rotation using legumes in Wang Chai watershed in Thailand and Than Ha Watershed in Vietnam substantially reduced nitrogen requirements for rainfed sugarcane in Wang Chai and maize in Than Ha watersheds. The IPM practices, which brought into use local knowledge using insect traps of molasses, light traps, and tobacco waste, led to extensive vegetable production in Xiaoxingcun (China) and Wang Chai (Thailand) watersheds [Wani et al. 2006]. Improved land and water management practices, along with integrated nutrient management comprising application of inorganic fertilizers and organic amendments (such as crop residues, vermicompost, farm manures, and Gliricidia loppings) as well as crop diversification with legumes, not only enhanced productivity but also improved soil quality. Increased C sequestration of 7.4 t/ha in 24 years was observed with improved management options in a long-term watershed experiment at ICRISAT [Wani et al. 2003a]. By adopting a fuel-switch for C, women SHGs in Powerguda (a remote village of Andhra Pradesh, India) have pioneered the sale of C units (147 t CO2 C) to the World Bank from their 4500 Pongamia trees, the seeds of which are collected for producing saplings for distribution and promotion of the biodiesel plantation. A normalized difference vegetation index (NDVI) estimation from the satellite images showed that within 4 years, vegetation cover could increase by 35% in Kothapally. The IGNRM options in the watersheds reduced loss of TABLE 9.8 Seasonal Rainfall, Runoff, and Soil Loss from Different Benchmark Watersheds in India and Thailand Watershed Tad Fa (Khon Kaen, NE Thailand) Kothapally (Andhra Pradesh, India) Ringnodia (Madhya Pradesh, India) Lalatora (Madhya Pradesh, India)
Runoff (mm)
Soil Loss (t/ha)
Seasonal Rainfall (mm)
Treated
Untreated
Treated
Untreated
1284
169
364
4.21
31.2
743
44
67
0.82
1.9
764
21
66
0.75
2.2
1046
70
273
0.63
3.2
Total Uptake of Nutrients
Treatment
Grain Yield
Total Dry Matter
kg ha−1
kg ha−1
kg ha−1 N
g ha−1
P
K
S
B
Zn
Maize 2002 Farmer inputs (FI) FI + SBZn LSD (0.05)
2730 4560 (67%) 419
6200 8850 (42%) 633
59.5 86.4 8.8
15.0 20.8 4.1
45.2 57.1 5.7
4.5 7.0 0.7
16.4 19.2 3.8
111.8 191.8 25.4
2003 FI FI + SBZn FI + SBZn + NP LSD (0.05)
2790 4130 (48%) 4880 (74%) 271
6370 9040 (41%) 10,377 (62%) 580
48.3 73.9 108.1 8.4
48.3 73.9 108.1 8.4
39.0 47.2 55.6 6.3
4.4 6.9 9.3 0.7
8.7 17.1 19.4 3.6
113.1 228.1 266.7 41.0
2004 FI FI + SBZn FI + SBZn + NP LSD (0.05)
2430 3110 (27%) 4230 (74%) 417
5820 7060 (21%) 9470 (62%) 1054
60.0 69.4 93.0 13.4
60.0 69.4 93.0 13.4
59.9 63.9 85.8 13.9
5.3 5.7 9.0 1.3
19.0 23.6 42.1 7.8
89.6 165.1 191.9 38.3
457
(continued)
New Paradigm to Unlock the Potential of Rainfed Agriculture
TABLE 9.9 Yield and Nutrient Uptake of Maize, Castor, Groundnut, and Mung Bean in Response to Fertilization in Andhra Pradesh, India, 2002–2004
458
TABLE 9.9 (Continued) Yield and Nutrient Uptake of Maize, Castor, Groundnut, and Mung Bean in Response to Fertilization in Andhra Pradesh, India, 2002–2004 Total Uptake of Nutrients
Treatment
Grain Yield
Total Dry Matter
kg ha−1
kg ha−1
kg ha−1 N
g ha−1
P
K
S
B
Zn
Castor 590 890 (50%) 143
1400 2070 (47%) 360
23.2 34.2 6.9
3.1 5.1 1.4
22.1 30.3 6.6
2.2 3.6 0.7
18.1 26.5 4.9
40.0 62.2 14.2
2003 FI FI + SBZn FI + SBZn + NP LSD (0.05)
690 1000 (44%) 1190 (72%) 186
1610 2270 (40%) 2770 (72%) 403
27.5 37.9 46.4 8.0
6.3 7.6 7.5 1.4
14.4 24.3 26.6 6.4
2.6 3.9 4.7 0.8
11.3 15.7 22.2 4.6
47.8 70.4 79.4 13.7
2004 FI FI + SBZn FI + SBZn + NP LSD (0.05)
990 1240 (25%) 1370 (38%) 285
2220 2710 (22%) 3350 (27%) 484
33.8 54.2 54.4 13.0
5.3 7.4 7.7 2.2
31.7 32.1 38.9 13.2
2.4 3.8 4.3 0.9
18.1 23.3 30.6 4.2
41.0 73.0 86.6 18.2
World Soil Resources and Food Security
2002 Farmer inputs FI FI + SBZn LSD (0.05)
2002 Farmer inputs ( FI) FI + SBZn LSD (0.05)
700 940 (34%) 103
2690 3420 (27%) 145
74.9 95.1 4.1
7.3 11.3 2.4
29.3 41.9 3.7
4.4 6.4 0.7
40.1 52.1 3.1
50.2 80.9 5.1
2003 FI FI + SBZn FI + SBZn + NP LSD (0.05)
560 810 (44%) 980 (75%) 59
2920 4150 (42%) 4740 (62%) 183
57.7 86.3 114.9 5.6
6.6 7.2 10.6 1.2
27.5 38.1 39.5 3.6
3.7 5.5 6.5 0.5
38.6 56.8 68.4 3.9
59.0 151.5 116.8 13.2
2004 FI FI + SBZn FI + SBZn + NP LSD (0.05)
920 1190 (29%) 1280 (39%) 96
4080 4930 (20%) 5060 (24%) 262
107.8 124.1 139.4 8.4
9.2 10.8 15.4 2.3
47.6 56.9 60.9 6.3
6.8 6.3 7.0 0.7
78.9 65.1 106.8 7.6
87.3 141.5 129.6 52.0 (continued)
New Paradigm to Unlock the Potential of Rainfed Agriculture
Groundnut
459
460
TABLE 9.9 (Continued) Yield and Nutrient Uptake of Maize, Castor, Groundnut, and Mung Bean in Response to Fertilization in Andhra Pradesh, India, 2002–2004 Total Uptake of Nutrients
Treatment
Grain Yield
Total Dry Matter
kg ha−1
kg ha−1
kg ha−1 N
g ha−1
P
K
S
B
Zn
Mung bean 770 1110 (44%) 145
1500 2110 (40%) 280
36.7 53.3 8.2
4.6 7.4 1.0
25.4 36.3 5.5
2.3 4.0 0.4
20.4 30.4 5.6
45.6 69.6 5.6
900
2900
54.7
6.9
52.1
3.0
37.6
59.8
FI + SBZn FI + SBZn + NP LSD (0.05)
1390 (54%) 1540 (71%) 160
4840 (66%) 5420 (86%) 417
87.9 103.9 14.2
13.7 13.2 2.1
80.4 95.3 16.6
7.8 6.4 1.0
73.0 79.9 9.4
129.2 208.4 23.8
2004 FI FI + SBZn FI + SBZn + NP LSD (0.05)
740 920 (24%) 1160 (56%) 131
2800 3200 (14%) 4050 (44%) 580
59.6 58.7 71.6 17.4
9.0 8.0 9.0 2.2
57.7 55.3 66.7 11.8
3.1 4.8 5.7 1.1
40.2 66.6 77.8 15.0
53.5 69.1 79.7 16.8
2003 FI
World Soil Resources and Food Security
2002 Farmer inputs FI FI + SBZn LSD (0.05)
New Paradigm to Unlock the Potential of Rainfed Agriculture
461
NO3–N in runoff water (8 vs 14 kg nitrogen per ha). Reduced runoff and erosion reduced risk of downstream flooding and siltation of water bodies that directly improved environmental quality in the watersheds [Pathak et al. 2005; Sahrawat et al. 2005; Wani et al. 2005]. 9.4.10.6 Conserving Biodiversity Conserving biodiversity in the watersheds was engendered through participatory NRM. The index of surface percentage of crops (ISPC), crop agrobiodiversity factor (CAF), and surface variability of main crops changed as a result of integrated watershed management interventions. Pronounced agrobiodiversity impacts were observed in Kothapally watershed, where farmers now grow 22 crops in a season with a remarkable shift in cropping pattern from cotton (200 ha in 1998 to 100 ha in 2002) to a maize/pigeonpea intercrop system (40 ha in 1998 to 180 ha in 2002), thereby changing the CAF from 0.41 in 1998 to 0.73 in 2002. In Thanh Ha, Vietnam the CAF changed from 0.25 in 1998 to 0.6 in 2002 with the introduction of legumes [Wani et al. 2005]. Similarly, in Bundi watershed in Rajasthan, in rehabilitated degraded common land above- and below-ground biodiversity (flora and fauna) has been recorded [Dixit et al. 2005].
9.4.11 Scaling-Up Most farming problems require integrated solutions, with genetic, management- related, and socioeconomic components. In essence, plant breeders and NRM scientists must integrate their work with that of private and public sector change agents to develop flexible cropping systems that can respond to rapid changes in market opportunities and climatic conditions. ICRISAT, in partnership with NARSs, has conceived, developed, and successfully evaluated an innovative farmers’ participatory consortium model for integrated watershed management. The model includes the consortium approach and adopts the concept of convergence in every activity in the watershed [Sreedevi and Wani 2009]. Watersheds are only management units for sustainable development of NRs and agriculture is the backbone of rural development. Watersheds need to be used as planning units for developing area plans by adopting a bottom-up approach for sustainable inclusive growth using water management as an entry point activity. Watershed management is just a beginning for holistic area development and improving livelihoods and not an end in itself. The watershed plans can be converged to make district and state plans for development of rainfed and drought-prone districts to reduce poverty. These plans can be used for implementing various programs such as Mahatma Gandhi Rural Employment Guarantee Scheme (MGREGS), food for work, watersheds, various crop missions (e.g., pulses mission, oil seeds mission, etc.), food security mission, Millennium Development Goal area plans, rural knowledge centers, etc. It calls for convergence of actors and actions at village, district, state, and country levels but it should not result in a race for defending operational territories. The new paradigm for upgrading rainfed agriculture can double the productivity in Asia and also reduce poverty without causing further degradation of natural resource base. Based on the success of the participatory consortium watershed
462
World Soil Resources and Food Security
management model at Kothapally, the Andhra Pradesh Rural Livelihoods Program (APRLP), Sir Dorabji Tata Trust, Mumbai, the World Bank’s Sujala Project in Karntaka, India, and the Asian Development Bank (ADB) have selected this model for scaling up the benefits in Andhra Pradesh, Madhya Pradesh, and Rajasthan in India, and Northeast Thailand, Northern Vietnam, and South China. As most of NRM technologies are agroecoregion- and site-specific, the representative benchmark watersheds allow transferring the findings from benchmark nucleus watersheds to the satellite watersheds in the similar target ecoregion. In the target ecosystems, project implementing agencies (PIAs) are selected based on their strengths and available current knowledge base. Nucleus watersheds were selected for development and critical monitoring as the sites for undertaking participatory action research, as sites of learning and training in a district, and as sites to study the processes to select different partners in the consortium. An innovative model with a consortium of institutions, as opposed to single institution approach, for technical backstopping was initiated (Figure 9.14) for project implementation [Wani et al. 2003b]. All the partners have worked in partnership with another institution to manage the watershed sustainably. A successful partnership based on a strong commitment with state and local agencies, community leaders, and people is desirable. Involvement of the state government departments, agricultural research and education institutions in the area, and, most importantly, the policymakers along with the farmers is critical from the beginning. To establish and operationalize the consortium, the transaction costs (time and financial resources) are more; however, once it is established, the scaling-out process is quite rapid, economical, and impact-oriented [Wani et al. 2008b]. To promote community participation in the watershed for site selection as well as implementation and assessment of activities, various committees/groups were formed. It was Consortium approach FTCs
AP Govt depts.
ce Fa
ien
ce
w
NRSA
an
Sc CRIDA
Increased productivity & incomes Improved livelihoods
itICRISAT h a Hu
m
KVKs
ANGRAU
NGOs
FIGURE 9.14 Farmer participatory consortium approach for integrated watershed development in Andhra Pradesh.
New Paradigm to Unlock the Potential of Rainfed Agriculture
463
recognized that to shift the community participation from a contractual to a consultative and collegiate mode, tangible private economic benefits to individuals were a must. Such tangible benefits to individuals could come from in situ rainwater conservation and translation through increased farm productivity by adopting an IGNRM approach. The farmers need to be graduated slowly after initiation to take on or participate in collective action. First, in rainfed areas in the tropics where subsistence agriculture is the rule, it is urgent that the immediate needs of individuals be met through the program rather than fixing the lofty goals of community participation for common societal benefits. Adopting the principle that “users pay” and providing no subsidies for investments on individuals’ farms for technologies, inputs, and conservation measures force development researchers to provide demand-driven options rather than supply-driven options. Once the individuals realize the benefits of soil and water conservation, they came forward to participate in community activities in the watershed through various organized groups and also sustain the initiatives that benefit them to improve their livelihoods. This approach builds the ownership, accountability, and sustainability vs. the target-driven conventional approach, which has not worked in tropical rainfed areas.
9.4.12 Unlocking the Potential of Rainfed Agriculture—A Beginning Is Made in India Lately, increasing attention is being paid to management of green water (soil moisture) resources to upgrade rainfed agriculture. In the past few years, there has been an increased priority on developing policies and building capacities in favor of increased investments in water management in rainfed agriculture. There is, thus, growing evidence of the importance of water investments in rainfed agriculture, and governance and management is gradually being redirected in certain regions of the world toward water management for upgrading rainfed agriculture as a key strategy for reducing poverty and increasing agricultural production [World Bank 2005]. It is further increasingly clear that water management for rainfed agriculture requires a landscape perspective and involves cross-scale interactions from the farm household scale to the watershed scale. In India, the initiation of IWMP by converging all watershed schemes under Department of Land Resources (DoLR) in the Ministry of Rural Development (MoRD), establishing the NRAA, the renaming the National Rural Employment Guarantee Scheme (NREGS) as the Mahatma Gandhi Employment Guarantee Scheme of MoRD by the Ministry of Agriculture as well as their establishing the National Food Security Mission (NFSM), Rashtriya Kisan Vikas Yojana (RKVY), and the pulses and oil seeds production enhancement initiatives, and the Ministry of Water Resources encouraging more crop per drop through farmers’ participatory action research trials are examples of upgrading the rainfed agriculture in India. All these programs are targeted to increase productivity of rainfed crops and improve livelihoods of the rural poor. The Indian Council of Agricultural Research (ICARGoI) has initiated a network project on climate change impacts on agriculture in India, as well as adaptation and mitigation strategies in the country [Venkateswarlu and Shankar 2009].
464
World Soil Resources and Food Security
The initiation of a mission project to upgrade rainfed agriculture in 25 rainfed districts of Karnataka, with technical backstopping provided by an ICRISAT-led consortium, is an example of science-led development to unlock the potential of rainfed agriculture [ICRISAT 2010]. The Department of Agriculture is converging all the schemes targeted for rainfed agriculture through this mission and has pooled together the expertise of state agricultural universities and different departments of agriculture, community-based organizations, and input suppliers to benefit the farmers.
ACKNOWLEDGMENTS We gratefully acknowledge the help of all the consortium partners in Asia, including the farmers who have enabled us to contribute this chapter. We also thank Asian Development Bank, Manila, Philippines, Sir Dorabji Tata Trust, Mumbai, and Sir Ratan Tata Trust, Mumbai, for their financial support for undertaking activities in various regions on which this chapter is based.
REFERENCES Agarwal, A. 2000. Drought? Try capturing the rain. Briefing paper for members of parliament and state legislatures—An occasional paper from the Centre for Science and Environment (CSE), New Delhi: CSE. Bhattacharyya, T., Ray, S.K, Pal, D.K., et al. 2009. Soil carbon stocks in India—Issues and priorities. 2009. Journal of the Indian Society of Soil Science 57(4):461–468. Dixit, S., Tewari, J.C., Wani, S.P., et al. 2005. Participatory biodiversity assessment: Enabling rural poor for better natural resource management. Global Theme on Agroecosystems Report no. 18. Patancheru 502324. Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. Falkenmark, M. 1986. Fresh water—Time for a modified approach. Ambio 15(4):192–200. Falkenmark M., Karlberg L., and J. Rockström. 2009. Present and future water requirements for feeding humanity. Food Sec. 1:59–69. Food and Agriculture Organization (FAO). 2002. Agriculture: Towards 2015/30. Technical interim report. http://www.fao.org/es/esd/at2015/toc-e.htm. FAOStat. 2010. AQUASTAT: FAO’s Information System on Water and Agriculture. http:// www.fao.org/nr/aquastat. FAOStat. 2010. FAOStat. http://faostat.fao.org/site/567/DesktopDefauly.aspx?pageID=567# ancor. Fisher, G., Harrij, V.V., Eva H., et al. 2009. Potentially obtainable yields in the semiarid tropics. Global Theme on Agroecosystems Report no. 54. Patancheru 502 324, Andhra Pradesh, India; International Crops Research Institute for the Semi-Arid Tropics. GoI. 2005. Serving farmers and saving farming–2006: Year of agricultural renewal. Third Report, National Commission on Farmers, Ministry of Agriculture. New Delhi: GoI. GoI. 2008. Common guidelines for watershed development projects. New Delhi: Ministry of Agriculture, Government of India. Hilhost, T., and F. Muchena. 2000. Nutrients on the move: Soil fertility dynamics in African farming systems. London: International Institute for Environment and Development. Holmen, H. 2004. Why no green revolution in Africa? Currents 34:12–16. ICRISAT. 2005. Identifying systems for carbon sequestration an increased productivity in semiarid tropical environments. A Project Completion Report, National Agricultural Technology Project. New Delhi: Indian Council of Agricultural Research.
New Paradigm to Unlock the Potential of Rainfed Agriculture
465
ICRISAT. 2010. Bhoo Chetana—Boosting rainfed agriculture in Karnataka. In Mission mode: Bridging yield gaps in dryland crops. Karnataka State Department of Agriculture, Government of Karnataka and Global Theme on Agroecosystems. Patancheru, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. IIASA. 2002. Climate change and agricultural vulnerability. Special report for the UN World Summit on Sustainable Development. Johhanesburg, 2002. http://www.iiasa.ac.at/ Research/LUC/JB-Report.pdf. IFDC. 2006. Agricultural production and soil nutrient mining in Africa: Implications for resource conservation and policy development. IFDC Technical Bulletin. Muscle Shoals, AL:IFDC. Intergovernmental Panel on Climate Change (IPCC). 1990. Climate change, eds. J.T. Houghton, G.J. Jenkinson, and J.J. Ephraums. Cambridge: Cambridge University Press. IPCC. 2007. Climate change-impacts: Adaptation and vulnerability. In Technical Summary of Working Group II to Fourth Assessment Report Intergovernmental Panel on Climate Change, ed. M.L. Parry, O.F. Canziani, J.P. Paultikof, P.J. van der Linden, and C.E. Hanon, 23–78. Cambridge: Cambridge University Press. Irz, X., and Roe, T. 2000. Can the world feed itself? Some insights from growth theory. Agrekon 39:513–528. Joshi, P.K., Jha, A.K., Wani, S.P., et al. 2008. Impact of watershed program and conditions for success: A meta-analysis approach. In Global theme on agroecosystems, Report no. 46. Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. Katyal, J.C., and R.K. Rattan. 2003. Secondary and micronutrients: Research gaps and future needs. Fert. News 48(4):9–14, 17–20. Kerr, J., Pangare, G., and V. Lokur. 2002. Watershed development projects in India. Research Report 127. Washington, DC: IFPRI. Lal, R. 2004. Soil carbon sequestration to mitigate climate change. Geoderma 123:1–22. Lee, K.K., and S.P. Wani. 1989. Significance of biological nitrogen fixation and organic manures in soil fertility management. In Soil fertility and fertility management in semiarid tropical India, ed. C.B. Christianson, 89–108. Muscle Shoals, AL: IFDC. Lorenz, K., and R. Lal. 2005. The depth distribution of soil organic carbon in relation to land use and management and the potential of carbon sequestration in subsoil horizons. Advances in Agronomy 88:35–66. Molden, D., Frenken, K., Barker, R., et al. 2007. ����������������������������������������� Trends in water and agricultural development In Water for food, water for life—A comprehensive assessment of water management in agriculture, ed. D. Moden, 57–89. Colombo, Sri Lanka: International Water Management Institute (IWMI). Mwale, F. 2003. Drought impact on maize production in Malawi. Unesco-IHE Report. Delft: the Netherlands. Nagavallemma, K.P., Wani, S.P., Lacroix, S., et al. 2004. Vermicomposting: Recycling wastes into valuable organic fertilizer. Global Theme on Agroecosystems Report No. 8. Patancheru 502 324, Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropic. Nagavallemma, K.P., Wani, S.P., Lacroix, S., et al. 2005. Vermicomposting: Recycling wastes into valuable organic fertilizer. Journal of Agriculture and Environment for International Development 99:188–204. Oweis, T., and Hachum, A. 2001. Reducing peak supplemental irrigation demand by extending sowing dates. Agricultural Water Management 50:109–123. Pathak, P., Sahrawat, K.L., Rego, T.J., et al. 2005. Measurable biophysical indicators for impact assessment: Changes in soil quality. In Natural resource management in agriculture: Methods for assessing economic and environmental impacts, eds. B. Shiferaw, H.A. Freeman, and S.M. Swinton, 53–74. Wallingford, UK: CAB International.
466
World Soil Resources and Food Security
Pathak, P., Sahrawat, K.L., Wani, S.P., et al. 2009. Opportunities for water harvesting and supplemental irrigation for improving rain-fed agriculture in semi-arid areas. In Rain-fed agriculture: Unlocking the potential. Comprehensive Assessment of Water Management in Agriculture Series, eds. S.P. Wani, J. Rockström, and T. Oweis, 197–221. Wallingford, UK: CAB International. Ranga Rao, G.V., Rupela, O.P., Wani, S.P., et al. 2007. Bio-intensive pest management reduces pesticide use in India. Pesticides News. Journal of Integrated Pest Management. 76:16–17. Rao, K.V., Venkateswarlu, B., Sahrawat, K.L., Wani, S.P., Mishra, P.K., Dixit, S., Reddy, K.S., Kumar, M., and Saikia, U.S. (eds.). 2010. Proceedings of National Workshop-cumBrainstorming on Rainwater Harvesting and Reuse through Farm Ponds: Experiences, issues and strategies. April 21–22, 2009, CRIDA, Hyderabad. Rego, T.J., Sahrawat, K.L., Wani, S.P., et al. 2007. Widespread deficiencies of sulfur, boron and zinc in Indian semi-arid tropical soils: On-farm crop responses. Journal of Plant Nutrition 30:1569–1583. Rego, T.J., Wani, S.P., Sahrawat, K.L., et al. 2005. Macro-benefits from boron, zinc, and sulphur application in Indian SAT: A step for grey to green revolution in agriculture. Global Theme on Agroecosystems Report no. 16. Patancheru, Andhra Pradesh, India: ICRISAT. Rockström, J. 2003. Water for food and nature in the tropics: Vapour shift in rain-fed agriculture. Invited paper to the Special issue 2003 of Royal Society Transactions B Biology, Theme Water Cycle as Life Support Provider 358(1440):1997–2009. Rockström, J., and Falkenmark, M. 2000. Semiarid crop production from a hydrological perspective: Gap between potential and actual yields. Critical Reviews in Plant Science 19(4):319–346. Rockström, J., Falkenmark, M., Karlberg, L., et al. 2009. Future water availability for global food production: The potential of green water for increasing resilience to global change. Water Resources Research 45:W00A12, doi:10.1029/2007WR006767. Rockström, J., Hatibu, N., Oweis, T., et al. 2007. Managing water in rain-fed agriculture. In Water for food, water for life: A comprehensive assessment of water management in agriculture, ed. D. Molden, 315–348. London: Earthscan and Colombo, Sri Lanka: International Water Management Institute (IWMI). Rosegrant, M., Ximing, C., Cline, S., et al. 2002. The role of rain-fed agriculture in the future of global food production. EPTD Discussion Paper No. 90, Washington, DC: IFPRI, Environment and Production Technology Division. Ryan, J.G., and Spencer, D.C. 2001. Challenges and opportunities shaping the future of the semi-arid tropics and their implications. Patancheru, Andhra Pradesh, India: ICRISAT. Sahrawat, K.L., Bhattacharyya, T., Wani, S.P., et al. 2005. Long-term lowland rice and arable cropping effects on carbon and nitrogen status of some semi-arid tropical soils. Current Science 89(12):2159–2163. Sahrawat, K.L., Wani, S.P., Rego, T.J., et al. 2007. Widespread deficiencies of sulphur, boron and zinc in dryland soils of the Indian semi-arid tropics. Current Science 93(10):1–6. Sanchez, P., Swaminathan, M.S., Dobie, P., and Yuksel, N. 2005. Halving hunger: It can be done. Summary version of the report of the Task Force on Hunger. New York: The Earth Institute at Columbia University. SEI. 2005. Sustainable pathways to attain the millennium development goals—Assessing the role of water, energy and sanitation. Document prepared for the UN World Summit, Sept 14, 2005, New York, NY. Stockholm: SEI. http://www.sei.se/mdg.htm. Shambu Prasad, C., Hall, A.J., and Wani, S.P. 2005. Institutional history of watershed research: The evolution of ICRISAT’s work on natural resources in India. Global Theme on Agroecosystems Report No. 12. Patancheru, Andhra Pradesh, India: ICRISAT.
New Paradigm to Unlock the Potential of Rainfed Agriculture
467
Shiferaw, B., Anupama, G.V., Nageswara Rao, G.D., et al. 2006. Socioeconomic characterization and analysis of resource-use patterns in community watersheds in semi-arid India. Journal of SAT Agricultural Research 2(1). Singh, P., Aggarwal, P.K., Bhatia, V.S., et al. 2009. Yield gap analysis: Modeling of achievable yields at farm level in rain-fed agriculture: Unlocking the potential, eds. S.P. Wani, J. Rockstorm, and T. Oweis, 81–123. Wallingford, UK: CAB International. Sreedevi, T.K., and Wani, S.P. 2009. Integrated farm management practices and up-scaling the impact for increased productivity of rain-fed systems. In Rain-fed agriculture: Unlocking the potential, ed. S.P. Wani, J. Rockström, and T. Oweis, 222–257. Comprehensive Assessment of Water Management in Agriculture Series. Wallingford, UK: CAB International. Sreedevi, T.K., Wani, S.P., and Pathak, P. 2007. Harnessing gender power and collective action through integrated watershed management for minimizing land degradation and sustainable development. Journal of Financing Agriculture 36:23–32. Sreedevi, T.K., Wani, S.P., Sudi, R., et al. 2006. On-site and off-site impact of watershed development: A case study of Rajasamadhiyala, Gujarat, India. Global Theme on Agroecosystems Report No. 20. Patancheru, Andhra Pradesh, India: ICRISAT. Srinivasa Rao, Ch., Vittal, K.P.R., Venkateswarlu, B., et al. 2009. Carbon stocks in different soil types under diverse rainfed production systems in tropical India. Communications in Soil Science and Plant Analysis 40:2338–2356. Srinivasa Rao, Ch., Wani, S.P., Sahrawat, K.L., et al. 2010. Effect of balanced nutrition on yield and economics of vegetable crop in participatory watersheds in Karnataka. Indian Journal of Fertilizers 6:39–42. Stoorvogel, J.J., and E.M.A. Smaling. 1990. Assessment of soil nutrient depletion in subSaharan Africa: 1983–2000, 1: Main report. Report No. 28. Wageningen, the Netherlands: Winand Staring Centre. Syers, J.K., Lingard J., Pieri J., et al. 1996. Sustainable land management for the semiarid and sub-humid tropics. Ambio 25:484–491. Thirtle, C., Beyers, L., Lin, L., et al. 2002. The impacts of changes in agricultural productivity on the incidence of poverty in developing countries. DFID Report No. 7946. London: Department for International Development (DFID). Twomlow, S., Shiferaw, B., Cooper, P., et al. 2006. Integrating genetics and natural resource management for technology targeting and greater impact of agricultural research in the semi-arid tropics. Patancheru Andhra Pradesh, India: ICRISAT. UNEP. 1997. World atlas of desertification, 2nd ed. Nairobi, Kenya: UNEP. UNStat. 2005. www.unstat.com. Velayutham, M., Pal, D.K., and Bhattacharyya, T. 2000. Organic carbon stock in soils of India. In Global climate change and tropical ecosystems. Advances in Soil Science, eds. R. Lal, J.M. Kible, and B.A. Stewart, 71–95. Boca Raton, FL: CRC Press. Venkateswarlu, B., Ramakrishna, Y.S., Reddy, S., et al. 2008. Rainfed farming—A profile of doable technologies. Technical Bulletin. Hyderabad, Andhra Pradesh, India: CRIDA. Venkateswarlu, B., and Shanker, A.K. 2009. Climate change and agriculture: Adaptation and mitigation strategies. Indian Journal of Agronomy 54(2):226–230. Walker, T. 2010. Challenges and opportunities for agricultural R & D in the semi-arid tropics. Internal document for strategic planning. Patancheru, Andhra Pradesh, India: ICRISAT. Wani, S.P., Balloli, S.S., Kesava Rao, A.V.R., et al. 2004. Combating drought through integrated watershed management for sustainable dryland agriculture. In Regional workshop on agricultural drought monitoring and assessment using space technology, 39–48. May 4, 2004. Hyderabad, India: National Remote Sensing Agency (NRSA).
468
World Soil Resources and Food Security
Wani, S.P., Joshi, P.K., Raju, K.V., et al. 2008a. Community watershed as a growth engine for development of dryland areas. A comprehensive assessment of watershed programs in India. Global Theme on Agroecosystems Report No. 47. Andhra Pradesh, India: International Crops Research Institute for the Semi-Arid Tropics. Wani, S.P., Joshi, P.K., Ramakrishna, Y.S., et al. 2008. A new paradigm in watershed management: A must for development of rain-fed areas for inclusive growth. In Conservation farming: Enhancing productivity and profitability of rain-fed areas, eds. A. Swarup, S. Bhan, and J.S. Bali, 163–178. New Delhi: Soil Conservation Society of India. Wani, S.P., Maglinao, A.R., Ramakrishna, A. et al. 2003. Integrated watershed management for land and water conservation and sustainable agricultural production in Asia. Proceedings of the ADB-ICRISAT-IWMI annual project review and planning meeting. December 10–14, 2001, Hanoi, Vietnam. Patancheru, Andhra Pradesh, India: ICRISAT. Wani, S.P., Pathak, P., Jangawad, L.S., et al. 2003a. Improved management of Vertisols in the semi-arid tropics for increased productivity and soil carbon sequestration. Soil Use and Management 19:217–222. Wani, S.P., Pathak, P., Sreedevi, T.K., et al. 2003b. Efficient management of rainwater for increased crop productivity and groundwater recharge in Asia. In Water productivity in agriculture: Limits and opportunities for improvement, eds. J.W. Kijne, R. Barker, and D. Molden, 199–215. Wallingford, UK: CAB International and Colombo, Sri Lanka: IWMI. Wani, S.P., Pathak, P., Tam, H.M., et al. 2002. Integrated watershed management for minimizing land degradation and sustaining productivity in Asia. In Integrated land management in dry areas. Proceedings of a joint UNU-CAS international workshop, ed. Z. Adeel, 207–230. September 8–13, 2001, Beijing, China. Wani, S.P., Ramakrishna, Y.S., Sreedevi, T.K., et al. 2006. Issues, concepts, approaches and practices in integrated watershed management: Experience and lessons from Asia. In: Integrated management of watersheds for agricultural diversification and sustainable livelihoods in Eastern and Central Africa, 17–36. Proceedings of the international workshop held at ICRISAT, December 6–7, 2004, Nairobi, Kenya. Patancheru, Andhra Pradesh, India: ICRISAT. Wani, S.P., Rego, T.J., Rajeswari, S., et al. 1995. Effect of legume-based cropping systems on nitrogen mineralisation potential of Vertisol. Plant and Soil 175:265–274. Wani, S.P., Rockström, J., and Oweis, T. (eds.). 2009a. Rain-fed agriculture: Unlocking the potential. Comprehensive Assessment of Water Management in Agriculture Series. Wallingford, UK: CAB International. Wani, S.P., Sahrawat, K.L., Sreedevi, T.K., et al. 2007. Carbon sequestration in the semiarid tropics for improving livelihoods. International Journal of Environmental Studies 64(6):719–727. Wani, S.P., Singh, P., Dwivedi, R.S., et al. 2005. ��������������������������������������������� Biophysical indicators of agro-ecosystem services and methods for monitoring the impacts of NRM technologies at different scale. In Natural resource management in agriculture: Methods for assessing economic and environmental impacts, eds. B. Shiferaw, H.A. Freemen, and S.M. Swinton, 97–123. Wallingford, UK: CAB International. Wani, S.P., Singh, H.P., Sreedevi, T.K., et al. 2003c. Farmer-participatory integrated watershed management: Adarsha Watershed, Kothapally India. An innovative and up-scalable approach. A Case Study. In Research towards integrated natural resources management: Examples of research problems, approaches and partnerships in action in the CGIAR, eds. R.R. Harwood and A.H. Kassam, 123–147. Washington, DC: Interim Science Council, Consultative Group on International Agricultural Research. Wani, S.P., Sreedevi, T.K., Rockström, J., et al. 2009. Rain-fed agriculture—Past trend and future prospects. In Rain-fed agriculture: Unlocking the potential, eds. S.P. Wani, J. Rockström, and T. Oweis, 1–35. Comprehensive Assessment of Water Management in Agriculture Series. Wallingford, UK: CAB International.
New Paradigm to Unlock the Potential of Rainfed Agriculture
469
Wani, S.P., Sreedevi, T.K., Vamsidhar Reddy, T.S., et al. 2008b. Community watersheds for improved livelihoods through consortium approach in drought prone rainfed areas. Journal of Hydrological Research and Development (23)55–77. World Health Organization (WHO). 2000. Gender, health and poverty. Factsheet No. 25, http://www.who.int/mediacenter/factsheets/fs251/en/. World Bank. 2000. Spurring agricultural and rural development. In Can Africa claim the 21st century?, 170–207. Washington, DC: World Bank. World Bank. 2005. Agricultural growth for the poor: An agenda for development. Washington, DC: The International Bank for Reconstruction and Development/The World Bank.
10 Land Degradation
Freddy O. Nachtergaele, Monica Petri, and Riccardo Biancalani
CONTENTS 10.1 10.2 10.3 10.4
Introduction................................................................................................... 471 Global Inventories of Land Degradation....................................................... 472 The Capability of Ecosystems to Provide Goods and Services.................... 475 Land Degradation: The Change in the Provision of Ecosystem Goods and Services................................................................................................... 478 10.4.1 Biomass Health.................................................................................. 478 10.4.2 Soil Health Indicators and Trends..................................................... 478 10.5 Water Quantity and Quality........................................................................... 479 10.6 Biodiversity Changes.....................................................................................480 10.6.1 Changes in (Rural) Economic Productivity.......................................480 10.6.2 Changes in Social and Cultural Provisions of Ecosystems............... 481 10.7 Analysis of Results........................................................................................ 481 10.8 The Impact of Land Degradation.................................................................. 488 10.9 National and Local Land Degradation Studies: Indicators and Monitoring..................................................................................................... 488 10.10 The Causes of Land Degradation................................................................ 490 10.11 The Cost of Land Degradation..................................................................... 491 10.12 Conclusions.................................................................................................. 492 References............................................................................................................... 496
10.1 INTRODUCTION There are numerous terms for and definitions of land degradation that are a source of confusion, misunderstanding, and misinterpretation. A wide range of terms is used in the literature, often with distinct disciplinary-oriented meaning, and leading to misinterpretation among disciplines. Some common terms used are soil degradation, land degradation, and desertification. While there is a clear distinction between soil and land (the term land refers to an ecosystem comprising land, landscape, terrain, vegetation, water, climate), there is no clear distinction between the terms land degradation and desertification. Desertification refers to land degradation in arid, semiarid, and subhumid areas resulting from various factors, including climatic variations and human activities [UNCCD 1994]. To compound the confusion even further, the term desertification is also used by some scientists when land is damaged beyond (reasonable) repair. In the context of the UN Food and Agriculture Organization (FAO)’s State of Land and Water (SOLAW), land degradation is 471
472
World Soil Resources and Food Security
defined as the reduction in the capacity of the land to provide ecosystem goods and services over a period of time for its beneficiaries. Ecosystem goods refer to absolute quantities of land products having an economic or social value for present and future generations. They include animal and vegetal production, land availability and soil health, and water quality and quantity. Ecosystem services concern more qualitative characteristics and their impact on the beneficiaries and the environment such as biodiversity, maintaining hydrological and nutrient cycles, etc. None of these can be measured or valued in a simple way. The land degradation definition includes an explicit reference to a time period over which degradation is assessed. Georeferenced data on economic goods produced rarely go back more than 50 years. Land cover and vegetation and similar remotely sensed data have the same time limitations. The establishment of a baseline is therefore a must, if comparable assessments are to be made. The explicit mention of the beneficiaries of goods and services of ecosystems in the definition draws the attention to the fact that the view of beneficiaries may change over time and that all beneficiaries may not have the same evaluation of the value of a particular good or service. Note that land degradation is not limited to biophysical effects nor is it limited to human-induced phenomenon, as was done previously. This makes land degradation assessment a particularly complex endeavor. The threat to sustainable development caused by land degradation was explicitly recognized by the 1992 Earth Summit and the 2002 World Summit on Sustainable Development. A unique UN Convention to Combat Desertification (UNCCD) was created in 1994 to specifically coordinate efforts to reduce land degradation in drylands and promote sustainable development. Global assessments of soil and land degradation started more than 35 years ago, but have not until now achieved a clear answer on where land degradation takes place, what impact it has on the population, and what the cost to governments and land users would be if the decline in soil, water, and vegetation resources continued unabated. Although institutional, socioeconomic, and biophysical causes of land degradation have been identified locally in many case studies, these have not been inventoried systematically at district, national, or regional levels. The Land Degradation Assessment in Drylands (LADA) project was launched by the Global Environmental Facility (GEF) to remedy this situation. The project was implemented by the UN Environment Programme (UNEP) and executed by the FAO. Many of the findings discussed in this chapter are outputs from this project. Much of the investment in land reclamation and rehabilitation during recent years has been driven by donor interest to fund action on the ground rather than research and understanding of the problem. Even the impact of what works and what does not in combating land degradation is hardly known or scientifically documented. In this respect the World Overview of Conservation Approaches and Technologies (WOCAT) program constitutes a notable exception with its systematic collection of information on sustainable soil and water conservation practices and their effects.
10.2 GLOBAL INVENTORIES OF LAND DEGRADATION Until LADA, the only comprehensive source of land degradation information had been the Global Assessment of Human Induced Soil Degradation (GLASOD), which
Pressures Status Axis Axis 1: Biomass status Land cover/organic carbon Axis 2: Soil health status Soil suitability for actual land use
Natural Bush invasion (+) and fire (−) Drought increase (−) Podzolization, ferallitization, salinization, sodification, etc. (−) Steep topography (−) High rainfall intensity (−) Low soil resistance to erosion (−)
Ecosystem type (desert/polar) Bush invasion (−)
None
Process Axis Axis 1: Biomass change Greenness change/deforestation rate Axis 2: Soil health change Compaction Water erosion Sealing, water erosion Water erosion Nutrient mining Pollution, salinization
Water use (−) or (+)
Axis 3: Water resource change
Land use (change) Legal protected area (+) Land use choice Inputs/management trend Irrigation (+) Economic supply/demand (+) or (−)
Axis 4: Biodiversity threat
Policies leading to HDI decline (−) Distance to markets (−) or (+)
Axis 6: Social services change Human development index trend
Axis 5: Economic value change Crop + livestock gross value Forest gross value
473
Axis 3: Water quantity status Amount per hectare per year Axis 4: Biodiversity status Land cover Axis 5: Economic provision status Crop value Livestock value Forest value Urban (100) Axis 6: Social provision status Accessibility Tourism Protected areas
Low land cover (−) Low soil nutrient stock (−) Low water availability Drought increase Drought increase
Human-Induced Deforestation (−) Other land use change (−) or (+) Mechanization (−) Overstocking (−) Land management (−) or (+) Subsistence management (−) Use of very high inputs (−) Irrigation (−)
Land Degradation
TABLE 10.1 Ecosystem Goods and Service Status, Pressures, and Consequent Land Degradation Process
474
World Soil Resources and Food Security
assessed—as implied in its title—not the full range of ecosystem goods and services, but was limited to a soil degradation evaluation only [Oldeman et al. 1990]. The assessment was based on expert judgments by single experts in a country or a region and suffered from inconsistency and lack of reproducibility [Sonneveld and Dent 2007]. Nevertheless, the GLASOD inventory remains a unique first global assessment that made policymakers aware of the widespread extent and impact of the problem. Under the LADA project, two different global assessments were undertaken:
(a)
1. Global Land Degradation Assessment (GLADA), which identified the Normalized Difference Vegetation Index (NDVI)-based vegetation greenness trends over the period 1981–2006. It defined critical areas as those where both the greenness and the rain-use efficiency were declining. In contrast to GLASOD, the methodology used hard (measured) data and it did restrict itself to a well-defined time period. The results indicated that the decline in greenness affected areas where 1 billion people were living and would result in a net loss of about 35 million tons of carbon/year. Most affected areas were in tropical Africa south of the equator and southeast Africa, southeast Asia, south China, north-central Australia, drylands and steep-lands of Central America and the Caribbean, Southeast Brazil and the Pampas, and boreal forest [Bai et al. 2008a; UNEP 2007]. Results were criticized because, although the observations were objective, the statistical methods used were debatable and the method focused directly only on a single ecosystem (biomass/carbon). The analysis was later improved by de Jong et al. [2010] without solving all problems related to it. 2. GLADIS (Global Land Degradation Information System) is based on an assessment of the status and the trends of ecosystem goods and services, including the impact of these changes on the population living in these areas. GLADIS is documented and online. It summarizes findings in radar diagrams that cover each of the major ecosystem goods and services (biomass, soil health, water Trend in ecosystems providing G&S
(b)
Ecosystems’ services provisions Crops and intensive grazing
Biomass
Biomass
100
Social
50
Economic
100
Soil
Water Biodiversity
Social
50
Economic
Soil
Water Biodiversity
FIGURE 10.1 Difference in the status of ecosystem goods and services between two different land use systems.
475
Land Degradation
quantity and quality, biodiversity, economics, social and cultural). Results are available for each area of the globe, per country or per land use system within a country. Table 10.1 and Figure 10.1 summarize the indicators used for each axis in the diagram. Each axis has a scale from 0 to 100, where 0 is the worst situation and 100 is the best. (Provisional results and maps are available from http://lprapp11.fao.org:8080/gladis_beta/.) Major changes in ecosystem goods and services were studied over a period of roughly 15–25 years.
10.3 T HE CAPABILITY OF ECOSYSTEMS TO PROVIDE GOODS AND SERVICES Before tackling the issue of degradation, which implies a change in the capacity to provide goods and services, it is required to establish a baseline status of degraded land or what the actual capacity of ecosystems is in this respect. Based on biophysical and socioeconomic parameters, the picture that emerges is that most developing nations are indeed at a great disadvantage in this respect. In fact, a combination of unfavorable climates (that are too dry, too wet, or too cold and affect the status of biomass, biodiversity, and water supply), poor soils unsuitable for agricultural use, and poor economic conditions based on subsistence agriculture, coupled with problems of markets and cultural attraction points that prevail in many developing nations. These conditions have been systematically evaluated at GLADIS and are illustrated in Figure 10.2 and Tables 10.2 and 10.3 for the average biophysical conditions prevailing in selected countries and land-use systems, respectively. (a)
Countries with high averages for ecosystem G&S provisions
Countries with low averages for ecosystem G&S provisions
(b)
Biomass
Biomass
100
Special
100
Soil
50
Special
0
0
Economic
Water
Biodiversity Country French Guiana Malaysia
Soil
50
Economic
Water
Biodiversity 0.6 0.7
Equatorial Guinea 0.6
Country Mauritania Burkina Faso Somalia
FIGURE 10.2 Capacities of ecosystems to provide goods and services.
0.25 0.21 0.19
476
TABLE 10.2 Sample Country Averages for Ecosystem Goods and Services Provisions
Country
Carbon Above Ground/Axis 1 Status
Soil Constraints/ Axis 2 Status
Water/Axis 3 Status
Biodiversity/ Axis 4 Status
Economic Value/Axis 5 Status
Social and Cultural Goods and Services/Axis 6 Status
Ecosystems Service Status Index
68.0 79.5 63.8 68.5 73.7 64.0 64.4 65.7
Countries with Better Provisions from Ecosystems 88.3 79.8 58.8 86.2 77.0 74.1 90.7 77.5 60.1 86.4 72.0 72.5 87.5 81.0 65.0 94.8 64.0 86.4 87.9 79.9 55.4 78.8 81.0 69.3
61.4 27.0 56.3 43.4 39.3 43.2 63.3 55.4
95.7 90.9 96.3 80.9 91.2 88.9 97.8 75.8
0.76 0.75 0.75 0.73 0.73 0.73 0.73 0.73
Eritrea Mauritania Sudan Niger Mali Chad Burkina Faso Somalia
6.5 2.5 10.7 4.6 7.0 8.4 16.5 12.8
Countries with Worst Provisions from Ecosystems 76.2 6.7 28.8 93.7 0.4 37.0 73.3 3.2 35.2 85.6 2.2 33.4 77.0 7.7 33.2 75.4 3.6 35.6 44.9 12.0 23.3 64.2 4.6 24.9
8.2 6.9 12.9 9.2 12.1 12.2 20.9 10.0
690.1 26.2 41.3 36.1 45.1 34.2 69.5 35.5
0.24 0.24 0.24 0.23 0.22 0.21 0.20 0.17
Note: Sample average values of ecosystem services status index and status of good and services in biomass, soil, water, biodiversity, productivity, and socio-cultural provision by countries. The land degradation indicators are estimated using the GLADIS method explained in Sections 10.4 to 10.7.
World Soil Resources and Food Security
Slovenia Belize Austria Dem. People’s Rep of Korea Costa Rica Finland Japan Malaysia
Land-Use Systems Forestry–protected Wetlands–mangrove Forestry–unmanaged Grasslands–protected Grasslands–unmanaged Agriculture with large scale irrigation Agriculture–large-scale irrigation/ high livestock presence Grasslands–extensive pastoralism Urban land Rainfed agriculture (subsistence/ commercial) Grasslands with high livestock presence Agriculture crops with high livestock Grasslands with moderate livestock presence Shrubs with high livestock presence
Biodiversity/ Axis 4 Status
Economic Value/Axis 5 Status
Social and Cultural Goods and Services/ Axis 6 Status
Ecosystems Service Status Index
63.3 71.5 59.9 47.9 54.9 57.5 59.3
89.9 80.0 99.8 59.3 59.4 40.0 30.1
26.9 5.8 21.8 5.1 0.3 79.9 75.3
100.0 59.4 48.5 100.0 42.2 86.8 85.0
0.79 0.74 0.67 0.61 0.54 0.53 0.52
100.0 0.0 34.8
49.0 100.0 58.5
69.9 10.0 30.1
7.2 100.0 47.4
57.0 100.0 72.8
0.51 0.50 0.43
31.3
50.9
44.5
30.2
48.0
73.2
0.43
11.1 31.6
41.6 50.4
56.0 41.9
10.2 49.9
53.3 19.3
79.6 70.3
0.39 0.36
16.8
51.0
33.4
10.6
48.3
67.8
0.30
Carbon Above Ground/Axis 1 Status
Soil Constraints/ Axis 2 Status
Water/Axis 3 Status
75.7 100.0 65.9 30.9 30.6 17.0 15.6
100.0 100.0 100.0 100.0 100.0 35.7 41.0
32.0 0.0 16.7
477
Source: Nachtergaele, F.O., and Petri, M., Mapping Land Use Systems at a Global and Regional Scale for Land Degradation Assessment, LADA Technical Report #8, FAO, Rome, 2008. Note: Average values of ecosystem services status index and status of good and services in biomass, soil, water, biodiversity, economics, and socio-cultural setting by land use system. The land degradation indicators are estimated using the GLADIS method explained in Sections 10.4 to 10.7.
Land Degradation
TABLE 10.3 Status of Ecosystem Goods and Services as a Function of Land-Use Systems (Global Averages)
478
World Soil Resources and Food Security
10.4 L AND DEGRADATION: THE CHANGE IN THE PROVISION OF ECOSYSTEM GOODS AND SERVICES In a similar way as done for the status of ecosystem goods and services, the changes can be presented in a radar diagram, with the additional complication that changes can be positive as well as negative. The approach has adopted a rating of 50 for situations where no change is occurring. Ratings above 50 are considered positive, while below that value they are considered negative and degrading.
10.4.1 Biomass Health Organic carbon (OC), as present in the vegetation and soils, is the single most integrating factor indicating environmental health. This generalization should be qualified because the accumulation of OC may occur in undesirable form (e.g., OC in an invading unpalatable bush that displaces useful grasses for livestock) and such differences can hardly be correctly presented at the global level. The approach follows the trend in vegetation health by normalized account deforestation rates as determined by the Forest Resources Assessment [FAO 2005]. Soil carbon is considered under soil health. Presence and trends of bush- and forest-fires, although feasible, is not incorporated in the evaluation at this stage.
10.4.2 Soil Health Indicators and Trends Soil has received by far the most attention in land degradation studies (the original definition of the phenomenon was described as the loss of soil productivity, and the GLASOD approach also focused exclusively on soils). The trend in soil health is determined by two major degradation processes taking place in the soil: (1) those related to physical pressures resulting in a loss of soil mass and structure (absolute soil loss, compaction); and (2) the long term chemical wellbeing of the soil in terms of nutrients (nutrient mining) and the absence of toxicity accumulations (pollution, salinization).
1. Soil sealing. Soil sealing is prevalent in cities and roads. These areas can be easily identified, but soil crusting is a different phenomenon at the surface of the soil, which depends on soil texture, aggregate stability, topography, and rainfall characteristics [Valentin et al. 2004]. This appears to be too complex to be simulated at this scale and has not been taken into consideration. 2. Soil erosion. Soil erosion by water is probably the most well-known and popular aspect of land degradation and the most researched one. It may result in severe landscape deterioration by gullies and ravines or the silting up of dams downhill. Not all effects of soil erosion are negative and they may indeed improve downstream (or downwind) soil fertility. The farmers in the valley benefit from degrading farming practices uphill and deltas wouldn’t exist without erosion in the upland mountains. But it is the overall effect of accelerated erosion that is largely negative, as most of the
Land Degradation
479
displaced soil and nutrients end up in rivers or dams and, finally, are largely lost in the sea. Extensive research on soil erosion by water has been carried out and formulas exist that allow the prediction of soil loss as a function of slope, soil type, rainfall, effective land cover, and protecting practices [Wishmeier and Smith 1978]. The formula can be simulated on the basis of existing global databases and is included in this study. Landslides are point events that are not considered at this scale. 3. Wind erosion. Wind erosion may result in huge dust storms that deposit the finest particles thousands of kilometers from the source. It is much more difficult to predict than water erosion due to the lack of global base data, therefore it was not possible to include this phenomenon at this scale. 4. Soil compaction. Soil compaction due to high livestock densities mainly in dry areas and mechanization in agriculture. The effects on soil health are variable but they often results in the creation of impermeable soil layers close to the surface. 5. Soil nutrient mining. Nutrient depletion of soils is a widespread soil degradation phenomenon that occurs as a consequence of soil erosion (it is generally the topsoil, in which most soil nutrients are present, that erodes fastest) but also because of poor management practices such as slash and burn and other subsistence agricultural practices that do not replenish the nutrients taken out of the soil by the crops. 6. Soil pollution. An aspect of importance to soil chemical health is the absence of toxic substances in the soil. These are associated with very high input and management levels, for instance cadmium toxicity related to high phosphorus applications, which the use of excessive nitrogen (N) fertilizer often leads to pollution of groundwater. In the present approach, all the areas concerned with very high management are considered to be suffering from an ongoing moderate degradation. The areas concerned are derived from the input-management level and concern, in particular, Western Europe. 7. Soil salinization. Increased salinization occurs generally because of inadequately maintained irrigation systems or of inadequate management (inadequate drainage provisions). In order to characterize the salinization process, the extent of the irrigation and the overall national salinization estimates are used in combination to arrive at a georeferenced estimate for ongoing salinization in each irrigated area within a country.
10.5 WATER QUANTITY AND QUALITY Meteorological water, surface water, and groundwater are sources that are not known to the same extent. For meteorological water, we have information on the length of the growing period [FAO 1978], which takes into account the rainfall and the moisture storage in the soil. For surface and groundwater resources, national figures generated by AQUASTAT subnational data are scarce. No global overview map is available on water quality, neither for surface water nor for groundwater. Trends in water resource availability can be estimated from the actual water use as a percentage of the total renewable water resources. These trends are also influenced
480
World Soil Resources and Food Security
by the trends in meteorological water as described by the aridity index. Competing uses make high demands on water resources. These include agriculture (for irrigation), industrial uses (e.g., electricity generation), and human consumption (drinking water, etc.). An index has been calculated per country reflecting water use. It refers to freshwater withdrawal (excluding reused wastewater and desalinated water, but including fossil groundwater) as a percentage of the renewable freshwater resources. It is recognized that once the threshold of 75% is exceeded, the environmental cycle of water becomes problematic [FAO 2008]. Most affected countries include nearly all dryland countries. Countries with considerable water demands (United States, Europe) are rated as moderately under pressure. Drought has been recognized as one of the major driving forces of land degradation affecting the health and recovery of vegetation. As with any reduction in vegetation, it enhances the risk of wind and water erosion. Trends in meteorological water have been superimposed on the trends determined above to refine the estimates of changes in the water availability.
10.6 BIODIVERSITY CHANGES The UN Convention on Biological Diversity (UNCBD) did not create a widely accepted and globally available set of measures to assess biodiversity—unlike the Framework Convention on Climate Change (UNFCCC), for example—and consequently, there is no generally acceptable global overview map of biodiversity. An attempt to classify global biodiversity risk has been undertaken by Lee and Jetz [2008], combining the threat from the human-induced climate and land-use changes. The work is based on conservation risk areas that are subjected to past or future landcover transformation, divided by the area currently protected [Hoekstra et al. 2005]. Furthermore, the model analyzes four UN Millennium Development Goals (MDG) socioeconomic scenarios to examine changes in global land-use and climate under diverse social and political futures (including information on current and future distributions for 18 different land-cover types for the years 2000 (now), 2050, and 2100. Past global habitat loss was evaluated using a modified version of GLC2000 [Global Land Cover 2000]. Land-cover transformations across 55 terrestrial biome combinations were estimated using Terrestrial Ecoregions of the World [Olson and Dinerstein 1998]. Governance estimations were based on estimations of the proportion of area potentially transformed in 174 countries using protected areas data and World Bank indicators assessing the relationship between quality of governance and wealth. Global endemicity and endangerment of terrestrial vertebrates were also assessed using WildFinder: Online Database of Species Distribution [WWF 2006], and the International Union for Conservation of Nature and Natural Resources (IUCN) Red List of Threatened Species [IUCN 2006].
10.6.1 Changes in (Rural) Economic Productivity The economic implications of the provisioning of goods and services by ecosystems are as important as the biophysical provisioning itself, but are often forgotten or ignored when studying land degradation. Nonindustrial economic provisioning
Land Degradation
481
statuses and trends are not easily determined as one has to rely on country statistics of sometimes dubious accuracy. In addition, some countries have broken up during the period considered (most notably Yugoslavia and the Soviet Union), making statistical comparison difficult, if not impossible. In the LADA approach, the crop, livestock, and forest outputs in those areas are considered. Changes are generally determined between 1990–95 and 2003.
10.6.2 Changes in Social and Cultural Provisions of Ecosystems Cultural and social provisions of ecosystems have been evaluated using data as diverse as accessibility to markets, tourism (as an indicator for quality and attractiveness of the environment), and the presence of protected areas. Trends in social provisions were evaluated by considering the trends in the human development index. There are little or no data available on the evolution and trends of accessibility, although the globalization of the economy and the technological progress of the past decades have obviously resulted in an enormous reduction in the amount of time needed to reach any given destination. The same is true for the evolution of protected areas, where it is difficult to judge if the increase in area with respect to data generated earlier is real, or simply show the effect of better information on the subject. This implies that we cannot use the trend in these two status indicators to evaluate pressures and changes. The only trend that could be followed is the tourist trend, which may well reflect the enhanced access to cheap travel rather than an increasing value of the landscape and ecosystem. Therefore, to indicate trends in social and cultural services, the Human Development Index (HDI) calculated by UNDP is used, which is only available at the national level, but provides a more balanced evaluation of the social situation than more economic indicators like the GDP.
10.7 ANALYSIS OF RESULTS Based on the six factors in the radar diagram, an index of overall land degradation can be calculated by summing up all axes and subtracting them from the maximum (600). Additionally, one can also determine an environmental degradation index (EDI) that only considers the biophysical factors. These indexes were calculated in such a way that they range from 0 to 1, where 0 is the best and 1 is the worst situation, and 0.4–0.5 corresponds with a no-change or slightly improved situation. This calculation corresponds with stakeholders optimizing the total index by giving equal weight to each major good and service. From this analysis it appears that most areas of the world are moderately improving (LDI = 0.50–0.75) or moderately degrading (LDI = 0.25–0.50), while a large area is also occupied by unchanging wasteland (LDI = 0.40–0.50). Very few areas are improving a lot or degrading severely using this approach. But such an approach implies that a single major output of an ecosystem can reach very low values and still be rated moderately well as far as degradation is concerned, provided the total output remains relatively high. Under these circumstances, for instance, a high economic output coupled with social progress would justify the depletion of water resources beyond replenishment. In order to take this into account, a correction is made so that if a single good service is damaged
482
1,750 3,500
7,000 Km
Wastelands
Geographic coordinates
Water
Very low (< 0.25)
Low (0.25 to 0.50)
N
Moderate (0.50 to 0.75)
High (> 0.75)
FIGURE 10.3 Land degradation index. (From Nachtergaele, F.O. et al., Global Land Degradation Information System (GLADIS), beta version. An information database for Land Degradation Assessment at Global Level, 2010.)
World Soil Resources and Food Security
0
483
Land Degradation
severely (defined as those that imply a near irreversible loss of biomass, soil, water biodiversity, economic output, or social progress), the overall index is set at severe. These critical values correspond roughly to a limit of 25 (37.5 for soil) in each original axis. When analyzed on a regional and country basis, Figures 10.3 and 10.4 show that actual ongoing degradation processes over the past 20–30 years were the most pressing in steeplands, mainly because of soil erosion, and also in parts of West and Central Europe due to intensive agriculture leading to soil pollution and ongoing urbanization. Around the Mediterranean and in Eastern Brazil, high pressures on biodiversity are noted. In the Near East, water depletion is a major problem, while in Central Asia and Northern India and Pakistan, pressures were due to a combination of factors, in the latter certainly mainly due to water depletion. The African continent does show a moderate rate of degradation, with the Sahel showing up as a particularly improving area due to improved rainfall and better economic performance over the past two decades. South America shows an overall favorable picture except in the easternmost tip of Brazil, where biodiversity is under stress (Table 10.4). Comparing the LDI with the EDI (Figure 10.3 vs. Figure 10.8) clearly shows that the inclusion of the socioeconomic considerations improves the overall land degradation pictures and a number of areas that were moderately degrading environmentally are better rated when the socioeconomic trends are considered too (most notably Brazil and parts of China because of economic development), although some become worse (Central Asia because of slightly declining social provisions). Figure 10.5 shows what particular good or service is particularly under threat and where, indicating soil constraints as the most prevalent, followed by biodiversity loss, and water depletion. (a) Comparison of ecosystem G&S provisions
(b) Comparison of ecosystem G&S provisions
Biomass
Biomass
100
Social
50
100
Soil
Social
0
Economic
50
Soil
0
Water
Biodiversity Forest - virgin (ESSI 0.67) Sparsely vegetated areas - moderate or higher livestock density (ESSI 0.24)
Economic
Water
Biodiversity Forest - with agricultural activities (ESSI 0.66) Crops and mod. intensive livestock density (ESSI 0.37)
FIGURE 10.4 Changes in the environmental capacity to provide goods and services (1990–2005).
484
TABLE 10.4 Sample of Average Country Land Degradation Indices Trend in Soil Health/Axis 2 Process
Cyprus Lebanon Albania Greece Spain Lesotho
48.2 53.5 53.9 54.0 54.7 48.1
38.4 33.9 39.6 35.1 39.4 32.9
Bangladesh Togo Brazil Benin Burkina Faso
50.8 54.8 51.0 58.6 61.8
40.6 42.9 53.0 43.1 39.6
Trend in Productivity Value/Axis 5 Process
Trend in Social and Cultural Provisions/Axis 6 Process
Land Degradation Index
Most Degraded Countries 49.4 21.0 48.3 21.0 89.9 22.0 66.5 22.2 47.1 22.7 96.5 41.3
52.7 55.7 62.7 52.0 56.8 47.5
49.0 59.7 49.0 50.0 50.0 54.0
0.75 0.75 0.75 0.73 0.73 0.70
Least Degraded Countries 84.5 32.6 97.0 35.7 93.2 32.5 99.3 36.3 99.3 36.3
71.6 62.2 94.7 79.5 96.7
67.0 63.0 52.0 68.0 65.0
0.44 0.43 0.41 0.38 0.36
Trend in Water Stress / Axis 3 Process
Biodviersity Risk Resilience/Axis 4 Process
Source: Nachtergaele, F.O., and Petri, M., Mapping Land Use Systems at a Global and Regional Scale for Land Degradation Assessment, LADA Technical Report #8, FAO, Rome, 2008. Note: Sample average values of whole land degradation index and land degradation in biomass, soil, water, biodiversity, productivity and socio-cultural provision by countries. The land degradation indicators are estimated using the GLADIS method explained in Sections 10.4 to 10.7.
World Soil Resources and Food Security
Country
Greenness and Deforestation Trend/Axis 1 Process
Land Degradation
0
1,750 3,500
Wastelands
7,000 Km
Geographic coordinates
Water
High degr. (> 0.75)
Moderate degr. (0.75 to 0.50)
N
Stable
(0.50 to 0.40)
Mod. improvement (< 0.40)
485
FIGURE 10.5 Main threatened ecosystem goods and services. (From Nachtergaele, F.O. et al., Global Land Degradation Information System (GLADIS), beta version. An information database for Land Degradation Assessment at Global Level, 2010.)
486
World Soil Resources and Food Security
TABLE 10.5 Environmental Degradation by Major Land-Use System
Land-Use Systems Shrubs with high livestock presence Irrigated agriculture with high livestock presence Irrigated agriculture Agriculture crops and high livestock presence Grasslands with high livestock presence Shrubs with moderate livestock presence Grasslands– protected Grasslands– extensive pastoralism Rainfed agriculture Forestry–unused Shrubs–extensive pastoralism Wetlands–protected Wetlands–mangrove
Greenness and Deforestation Trend/Axis 1 Process
Trend in Soil Health/ Axis 2 Process
Trend in Water Stress/Axis 3 Process
Biodiversity Risk Resilience/ Axis 4 Process
Environmental Degradation Index
56.5
35.2
68.0
33.5
0.67
56.6
37.0
53.1
32.9
0.64
53.4 54.1
38.9 36.2
58.7 75.2
33.1 33.2
0.62 0.59
53.3
38.7
71.2
34.7
0.58
54.1
38.8
68.9
34.8
0.57
52.3
37.5
85.9
35.5
0.56
52.6
38.8
86.3
35.7
0.56
51.8 50.9 52.4
47.3 41.7 44.2
80.7 92.3 81.1
32.4 34.9 40.3
0.53 0.51 0.50
51.8 50.4
45.1 49.1
91.4 94.4
36.0 42.9
0.47 0.46
Source: Nachtergaele, F.O., and Petri, M., Mapping Land Use Systems at a Global and Regional Scale for Land Degradation Assessment, LADA Technical Report #8, FAO, Rome, 2008. Note: Average values of environmental degradation index and environmental degradation in biomass, soil, water, and biodiversity by land use system. The land degradation indicators are estimated using the GLADIS method explained in Sections 10.4 to 10.7.
Greenness and Deforestation Trend/Axis 1 Process
Trend in Soil Health/Axis 2 Process
French Guiana Libyan Arab Jamahiriya Western Sahara Suriname Russian Federation Namibia
53.7 49.9 50.0 52.4 52.6 53.2
49.4 49.6 50.0 49.2 38.1 42.4
Least Impacted Countries 99.4 48.0 50.0 98.7 92.9 87.8
Swaziland Sierra Leone India Haiti Malawi Rwanda Burundi
46.5 51.3 57.8 51.0 47.8 41.2 45.2
35.5 29.8 32.1 16.7 37.7 27.1 29.8
Most Impacted Countries 66.3 99.1 46.6 85.1 86.3 95.9 94.3
Country
Trend in Water Stress/Axis 3 Process
Trend in Productivity Value/Axis 5 Process
Trend in Social and Cultural Provisions/ Axis 6 Process
Land Degradation Impact Index
38.0 44.0 44.8 38.0 35.0 32.9
50.1 64.1 49.8 60.7 52.3
47.0 50.4 50.0 47.0 41.6 59.8
0.001 0.001 0.001 0.002 0.002 0.003
37.3 33.9 32.1 36.0 40.0 34.0 35.9
49.7 53.4 73.7 50.6 73.5 57.3 48.4
55.0 65.0 64.0 61.0 65.0 62.0 60.0
0.115 0.124 0.126 0.134 0.134 0.233 0.256
Biodiversity Risk Resilience/Axis 4 Process
Land Degradation
TABLE 10.6 Sample Country Average Index Illustrating the Impact of Land Degradation
Note: Example average values of land degradation impact index and land degradation index in biomass, soil, water, biodiversity, productivity and socio-cultural provision by countries. The land degradation indicators are estimated using the GLADIS method explained in Sections 10.4 to 10.8.
487
488
World Soil Resources and Food Security
Considered by land-use systems, cropped agriculture in combination with irrigation and livestock-rearing puts the most pressures on the ecosystem, as illustrated in Table 10.5, where the trends in biophysical pressures are combined in an environmental degradation index. Higher livestock numbers and more intensive cropped agriculture results in a higher environmental degradation rate. Forestry, protected areas, and wetlands—if unused—show the lowest environmental degradation. However, these changes and trends labeled degradation mask the fact that they have to be seen against an original state of the ecosystem (Figure 10.2). Given that many developing nations start off from an extremely poor biophysical and socioeconomic resource base, trends are likely positive, or less negative, in these areas. On the other hand, industrial nations are often concentrated in more favorable climates, having overall more fertile soils and a better socioeconomic base to start from. Consequently, problems with overuse of the ecosystem capacities by intensification, particularly affecting water and biodiversity resources, result in a relatively bleak negative picture as far as overall degradation is concerned in these industrial countries.
10.8 THE IMPACT OF LAND DEGRADATION An analysis simply based on the changes in the provisioning capacity of ecosystems is incomplete if it is not matched with an analysis of their impact on the affected people. This has been done in GLADIS by using the degradation trend as given in Figure 10.3 and matching it with the number of people affected and their poverty level. It is notable that the picture changes considerably when using this approach (Table 10.6) as little changes in poor and densely populated areas have a significant effect, while moderate to strong degradation trends in more affluent or desolate areas have a much smaller effect. Consequently, nearly the whole of Africa (with the exception of the wastelands, but in particular the Sahel, Central Africa, and Ethiopia) and southeast Asia (India and parts of China in particular) are most affected (Figure 10.6). Note: At the moment of going to press, the authors are revising the outcome of the GLADIS system on the basis of comments received from users. Although the overall methodology and main global results will remain fundamentally unchanged, the appearance of the maps and the details of the country figures will be revised. Hence, the data showed here are not supposed to be used for operational purposes.
10.9 N ATIONAL AND LOCAL LAND DEGRADATION STUDIES: INDICATORS AND MONITORING There is a plethora of methods, indicators, and punctual studies concerning specific aspects of land degradation at local and national levels. Barry and colleagues inventoried more than 900 different land degradation indicators in use in a sample of UNCCD countries. Efforts to harmonize these were undertaken by the UNCCD in a scientific conference [UNCCD 2009], but the political will to accept and report
Land Degradation
0
1,750 3,500
7,000 Km
Reduced threat
Geographic coordinates
Water
Biomass
Soil
Water
N
Biodiversity
Economic
Social
489
FIGURE 10.6 Impact of land degradation processes. (From Nachtergaele, F.O. et al., Global Land Degradation Information System (GLADIS), beta version. An information database for Land Degradation Assessment at Global Level, 2010.)
490
World Soil Resources and Food Security
on a number of indicators, using standard methods, proved to be lacking in most countries. Four methods deserve to be highlighted in this respect:
1. The WOCAT/LADA approach [Liniger et al. 2009] inventories in a participatory way at provincial levels the main parameters that describe the state, cause, and impact of degradation, and at the same time inventories the type and extent of sustainable land management interventions. The method has been standardized and tested in six countries (Argentina, China, Cuba, Senegal, South Africa, and Tunisia). This method allows us to obtain a baseline for future monitoring as illustrated in Figure 10.5. 2. The coupled human–environment (H–E) approach [Lebel et al. 2006] promotes the integrated consideration of biophysical and socioeconomic parameters linking institutional and policy considerations with land degradation, considering threshold tipping points beyond which systems can no longer be restored. This integrated approach has been applied, particularly at the local level, in drylands [Mortimer 2009]. 3. Remote sensing approaches: these have the significant advantage that data are continuously collected in an objective way and, as such, are ideally suited for monitoring purposes. Moreover, the resolution and detail of data available has increased during the past decade at a very fast pace. A disadvantage is that the observations are limited to land cover and derivatives that limit their scope somewhat. Examples are land cover change studies such as [Wessels et al. 2004]. Lately, soil properties have also been investigated in combination with ground truthing with some success [Sanchez et al. 2009]. 4. Local sampling techniques and surveys are quite objective and the most detailed of all, but they are generally rather expensive (with the notable exception of those promoted by LADA-local [McDonagh et al. 2008]. Another disadvantage of the sampling approach is the difficulty of extrapolating results, and the analytical variability often exceeds the changes of the parameter in time.
An example of punctual applied academic studies concerns soil nutrient decline: soil erosion and the removal of the harvest and crop residues depletes the soils. Several studies in the 1990s indicated soil nutrient depletion, particularly in sub-Saharan Africa [Elwell and Stocking 1988]. Nutrient budgets were made [Stoorvogel and Smaling 1990] and calculations performed for sub-Saharan Africa with 950,000 km2 affected [Henao and Baanante 2006]. These kinds of studies also have their limitations due to different interpretations of the same data and the assumed economic costs [Hartemink and Van Keulen 2005].
10.10 THE CAUSES OF LAND DEGRADATION It is undeniable that there are two groups of distinct causes for land degradation that have some areas of overlap. These are summarized in Table 10.1 and are (1) natural causes and (2) human-induced causes. Natural causes involve the inherent capacity
Land Degradation
491
of the ecosystem to provide goods and services. These include climatic ones such as drought, inherent climatic factors determining the capacity to generate biomass and provide ground cover and biodiversity, soil- and terrain-related causes such as slope and soil vulnerability to water and wind erosion, and water availability. Humaninduced causes are largely determined by land use and land-use change, economic factors related with the possibility of investment in the land and access to market, and social factors assuring the availability of infrastructure and the accessibility to land that allows farmers to produce at maximum capacity. A number of direct causes are seemingly natural but may have wholly or partly indirectly human causes (bush invasion, forest fires, floods, landslides, and droughts as a result of humaninduced climate change). Behind the direct obvious causes of human-induced land degradation, there often exist other, more deeply rooted, drivers that have to do with population pressure, poverty, lack of markets and infrastructure, poor governance and weak institutional frameworks, inadequate education, etc. Although undoubtedly correct, it is very difficult to prove a cause-and-effect relationship in a statistically significant way and relationships maintained often reflect more opinion than reality. Most of the relationships are based on a few local studies that have little value outside the study area. A good example is the Machakos study [Tiffen et al. 1994], which found a positive correlation between population density and less land degradation. Apparently present day conditions do not confirm these effects were sustainable, nor could they be extrapolated. Some [e.g., Bot et al. 2000] found on a global basis an opposite effect. The conclusion must be that, given that ecosystems produce a range of services and goods including economic and social ones, a cause that affects one service negatively may well affect another positively and a sensible trade-off, depending on the views of the stakeholders, should be reached depending on the local situation.
10.11 THE COST OF LAND DEGRADATION In the wake of the GLASOD study, a hot debate developed on the cost of land degradation. The major alarmist argument made the point that: Soil erosion is a major environmental threat to the sustainability and productive capacity of agriculture. During the last 40 years, nearly one-third of the world’s arable land has been lost by erosion and continues to be lost at a rate of more than 10 million hectares per year. With the addition of a quarter of a million people each day, the world population’s food demand is increasing at a time when per capita food productivity is beginning to decline. [Pimentel et al. 1995]
More recently, a rigorous study on soil erosion and food security and associated costs stated that: Production loss estimates that vary across crops, soils and regions but average 0.3% yr-1 at the global level, assuming that farmers’ practices do not change. These losses correspond to an estimated economic value of $520 million yr−1. Reducing production losses by limiting soil erosion would, therefore, go a long way to attain food security, especially in the developing countries of the tropics and subtropics. [den Biggelar et al. 2003]
492
World Soil Resources and Food Security
Note that most of these studies estimated costs of soil erosion, not of land degradation, which may be magnitudes higher when one also considers biomass, water, and biodiversity. Moreover, the studies are largely limited to productivity losses for which there is an overall problem of a lack of consistent relations between soil losses and productivity [Eswaran et al. 2001]. Unless the environmental cost (loss of carbon, decline in water resources, loss of cultural services, etc.) is correctly valued, it is clear that economic valuation results will largely underestimate the costs. Unfortunately, there is no widespread agreement about valuing ecosystem goods and services and, until that is achieved, no progress will be made in correctly estimating the real global or national cost of land degradation. On a more practical local level, Scherr [2003] argued that “Whatever the costs of land degradation, three kinds of contextual information are required.”
1. So long as degradation is reversible at an economically acceptable cost and other investment opportunities are more attractive, prevention is not always preferable, or even cheaper, than cure. 2. Even if the economic impacts of degradation are high, it may not be necessary to take direct policy actions apart from generally supportive measures for the agricultural economy. 3. From a policy perspective it may be wise to invest in soil protection and rehabilitation in areas with the greatest long-term significance for agricultural supply, rural poverty alleviation, or economic growth.
In addition, new economic options for use of degraded lands—growing biofuel crops or for carbon sequestration and carbon trading—will have additional spinoff environmental benefits. Scenarios are very uncertain in this respect, however, in the face of volatile and uncertain markets and the absence of international binding agreements.
10.12 CONCLUSIONS Land degradation is more than an environmental problem alone and should be considered in a holistic way, taking into account all ecosystem goods and services, biophysical as well as socioeconomic. Results should refer to a given time period and solutions should require a full consultation with stakeholders and imply trade-offs. Degraded lands, based on the capacity of the globe’s ecosystem to deliver goods and services, are highly variable. Degraded land occurs most in drylands and steep lands, which deserve special attention. The capacity to deliver ecosystem services is also, generally, significantly less in developing countries compared with industrial nations (Figure 10.7). Degradation of this capacity takes many forms and affects soils, biomass, water, biodiversity, economics, and social services derived from the ecosystem. This decline (degradation) appears to be proportional with the present capacity of the system. In other words, ecosystems with lower capacities decline less than ecosystems with higher capacities.
493
Land Degradation (a)
(b)
Trends of ecosystems services
Trends of ecosystems services
Biomass 100 Social
Biomass
100
Soil
50
Social
0
0
Economic
Water
Biodiversity Country Japan Nepal Spain
Soil
50
LDI 0.31 0.31 0.27
Economic
Water
Biodiversity Country Bangladesh Brazil Burkina Faso
0.56 0.58 0.64
FIGURE 10.7 Comparison between sample countries with low (a) and high (b) trend in ecosystems and services.
However, the impact of this degradation is most felt in areas with large populations and/or high poverty. This implies that, even when starting from a low resource base, the lower rate of degradation in these areas has a much greater impact compared to ecosystems with a higher capacity and a higher rate of degradation, but fewer poor people (Figures 10.9 and 10.10). Agricultural land uses (cropping, livestock-rearing) have a much higher risk to be degrading than nonagricultural ones. Land use and associated inputs and management are indeed the main direct causes of land degradation. Land use itself is determined by natural conditions and cultural and socioeconomic aspects including institutional settings, infrastructure, education, and market availability. There are quite a number of natural factors that cause land degradation, but often they are strongly interlinked, indirectly or directly, with human actions. Consequently, it is difficult to distinguish between the two. At a sub-national level, a harmonized and tested survey methodology developed by WOCAT and LADA is available and could become a harmonized, relatively lowcost way to quickly obtain standardized UNCCD indicators for country reporting on the impact of land degradation. At a local level, various approaches have been promoted. Among them, those based on integrated human–environment considerations, and those developed by LADA that use simplified sampling and socioeconomic surveys, appear to be the most promising. Remote sensing techniques have a definite role to play, particularly in monitoring land degradation, because they provide high resolution information on a continuous timescale. In addition, they are ideal to follow land-cover changes that are linked to land-use changes that are the major cause of degradation. However, until now, no
494
2,250
4,500
9,000 Km
Wastelands or no pop. or no pov.
Geographic coordinates
Water
High
Moderate
N
Low
Very low
FIGURE 10.8 Environmental Degradation Index – EDI. (From Nachtergaele, F.O. et al., Global Land Degradation Information System (GLADIS), beta version. An information database for Land Degradation Assessment at Global Level, 2010.)
World Soil Resources and Food Security
0
495
Land Degradation
None Light Moderate Severe Very severe Wastelands Urban Water
FIGURE 10.9 Percentages of population in degraded area by status of ecosystems services provisions. (a) Water
% of area degraded
Urban Wasteland or no pop Very severe Severe Moderate Light None
High
Medium Poverty
Low
% Population in degraded areas
(b)
Urban Wastelands Very severe Severe Moderate Light None
High
Medium Poverty
Low
FIGURE 10.10 Extent of land degraded by poverty class (a) and population in degraded areas by poverty class (b).
496
World Soil Resources and Food Security
unique remote-sensing–based methodology has proven to be able to go much beyond land cover parameters. The cost of land degradation has been hotly debated since the publication of the first global inventory. Given the widely different definitions of land degradation and the limited information on economic losses due to declining environmental services other than productivity, one can only state that the cost of land degradation has a significant impact in most developing countries, given that their overall capacity to generate ecosystem goods and services are significantly less than in most industrial nations. An overall observation concerning global level assessment is that there is an insufficiency of quantitative reliable data, particularly on available water resources and its trends and on economic factors that are often based on statistics of dubious quality. Therefore, apart from the complexity in interpretation highlighted here, the overall reliability of the input data remains another concern.
REFERENCES Bai, Z.G., D.L. Dent, L. Olsson, and M.E. Schaepman. 2008a. Proxy global assessment of land degradation. Soil Use Manag. 24:223–234. Bai, Z.G., D.L. Dent, L. Olsson, and M.E. Schaepman. 2008b. GLADA. Global Assessment of Land Degradation Improvement 1. Identification by remote sensing. LADA/ISRIC/ FAO. LADA technical report N.12. Bot, A.J., F.O. Nachtergaele, and A. Young. 2000. Land resource potential and constraints at regional and country level. World Soil Resources Report #90. Rome: FAO. de Jong, R., S. de Bruin, A. de Wit, M.E. Schapman, and D.L. Dent. 2011. Analysis of monotonic 1 greening and browning trends from global NDVI time-series. Remote Sensing of Environment 115(2):692–702. den Biggelaar, C., R. Lal, K. Wiebe, H. Eswaran, V. Breneman, and P. Reich. 2003. The global impact of soil erosion and productivity II: Effects on crop yields and production over time. Advances in Agronomy 81:49–95. Elwell, H.A., and M.A. Stocking. 1988. Loss of nutrients by sheet erosion is a major cost. Zimbabwe Science News 22:7–8, 83–85. Eswaran, H., R. Lal, and P. Reich. 2001. Land degradation: An overview. In Responses to land degradation. Proceedings of the Second International Conference on Land Degradation and Desertification, eds. E.M. Bridges, I.D. Hannam, L.R. Oldeman, F.W.T. Pening de Vries, S.J. Scherr, and S. Sombatpanit. Thailand: Khon Kaen; New Delhi: Oxford Press. Eswaran, H., P. Reich, and F. Beinroth. 2001. Global desertification tension zones. In Sustaining the global farm, eds. D.E. Stott, R.H. Mohter, and G.C. Steinback. Proceedings of the Tenth Annual ISCO Conference, May 24–29, 1999, Purdue University, West Lafayette, Indiana. Food and Agriculture Organization (FAO). 1978. Report on agro-ecological zones project. World Soil Resources Report 48, Rome: FAO. FAO. 2005. Global forest resources assessment 2005. Progress towards sustainable forest management. Forestry Paper 147. Rome: FAO. FAO. 2008. Aquastat: FAO’s information system on water and agriculture. http://www.fao.org/ nr/water/aquastat/main/index.stm. Global Land Cover 2000 database. European Commission, Joint Research Centre, 2003. http:// bioval.jrc.ec.europa.eu/products/glc2000/glc2000.php. Hartemink, A., and H. van Keulen. 2005. Soil degradation in sub-Saharan Africa. Land Use Policy 22:1.
Land Degradation
497
Henao, J., and C. Baanante. 2006. Agricultural production and soil nutrient mining in Africa— Implications for resource conservation and policy development. Muscle Shoals, AL: IFDC. Hoekstra, J.M., T.M. Boucher, T.H. Ricketts, and C. Roberts. 2005. Confronting a biome crisis: Global disparities of habitat loss and protection. Ecol. Letters 8:23–29. International Union for Conservation of Nature and Natural Resources (IUCN). 2006. Red List of Threatened Species. http://www.iucnredlist.org. Lebel, L., J.M. Anderies, B. Campbell, C. Folke, S. Hattfield-Dodds, T.P. Hughes, and J. Wilson. 2006. Governance and the capacity to manage resilience in regional socioecological systems. Ecology and Society 11:19. Liniger, H., G. van Lynden, F. Nachtergaele, and G. Schwilch (eds.). 2009. CDE-FAO-ISRICmapping land degradation and sustainable land management. LADA Technical Report #9. Rome: FAO. McDonagh, J., and S. Bunning. 2009. Field manual for local level land degradation assessment in drylands. LADA Technical Report #11. Rome: FAO. Mortimer, M. 2009. Dryland opportunities, a new paradigm for people, ecosystems and development. IUCN, IIED and UNDP. Gland, Switzerland: IUCN. Nachtergaele, F.O., M. Petri, R. Biancalani, G. van Lynden, and H. van Velthuizen. 2010. Global Land Degradation Information System (GLADIS) version 0.5. An information database for Land Degradation Assessment at Global Level. http://www.fao.org/nr/lada/ index.php?option=com_docman&task=cat_view&gid=26&Itemid=165&lang=en. Nachtergaele, F.O., and M. Petri. 2008. Mapping land use systems at a global and regional scale for land degradation assessment. LADA Technical Report #8. Rome: FAO. Oldeman, L.R., R.T.A. Hakkeling, and W.G. Sombroek. 1990. Global assessment of soil degradation. Wageningen, the Netherlands: International Soil Reference Information Centre. Olson, D.M., and E. Dinerstein. 1998. The global 200: A representation approach to conserving the earth’s most biologically valuable ecoregions. Conserv. Biol. 12:502–515. Pimentel, E., C. Harvey, P. Resosudarmo, K. Sinclair, D. Kurz, M. McNair, S. Crist, L. Shpriz, L. Fitton, R. Saffouri, and R. Blair. 1995. Environmental and economic costs of soil erosion and conservation benefits. Science 24:1117–1123. Sanchez, P.A., S. Ahamed, F. Carre, A.E. Hartemink, J. Hempel, J. Huising, P. Lagacherie, A.B. McBratney, N.J. McKenzie, M. de Lourdes Mendonca-Santo, B. Minasny, L. Montanarella, P. Okoth, C.A. Palm, J.D. Sachs, K.D. Shepherd, T.-G. Vagen, B. Vanlauwe, M.G. Walsh, L.A. Winowiecki, and G.-L. Zhang. 2009. Digital soil map of the world. Science 325:680–681. Scherr, S.J. 2003. Productivity related economic impacts of soil degradation in developing countries. An evaluation of regional experience. In Land quality, agricultural productivity and food security at local, national and global levels, ed. K. Wiebe, 223–261. Cheltenham Glos, UK: Edward Elgar. Sonnevald, B.G.J.S., and D.L. Dent. 2009. How good is GLASOD? J. Env. Manag. 90:274–283. Stoorvogel, J.J., and E.M.A. Smaling. 1990. Assessment of soil nutrient decline in sub- Saharan Africa, 1983–2000. Report 28. Wageningen, the Netherlands: Winard Staring Centre-DLO. Tiffen, M., M. Mortimore, and F. Gichuki. 1994. More people, less erosion: Environmental recovery in Kenya. Chichester, UK: Wiley & Sons. United Nations Convention to Combat Desertification (UNCCD). 1994. Elaboration of an international convention to combat desertification in countries experiencing serious drought and/or desertification, particularly in Africa. U.N. Doc. A/Ac.241/27, 33 I.L.M. 1328. Bonn, Germany: UNCCD.
498
World Soil Resources and Food Security
United Nations Development Programme (UNDP). 2009. Overcoming barriers: Human mobility and development. Human Development Report 2009. New York: UNDP. United Nations Environment Programme. (UNEP). 2007. Global environment outlook 4 (GEO-4). Nairobi, Kenya: UNEP. UNEP. 2008. Carbon in drylands: Desertification, climate change and carbon finance. UNEP/ UADP/UNCCD Tech. Note, CRIC 7. November 3–14, 2008, Istanbul, Turkey. Nairobi, Kenya: UNEP. UNEP/FAO. 1999. Guidelines for land use planning. Rome: FAO. Valentin, C., J.L. Rajkot, and D. Mitja. 2004. Response of soil crusting, runoff and erosion to fallowing in the sub-humid and semi-arid regions of West Africa. Agric. Ecosyst. Env. 104:287–302. Wishmeier, W.H., and D.D. Smith. 1978. Predicting rainfall losses—A guide to conservation planning. Agricultural Handbook No. 573. Washington, DC: USDA. Wessels, K.J., S.D. Prince, P.E. Frost, and D. Van Zyl. 2004. Assessing the effects of humaninduced land degradation in the former homelands of northern South Africa with a 1 km AVHRR NDVI time-series. Remote Sensing of Environment 91:47–67. World Wildlife Fund (WWF). 2006. WildFinder: Online database of species distributions, V. 6 (January). http://gis.wwfus.org/wildfinder/.
Do We Stand 11 Where 20 Years after the Assessment of Soil Nutrient Balances in Sub-Saharan Africa? E. M. A. Smaling, J. P. Lesschen, C. L. van Beek, A. de Jager, J. J. Stoorvogel, N. H. Batjes, and L. O. Fresco CONTENTS 11.1 Nutrient Stocks, Flows, and Balances...........................................................500 11.2 Calculating Nutrient Flows and Balances: Continent, Country, and District (NUTBAL)....................................................................................... 503 11.2.1 Assessment of Soil Nutrient Balances in SSA (1990)....................... 503 11.2.1.1 Mineral Fertilizers (IN1).................................................... 503 11.2.1.2 Manure (IN2)......................................................................504 11.2.1.3 Deposition (IN3).................................................................504 11.2.1.4 Biological N Fixation (IN4)................................................504 11.2.1.5 Sedimentation (IN5)...........................................................504 11.2.1.6 Harvested Product (OUT1).................................................504 11.2.1.7 Crop Residues (OUT2)........................................................504 11.2.1.8 Leaching (OUT3)................................................................504 11.2.1.9 Gaseous Losses (OUT4)...................................................... 505 11.2.1.10 Erosion (OUT5)................................................................ 505 11.2.2 Updating the 1990 Methodology.......................................................506 11.2.2.1 Mineral Fertilizer (IN1)...................................................... 507 11.2.2.2 Organic Inputs (IN2)........................................................... 507 11.2.2.3 Atmospheric Deposition (IN3)............................................ 507 11.2.2.4 N Fixation (IN4)................................................................. 507 11.2.2.5 Sedimentation (IN5)........................................................... 507 11.2.2.6 Harvested Product (OUT1)................................................. 508 11.2.2.7 Crop Residues (OUT2)........................................................ 508 11.2.2.8 Leaching (OUT3)................................................................ 508 499
500
World Soil Resources and Food Security
11.2.2.9 Gaseous Losses (OUT4)...................................................... 508 11.2.2.10 Erosion (OUT5)................................................................ 508 11.2.3 Applications at District Level............................................................ 511 11.3 Calculating, Monitoring, and Manipulating Nutrient Flows and Balances at Farm and Plot Level (NUTMON).............................................. 512 11.3.1 Farm-NUTMON................................................................................ 512 11.3.2 Calculating Farm-Level Nutrient Balances....................................... 513 11.3.2.1 Inputs–Outputs................................................................... 513 11.3.2.2 Internal Flows..................................................................... 514 11.3.2.3 Farm Income....................................................................... 517 11.3.2.4 Farm Level Heterogeneity.................................................. 517 11.3.3 From NUTMON to MonQI............................................................... 520 11.3.3.1 Joint Learning..................................................................... 520 11.3.3.2 Environment........................................................................ 521 11.3.3.3 Livelihoods......................................................................... 521 11.3.3.4 Agro-Food Chains and Certification.................................. 521 11.4 Studies on Soil Nutrient Dynamics Beyond the NUTMON Family............. 522 11.4.1 Nutrient Balances.............................................................................. 522 11.4.2 C and Nutrient Stocks........................................................................ 525 11.4.3 Integrated Nutrient Management....................................................... 526 11.5 Where Do We Stand, Where Do We Go?...................................................... 527 Acknowledgment.................................................................................................... 531 References............................................................................................................... 531
11.1 NUTRIENT STOCKS, FLOWS, AND BALANCES To feed 9 billion people in 2050, recent estimates indicate that global food production will have to increase by 70% [FAO 2009]. Food security can be realized by expanding the cultivated area and by increasing production per unit land, labor, or capital. Further down the production–consumption chain, increasing efficiency and recycling (including postharvest and waste management) and dietary change also are important. To increase agricultural production, area expansion is still possible, mainly in the range of Ukraine–Russia–Kazakhstan, in sub-Saharan Africa (SSA), and in Latin America, but it will negatively affect the services provided by natural ecosystems. The recent agricultural area expansions in Latin America (soybean for savannah and Amazon forest) and Southeast Asia (oil palm for rainforest and peat lands) clearly show the dilemmas: a booming international market for soy and palm oil and for soy meal as animal feed [Smaling et al. 2008], go at the expense of natural vegetation, plant and animal biodiversity, climatic stability, and above- and below-ground carbon (C) stocks. Hence, raising productivity is the main feasible option on the upstream part of the food security chain. This is still possible in most agricultural systems, but the immediate large increases in cereal production of the Green Revolution will not be repeated easily. As Conway [1997] expressed, “What is needed is a doubly green revolution that raises yields while reducing environmental impact.”
20 Years after the Assessment of Soil Nutrient Balances
501
In the quest for sustainable food production, much attention is given to improved varieties and biotechnology, irrigation, and crop protection measures. Soil fertility management seems of lesser priority. Nonetheless, low soil nutrient stocks and soil nutrient depletion are regarded as fundamental root causes of hunger and poverty [Sanchez 2002]. This is because crop and grass yields and dietary values are reflections of the nutrient status of the soil. Soil organic C (SOC), although not a nutrient, may be regarded as a proxy of the nutrient stocks in tropical soils. In soil fertility/crop production models such as Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS), SOC, as a proxy for soil organic matter, comes out as a key soil characteristic explaining crop response to soil fertility [e.g., Janssen et al. 1990; Smaling and Janssen 1993; Samaké 2003]. Estimates of the global mean SOC stocks in the top meter of soil are approximately 110 ton C ha−1, but the mean stock in Africa is only 57–60 tons C ha–1 and in West Africa even less, 42–45 tons C ha−1 [Batjes 2001]. Global amounts of soil nitrogen are estimated at approximately 135 × 109 tons of N for the upper 1 m [Batjes 1996]. A rapid analysis of the ISRIC-WISE (World Inventory of Soil Emission Potentials) database shows that European agricultural soils have an average SOC content that is twice the level of those in SSA [Batjes 2002]. This is not the result of land use history, but largely of differences in soil age and climatic conditions. With the exception of soils of recent volcanic origin, most African agricultural soils are derived from 2 billion-year-old granites, whereas many European agricultural soils are developed in periglacial and holocene sediments, as are many soils of the fertile deltas of Asia. Where there are stocks, there are flows. All soils gain and lose nutrients over time. Fertilizers, for example, represent nutrient inputs applied to realize nutrient outputs in harvested crop parts. Nutrients in sediments in lower reaches of river basins are inputs that relate to nutrients lost further upstream due to erosion. Hence, soil nutrient stocks change due to the combined effect of positive and negative flows. This is true at the scale of individual agricultural plots, but nutrient loss and accumulation occur at all scales, up to a global scale through trade in agricultural commodities. Countries with a net loss of NPK in agricultural commodities correspond to the major food exporting countries—the United States, Australia, and some Latin American countries. In the case of the United States, for example, exports of NPK were 3.1 million tons in 1997 and are expected to reach 4.8 million tons in 2020 [Grote et al. 2005]. West Asia and North Africa, China, and SSA are net importers of NPK in agricultural commodities, but the nutrients imported are commonly concentrated in the cities, creating waste disposal problems rather than alleviating deficiencies in rural soils. Calculating or estimating nutrient flows allows the drafting of a nutrient balance. For SSA, a continental nutrient balance (NUTBAL) study reveals that net flows were negative, i.e., 22 kg N, 2.5 kg P, and 15 kg K are lost annually per hectare over the 1982–1984 period [Stoorvogel and Smaling 1990; Stoorvogel et al. 1993]. This study triggered substantial debate on soil fertility management in SSA and the role of fertilizers, culminating in involvement of many donor agencies, as well as political commitments on fertilizer use at the Africa Fertilizer Summit in Abuja in 2006 [Sanginga and Woomer 2009]. Furthermore, a plethora of nutrient balance studies at different spatial scales emanated inside and out of Africa. Without the
502
World Soil Resources and Food Security
pretension to be exhaustive, this chapter summarizes 20 years of work on nutrient balances, mainly in SSA, but where appropriate with comparisons to other parts of the globe. An earlier review in Advances in Soil Science addressed NUTBAL and the pros and cons of the input–output approach at the subcontinental scale, where calculations are always based on secondary data and empirical models [Smaling and Oenema 1997]. The model was criticized for its methodological limitations [Faerge and Mahid 2004], and for its limited value for intervention and action [Scoones and Toulmin 1998]. After NUTBAL, the focus shifted to subnational and local scales and from calculating and modeling to measuring and monitoring. The acronym changed accordingly to NUTMON [Smaling and Fresco 1993], with integrated soil fertility management (ISFM) [Sanginga and Woomer 2009; Vanlauwe et al. 2010] and, more broadly, integrated nutrient management (INM), participatory learning and action [Defoer 2002] and resource flow mapping [De Jager 2007] as the actionoriented components. Ten years later, the approach considers other factors than just soil fertility management leading to monitoring for quality improvement (MonQI). It is the successor of the NUTMON field tool, as described by Vlaming et al. [2001], and is a farm monitoring tool that facilitates structured interviews with farmers concerning their daily management of crop and livestock, data entry, data storage and data checking, and data processing and presentation. Currently, MonQI is used by a blend of users from state-of-the-art science toward agencies for certification of niche markets. The timeline of the nutrient balance model development is shown in Figure 11.1. The highlights of the research performed under the NUTBAL and NUTMON umbrellas are described in Chapters 2 and 3. Studies outside the NUTMON family are summarized in Chapter 4. The final chapter provides an assessment of the current state of knowledge and possible future research and development pathways that follow from the analysis.
Use at continental and national levels FARM-NUTMON: Down-scaling to farm level, extension of compartments. Extension and improvement of calculation algorithms. Recoding of software, flexible programming.
Household economics
Tailor-made individual Incorporation of reporting module. optimal soil fertility management, water usage and C management.
Pesticide registration, value chain assessment
1990
1995
2000
2005
2010
FIGURE 11.1 Development of NUTMON family nutrient balance models over the past 20 years.
503
20 Years after the Assessment of Soil Nutrient Balances
11.2 C ALCULATING NUTRIENT FLOWS AND BALANCES: CONTINENT, COUNTRY, AND DISTRICT (NUTBAL) Inspired by Frissel [1978], Pieri [1985], and Van der Pol [1992], the subcontinental NUTBAL study adopts a clearly defined nutrient balance and quantified nutrient flows. It constitutes the basis for many subsequent nutrient balance studies as shown by several reviews [FAO 2003; Schlecht and Hiernaux 2004; Cobo et al. 2010]. In this chapter, the original 1990 NUTBAL methodology is summarized, followed by an overview of the updates up to the latest version described by Lesschen et al. [2007].
11.2.1 Assessment of Soil Nutrient Balances in SSA (1990) The initial NUTBAL study [Stoorvogel and Smaling 1990] assesses the state of soil nutrient depletion in SSA for 1982–1984 with a projection for 2000. It provides data on the net balance of the macronutrients N, P, and K from the rootable soil layer on a country basis. Production figures (1982–1984) and projections (2000) for major crops per country were provided by the UN Food and Agriculture Organization (FAO). These statistics are further specified for six largely climate-based land/water classes (LWC): low, uncertain, and good rainfall areas, problem areas, and naturallyflooded and irrigated areas. In addition to this, three broad soil fertility classes are used: low, medium, and high. Soil fertility dynamics is captured by five inputs (IN) and five outputs (OUT) as shown in Figure 11.2. The various model components, and underlying assumptions, are detailed below. 11.2.1.1 Mineral Fertilizers (IN1) The FAO database contains information on actual total fertilizer consumption per crop per country for 1982–1984 and projections for 2000. However, these data are not specified per LWC. Hence, literature data on the regional distribution of fertilizers within a country is used, or weighting factors are used for each land-use system (LUS). Animals IN2 Manure IN1 Mineral fertilizers IN3 Deposition IN4 Biological N fixation IN5 Sedimentation
Crops
Soil organic and mineral N, P, & K
OUT2 Crop residues OUT1 Harvested products OUT3 Leaching OUT4 Gaseous losses OUT5 Erosion
FIGURE 11.2 Nutrient flows to and from the soil. (From Stoorvogel, J.J., et al. Fert. Res., 35, 227–235, 1993. With permission.)
504
World Soil Resources and Food Security
11.2.1.2 Manure (IN2) NUTBAL only considers soil nutrient balances of arable land and does not consider the balance of extensive grazing lands. Two forms of manuring are distinguished: 1) manure collection from stables, kraals, and other storage places, and application to arable fields prior to planting, with fixed quantities (0, 500, 1000, or 1500 kg ha−1) per land use system, and 2) on-the-spot manuring by livestock feeding on crop residues with interaction with OUT2. 11.2.1.3 Deposition (IN3) Input by dry deposition, which mainly occurs in West Africa under the influence of the Harmattan dust storms, is determined by an interpolation of available measurements. For wet deposition, a regression with mean annual rainfall is carried out on the basis of literature data. 11.2.1.4 Biological N Fixation (IN4) Based on information from literature, three stipulations are presented, depending on total N demand by crops: 1) for symbiotic N fixation in leguminous species, 2) for chemoautotrophic N fixation in wetland rice and 3) for nonsymbiotic fixation. 11.2.1.5 Sedimentation (IN5) For the naturally flooded LWC, it is assumed that the nutrient balance is in equilibrium due to sedimentation. For the irrigated area LWC, the nutrient content of the irrigation water is also considered as an input factor. Based on literature, an annual input of 10 kg N per ha is assumed. 11.2.1.6 Harvested Product (OUT1) Based on literature, average values for the nutrient content of each crop are compiled (in kg nutrient per ton harvested product). In order to obtain an estimate of OUT1, these data are combined with FAO production figures. 11.2.1.7 Crop Residues (OUT2) An estimate of the amount of crop residues removed from the arable field is obtained from the literature. For each LUS, i.e., a unique combination of crop and LWC, a removal factor is assigned: complete removal of residues (e.g., used for fuel, roofing, or manufacturing), or incomplete (e.g., grazing or burning). Average values of the amount of nutrients in crop residues per harvested ton are compiled. 11.2.1.8 Leaching (OUT3) Leaching is a significant loss mechanism for some nutrients. Based on a literature review and expert consultations, the following regression equation is developed for N:
OUT3N = 2.3 + (0.0021 + 0.0007 × F ) × R + 0.3 × (IN1 + IN2) – 0.1 × U,
in which F is a soil fertility class (1 - low; 2 - moderate; 3 - high), R is rainfall (mm), and U is total nutrient uptake by the crop.
20 Years after the Assessment of Soil Nutrient Balances
505
11.2.1.9 Gaseous Losses (OUT4) Nitrogen losses through denitrification are expected to be highest in wet climates, on highly fertilized and clayey soils. Ammonia volatilization is linked to the amount of mineral and organic fertilizer and plays a role in alkaline environments. However, such soils are not very common in SSA, and therefore volatilization and denitrification are not treated separately. Based on scarce literature data the following regression equation is developed for N:
OUT4 = Base + 2.5 × F + 0.3 × (IN1 + IN2) – 0.1 × U,
in which Base is a constant value covering relative wetness of the soils specific for LWCs. 11.2.1.10 Erosion (OUT5) Soil loss estimates are based on the LUS descriptions within the LWCs. Additionally, a nutrient content is assigned to each soil fertility class. As the finest soil particles are the first to be dislodged during erosion, an enrichment factor is established, which is set at 2.0 for all three nutrients. As topsoil erodes, the roots of crops start to enter layers that were previously beyond the root zone. Hence, part of what is lost on top is gained at the bottom of the soil profile. These contributions are set at 25% percent of the calculated losses for P and K. The NUTBAL study presents N, P, and K balances for LUS and LWCs for most countries in SSA; negative balances are observed throughout the subcontinent (Figure 11.3). Densely populated regions in the Rift Valley (Kenya, Ethiopia, Rwanda, and Malawi) have the most negative values, owing to high ratios of cultivated land to
Low Moderate High Very high
FIGURE 11.3 Nitrogen depletion in sub-Saharan Africa. Low < 10; moderate 10–20; high 20–40; very high > 40 kg ha−1 yr−1. (After Stoorvogel, J.J., and E.M.A. Smaling, Assessment of Soil Nutrient Depletion in Sub-Saharan Africa: 1983–2000. Report 28, Winand Staring Centre, Wageningen, the Netherlands, 1990; Stoorvogel, J.J., et al. Fert. Res., 35, 227–235, 1993.)
506
World Soil Resources and Food Security
total arable land, relatively high crop yields (OUT1) and soil erosion (OUT5), as well as relatively high nutrient stocks. For SSA as a whole, the nutrient balances for 1982–1984 and projections for 2000 are −22 and −26 kg N ha−1 yr−1; −2.5 and −3.0 kg P ha−1 yr−1; and −15 and −19 kg K ha−1 yr−1, respectively. The projection for 2000 being more negative is partly attributed to optimistic FAO estimates for crop production in 2000 and the expected decrease in fallow areas in 2000.
11.2.2 Updating the 1990 Methodology The FAO, who commissioned NUTBAL in 1990, also facilitated an overhaul of the approach [FAO 2004]. The approach is broader and includes stocks, flows, and balances at macrolevel (national), mesolevel (district or province), and microlevel (village or farm). Nutrient balances are calculated at different levels in Ghana, Kenya, and Mali with important cash crops. The methodology can be applied to all SSA countries by using continental GIS maps and FAOSTAT data. The calculation is performed for N, P, and K based on averaged data for the period 1997–1999. The updated TABLE 11.1 Changes in Calculation of Nutrient Stocks and Flows When Moving from NUTBAL to NUTMON Land use systems Nutrient stocks IN1: mineral fertilizer IN2: organic inputs IN3: atmospheric deposition
IN4: nitrogen fixation IN5: sedimentation OUT1: crop products OUT2: crop residues OUT3: leaching OUT4: gaseous losses OUT5: erosion
The spatial distribution of land use systems is modeled through a biophysical land suitability assessment based on Ecocrop WISE database with soil nutrient concentrations per soil type of the 1:5,000,000 FAO soil map [Batjes 2002] Fertilizer use data per crop [IFA/IFDC/FAO 2000] and total consumption from FAOSTAT (Reference) Livestock density maps [Wint et al. 2000] in combination with literature data on nutrient contents Dry deposition: Improved map for Harmattan deposition (source) New regression on nutrient concentrations and IIASA rainfall map [Leemans and Cramer 1991] Percentage of leguminous crop production and related to rainfall [Leemans and Cramer 1991] Sedimentation calculated using the LAPSUS model [Schoorl et al. 2002] No changes No changes New regression model based on review by De Willigen [2000] New regression model based on data from IFA/FAO [2001] Erosion calculations using the LAPSUS model [Schoorl et al. 2002]
Source: Stoorvogel, J.J. and Smaling, E.M.A., Assessment of Soil Nutrient Depletion in Sub-Saharan Africa: 1983–2000. Report 28, Winand Staring Centre, Wageningen, the Netherlands, 1990; FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, 2004; Lesschen, J.P., et al. Nutr. Cycl. Agroecos., 78, 111–131, 2007.
20 Years after the Assessment of Soil Nutrient Balances
507
methodology is based on NUTBAL, but with a substantial number of improvements. Although LWCs in NUTBAL are to some extent spatially explicit, the revised version allows taking into account the spatial variation in soils, climate, and nutrient balances within a country. Updated procedures to calculate nutrient flows are used (Table 11.1). Nutrient stocks are quantified for each soil unit instead of using three soil fertility classes. A disaggregation procedure is developed to create a land–use map for all SSA countries, which shows the most likely crop distribution at a grid resolution of 1 km. The methodology is based on the principles of qualitative land evaluation, where land qualities are matched with land-use requirements to assess the suitability of land for a given use [FAO 1976]. The methodology involves three key steps: 1) identification of land units with similar topography, climate, and soil conditions; 2) matching properties of the land units with crop requirements; and 3) disaggregating harvested areas from FAOSTAT over the land units. The final land-use map is combined with other spatial data needed for the nutrient balance calculation. 11.2.2.1 Mineral Fertilizer (IN1) The FAOSTAT database provides figures for total fertilizer consumption per country. Data of the fertilizer use per crop studies [IFA/IFDC/FAO 2000] are used to derive the fractions of the total fertilizer consumption per nutrient for each crop. 11.2.2.2 Organic Inputs (IN2) Livestock density maps are available for the major livestock classes, i.e., cattle, small ruminants, and poultry [FAO 2000]. The livestock densities are multiplied by the excretion per animal per year and the nutrient content of the manure, for which updated figures are established. These amounts are corrected for country-dependent differences in management for both grazing and application of manure from storage. 11.2.2.3 Atmospheric Deposition (IN3) Updated factors for nutrient contents in both rainfall and dust are used, and the IIASA rainfall map is used for wet deposition [Leemans and Cramer 1991]. Based on several literature sources and windstream patterns, a new interpolated map of Harmattan dust is used for dry deposition. 11.2.2.4 N Fixation (IN4) For symbiotic N fixation, updated factors are used. For wetland rice, a fixed amount of 15 kg N ha−1 yr−1 is assumed for N fixation by cyanobacteria. For the nonsymbiotic N fixation and N fixing trees, a regression equation is developed based on annual rainfall. 11.2.2.5 Sedimentation (IN5) This flow consists of two parts: input of nutrients by irrigation water, and input of sediment as a result of erosion. For the nutrient input by irrigation water, the worldwide map of irrigation areas [Döll and Siebert 2000] is combined with the assumptions about nutrient content and amount of irrigation water of NUTBAL. The input by sedimentation is calculated by the LAPSUS model [Schoorl et al. 2002], which provides a feedback between IN5 and OUT5.
508
World Soil Resources and Food Security
11.2.2.6 Harvested Product (OUT1) The NUTBAL approach is used with updated crop production statistics. 11.2.2.7 Crop Residues (OUT2) The NUTBAL approach is followed with updated crop and country-dependent removal factors. 11.2.2.8 Leaching (OUT3) A new regression model for N is used:
OUT3N = (0.0463 + 0.0037 × (R/(C × L))) × ((IN1 + IN2) + D × NOM − U),
in which C is clay content in the topsoil (%), R is rainfall (mm), L is the layer thickness or rooting depth (m), D is the decomposition rate (set at 1.6 % per year), and NOM is the amount of N in soil organic matter (kg N ha–1). This N-leaching regression model is based on 43 measurements and accounted for 67% of the variance [De Willigen 2000]. The equation is slightly edited for perennial crops to prevent overestimation of N leaching. 11.2.2.9 Gaseous Losses (OUT4) A new regression model is used:
OUT4 = (0.025 + 0.000855 × R + 0.01725 × (IN1 + IN2) + 0.117 × SOC) + 0.113 × (IN1 + IN2),
in which SOC is the organic C content (%). The equation is based on a larger data set [IFA/FAO 2001] and consists of one regression model for the N2O and NOx losses through denitrification, and a direct loss factor for volatilization of NH3. The regression model has a R2 of 0.70. 11.2.2.10 Erosion (OUT5) To assess nutrient loss by erosion, the LAPSUS model [Schoorl et al. 2002] is used. This model simulates erosion and sedimentation at the landscape scale, which has several advantages: quantitative data is generated, erosion is considered at the landscape scale, and sedimentation is taken into account. Main input data of the LAPSUS model are the topographical potentials derived from a digital elevation model [USGS 1998] and rainfall surplus, derived from the rainfall map [Leemans and Cramer 1991]. Other input data are soil depth and erodibility, which are based on the soil map [FAO/UNESCO 1997], and a land cover map [USGS et al. 2000]. With these inputs the model simulates runoff and erosion–sedimentation for 1 year at a 1 km2 resolution. The loss or gain of nutrients is calculated by multiplying the erosion or sedimentation by the soil nutrient contents and an enrichment factor. Based on additional literature the enrichment factor is adjusted to 2.3 for N, 2.8 for P, and 3.2 for K. The revised NUTBAL study includes nutrient balance maps as shown in Figure 11.4 for Mali. Whereas the 1990 NUTBAL study indicates N depletion of 8 (1983)
509
20 Years after the Assessment of Soil Nutrient Balances
N depletion (kg/ha) > 100 50–100 20–50 5–20 0–5 Positive No data
N W
E S
FIGURE 11.4 Example of the spatially explicit nitrogen balance for Mali. (Data from FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities. FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004. With permission.)
to 11 (2000) kg ha−1 yr−1, the updated approach shows large differences within the country. The N balance is mainly positive in the central part of Mali, where rice and fallow are the main land uses, while the northern border of the agricultural zone, with mainly millet and sorghum cultivation, shows severe depletion. Projected N flows for Ghana, Kenya, and Mali are summarized in Figure 11.5. In Kenya, the input of mineral and organic fertilizer is relatively important, whereas 15
Ghana Kenya Mali
10
Nitrogen (kg/ha)
5 0 –5
–10 –15 –20 –25
IN1
IN2
IN3
IN4
IN5 OUT1 OUT2 OUT3 OUT4 OUT5 Nitrogen flows
FIGURE 11.5 Nitrogen flows for Ghana, Kenya, and Mali. (Data from FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities. FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004. With permission.)
510
World Soil Resources and Food Security
Ghana has a greater input by atmospheric deposition because of Harmattan dust. Outflows by leaching and gaseous losses are somewhat greater in Kenya where more mineral fertilizers are used. Most apparent are the N-losses due to widespread water erosion in Kenya. According to the updated NUTBAL study, Kenya has the greatest nutrient depletion for N and K, followed by Ghana and Mali. For P, Ghana and Mali show slightly negative balances, while the P balance for Kenya is neutral (Table 11.2). Kenyan farmers apply 30,000 tons of P with mineral fertilizer (IN1), which is 15 times that applied in Ghana and five times that applied in Mali. A comparison with the results of the 1990 NUTBAL study shows that the latter’s projections for 2000 is in agreement with the projections from the updated NUTBAL, particularly for Ghana and Mali. For Kenya, nutrient depletion is about 25% less severe compared to the projections for 2000 by the NUTBAL study. The 1990 NUTBAL study has no detailed uncertainty analysis, although uncertainty in the projections can be great due to the assumptions and simplifications used. This has been remedied in the revised version. Uncertainty in the nutrient balances can be attributed to various biases and errors. It is usually smaller for farmgate balances (dealing mainly with IN 1, IN2, and OUT1) than for soil surface balances that consider leaching and gaseous losses [Oenema et al. 2003]. Lesschen et al. [2007] apply the revised NUTBAL methodology [FAO 2004] to Burkina Faso, and this includes an uncertainty assessment. Cross and spatial correlations between the various sources of uncertainty and their scale-dependency are taken into account. In the case of Burkina Faso, the error margin in the projected soil nutrient balance is −20 (±15) kg N ha−1 yr−1, −3.7 (±2.9) kg P ha−1 yr−1, and −15 (±12) kg K ha−1 yr−1. Overall, however, uncertainty associated with the soil nutrient balances is relatively low, compared to uncertainties of all input data. According to the uncertainty analysis, most LUSs are being depleted in soil nutrients.
TABLE 11.2 Comparison between the Nutrient Balance Calculations of Stoorvogel and Smaling and the FAO FAO [2004]
Stoorvogel and Smaling [1990]
1997–1999 Ghana Kenya Mali
1982–1984
2000
N
P
K
N
P
K
N
P
K
−27 −38 −12
−4 0 −3
−21 −23 −15
−30 −42 −8
−3 −4 −1
−17 −29 −7
−35 −46 −11
−4 −1 −2
−20 −36 −10
Source: Stoorvogel, J.J. and Smaling, E.M.A., Assessment of Soil Nutrient Depletion in Sub-Saharan Africa: 1983–2000. Report 28, Winand Staring Centre, Wageningen, the Netherlands, 1990; FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004. Note: Values for the year 2000 are projections carried out in 1990.
511
20 Years after the Assessment of Soil Nutrient Balances
TABLE 11.3 Nutrient Balance for Nkawie District, Ghana Area Crop Cassava Maize Plantain Cocoa All crops
N
(ha) 11 838 11 455 11 725 48 493 110 262
P
K
(kg/ha) −68.3 −32.4 −8.7 −3.2 −18.0
−9.6 −6.3 −0.3 −0.1 −1.9
−59.0 −20.3 −35.6 −8.5 −20.3
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004.
11.2.3 Applications at District Level The updated NUTBAL study includes an assessment of nutrient balances at district levels in Ghana, Kenya, and Mali for major cash crops. Tables 11.3 and 11.4 show the nutrient balances for the four most important crops in Nwakie District (Ghana) and Embu District (Kenya), as an example. It is apparent that these cash crops have much better nutrient balances than the food crops. Figure 11.6 shows the dominant role of cotton in the N balance of the Koutiala region of Mali. Soil N outputs under sorghum and millet are smaller than total N outputs under cotton, even though the latter receives large inputs through fertilizers. However, in the common rotational systems, millet and sorghum scavenge on the fertilizer applied to cotton in the preceding year. The study shows that it is possible to construct a nutrient balance at district level, also offering entry points for farming and land-use strategies.
TABLE 11.4 Nutrient Balance of the Tea-Coffee-Dairy Zone, Embu District, Kenya Area Crop Maize Beans Coffee Tea All crops
N
(ha) 5 143 2 748 8 813 1 092 20 678
P
K
(kg/ha) −174.2 −142.0 −39.1 −16.3 −95.6
−31.2 −25.9 −7.6 −1.4 −14.9
−73.0 −23.8 −7.3 −2.3 −33.1
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004.
512
World Soil Resources and Food Security 40
Cotton Millet Sorghum
Nitrogen (kg/ha)
30 20 10 0
–10 –20 –30
IN1
IN2
IN3
IN4
IN5
OUT1 OUT2 OUT3 OUT4 OUT5
Nutrient flows
FIGURE 11.6 Nitrogen flows for the three major crops, Koutiala Region. (Data from FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities. FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004. With permission.)
11.3 C ALCULATING, MONITORING, AND MANIPULATING NUTRIENT FLOWS AND BALANCES AT FARM AND PLOT LEVEL (NUTMON) 11.3.1 Farm-NUTMON The farm and plot levels offer new dimensions to nutrient balance research. Nutrient stocks and flows can be calculated and estimated, but also monitored. Next, internal flows between farm and plot compartments can play a large role, in addition to the inputs and outputs of Figure 11.2. The internal flows become visible by partitioning farms and plots into several compartments or activity levels such as household, plots, livestock, feedstocks, and redistribution units (stables, compost heaps, latrines, etc.). The conceptual framework is shown in Figure 11.7. By using internal farm compartments, different farming strategies can be evaluated and best practices identified. Such best practices can be labeled INM and offer concrete options for action. INM technologies often combine one or more of the following categories: • Adding nutrients to the system (increasing INs), such as the application of mineral fertilizers and amendments, concentrates for livestock, organic inputs from outside the farm, and N-fixation in wetland rice and by leguminous species • Saving nutrients from being lost from the system (decreasing OUTs), such as erosion control, keeping crop residues inside the farming system, and planting deep-rooting species to reduce leaching losses • Recycling the volume of nutrients within the system so as to maximize nutrient use efficiency and system productivity (improving routing of internal flows)
513
20 Years after the Assessment of Soil Nutrient Balances
IN 1,2,3,4,5
Farm PPU
OUT 1,2
OUT 2,3,4
IN 1,2
RU IN 1,2
IN 2
SPU
Stock
Household
OUT 1,2
LEGEND External flows IN 1 IN 2 IN 3 IN 4 IN 5
in flows
mineral fertilizer organic manure wet + drt deposition biological N-fixation sedimentation
OUT 1 OUT 2 OUT 3 OUT 4 OUT 5 OUT 6
out flows
Internal flows
crop products crop residues animal products fresh manure FYM/compost household waste
harvested crop parts crop residues leaching denitrification water erosion human feces
OUT 1,2
OUT 6
FIGURE 11.7 Nutrient inflows, outflows, and internal flows used in NUTMON. PPU = primary production unit; SPU = secondary production unit; RU = redistribution unit; FYM = farmyard manure. (Data from Van den Bosch, H., et al., Agric. Ecos. Envir., 71, 62–80, 1998; Vlaming, J., et al., Monitoring Nutrient Flows and Economic Performance in Tropical Farming Systems (NUTMON)—Part 1: Manual for the NUTMON-Toolbox, Alterra, Wageningen, the Netherlands, 2001. With permission.)
Also, an economic component can be added to the nutrient monitoring, as IN1, IN2, OUT1, and OUT2 can be expressed in monetary units. Moreover, nutrient management is brought inside the wider domain of farm household production and consumption strategies. The resulting monitoring framework is known as NUTMON. The different features of NUTMON are described in a suite of papers [Smaling and Fresco 1993; Van den Bosch et al. 1998; De Jager et al. 1998a, 1998b], that form part of a larger series of documented nutrient balance studies [Smaling 1998; Smaling et al. 1999], and a characterization of major SSA farming systems on the basis of nutrient stocks, flows, and INM technologies [Smaling and Dixon 2006]. NUTMON has, since its inception, been used to i) calculate input–output balances [e.g., Van den Bosch et al. 1998], ii) understand actual farm practices under diverging agroenvironmental conditions [e.g., Van Beek et al. 2009], iii) perform impact assessments of interventions [e.g., De Jager 2007], iv) guide registration and certification procedures [e.g., De Jager 2007], and v) support joint learning and participatory research with different (groups of) stakeholders [e.g. Onduru et al. 1999, 2001].
11.3.2 Calculating Farm-Level Nutrient Balances 11.3.2.1 Inputs–Outputs To demonstrate differences in farm nutrient management, the results of five major NUTMON studies are brought together (Table 11.5). All studies in the table refer to mixed smallholder farming systems of, on average, 3 ha, but the objectives and monitoring periods of the projects differed. Table 11.6 provides the average nutrient balances recorded in the projects listed in Table 11.5. The results follow from a database
514
World Soil Resources and Food Security
TABLE 11.5 Characteristics of Major NUTMON Projects Project LEINUTS
NUTSAL
VARINUTS
Objective Identification of potentials of low-external input and sustainable agriculture to attain productive and sustainable land use in Kenya and Uganda Assessment and monitoring of nutrient flows and stocks and development of appropriate nutrient management strategies for semiarid areas in Kenya Spatial and temporal variation of soil nutrient stocks and management in sub-Saharan African farming systems
PIMEA INMASP
INM to attain sustainable productivity increases in East African farming systems
Main Cropping System
Monitoring Period
Maize, coffee, vegetables
1997
Maize, peas, sorghum, beans
1999, 2000
Maize, coffee
1997
Beans, barley, wheat Cassava, sorghum, millet, maize
2001 2002
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004.
that contains approximately 500 farm records. Each site is characterized by net farm income (NFI), i.e., the gross margins of all farm activities, excluding off-farm labor and income. Hence, the NFI is an indicator of farm profitability. In Table 11.6, NFIs are converted to US$ at the time of monitoring and cover one growing season. The range is considerable, and includes two sites with negative NFI, in the low-potential areas of Mbeere (Kenya) and Pallisa (Uganda). Next, full and partial nutrient balances are given. Full nutrient balances refer to the sum of all inputs minus the sum of all outputs. Partial balances only cover the easy-to-quantify flows IN1, IN2, OUT1, and OUT2. They are clearly incomplete from a biophysical standpoint, but more accurate than a full balance and offering a basis for strategic farm decisions. The partial balance is generally regarded as an indicator for the efficiency of nutrient use, whereas the full balance is an indicator for soil nutrient depletion. Figure 11.8 shows the breakdown of the full N balance for each site. Highest losses of N were estimated for erosion and leaching, but the uncertainties in these estimations are high. On the input side, most entries are easy to quantify and hence considered relatively accurate. 11.3.2.2 Internal Flows Strategic management of the interactions between the crop and livestock compartments, or the PPU and SPUs (Figure 11.7), is common in many farming systems, and a major INM technology. Grass, fodder crops, and crop residues are eaten by animals, whereas manure is returned to the grasslands and crop fields. Figure 11.9
Project
Country
LEINUTS LEINUTS NUTSAL2 NUTSAL NUTSAL NUTSAL NUTSAL Varinuts3 Varinuts Pimea4 Pimea INMASP5 INMASP INMASP INMASP INMASP INMASP INMASP INMASP INMASP
1
Kenya Uganda Kenya Kenya Kenya Kenya Kenya Kenya Burkina Faso Ethiopia Ethiopia Uganda Uganda Uganda Ethiopia Ethiopia Kenya Kenya Kenya Kenya
District Nyeri Pallisa Machakos Makueni Makueni Mwingi Kajiado Embu Manga Teghane GoboDeguat Wakiso Pallisa, Chelekura Pallisa, Akadot Solkua Wache Mbeere, Munyaka Kiambu, Kibichoi Kiambu, Ngaita Mbeere, Kamugi
No. of Farms 19 15 29 19 17 13 8 16 31 8 11 28 10 21 20 14 31 31 16 32
NFI (US$ farm−1) 172 147 28 89 107 28 478 581 943 1038 765 4131 886 –243 831 1013 184 272 1019 −933
Nfull
Pfull
Kfull
Npart
Ppart
Kpart
−10.4 −2.6 −50.0 1.1 −15.8 −17.2 −28.7 3.4 −23.5 −121.8 −0.7 −8.2 −23.7 −2.6 −3.7 −8.8 –45.6 111.5 −66.2 −11.6
3.1 36.7 5.9 −1.7 −1.3 −0.8 −2.1 7.2 8.2 −64.3 0.5 0.1 −0.4 2.3 0.0 −12.4 −8.4 18.6 −2.1 3.8
2.5 16.9 −28.8 −5.4 −1.8 −0.6 −3.7 19.7 −3.3 −36.2 7.5 −0.4 −7.6 −0.1 1.9 −28.2 −3.8 155.4 −9.2 −8.2
−1.4 12.2 68.0 36.7 4.2 0.3 0.2 32.7 36.6 −1.4 9.8 0.3 −0.1 4.7 5.8 0.6 9.9 178.3 31.7 4.3
−0.3 1.8 41.8 13.0 1.0 0.2 0.6 7.9 9.3 −0.2 1.0 0.1 1.2 −0.3 0.5 1.1 2.1 26.6 7.3 6.2
−0.7 16.3 44.8 8.5 8.0 0.2 1.2 25.2 4.9 −2.5 10.2 0.2 −3.9 −0.6 7.7 −1.9 3.2 173.2 12.4 0.9
515
Source: 1Onduru, D.D., et al., Exploring New Pathways for Innovative Soil Fertility Management in Kenya. Managing Africa’s Soils, No. 25, IIED, London, 2001; 2Gachimbi, L.N., et al., Land Use Policy, 22, 13–22, 2005; 3Onduru, D.D., et al., Participatory Research of Compost and Liquid Manure in Kenya. Managing Africa’s Soils, No. 8, IIED, London, 1999; 4 De Jager, A.H., et al., Internat. J. Sustain. Agric., 3, 189–205, 1999; 5 Van Beek, C.L., et al., Agricultural, Economic and Environmental Performance of Four Farmer Field Schools in Kenya, INMASP Project Reports No. 18, Alterra, Wageningen, the Netherlands, 2005. Note: NFI = Net farm income. The subscripts full and part refer to full balances including hard to quantify flows and partial balances consisting only of easy to quantify flows, respectively (see text), and are expressed in kg ha−1 season−1. The upper seven sites were monitored twice, the others once.
20 Years after the Assessment of Soil Nutrient Balances
TABLE 11.6 Overview of Datasets of Previous NUTMON Studies (1997–2002) and Main Results at Farm Level
516
World Soil Resources and Food Security
Embu Manga Teghane GoboDeguat Matuu Kiomo Kibwezi Kasikeu Enkorika Pallisa Kabarole Nyeri Machakos Wakiso Chelekura Siakago Ngaita Kibichoi Gachoka Wache Solkua Akadot –300
–200
–100
0
Mineral fertilizer Organic fertilizer Inflow grazing Atmospheric deposition
100 kg N ha–1
200
Biological N fixation Harvest crop products Removed crop residues Outflow grazing
300
400
Leaching Gaseous losses Erosion Human feces
FIGURE 11.8 Breakdown of full N balances for the NUTMON project sites.
and Table 11.7 (for each project), show the magnitude of the N flows from crop to livestock and vice versa, averaged for the dataset of Table 11.6. Clearly, the flow of N from crops to livestock exceeds the flow of N from livestock to crop. At the same time, external inputs of N were greater for the livestock compartment than for the crop compartment. In other words, more N is imported as concentrates and fodder from external markets than N imports through fertilizers. However, exports of N 9 13
b
Crop 14 a
39 Livestock Farm 8 24
External (markets)
FIGURE 11.9 Average N flows between crop and livestock at farm level and interactions with external compartments (markets, kg farm−1).
517
20 Years after the Assessment of Soil Nutrient Balances
TABLE 11.7 Internal N Flows (kg) between PPUs and SPUs, Averaged per Project Project INMASP LEINUTS NUTSAL PIMEA Varinuts
From Crop to Livestock 9 55 61 6 37
From Livestock to Crop 3 19 28 <1 8
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, Rome, FAO, 2004.
from the livestock and crop compartments were about equal (8 and 9 kg N farm−1, respectively). The relatively high N flows from crop to livestock may have several reasons, including poor N efficiency of livestock, difficulties in collecting livestock droppings (depending on livestock management system), and poor composting and/ or storage facilities. Using the data of Figure 11.9, the compartmental N balances equaled −21 kg N farm−1 yr−1 for the average PPU/crop compartment and +41 kg N farm−1 yr−1 for the average SPU/livestock compartment. Nevertheless, cash expenses on crops exceeded cash expenses on livestock on average by 90% and include seeds, fertilizers, traction, labor, pesticides, etc. 11.3.2.3 Farm Income Most often, economic farm performance is evaluated using NFI (Table 11.6). It is, however, increasingly recognized that sustainable nutrient management is also related to socioeconomic farm conditions [De Jager 2005]. To demonstrate these relations, Pearson’s correlations are shown for N and P flows and a few social and economic characteristics as shown in Tables 11.8a and 11.8b, respectively. The tables demonstrate the multiple interactions between different flows and the complexity of unraveling causes and effects of nutrient management at farm level. The analysis of the dataset does not reveal a consistent picture of relations, which is most likely due to the large variations in socioeconomic circumstances of the farm households included in the database. The N-total balance and NTot/NStock ratio (quantifying the full N balance as a fraction of the soil N stock) are negatively related to the NFI (Table 11.8a). The number of household members and the market share have little effect on farm performance and nutrient balance. However, the number of tropical livestock units (TLUNO) has considerable effects on several balance entries and on the NFI. With regard to P (Table 11.8b), most strikingly there are no correlations with NFI. However, significant correlations exist with respect to the presence of livestock. Clearly, livestock increases P inputs and the partial P balance. 11.3.2.4 Farm Level Heterogeneity Table 11.6 lumps project output and therefore does not show relevant diversity observed at the field level where the projects took place. Table 11.9 shows more
518
World Soil Resources and Food Security
TABLE 11.8a Pearson’s Correlations for N Fluxes (p < 0.0001, n > 540) NSTOCK Ntot/Nstock N_IN1 N_IN2 N_IN2B N_IN3 N_IN4 N_OUT1 NTOT NFI NPART N_OUT2 N_OUT2B N_OUT3 N_OUT4 N_OUT5 N_OUT6 MARKETSH TLUNO NOHHM
NSTOCK Ntot/Nstock N_IN1 0.3 N_IN2 0.2 0.4 N_IN2B 0.2 N_IN3 –0.2 N_IN4 0.3 N_OUT1 –0.5 –0.3 NTOT –0.3 0.7 0.2 –0.4 –0.3 NFI –0.4 0.2 –0.3 –0.3 NPART 0.2 0.2 0.7 0.5 0.6 0.3 N_OUT2 0.3 N_OUT2B –1.0 –0.2 –0.5 N_OUT3 –0.2 0.7 –0.4 –0.4 0.9 –0.3 N_OUT4 –0.5 0.3 –0.3 –0.4 –0.4 0.8 –0.2 –0.2 0.7 N_OUT5 –0.3 0.2 –0.4 –0.3 0.6 0.3 0.5 N_OUT6 –0.2 0.3 0.4 MARKETSH TLUNO –0.5 0.2 0.2 0.2 –0.2 0.3 0.3 –0.2 –0.4 –0.2 NOHHM –0.4
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004. Note: NStock refers to the total N content in the upper 30 cm of soil; Ntot is the total N balance; NFI is the net farm income; Npart is the partial N balance; Marketsh is an indicator for the market orientation of the farm; TLUNO is number of tropical livestock units at farm level; and NOHHM is the total number of household members.
TABLE 11.8b Pearson’s Correlations for P Fluxes (p < 0.0001, n > 540) NFI MARKETSH TLUNO NOHHM PSTOCK PTOT PPART P_IN1 P_IN2 P_IN2B P_IN3 P_OUT1 P_OUT2 P_OUT2B P_OUT5 P_OUT6
NFI
MARKETSH TLUNO NOHHM PSTOCK PTOT –0.8 PPART 0.3 0.2 P_IN1 0.3 0.9 P_IN2 0.2 0.5 0.2 P_IN2B 0.2 P_IN3 –0.2 0.2 P_OU –0.2 P_OUT2 P_OUT2B –0.2 –0.9 P_OUT5 –0.8 –0.3 P_OUT6 –0.4 –0.3 –0.2 0.2 0.2
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004.
Characteristic Altitude (m) Rainfall (mm) Mean temp (°C) Main land use Main soil types Total soil N (g kg−1) N balance (kg ha−1 yr−1) P balance (kg ha−1 yr−1) K balance (kg ha−1 yr−1)
AEZ1
AEZ2
AEZ3
AEZ4
AEZ5
1 770 1 750 16.8 Tea/dairy Andosol/Nitisol 7.5 −143 −4.0 −11.7
1 590 1 400 18.2 Tea/coffee/dairy Nitisol 6.4 −197 −11.6 −30.3
1 320 1 200 20.2 Coffee/maize Nitisol 4.4 −143 −3.6 −31.2
980 900 21.4 Tobacco/food crops Luvisol 2.1 −30 8.8 1.8
830 800 22.6 Livestock/shifting cultivation Lixisol 0.9 −27 −1.9 6.6
20 Years after the Assessment of Soil Nutrient Balances
TABLE 11.9 Land Characteristics and Nutrient Balances of Agroecological Ones in Embu District, Kenya
Source: FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, Rome, FAO, 2004.
519
520
World Soil Resources and Food Security
TABLE 11.10 Partial N Balance for Three Soil Fertility Management Classes in Two Villages, Southern Mali M’Peresso Number of farms IN1 IN2 OUT1 OUT2 Partial balance
Noyaradougou
Class 1
Class 2
Class 3
Class 1
Class 2
Class 3
3 15.9 23.8 21.6 19.4 –1.3
10 16.4 16.4 18.8 13.3 0.7
7 13.4 14.5 17.3 13.1 –2.5
8 42.9 11.5 28.2 17.3 8.9
5 42.1 8.1 24.0 15.0 11.2
7 40.6 11.8 22.6 17.3 12.5
Source: Kanté, S., Gestion de la Fertilité des Sols par Classe d’Exploitation au Mali-Sud. PhD thesis, Wageningen University, the Netherlands, 2001; FAO, Scaling Soil Nutrient Balances—Enabling Mesoscale Approaches for African Realities, FAO Fertilizer and Plant Nutrition Bulletin 15, FAO, Rome, 2004.
detail for the farms monitored during the VARINUTS project in Embu, Kenya [FAO 2004]. In each agroecological zone (AEZ), three farms were monitored. Although the choice of farms does not pretend to be fully representative of each AEZ, it is clear that the wetter AEZs in the sloping areas with little fallow land have the more negative nutrient balances, but also have the larger N stocks. In Southern Mali, partial nutrient balances were calculated for two villages where cotton, millet, and sorghum were the major crops (Table 11.10). Villagers grouped each other into three classes related to nutrient management. M’Peresso had the higher livestock density (0.36 vs. 0.12 TLU ha−1), explaining the higher IN2 values. Noyaradougou compensates this by higher fertilizer use. M’Peresso has average soil N stocks of 600 kg ha−1 in the upper 40 cm of soil, whereas Noyaradougou has 900 kg ha−1 [Kanté 2001; FAO 2004].
11.3.3 From NUTMON to MonQI In the past years, the scope of application of NUTMON has been broadened and the model renamed MONitoring for Quality Improvement (MonQI). At present, when the MonQI toolbox is applied, generally nine consecutive steps are taken, as shown in Figure 11.10. To facilitate the choice for a certain application, four different profiles were determined, which consist of different sets of software modules. 11.3.3.1 Joint Learning This profile is used in situations where the main objective is to relate social empower ment and action learning. The main feature is that it uses the reporting module, which allows farmers to compare their farm performance with the results of other members of a group and/or the average of a group. The tool produces reports per individual farm, as well as at the group or village level. Data are collected by farmers
521
20 Years after the Assessment of Soil Nutrient Balances 2. Training of enumerators by MonQl R&D team
3. Check and modify background data to local conditions
1. Adapt MonQl configuration to project objectives
4. Farm interviews using structured questionnaires
5. Data entry
9. Follow up activities
8. Data interpretation and reporting
6. Data debugging 7. Data processing
FIGURE 11.10 The nine steps of the MONQI methodology.
in their own language, using the units of measure they are accustomed to. This profile is used in the INMASP project (Table 11.6) and demonstrates the impact of providing quantitative data (in charts) to stimulate farmers to learn from each other [De Jager et al. 2008]. 11.3.3.2 Environment The environment profile is used when the objective is to understand environmental impacts of smallholder agriculture through the leaching of nutrients and pesticides. It uses the pesticide monitoring and evaluation module and hard to quantify estimation of nutrient losses. Currently, this profile is mostly used in China and Vietnam to identify agricultural practices with risks for environmental trade-off effects [Peeters et al. 2007]. 11.3.3.3 Livelihoods The livelihoods profile is mostly used to assess food security, household assets, and income. This profile was used after the tsunami of 2004 to monitor asset development and the impact of aid. Another example of this profile is the application of MonQI as an early warning indicator in Somaliland for food insecurity through monitoring asset developments of smallholders. 11.3.3.4 Agro-Food Chains and Certification The agro-food chain profile is used for tracking and tracing the use of inputs (most often pesticides) or monitoring on-farm production management in the agro-food chain with the objective to meet quality standards of the EU or for certification for niche markets. This profile is currently used for the certification of watermelon in
522
World Soil Resources and Food Security
Vietnam and to monitor adherence to a private-sector-set code of conduct in the floriculture sector in Ethiopia. In a study in Vietnam, for example, in order to get detailed and quantitative information on pesticide use on the watermelon variety Happy Sweet, farm management was monitored in detail using an additional module called MonQI-P on environmental risks associated with the use of pesticides [Peeters et al. 2007]. In MonQI-P, the tailor-made reporting tool of profile 1 is further extended with calculations of recorded amounts of pesticides into active ingredients. In this way, the pesticide protocol of the buying company can be compared with farmers’ field operations against the background of international health and sanitation standards and export quality restrictions. The MonQI profiles overlap and, through the flexible activation of modules, mixed profiles are also possible. A mixed profile is used for monitoring sustainability of commercial tea production in Kenya. In the coming years, further modifications of the software are foreseen such as the monitoring of C and water management and optimal resource distribution under limited availability (Figure 11.1).
11.4 S TUDIES ON SOIL NUTRIENT DYNAMICS BEYOND THE NUTMON FAMILY Soil fertility dynamics can be studied through expert judgment by farmers when no quantitative assessments are available using soil color, workability, previous harvests, etc. Vernacular names then often shed light on the perception and appreciation of soil quality differences [e.g., Stroosnijder and van Rheenen 2001]. Remarkably, on a world scale, the Global Assessment of Human-Induced Soil Degradation (GLASOD) is also largely based on expert judgment [Oldeman et al. 1991]. The NUTMON approach is semiquantitative, using an input–output model with different terms, providing a net balance for the system under study. A considerable number of nutrient balance research papers not linked to NUTMON have been published over the past 20 years, and are summarized in Section 11.l. Calculating and assessing changes in soil fertility do not necessarily require a nutrient balance approach. Changes can also be measured or assessed by sampling soils over certain time intervals (chronosequence), or by sampling soils at the same time but in different LUSs (biosequence). Results of these approaches are given in Section 11.2. A set of cases on the manipulation of nutrient stocks and balances (INM) is given in Section 11.3.
11.4.1 Nutrient Balances At the national level, the Lesschen et al. [2007] study covers Burkina Faso, with a nutrient balance of −20, −3.7, and −15 kg N, P, and K ha−1 yr−1, respectively. Folmer et al. [1998] estimated soil fertility loss in Mozambique by combining land units (based on soil fertility, precipitation, and erosion) and land-use types (characterized by crops, scale, and occupation percentage of the land) into LUSs. On average, nutrient balances for 1997 were estimated at −33, −6.4, and −25 kg N, P, and K ha−1 yr−1. Haileslassie et al. [2005] assessed soil nutrient depletion and its spatial variability for Ethiopia and its regional states. At the national level, full nutrient balance results indicated an annual
523
20 Years after the Assessment of Soil Nutrient Balances 200
N balance
100 0 –100
Scale n s P 132 11 F 133 10 VW 6 4 DR 86 4 N 251 6 C 13 3
–200
P balance
40 20 0 Scale n s P 106 10 F 133 9 VW 6 4 DR 81 3 N 251 6 C 13 3
–20 –40
200
K balance
100 0 –100
Scale n s P 46 5 F 127 9 VW 3 2 DR 86 4 N 251 6 C 13 3
–200 –300 –400
P
F
VW DR Main spatial scale
N
C
FIGURE 11.11 Nutrient balances at different spatial scales in Africa. P = plot; F = farm; VW = village and watershed; DR = district and region; N = nation; C = continent. Only data expressed as kg ha−1 year−1 and derived from full nutrient balances studies were plotted for the comparison. The number of observations (n) and studies (s) per category are shown in the rectangles. (Data from Cobo, J.G., et al., Agric. Ecos. Envir., 136, 1–15, 2010. With permission.)
524
World Soil Resources and Food Security
depletion rate of 122 kg N ha−1, 13 kg P ha−1, and 82 kg K ha−1. Soil nutrient stocks in all regional states were decreasing with the exception of areas under permanent and vegetable crops. Soil erosion was the major cause of nutrient depletion in Ethiopia, accounting for 70%, 80%, and 63% of the N, P, and K losses, respectively. For Asia, results are different. Surface N balances for China’s crop production systems were estimated using statistical data collected from 1980 to 2004 at the national and provincial scales and from 1994 to 1999 at the county level. The total cultivated land in China had a positive N balance of 143 kg ha−1 yr−1 in 2004, which is expected to increase to 169 kg ha−1 yr−1 in 2015. The N balance surplus in the more developed southeastern provinces was the largest [Sun et al. 2008]. Pathak et al. [2010] studied nutrient balances for the different states of India. Removals of N, P, and K by major agricultural crops in the country were 7.7, 1.3, and 7.5 Mt, respectively, but yet there were positive balances of N (1.4 million tons) and P (1.0 million tons), and a negative balance of K (3.3 million tons) for the 2000–2001 season. Cobo et al. [2010] recently published a comprehensive overview for 57 peer-reviewed studies in Africa. Most nutrient balances are calculated at the plot and farm scales and generated in East Africa. The analysis confirms the trend of nutrient mining in Africa for N and K, where more than 75% of selected studies have negative balances, whereas for P, 56% of the studies show negative mean balances (Figure 11.11). Other significant results derived from the analysis of the dataset are 1) LUSs of wealthier farmers mostly present higher N and P balances than systems of poorer farmers; 2) plots located close to homesteads also show higher balances than plots located relatively further away; and 3) partial nutrient balances are significantly higher than full balances calculated for the same systems. The data do not reveal a major trend in the magnitude of N, P, and K balances by increasing the spatial scale from the plot to the continental level. This is in apparent contradiction to Haileslassie et al. [2007], Schlecht and Hiernaux [2004], and Onduru and Du Preez [2007] who claim a trend of increasingly negative nutrient balances with increasing scales of observation. However, a limitation of the results in Figure 11.11 is the diversity of farming systems assessed and the inclusion of sublevels within main scales, which could increase the variability. The review of Cobo et al. [2010] illustrates the high diversity of methods used to calculate nutrient balances and highlighted the main pitfalls, especially in the case of upscaling nutrient flows and balances. According to the authors, the major generic problems are the arbitrary inclusion and exclusion of flows from the calculations, short evaluation periods, difficulties on the setting of spatialtemporal boundaries, inclusion of lateral flows, and the linking of the balances to soil nutrient stocks. Main challenges during scaling-up are related to the type of aggregation and internalization of nutrient flows, as well as issues of nonlinearity and spatial variability, resolution, and extent, which have not been properly addressed yet. Urban and periurban agriculture have nutrient balances that can differ largely from those in rural areas. Grote et al. [2005] point out that urban environments are sinks of nutrients through the large-scale import of food stuffs. The waste that is created with it can be usefully recycled back into the agricultural system, but can also pose overuse and health problems [Drechsel and Kunze 2001]. A recent study in Niamey (Niger) shows positive partial nutrient balances in ten vegetable gardens of 290–1133 kg N ha−1, 125–223 kg P ha−1, and 312–351 kg K ha−1 [Diogo et al. 2010].
20 Years after the Assessment of Soil Nutrient Balances
525
For Asia at the field level, Regmi et al. [2002] did a nutrient balance assessment, based on long-term rice-rice-wheat trials in Central Terai, Nepal, and found nutrient balances of 3, −4, −12 (control treatments), 29, 23, −62 (NPK), and 96, 33, −16 (manure) kg ha−1 yr−1 for N, P, and K, respectively. Even in an intensive high-input system, K withdrawal is apparently not matched by stocks and inputs, and is the major cause of the slight but consistent yield declines that were observed throughout the 20-year period. Dobermann et al. [1996a, 1996b] provide accounts of P and K balances in intensive wetland rice systems in Southeast Asia, and also find negative K balances in most systems. Finally, the use of nutrient balances is not restricted to agricultural systems. Various studies applied the concept to natural ecosystems [e.g., Stoorvogel et al. 1997; Klinge et al. 2004].
11.4.2 C and Nutrient Stocks Various studies of SOC stocks and changes exist for SSA at a national level, including Kenya [Batjes 2004], Senegal [Batjes 2001; Woomer et al. 2004], Congo Republic [Schwartz and Namri 2002], and Central Africa [Batjes 2008] using soil survey legacy data and conventional GIS-based mapping approaches. Other studies apply dynamic, process-based models to assess changes in SOC subsequent to defined changes in land use and management [Tschakert et al. 2004; Kamoni et al. 2007; Tan et al. 2009]. SOC stocks for Senegal, for example [Batjes 2001], range from 10 tons C ha−1 for an Arenosol under sparse vegetation to 72 tons for a Ferralsol under forest, whereas the highest value was 300 tons C ha−1 for a Fluvisol under rice and short grassland. This study accounts for differences in bulk density and gravel content, with soil type and land-use management. Alternatively, Windmeijer and Andriesse [1993] report average SOC content for the West African Equatorial Forest, Guinea savanna, and Sudan savanna to be 24.5, 11.7, and 3.3 g kg−1, respectively. Pieri [1989] mentions annual loss rates of SOC content in cultivated fields in Senegal (3.2%– 7.0%), Burkina Faso (1.5%–6.3%), and Chad (0.5%–2.8%). At the district and province levels, a survey by the Burkina Faso Bureau National des Sols (unpublished) for the southwestern Houet Province reports 11 to 30 tons C ha −1 , and 1000 to 2500 kg N ha−1. In the Sanmatenga Province, on the Mossi Plateau, these values are 17 to 25 ton C ha−1, and 1100 to 1700 kg N ha−1 [Stroosnijder and Van Rheenen 2001]. In the Embu District in Kenya, total N content ranges from approximately 7.5 g kg−1 on the tea areas on the higher slopes of Mount Kenya to 0.9 g kg−1 in the semiarid lowlands to the east [Smaling et al. 2002]. In Zimbabwe, SOC stocks under reference woodlands are largest (53.3 t ha−1) in a red clay soil, followed by a granitic sand (22.8 t ha−1), and a Kalahari sand (19.5 t ha−1) [Zingore et al. 2005]. When looking at chronosequences and biosequences in SSA, Hartemink [2003] mentions that SOC contents under Nigerian forest were 26–35 g kg−1, against 13–19 g kg−1 for cocoa fields of 10–55 years of age, and that SOC equilibrium under cocoa is below that observed for soils under natural forest. Data from Cameroon [Kanmegne 2004] and Madagascar [Brand and Pfund 1998] show soil N losses as high as 3000 and 5000 kg ha−1 when forest is converted to agricultural land. Prolonged fallow [Nye and Greenald 1960] can help restore soil C stocks, albeit seldom to the original level observed under natural vegetation. Samaké et al. [2005] mention SOC values
526
World Soil Resources and Food Security
in unfertilized bush fields in the Malian Dogon between 1.9 (1 year fallow) and 3.4 g kg−1 (7 years fallow), and total N of 0.17 and 0.25 g kg−1. For land under 3–5 years of millet cultivation, SOC values were 1.5 and 0.14 g kg−1. In terms of distance from the homestead, values ranged from 1.0 to 5.5 g kg−1 C and from 0.12 to 0.23 g kg−1 N, when 2000 m vs. 50 m away from the homestead. Bado et al. [1997] found that 10 years of continuous maize cultivation in West Burkina Faso led to SOC levels of 1.9 g kg−1 (no fertilizer) and 3.6 g kg−1 (N fertilizer + manure), but fell short of the initial fertility of 4.5 g kg−1. Corresponding values for N were 0.17, 0.27, and 0.29 g kg−1. Pieri [1989] found annual loss rates of SOC in cultivated fields in Senegal (3.2%–7.0%), Burkina Faso (1.5%–6.3%), and Chad (0.5%–2.8%). In Asia, Maskey et al. [2001] show that SOC in Nepal declines upon cultivation. In the upper 30 cm, uncultivated land in the Central Hills had 12.6 g kg−1, lands cultivated for <10 years 9.6 g kg−1, and those cultivated for >50 years 6.9 g kg−1. For the Terai lowlands, bordering India, these values were 7.9, 5.9, and 4.0 g kg−1. The picture in China is mixed. A countrywide survey of SOC dynamics in trial sites in China (>1000 observations), monitored between 1985 and 2006, reveals that for dryland crops, average SOC went up from 10.1 to 10.8 g kg−1, and in paddy rice from 15.7 to 17.4 g kg−1 [Pan et al. 2010]. A long-term experiment in Northeastern China [1990–2005] shows that N stocks (0–20 cm) under control plots went down from 2625 to 2068 kg ha−1, under N (150 kg ha−1) from 2396 to 2192 kg ha−1, under N + manure (205 kg ha−1) from 2429 to 2455 kg ha−1, and under NPK from 2577 to 2302 kg ha−1 [Qiang Ma et al. 2010]. In a long-term double cropping (maize, wheat) experiment on the red clays of Southern China, SOC contents under the control treatment were stable: 8.5 (1991–1998) to 8.7 g kg−1 (1999–2006), and increased under NPK (300-53-104) from 9.2 to 10.3, and under 60 tons ha−1 pig manure from 10.9 to 14.4 g kg−1. Soil N, however, declined over these periods, apart from the treatments that included manure [Zhang et al. 2009]. For Brazil, Bustamante et al. [2006] estimate that soils of the Cerrado region have an average stock of 117 ton C ha−1. Comparing the dynamics of SOC in a long-term experiment in the South of Brazil, Mielniczuk et al. [2003] estimate that no-tillage reduced the decomposition rate from 3.2% to 1.7% yr−1. In the Cerrado region, Silva et al. [1994] found SOC to drop by 41% on clayey and by 80% on sandy soils after 5 years of cultivation. However, Freitas et al. [2000] and Roscoe and Buurman [2003] did not observe changes in the SOC stocks of a clay soil after 25 years of maize–bean successions with conventional tillage. Lilienfein and Wielcke [2003] report no significant changes in C content of a Ferralsol Oxisol after 12 years of maize–soybean rotation under conventional tillage. A recent, extensive review of SOC changes under different land management systems in the Brazilian savanna is given by Batlle-Bayer et al. [2010].
11.4.3 Integrated Nutrient Management Implementation of so-called best practices can help sustain and improve SOC levels. Such practices should be directed at increasing inputs of organic matter into the soil and on decreasing organic matter decomposition. They typically include a judicious combination of various practices such as organic residue and fallow management, water conservation and management, soil fertility management including
20 Years after the Assessment of Soil Nutrient Balances
527
use of chemical fertilizers, organic manures, and liming, and introduction of agroecologically adapted crop and plant species, including agroforestry [Batjes and Sombroek 1997; Paustian et al. 1998; Bruce et al. 1999b; Lal 2008]. Nonetheless in SSA, examples of successful INM in a research environment abound, but the exact spread of the successes is limited [Gabre-Madhin and Haggblade 2004]. Zougmoré et al. [2002] show in erosion-prone central Burkina Faso that stone rows alone failed to completely stop soil fertility decline, but in a later study, Zougmoré et al. [2003] find that stone rows and Andropogon grass strips together reduced runoff by 59%. Adding fertilizer N or organic N caused an even stronger reduction of runoff by 67% and 84%, respectively, next to a strongly positive effect on crop yields by fertilizer (65%) or compost and manure (142%). Stone rows alone or grass strips alone improved water storage but not yields, but the combination of stone rows + compost was the best treatment giving 2300–2800 kg ha−1. Planting pits (zaï) can turn hardened ironstone plateaus into relatively productive land again while reducing runoff. Experimentation in Yatenga (central Burkina Faso) gave 200 kg ha−1 sorghum with pits alone, 700 kg ha−1 with dry dung, 1400 kg ha−1 with mineral fertilizers, and 1700 kg ha−1 with both dung and fertilizers [Reij and Thiombiano 2003]. Microdosing of fertilizer also holds promise. Twomlow et al. [2010] present a review of 1200 pairedplot trials in Zimbabwe consistently showing that microdosing of 17 kg N ha−1 can increase grain yields by 30%–50% across a broad spectrum of soil, farmer management, and seasonal climate conditions. In order for a household to make a profit, farmers need to obtain 4–7 kg of grain for every kg of N applied, but they commonly obtain 15–45 kg. The benefits of leguminous species in farming systems are widely known [e.g., Giller 2001; Sanginga and Woomer 2009]. An African reality is that smallholder farmers tend to spread risks in land management. This can manifest itself in multiple cropping systems, multistory cropping, or integrated crop–livestock systems. As to soils, farmers tend to cherish some parts of their fields at the expense of others, which is also a way to spread drought risks. Samaké et al. [2005] found that N stocks (0–15 cm) in concentric rings, at 10–200 m from homesteads in the Malian Dogon were 600 kg ha−1, whereas in the outer rings, 500–2000 m away from the homesteads, it was only 300 kg ha−1. Similar spatial soil fertility variation-driven farm nutrient management was found by Prudencio [1993] in Burkina Faso, Elias et al. [1998] in Ethiopia, Zingore et al. [2007] in Zimbabwe, Ebanyat et al. [2009] in Uganda, and Vanlauwe et al. [2006] and Tittonell et al. [2008] in West Kenya. Combination of spatial management with INM technologies lead Giller et al. [2006] to conclude that spatial patterns of resource use in SSA are consistent across different farming systems. Farmers preferentially allocate manure, mineral fertilizers, and labor to fields close to the homestead, resulting in strongly negative soil fertility gradients away from the homestead. Giller et al. [2006] also state that livestock in SSA farming systems are the central means of concentration of nutrients, which is confirmed by the NUTMON analysis in Chapter 3.
11.5 WHERE DO WE STAND, WHERE DO WE GO? The NUTBAL study [Stoorvogel et al. 1993] and its improved version [FAO 2004; Lesschen et al. 2007] reveal that nutrient balances in SSA have negative values for N,
528
World Soil Resources and Food Security
P, and K (Table 11.2). This means that, on average, soil fertility on the subcontinent declines. On top of that, C and nutrient stocks in the soils of SSA are only 60% of the estimated global average. Comparing the N depletion projection of NUTBAL in 2000 (26 kg ha−1 yr−1) with the Batjes [2001] estimate of African SOC stocks (60,000 kg ha−1) suggests, when taking an approximate C/N ratio of 12, an annual N loss of 0.5% of the total N stocks in soils. The NUTBAL studies raise awareness of soil fertility problems at a macrolevel (FAO, research community, donor community), and they show in which regions nutrient depletion is most severe. It is the mesolevel studies, however, that provide clearer options for national policies, public services, and private sector investment, i.e., through the development of vertical supply chains and establishment of agro-input dealer outposts. The district examples in Ghana, Kenya, and Mali, for example, show the positive impact of cash crops on the nutrient flows and balances (Figure 11.6; Tables 11.3 and 11.4). At the microlevel, NUTBAL was transformed into a farm-level monitoring tool: NUTMON. The tool has helped to not only better understand differences between African farming systems and their nutrient stocks and balances, but also increased insight into farmers’ soil fertility and farm management strategies. Many MSc and PhD students, inside or outside the African Network for Soil Biology and Fertility (http://webapp.ciat.cgiar.org/tsbf_institute/africa.htm) have used NUTMON over the past 20 years [e.g., Vanlauwe et al. 2002; Bationo 2004; Bationo et al. 2007; Sanginga and Woomer 2009]. The NUTMON studies showed a wide range of N, P, and K balances as well as related net farm income (Table 11.6). Out of 20 projects, 80% have a negative (full) balance for N, 40% for P, and 70% for K. The visualization of internal flows shows that, on average, the traffic of nutrients between crops and livestock is strongly in favor of the latter (Figure 11.9; Table 11.7). Analyzing 57 nutrient balance studies performed at the microlevel in SSA, Cobo et al. [2010] finds that 75% have a negative balance for N and K, and 56% for P (Figure 11.11). NUTMON allows manipulation of soil fertility by integrating INM technologies such as zaï, use of compost pits, biomass transfers, contour planting, and improving the efficiency within crop-livestock systems. Visualizing internal flows within the farm systems has been key at the microlevel, whereas this is all inside the black box at the macrolevel and mesolevel. Many INM technologies are potentially successful, and are known to have been adopted. The further rapid spreading of such successes should be a key priority. Still, going by analyses of success stories in African agriculture and natural resource management, more has been achieved on improved varieties (maize, rice), eradication of diseases (cassava viruses, rinderpest), and improved market opportunities (horticulture, floriculture) than through INM technologies [Gabre-Madhin and Haggblade 2004]. Successes in the microdosing of fertilizer and improved land management are notable [Reij and Smaling 2008], but there is no major leap ahead yet in fertilizer and manure use, and INM in general. As periurban systems show positive nutrient balances, even in SSA, it may be worthwhile to improve agricultural areas that are not too distant from major cities in such a way that they become breadbaskets, benefiting from fertilizers (IN1), city waste (IN2), crop rotations that include leguminous species (IN4), conservation tillage (less OUT3 and OUT4 and maintenance of SOC), erosion control (less OUT5), in relatively large-scale management units.
20 Years after the Assessment of Soil Nutrient Balances
529
NUTMON has meanwhile grown into MONQI, which allows for the inclusion of producer strategies within a real-world context of priority setting at the farm level, risk management, reliance on nonfarm income including remittances, realities of changing market prices, and other policy considerations. Smallholder farmers play a key role in rural development and various public and private interventions are in progress to assist smallholders in optimizing farm management (increase productivity and income), in the sustainable use of natural resources (water, pest, and soil fertility management), in improving market linkages, in diversifying sources of income, and in reducing risks. The need for actual farm management information and monitoring change and impacts as well as for generating information to assist learning and innovation processes in smallholder enterprises will increase and the relevance for further development of MONQI remains high. Uncertainties play a large role when estimating balances on the basis of a set of inputs and outputs. Lesschen et al. [2007] determined the uncertainties in calculating the nutrient balance for Burkina Faso, based on the improved version of NUTBAL [FAO 2004]. Uncertainty is, of course, not restricted to NUTBAL and SSA. A special issue of the European Journal of Agronomy [2003, 20] describes how nutrient balances in different EU countries are used to gain understanding and input for agricultural and environmental policies and measures. At this advanced level, a range of shortcomings and uncertainties exist [Öborn et al. 2003]. The development of a mineral bookkeeping system (MINAS) at farm level in the Netherlands, for example [Hanegraaf and Den Boer 2003], was not deemed solid enough by the European Commission and had to be replaced by a system of maximum fertilizer and manure application per unit of agricultural land [Schröder and Neeteson 2008]. Also, the comparison of soil C and nutrient stocks is hampered by uncertainties and by a lack of information on sampled soil depth and bulk density, which would allow transformation of data provided on a per soil mass basis to a (preferred) per hectare/volume basis. Batjes [1996], for example, estimated that world SOC stocks are approximately 700 × 109 ton for the upper 30 cm, but for the upper 1 m, values were about twice this amount and, for the upper 2 m, even 2400 × 109 ton C was reported. Overall, opportunities exist for increasing SOC in the soil, thereby also contributing to atmospheric C mitigation, but these are constrained by the available knowledge and access to resources; hence, the need for mechanisms that pay for environmental services [Ringius 2002] such as improved soil C and soil water management. In Asia, nutrient balances at the national level are estimated to be positive for N, but negative for K. Field studies confirm these findings. Reasons include the much higher and often subsidized use of (N and P) fertilizers in Asia and the relatively large percentage of nutrient-balanced irrigated systems. This review also shows that in China and India, average nutrient balances are positive for N and P [Sun et al. 2008; Pathak et al. 2010]. Under conservation agriculture in Latin America, nutrient balances can also remain relatively neutral. In SSA, the examples given in Chapter 4 show that conversion of forest leads to massive nutrient losses, whereas continuous cultivation only keeps SOC, at best, at 75% of its original value under high applications of fertilizers and manure. Asia and Latin America have two more advantages over SSA. In Asia, production per unit land has increased largely as a result of improved rice production methods (breeding, mechanization, pest control,
530
World Soil Resources and Food Security
and irrigation). In parts of Latin America, farms are often so large that production per unit labor can be maximized through large-scale mechanization and using other economies of scale. However, it is feared that the smallholders’ systems in mountain areas of the Andes and Central America experience strongly negative balances similar to Africa, although inherent soil fertility is higher on average. Trials in rice-based systems in Asia also show that (too) large applications of N have two negative externalities, i.e., a small percentage of the N in urea is actually taken up by the crop, whereas the remainder stays in the soil or is lost as OUT3 or OUT4, and the withdrawal of soil K and other nutrients will be much larger than in a situation where little urea is applied. Also, soils with high N but low P and K stocks may have quite negative N balances. Application of N fertilizers will then not be a sensible option until the soil N, P, and K ratio is more in balance with cropuptake ratios. Hence, a nutrient balance close to zero is preferred, but the road toward achieving that can be different and relates to a target or ideal soil fertility [Janssen and De Willigen 2006]. Finally, although awareness of nutrient balances has increased among agricultural scientists, the subject still features marginally in most debates about sustainable agricultural development. As a result, the great missing link between scientific nutrient balance assessments and agricultural policy remains unaddressed. Few countries, particularly in SSA, have developed comprehensive policies, including subsidies, credit, and marketing, to promote increased fertilizer use and INM. This remains of serious concern in view of even the most optimistic projections of nutrient needs to feed the world population in 2050. Initiatives such as the release of, and active participation of, key partners in the Web site and discussion platform on www.africafertilizer.org are highly necessary. Also, a next generation of fertilizers is in the making through the IFDC-triggered Virtual Fertilizer Research Center, aimed at tailoring type and amount of fertilizer to soil characteristics and crop needs so as to maximize fertilizer use efficiency and value–cost ratios. Welltrained agro-input dealers can play a key role in acceptance of fertilizers, also by women and first-time users. This is testified to by much of the work IFDC and TSBF-CIAT have done in SSA. Banking on fertilizers alone is, however, not the way ahead for resource-poor farmers. Not only can the price be prohibitive, the way it is offered on the market (in 50-kg bags), substandard product quality, and the risk of late availability during the growing season make resource-poor farmers look for broader INM options. The entire manipulation, including the components offered in the NUTMON toolbox, is geared at realizing high OUT1 at higher and more efficient use of inputs and the clever management of internal flows. This approach addresses the increased productivity, efficiency, and recycling spelled out in Chapter 1, and includes the add, save, and recycle components of INM introduced in Chapter 3. Nonetheless, a concerted effort is required to raise the visibility of soil nutrients as an essential ingredient in sustainable agriculture. Little political action has been taken so far, apart from pledging support during the 2006 Africa Fertilizer Summit, where it was agreed to raise fertilizer use in SSA to 50 kg ha−1 (from the current <10 kg ha−1). Under NEPAD, African governments also promised to spend 10% of GDP on agriculture. These promises still largely have yet to be implemented.
20 Years after the Assessment of Soil Nutrient Balances
531
More than 20 years of work on nutrient balances in Africa demonstrate clearly the importance of managing the entire nutrient balance, i.e., all the inputs and outputs and not just the partial management of single nutrients or partial nutrient losses such as erosion, or additions such as fertilizer. This is not rocket science and could benefit the African people in an immediate way.
ACKNOWLEDGMENT Martine Hoogsteen (Wageningen UR) is greatly thanked for her inspired support during parts of the writing process.
REFERENCES Bado, B.V., M.P. Sedogo, M.P. Cescas, F. Lompo, and A. Bationo. 1997. Effet à long terme des fumures sur le sol et les rendements du maïs au Burkina Faso. Cah. d’Etudes Rech. Fran. 6:571–575. Bationo, A. 2004. Managing nutrient cycles to sustain soil fertility in sub-Saharan Africa. Nairobi, Kenya: Academy Science Publishers and Cali, Colombia: TSBF-CIAT. Bationo, A., B. Waswa, J. Kihara, and J. Kimetu. 2007. Advances in integrated soil fertility management in sub-Saharan Africa: Challenges and opportunities. Dordrecht, the Netherlands: Springer. Batjes, N.H. 1996. Total carbon and nitrogen in the soils of the world. Eur. J. Soil Sci. 47:151–163. Batjes, N.H. 2001. Options for increasing carbon sequestration in West African soils: An exploratory study with special focus on Senegal. Land Degrad. Dev. 12:131–142. Batjes, N.H. 2002. Soil parameter estimates for the soil types of the world for use in global and regional modelling (version 2.0). ISRIC Report 2002/02. Wageningen, the Netherlands: ISRIC. Batjes, N.H. 2004. Soil carbon stocks and projected changes according to land use and management: A case study for Kenya. Soil Use Manage. 20:350–356. Batjes, N.H. 2008. Mapping soil carbon stocks of Central Africa using SOTER. Geoderma 146:58–65. Batjes N.H., and W.G. Sombroek. 1997. Possibilities for carbon sequestration in tropical and subtropical soils. Global Change Biology 3:161–173. Batlle-Bayer, L., N.H. Batjes, and P.S. Bindraban. 2010. Changes in organic carbon stocks upon land use conversion in the Brazilian Cerrado: A review. Agric. Ecos. Envir. 137:47–58. Brand, J., and J.L. Pfund. 1998. Site and watershed-level assessment of nutrient dynamics under shifting cultivation in eastern Madagascar. Agric. Ecos. Envir. 71:169–183. Bruce, J.P., M. Frome, E. Haites, H. Janzen, R. Lal, and K. Paustian. 1999. Carbon sequestration in soils. J. Soil Water Conserv. 54:382–389. Bustamante, M.M.C., M. Corbeels, E. Scopel, and R. Roscoe. 2006. Soil carbon storage and sequestration potential in the Cerrado Region of Brazil. In Carbon sequestration in soils of Latin America, eds. R. Lal, C.C. Cerri, M. Bernoux, J. Etchevers, E. Cerri, E., 285– 299, Binghamton, NY: The Haworth Press. Cobo, J.G., G. Dercon, and G. Cadisch. 2010. Nutrient balances in African land use systems across different spatial scales: A review of approaches, challenges and progress. Agric. Ecos. Envir. 136:1–15. Conway, G. 1997. The doubly green revolution: Food for all in the twenty-first century. London: Penguin Books.
532
World Soil Resources and Food Security
Defoer, T. 2002. Learning about methodology development for integrated soil fertility management. Agric. Syst. 73:57–81. De Jager A. 2005. Participatory technology, policy and institutional development to address soil fertility degradation in Africa. Land Use Policy 22:57–66. De Jager, A. 2007. Practice makes perfect—Participatory innovation in soil fertility management to improve rural livelihoods in East Africa. PhD. thesis, Wageningen University, the Netherlands. De Jager, A., S.M. Nandwa, and P.F. Okoth. 1998a. Monitoring nutrient flows and economic performance in African farming systems (NUTMON). I. Concepts and methods. In Nutrient flows and balances as indicators of productivity and sustainability in sub-Saharan African agro-ecosystems, ed. E.M.A. Smaling. Agric. Ecos. Envir. 71:37–48. De Jager, A., F.M. Kariuki, M. Matiri, M. Odendo, and J.M. Wanyama. 1998b. Monitoring nutrient flows and economic performance in African farming systems (NUTMON). IV. Monitoring of farm economic performance in three districts in Kenya. In Nutrient flows and balances as indicators of productivity and sustainability in sub-Saharan African agro-ecosystems, ed. E.M.A. Smaling. Agric. Ecos. Envir. 71:81–92. De Jager, A., H. van Keulen, F. Mainah, L.N. Gachimbi, J.K. Itabari, E.G. Thuranira, and A.M. Karuku. 2006. Attaining sustainable farm management systems in semi-arid areas in Kenya: Few technical options, many policy challenges. Internat. J. Sustain. Agric. 3:189–205. De Jager, A., D.D. Onduru, L.N. Gachimbi, F. N. Muchena, G. Gachini, and C.L. Van Beek. 2008. Farmer field schools for rural empowerment and life-long learning in integrated nutrient management: Experiences in Eastern and Central Kenya. In Innovation in Africa: Enriching farmers’ livelihoods, eds. P. Sanginga et al. London: Earthscan, UK. De Willigen, P. 2000. An analysis of the calculation of leaching and denitrification losses as practised in the NUTMON approach. Report 18. Wageningen, the Netherlands: Plant Research International. Diogo, R.V.C., A. Buerkert, and E. Schlecht. 2010. Horizontal nutrient fluxes and food safety in urban and peri-urban vegetable and millet cultivation of Niamey, Niger. Nutrient Cycl. Agroecos. 87(1):81–102. Dobermann, A., P.C.S. Cruz, and K.G. Cassman. 1996. Fertilizer inputs, nutrient balance, and soil nutrient-supplying power in intensive, irrigated rice systems. I. Potassium uptake and K balance. Nutr. Cycl. Agroecos. 46:1–10. Dobermann, A., K.G. Cassman, C.P. Mamaril, and J.E. Sheehy. 1996. Fertilizer inputs, nutrient balance and soil nutrient supplying power in intensive, irrigated rice system. III. Phosphorus. Nutr. Cycl. Agroecos. 46:111–125. Döll, P., and S. Siebert. 2000. A digital global map of irrigated areas. ICID J. 49:55–66. Drechsel, P., and D. Kunze. 2001. Waste composting for urban and peri-urban agriculture: Closing the rural-urban nutrient cycle in sub-Saharan Africa. Wallingford, UK: CAB International. Ebanyat, P., N. de Ridder, A. de Jager, R.J. Delve, M.A. Bekunda, and K.E. Giller. 2009. Impacts of heterogeneity in soil fertility on legume-finger millet productivity, farmers’ targeting and economic benefits. Nutr. Cycl. Agroecos. 87(2):209–231. Elias, E., S. Morse, and D.G.R. Belshaw. 1998. Nitrogen and phosphorus balances of Kindo Koisha farms in southern Ethiopia. Agric. Ecos. Envir. 71:93–113. Faerge, J., and J. Magid. 2004. Evaluating NUTMON nutrient balancing in sub-Saharan Africa. Nutr. Cycl. Agroecos. 69:101–110. FAO. 1976. A framework for land evaluation. FAO Soils Bulletin 32. Rome: FAO. FAO. 2000. Livestock distribution, production and diseases: Towards a global livestock atlas. Consultants’ report. Rome: FAO. FAO. 2003. Assessment of soil nutrient balance. Approaches and methodologies. FAO Fertilizer and Plant Nutrition Bulletin 14. Rome: FAO.
20 Years after the Assessment of Soil Nutrient Balances
533
FAO. 2004. Scaling soil nutrient balances—Enabling mesoscale approaches for African realities. FAO Fertilizer and Plant Nutrition Bulletin 15. Rome: FAO. FAO. 2009. Harvesting agriculture’s multiple benefits: Mitigation, adaptation, development and food security. FAO Policy brief. ftp://ftp.fao.org/docrep/fao/012/ak914e/ak914e00 .pdf. Rome: FAO. FAO/UNESCO. 1997. Digital soil map of the world and derived soil properties. Land and Water Digital Media Series 1. Rome: FAO. Folmer, E.C.R., P.M.H. Geurts, and J.R. Francisco. 1998. Assessment of soil fertility depletion in Mozambique. Agric. Ecos. Envir. 71:159–167. Freitas, P.L., P. Blancaneaux, E. Gavinelli, M.C. Larre-Larrouy, and C. Feller. 2000. Nivel e natureza do estoque organico de latossolos sob diferentes sistemas de uso e manejo. Pesqui. Agropecu. Bras. 35:157–170. Frissel, M.J. 1978. Cycling of mineral nutrients in agricultural ecosystems. London: Elsevier. Gabre-Madhin, E.Z., and S. Haggblade. 2004. Successes in African agriculture: Results of an expert survey. World Development 32:745–766. Gachimbi, L.N., H. van Keulen, E.G. Thuranira, A.M. Karuku, A. de Jager, S. Nguluu, B.M. Ikombo, J.M. Kinama, J.K. Itabari, and S.M. Nandwa. 2005. Nutrient balances at farm level in Machakos (Kenya), using a participatory nutrient monitoring (NUTMON) approach. Land Use Policy 22:13–22. Giller, K.E. 2001. Nitrogen fixation in tropical cropping systems, 2nd ed. Wallingford, UK: CABI Publishing. Giller, K.E., E.C. Rowe, N. de Ridder, and H. van Keulen. 2006. Resource use dynamics and interactions in the tropics: Scaling up in space and time. Agric. Syst. 88:8–27. Grote, U., E. Craswell, and P. Vlek. 2005. Nutrient flows in international trade: Ecology and policy issues. Envir. Sci. Policy 8:439–451. Haileslassie, A., J. Priess, E. Veldkamp, D. Teketay, and J.P. Lesschen. 2005. Assessment of soil nutrient depletion and its spatial variability on smallholders’ mixed farming systems in Ethiopia using partial versus full nutrient balances. Agric. Ecos. Envir. 108:1–16. Haileslassie, A., J. Priess, E. Veldkamp, and J.P. Lesschen. 2007. Nutrient flows and balances at the field and farm scale: Exploring effects of land-use strategies and access to resources. Agric. Syst. 94:459–470. Hanegraaf, M.C., and D.J. den Boer. 2003. Perspectives and limitations of the Dutch minerals accounting system (MINAS). Eur. J. Agron. 20:25–31. Hartemink, A.E. 2003. Soil fertility decline in the tropics (with case studies on plantations). Wageningen, the Netherlands: ISRIC and Wallingford, UK: CABI. IFA/FAO. 2001. Global estimates of gaseous emissions of NH3, NO, and N2O from agricultural land. Rome: FAO. IFA/IFDC/FAO. 2000. Fertilizer use by crop, 4th ed. Rome: FAO. Janssen, B.H., F.C.T. Guiking, D. van der Eyk, E.M.A. Smaling, J. Wolf, and H. van Reuler. 1990. Quantitative evaluation of the fertility of tropical soils (QUEFTS). Geoderma 46:299–318. Janssen, B.H., and P. de Willigen. 2006. Ideal and saturated soil fertility as benchmarks in nutrient management. II. Interpretation of chemical soil tests in relation to ideal and saturated soil fertility. Agric. Envir. Ecos. 116:147–155. Kamoni, P.T., P.T. Gicheru, S.M. Wokabi, M. Easter, E. Milne, K. Coleman, P. Falloon, and K. Paustian. 2007. Predicted soil organic carbon stocks and changes in Kenya between 1990 and 2030. Agric. Ecos. Envir. 122:105–113. Kanmegne, J. 2004. Slash and burn agriculture in the humid forest zone of southern Cameroon. Soil quality dynamics, improved fallow management and farmers’ perception. PhD thesis, Wageningen University, the Netherlands. Kanté, S. 2001. Gestion de la fertilité des sols par classe d’exploitation au Mali-Sud. PhD thesis, Wageningen University, the Netherlands.
534
World Soil Resources and Food Security
Klinge, R., A. R. Araujo Martins, J. Mackensen, and H. Fölster. 2004. Element loss on rain forest conversion in East Amazonia: Comparison of balances of stores and fluxes. Biogeochemistry 69:63–82. Lal, R. 2008. Soils and sustainable agriculture. A review. Agron. Sustain. Dev. 28:57–64. Leemans, R., and W. Cramer. 1991. The IIASA database for mean monthly values of temperature, precipitation and cloudiness on a global terrestrial grid. Research report RR-91–18. Laxenburg, Austria: International Institute of Applied Systems Analysis. Lesschen, J.P., J.J. Stoorvogel, E.M.A. Smaling, G.B.M. Heuvelink, and A. Veldkamp. 2007. A spatially explicit methodology to quantify soil nutrient balances and their uncertainties at the national level. Nutr. Cycl. Agroecos. 78:111–131. Lilienfein, J., and W. Wielcke. 2003. Element storage in native, agri-, and silvicultural ecosystems of the Brazilian savanna. I. Biomass, carbon, nitrogen, phosphorus and sulfur. Plant Soil 254:425–442. Maskey, S.L., S. Bhattarai, M.B. Peoples, and D.F. Herridge. 2001. On farm measurements of nitrogen fixation by winter and summer legumes in the Hill and Terai regions of Nepal. Field Crops Res. 70:209–221. Mielniczuk, J., C. Bayer, F.M. Vezzani, T. Lovato, F.F. Fernandes, and L. Debarba. 2003. Manejo de solo e culturas e sua relacao com os estoques de carbono e nitrogenio do solo. To picos em Ciencia do Solo 3:209–248. Nye, P.H., and D.J. Greenland. 1960. The soil under shifting cultivation. Technical Communication 51. Farnham Royal, Bucks (UK): Commonwealth Bureau of Soils. Öborn, I., A.C. Edwards, E. Witter, O. Oenema, K. Ivarsson, P.J.A. Withers, S.I. Nilsson, and A. Richert Stinzing. 2003. Element balances as a tool for sustainable nutrient management: A critical appraisal of their merits and limitations within an agronomic and environmental context. Eur. J. Agron. 20:211–225. Oenema, O., H. Kros, and W. de Vries. 2003. Approaches and uncertainties in nutrient budgets: Implications for nutrient management and environmental policies. Eur. J. Agron. 20:3–16. Oldeman L.R., R.T.A. Hakkeling, and W.G. Sombroek. 1991. World map of the status of human-induced soil degradation. Wageningen, the Netherlands: UNEP and ISRIC. Onduru, D.D., G. Gachini, G., A. de Jager, and J.M. Diop. 1999. Participatory research of compost and liquid manure in Kenya. Managing Africa’s Soils, No. 8. London: IIED. Onduru, D.D., A. de Jager, G. Gachini, and J.-M. Diop. 2001. Exploring new pathways for innovative soil fertility management in Kenya. Managing Africa’s Soils, No. 25. London: IIED. Onduru, D.D., and C. du Preez. 2007. Ecological and agro-economic study of small farms in sub-Saharan Africa. Agron. for Sustain. Dev. 27:197–208. Pan, G., X. Xu, P. Smith, W. Pan, and R. Lal. 2010. An increase in topsoil SOC stock of China’s croplands between 1985 and 2006 revealed by soil monitoring. Agric. Ecos. Envir. 136:133–138. Paustian, K., O. Andren, H.H. Janzen, R. Lal, P. Smith, H. Tiessen, M. van Noordwijk, and P.L. Woomer. 1998. Agricultural soils as a sink to mitigate CO2 emissions. Soil Use Manage. 13:230–244. Pathak, H., S. Mohanty, N. Jain, and A. Bhatia. 2010. Nitrogen, phosphorus, and potassium budgets in Indian agriculture. Nutr. Cycl. Agroecos. 86(3):287–299. Peeters, F., P. van den Brink, J. Vlaming, Phuong Mai Huynh, and R. van den Bosch. 2007. SUPHort Project: Environmental and human risk assessment of pesticide use in ‘Happy Sweet’ watermelon production in Vietnam. Content Report 5. Wageningen, the Netherlands: Alterra. Pieri, C. 1985. Bilans mineraux des systemes de cultures pluviales en zones arides et semiarides. Agron. Tropicale 40:1–20.
20 Years after the Assessment of Soil Nutrient Balances
535
Pieri, C. 1989. Fertilité des terres de savanes. bilan de trente ans de recherche et de développement agricoles au sud du Sahara. Paris: Ministère de la Coopération et CIRAD-IRAT. Prudencio, C.Y. 1993. Ring management of soils and crops in the West-african semiarid tropics—The case of the Mossi farming system in Burkina Faso. Agric. Ecos. Envir. 47:237–264. Qiang Ma, Wan-Tai Yu, Shan-Min Shen, Hua Zhou, Zi-Shao Jiang, and Yong-Gang Xu. 2010. Effects of fertilization on nutrient budget and nitrogen use efficiency of farmland soil under different precipitation years in Northeastern China. Nutr. Cycl. Agroecos. 88(3):315–327. Regmi, A.P., J.K. Ladha, H. Pathak, E. Pasuquin, C. Bueno, W. Dawe, P.R. Hobbs, D. Joshy, S.L. Maskey, and S.P. Pandey. 2002. Yield and soil fertility trends in a 20-year rice-ricewheat experiment in Nepal. Soil Sci. Soc. Am. J. 66:857–867. Reij, C.P., and T. Thiombiano. 2003. Development rural et environnement au Burkina Faso: la réhabilitation de la capacité productive des terroirs sur la partie nord du Plateau Central entre 1980 et 2001. Rapport de synthèse. Project funded by DGIS/Netherlands Embassy Ouagadougou. Reij, C.P., and E.M.A. Smaling. 2008. Analyzing successes in agriculture and land management in Sub-Saharan Africa: Is macro-level gloom obscuring positive micro-level change? Land use policy 25:410–420. Ringius, L. 2002. Soil carbon sequestration and the CDM: opportunities and challenges for Africa. Climatic Change 54:471–495. Roscoe, R., and P. Buurman. 2003. Tillage effects on soil organic matter in density fractions of a Cerrado Oxisol. Soil Tillage. Res. 70:107–119. Samaké, O. 2003. Integrated crop management strategies in Sahelian land use systems to improve agricultural productivity and sustainability: A case study in Mali. PhD thesis, Wageningen University, the Netherlands. Samaké, O., E.M.A. Smaling, M.J. Kropff, T.J. Stomph, and A. Kodio. 2005. Effects of cultivation practices on spatial variation of soil fertility and millet yields in the Sahel of Mali. Agric. Ecos. Envir. 109:335–345. Sanchez, P.A. 2002. Soil fertility and hunger in Africa. Science 295:2019–2020. Sanginga, N., and P.L. Woomer. 2009. Integrated soil fertility management in Africa: Principles, practices and developmental process. Nairobi: TSBF-CIAT. Schlecht, E., and P. Hiernaux. P. 2004. Beyond adding up input and outputs: Process assessment and upscaling in modelling nutrient flows. Nutr. Cycl. Agroecos. 70:303–319. Schoorl, J.M., A. Veldkamp, and J. Bouma. 2002. Modelling water and soil redistribution in a dynamic landscape context. Soil Sci. Soc. Am. J. 66:1610–1619. Schröder, J.J., and J.J. Neeteson. 2008. Nutrient management regulations in the Netherlands. Geoderma 144: 418–425. Scoones, I., and C. Toulmin. 1998. Soil nutrient balances: What use for policy? Agric. Ecos. Envir. 71:255–267. Schwartz, D., and M. Namri. 2002. Mapping the total organic carbon in the soils of the Congo. Global and Planetary Change 33:77–93. Silva, J.E., J. Lemainski, and D.V.S. Resck. 1994. Perdas de materia organica e suas relacoes com a capacidade de troca cationica em solos da regiao de cerrados do oeste baiano. Revista Brasileira de Ciencia do Solo 18:541–547. Smaling, E.M.A. 1998. Nutrient balances as indicators of productivity and sustainability in Sub-Saharan African agriculture. Agric. Ecos. Envir. 71 (special issue). Smaling, E.M.A., and B.H. Janssen. 1993. Calibration of QUEFTS, a model predicting nutrient uptake and yields from chemical soil fertility indices. Geoderma 59:21–44. Smaling, E.M.A., and L.O. Fresco. 1993. A decision-support model for monitoring nutrient balances under agricultural land use (NUTMON). Geoderma 60:235–256.
536
World Soil Resources and Food Security
Smaling, E.M.A., and O. Oenema. 1997. Estimating nutrient balances in agro-ecosystems at different spatial scales. In Methods for assessment of soil degradation, eds. R. Lal, W.H. Blum, C. Valentin, and B.A. Stewart, 229–252. Boca Raton, FL: CRC Press. Smaling, E.M.A., O. Oenema, and L.O. Fresco. 1999. Nutrient disequilibria in agroecosystems—Concepts and case studies. Wallingford, UK: CABI Publishing. Smaling, E.M.A., J.J. Stoorvogel, and A. De Jager. 2002. Decision making on integrated nutrient management through the eyes of the scientist, the land-user and the policy maker. In Integrated plant nutrient management in sub-Saharan Africa, eds. E. van Lauwe, J. Diels, N. Sanginga, and R. Merckx, 265–283. Wallingford, UK: CABI. Smaling, E.M.A., and J. Dixon. 2006. Adding a soil fertility dimension to global farming system approach, with cases from Africa. Agric. Ecos. Envir. 116:15–26. Smaling, E.M.A., R. Roscoe, J.P. Lesschen, A.F. Bouwman, and E. Comunello. 2008. From forest to waste: Assessment of the Brazilian soybean chain, using nitrogen as a marker. Agric. Ecos. Envir. 128:185–197. Stoorvogel, J.J., and E.M.A. Smaling. 1990. Assessment of soil nutrient depletion in subSaharan Africa: 1983–2000. Report 28. Wageningen, the Netherlands: Winand Staring Centre. Stoorvogel, J.J., E.M.A. Smaling, and B.H. Janssen. 1993. Calculating soil nutrient balances in Africa at different scales. 1. Supra-national scale. Fert. Res. 35:227–235. Stoorvogel, J.J., B.H. Janssen, and N. van Breemen. 1997. The nutrient budgets of a watershed and its forest ecosystem in the Taï National Park in Côte d’Ivoire. Biogeochemistry 37:159–172. Stroosnijder, L., and T. Van Rheenen. 2001. Agro-silvo-pastoral land use in Sahelian villages. Advances in Geoecology, Vol. 33. Wageningen, the Netherlands: Wageningen UR. Sun, B., R.-P. Shen, and A.F. Bouwman. 2008. Surface N balances in agricultural crop production systems in China for the period 1980–2015. Pedosphere 18:304–315. Tan, Z., S. Liu, L.L. Tieszen, and E. Tachie-Obeng. 2009. Simulated dynamics of carbon stocks driven by changes in land use, management and climate in a tropical moist ecosystem of Ghana. Agric. Ecos. Envir. 130:171–176. Tschakert, P., M. Khouma, and M. Sene. 2004. Biophysical potential for soil carbon sequestration in agricultural systems of the Old Peanut Basin of Senegal. J. Arid Envir. 59:511–533. Tittonell, P., B. Vanlauwe, M. Corbeels, and K.E. Giller. 2008. Yield gaps, nutrient use efficiencies and response to fertilisers by maize across heterogeneous smallholder farms of western Kenya. Plant Soil 313:19–37. Twomlow, S., D. Rohrbach, J. Dimes, J. Rusike, W. Mupangwa, B. Ncube, L. Hove, M. Moyo, N. Mashingaidze, and P. Mahposa. 2010. Micro-dosing as a pathway to Africa’s Green Revolution: Evidence from broad-scale on-farm trials. Nutr. Cycl. Agroecos. 88(1):3–15. US Geological Survey (USGS). 1998. Hydro 1K Africa. Elevation data. Sioux Falls, SD: USGS. US Geological Survey (USGS). 2000. University of Nebraska-Lincoln and European Commission’s Joint Research Centre. Africa land cover characteristics data base. Version 2.0. Seasonal land cover region legend. Sioux Falls, SD: USGS. Van Beek, C.L., A. de Jager, D.D. Onduru, and L.N. Gachimbi. 2005. Agricultural, economic and environmental performance of four farmer field schools in Kenya. INMASP project reports no. 18. Wageningen, the Netherlands: Alterra. Van Beek, C.L., D.D. Onduru, L.N. Gachimbi, and A. de Jager. 2009. Farm nutrient flows of four farmer field schools in Kenya. Nutr. Cycl. Agroecos. 83:63–72. Van den Bosch, H., J.N. Gitari, V.N. Ogaro, S. Maobe, and J. Vlaming. 1998. Monitoring nutrient flows and economic performance in African farming systems (NUTMON). III. Monitoring nutrient flows and balances in three districts in Kenya. Agric. Ecos. Envir. 71:62–80.
20 Years after the Assessment of Soil Nutrient Balances
537
Van der Pol, F. 1992. Soil mining: An unseen contributor to farm income in Southern Mali. Bulletin 35. Amsterdam, the Netherlands: The Royal Tropical Institute. Vanlauwe, B., J. Diels, N. Sanginga, and R. Merckx. 2002. Integrated plant nutrient management in sub-Saharan Africa. Wallingford, UK: CABI. Vanlauwe, B., P. Tittonell, and J. Mukalama. 2006. Within-farm soil fertility gradients affect response of maize to fertiliser application in western Kenya. Nutr. Cycl. Agroecos. 76:171–182. Vanlauwe, B., A. Bationo, J. Chianu, K.E. Giller, R. Merckx, U. Mokwunye, O. Ohiokpehai, P. Pypers, R. Tabo, K.D. Shepherd, E.M.A. Smaling, P.L. Woomer, and N. Sanginga. 2010. Integrated soil fertility management. Operation definition and consequences for implementation and dissemination. Outlook Agric. 39:17–24. Vlaming, J., H. van den Bosch, M.S. van Wijk, A. de Jager, A. Bannink, and H. van Keulen. 2001. Monitoring nutrient flows and economic performance in tropical farming systems (NUTMON)—Part 1: Manual for the NUTMON-Toolbox. Wageningen, the Netherlands: Alterra. Windmeijer, P.N., and W. Andriesse. 1993. Inland valleys in west Africa: An agro-ecological characterization of rice-growing environments. Publication 52. Wageningen, the Netherlands: ILRI. Wint, W., J. Slingenbergh, and D. Rogers. 2000. Livestock distribution, production, and diseases. Towards a global livestock atlas. Consultants’ report. Rome: FAO. Woomer, P.L., A. Toure, and M. Sall. 2004. Carbon stocks in Senegal’s Sahel Transition Zone. J. Arid Envir. 59:499–510. Zhang, W., M. Xu, B. Wang, and X. Wang. 2009. Soil organic carbon, total nitrogen and grain yields under long-term fertilizations in the upland red soil of southern China. Nutr. Cycl. Agroecos. 84:59–69. Zingore, S., C. Manyame, P. Nyamugafata, and K.E. Giller. 2005. Long-term changes in organic matter of woodland soils cleared for arable cropping in Zimbabwe. Eur. J. Soil Sci. 56:727–736. Zingore, S., H.K. Murwira, R.J. Delve, and K.E. Giller. 2007. Influence of nutrient management strategies on variability of soil fertility, crop yields and nutrient balances on smallholder farms in Zimbabwe. Field Crops Res. 101:296–305. Zougmore, R., Z. Gnankambary, S. Guillobez, and L. Stroosnijder. 2002. Effect of stone lines on soil chemical characteristics under continuous sorghum cropping in semiarid Burkina Faso. Soil Tillage Res. 66:47–53. Zougmore, R., A. Mando, J. Ringersma, and L. Stroosnijder. 2003. Effect of combined water and nutrient management on runoff and sorghum yield in semi-arid Burkina Faso. Soil Use Manage. 19:257–264.
Needs 12 Research for Credible Data on Soil Resources and Degradation Rattan Lal CONTENTS 12.1 Introduction................................................................................................... 539 12.2 Improving Credibility of Data on Soil Degradation...................................... 541 12.3 Adverse Impacts on Agronomic Production................................................. 543 12.4 Anomaly between the Data on Soil Degradation and Agronomic Production......544 12.5 Research Needs............................................................................................. 545 12.6 Conclusions.................................................................................................... 545 Acronyms................................................................................................................546 References...............................................................................................................546
12.1 INTRODUCTION The soil science profession is at a crossroads. While the conventional themes of managing soil resources for enhancing and sustaining agronomic production are more important now than ever before, there are other issues of global relevance that must also be addressed. Important among the other issues are (i) climate change and offsetting gaseous emissions by carbon (C) sequestration in agroecosystems, (ii) the production of feedstocks for the first, second, and third generation of biofuels, (iii) urban encroachment on prime agricultural lands, (iv) rising competition for irrigation water with increasing demands from industrial and urban uses, (v) declines in the available soil resources due to degradation by erosion, leaching, salinization, and depletion of soil organic carbon (SOC) and plant nutrients, (vi) environmental pollution through intensification of agricultural activities (i.e., tillage, use of chemicals, and irrigation), and (vii) waste disposal. Most of the future increase in world population, from 6.8 billion in 2008 to 9.2 billion in 2050, will occur in developing countries where the natural resources (soil and water) are already under great stress. Consequently, soil degradation processes that lead to declines in the quality of physical, chemical, and biological attributes (Figure 12.1) must be understood and reversed through adoption of restorative land uses and recommended management practices (RMPs). 539
540
World Soil Resources and Food Security
However, long-term planning for sustainable management requires credible data on the type, extent, and severity of soil degradation. Declines in soil physical quality include deterioration in soil structure (aggregation) and tilth, increased susceptibility to crusting and compaction, decreased water infiltration rates, and accelerated runoff and erosion. Declines in soil chemical quality entail depletion of plant nutrient reserves, elemental imbalances (toxicity), unfavorable pH, build-up of salt in the root zone, and a decline in effective cation exchange capacity (ECEC) because of the preferential removal of clay and humus by surface runoff and soil erosion. Declines in soil biological quality comprise depleted soil organic matter (SOM) reserves, reduced microbial biomass C, declines in activity and species diversity of soil fauna (notably earthworms), and disrupted biogeochemical cycles. These degradative processes, leading to an accelerated downward spiral, interact with one another and create positive feedbacks. The net impact of the downward spiral is a decline in
Soil degradation impact
Soil chemical quality
Soil biological quality
Reduction in aggregation
Nutrient depletion
Depletion of SOM
Increase in compaction and crusting
Elemental toxicity
Reduction in MBC
Decrease in water infiltration and AWC
Change in soil reaction
Decline in earthworm and biodiversity
Increase in runoff and erosion
Decline in ECEC
Disruption in biogeochemical cycles
Decline in rotting depth
Secondary salinization
Unsustainability of soil and water resources
Decline in ecosystem services
Soil physical quality
Decline in NPP
FIGURE 12.1 Decline in soil quality by degradation processes leading to a reduction in net primary production (NPP). The decline in NPP is exacerbated by a reduction in available water capacity (AWC), depletion of soil organic matter (SOM) and microbial biomass carbon (MBC), depletion of nutrient reserves, and the effective cation exchange capacity (ECEC).
Research Needs for Credible Data on Soil Resources and Degradation
541
net primary production (NPP), a reduction in ecosystem services, and unsustainable use of soil and water resources (Figure 12.1). Such degradation processes must be quantified at local, regional, and global scales for judicious planning of land use and choice of farming and cropping systems as well as sustainable management of soil and water resources.
12.2 IMPROVING CREDIBILITY OF DATA ON SOIL DEGRADATION Achieving global good security, a necessity now more than ever before because 1.02 billion food-insecure people worldwide concentrated mostly in South Asia and sub-Saharan Africa, requires restoration of degraded soils and reversal of the site-specific degradation processes. Developing countries, where most of the future increases in population are projected to occur, will account for >80% of the required increase in food production. These are also the regions where the credible data on soil resources are not available. Planning for sustainable development requires a thorough understanding of the state-of-the-soils to choose an appropriate land use and identify RMPs. Oldeman [1998] estimated the productivity loss for the period between World War II and 1990 at 12.7% for croplands, and 8.9% for combined croplands and pasturelands. The estimated loss on a regional basis for croplands ranged from 3.2% for Oceania, to 36.8% for Central America, and to 25% for Africa (Table 12.1). These estimates were based on the national average yield levels and the areal extent of degradation based on the Global Assessment of Human-Induced Soil Degradation (GLASOD) methodology. There are two databases of global assessments of soil and land degradation (Table 12.2). The third database on land desertification and degradation in arid climates is not discussed in this chapter. The GLASOD methodology estimated soil degradation in 1990 at 1.964 billion ha (Bha). The regional estimates included 747 million ha (Mha) (38% of the world total) in Asia, 494 Mha (25%) in Africa, 306 Mha (15.6%) in Latin America and the Caribbeans, 218 Mha (11.2%) in Europe, 103 Mha (5.2%) in Australia and the Pacific, and 96 Mha (5.0%) in North America. Another global study in 2008 estimated land area affected by degradation at 2.63 Bha (overall 3.5 Bha). The regional estimates, recalculated from the national level data provided by Bai et al. [2008], indicated degradation of 626 Mha (23.8%) in Asia, 523 Mha (19.9%) in Africa, 468 Mha (17.8%) in Latin America and the Caribbeans, 446 Mha (16.9%) in North America, 355 Mha (13.5%) in Europe, and 216 Mha (8.1%) in Australia and the Pacific (Table 12.2). The data, in which regional totals of area affected by degradation do not add up to the global total, show a global increase of 34% even with the conservative estimate of 2.63 Bha (Table 12.2). The 2008 estimates by Bai et al. indicated an increase in area affected by land degradation over that reported in 1990 by Oldeman et al. for all regions except in Asia. There is an important distinction between the terms land and soil. The latter is only one of the components of land, another being vegetation. The change in area under land degradation was −16.2% in Asia, 5.9% in Africa, 52.9% in Latin America and the Caribbeans, 364.6% in North America, 109.7% in Australia and the Pacific, and 62.8% in Europe (Table 12.2). The estimates in 2008 were based on the vegetation index, and were also related
542
World Soil Resources and Food Security
TABLE 12.1 Average Cumulative Loss of Productivity from Cropland Region Africa Asia South America Central America North America Europe Oceania World
Cropland
Crop and Pastureland
25.0 12.8 13.9 36.8 8.8 7.9 3.2 12.7
14.2 8.9 6.7 14.5 5.8 9.0 3.2 8.9
Source: Oldeman, L.R., Soil degradation: A threat to food security. ISRIC Report 98/01. Wageningen, the Netherlands: ISRIC, 1998.
to NPP. However, higher estimates reported by Bai et al. are partly due to the fact that they assessed degradation of land rather than of soil. To complicate the already complex situation, the terms degradation and desertification are also used widely and interchangeably. Regardless of the diverse but related approaches and use of different terminology, the reliability/credibility of the data were not extensively checked by validation through ground truthing. Yet, it is important that estimates are validated on a pilot scale for principal soils in major biomes or ecoregions. The major drawback, which creates skepticism and the crying wolf syndrome, is the lack of validation and of establishing the cause–effect relationship with productivity and other ecosystem
TABLE 12.2 Global Assessment of Soil Degradation in 1990 and 2006 Degrading Area (Mha) Region Asia Africa Latin America and Caribbeans North America Australia and Pacific Europe Total a
1990 [Oldeman et al. 1990] 747 494 306 96 103 218 1964
2008 [Bai et al. 2008]
Percent Change
626 523 468 446 216 355 2634 (3606)a
−16.2 5.9 52.9 369.6 109.7 62.8 34.1
The regional total, computed form the national statistics published by Bai et al. [2008], do not add up to the global total. The discussion in the text and the ratio shown in the last column are based on the lower estimates based on the regional total.
Research Needs for Credible Data on Soil Resources and Degradation
543
services. Improving the data on the state-of-the-soils is essential for acceptance of the risks of soil degradation by policymakers and planners, which is necessary to implement any long-term soil restoration programs.
12.3 ADVERSE IMPACTS ON AGRONOMIC PRODUCTION The data on extent and severity of soil degradation must be related to quantity and quality of agronomic production. The term quantity implies change in mean crop yields and total production at regional, national, and global levels. Crop yield, being a function of management and inputs, must be assessed for a range of technological options (i.e., from traditional to improved and innovative). Any masking impact of RMPs on degradation-induced declines in crop yields must be quantified. Soil degradation, especially that caused by the depletion of SOM content and deficiency of macronutrients and micronutrients, is also related to malnutrition in humans and the overall decline in public health. Indeed, soil degradation affects food security directly by reducing crop yields and decreasing agronomic production, and indirectly by reducing the nutritional value of the agricultural produce (protein content, concentration of micronutrients such as Zn, Cu, I, Se). Other indirect effects of soil degradation on human health are those related to pollution of soils, air, and water. Thus, both direct and indirect effects of soil degradation must be quantified for major cropping and farming systems on principal soils and representative agroecosystems. Soil degradation also exacerbates the adverse impacts of extreme events related to climate change by natural variability or human-induced factors. For example, frequency and duration of drought are accentuated by soil degradation due to erosion, salinization, or elemental imbalance. Among the three types of drought (meteorological, hydrological, and agronomical), soil degradation exacerbates the adverse effects of an agronomic drought. The latter is attributed to an increase in the loss of water by runoff and evaporation, and a decrease in plant available water capacity (AWC) in the root zone (Figure 12.1). In contrast to drought, extreme events may also reduce agronomic yields by inundation (or flooding) that causes anaerobic environments in the root zone. These impacts need to be quantified through implementation of properly designed long-term studies. Nutrient mining (refer to Chapter 11 by Smaling et al.) is a major problem in subSaharan Africa and also in South Asia. The adverse impact of the negative nutrient budget, assessed for soil-specific situations under different levels of nutrient inputs, must be assessed for a range of crops and cropping systems. The goal is to quantify the vulnerability of agronomic production to soil degradation by different processes under a range of management options. Agronomic yield can also be related to use efficiency of inputs (i.e., fertilizer, water, energy) under different severities of soil degradation. Agronomic production also affects farm income, poverty, and social well-being. The so-called poverty trap tightened by soil degradation must be quantified, especially for the resource-poor farmers of sub-Saharan Africa, South Asia, and other developing countries. To be credible, the statistics on the extent and severity of soil degradation must be related to some quantifiable and verifiable indicators.
544
World Soil Resources and Food Security
12.4 A NOMALY BETWEEN THE DATA ON SOIL DEGRADATION AND AGRONOMIC PRODUCTION The estimates of the land area affected by degradation indicate an increase in the extent of degradation at regional and global scales between 1990 and 2008. All other factors remaining the same, there should be a corresponding decline in agronomic production. Yet, there has been an increase in crop yields and total cereal and food production globally and in most regions (except sub-Saharan Africa). The data in Table 12.3 show that between 1990 and 2008, the global average cereal yield increased by 27% from 2783 kg/ha to 3539 kg/ha, and total cereal production increased by 28% from 1974 million Mg/yr to 2521 million Mg/yr. Thus, there is an apparent anomaly and discrepancy in the data on the extent of degradation and agronomic production. This discrepancy, in a critical need of an objective assessment, seemingly indicates the following:
1. All other factors have not been the same between 1990 and 2008 because of increases in inputs (i.e., irrigation, fertilizers, improved varieties) and differences in cropping systems. 2. The data on soil degradation are not credible and do not reflect the local and regional levels. 3. The data sets are not comparable because of differences in terminology and methodology (i.e., land vs. soil).
Providing that the areal extent of degradation has increased since 1990, there must be a strong “masking” impact of RMPs on productivity of degraded soils. If the use of inputs (i.e., fertilizers, irrigation, tillage, new varieties) can alleviate the soilrelated constraints exacerbated by erosion or nutrient depletion, then there is an urgent need to revisit the definition of the term soil degradation. Rather than including the categories of light and moderate, estimates may only focus on strong or even very strong categories. Assuming that the estimates of soil degradation are limited to strong and very strong categories, ground truthing and field measurements of crop yields for a range of management scenarios are crucial. TABLE 12.3 World’s Total Cereal Production and Average Cereal Yield Total Cereal Production
Average Cereal Yield
Year
(106 Mg/yr)
(%)
(kg/ha)
(%)
1992 2005 2008
1974 2268 2521
100 115 128
2783 3286 3539
100 118 127
Source: FAOSTAT. Rome: FAO, 2010. http://faostat.fao.org/site/567/DesktopDefauly.aspx?pageID=567.
Research Needs for Credible Data on Soil Resources and Degradation
545
12.5 RESEARCH NEEDS In view of the argument presented in Section 12.4, there is a strong need to implement coordinated, multidisciplinary, and interinstitutional studies on assessing: (1) the area affected by strong or very strong degradation processes; (2) severity of degradation in relation to the on-site and off-site impact; and (3) the impact on agronomic production, use-efficiency of inputs, and long-term sustainability. The methodology and terminology used must be standardized and used uniformly across the world. Such an evaluation must be done once every decade so that long-term trends can be established. Some researchable priorities include the following:
1. Assessing degradation on a soil basis (rather than a regional or national basis) in relation to the global data base published by FAO/UNESCO, USDA, ISRIC and other organizations. 2. Using remote sensing and geostatistical techniques to extrapolate the data to regional and global scales. 3. Quantifying loss in agronomic production and economic returns under different managements, cropping systems, and input scenarios. 4. Evaluating the impact of soil degradation on the nutritional quality of produce. 5. Determining the social impact of soil degradation on farm income, standards of living, education of children (especially of girls), and social and gender equity. 6. Relating soil degradation to environmental factors such as the quality of natural waters (surface and groundwater), biodiversity, emission of greenhouse gases, ecosystem C budget, etc. 7. Establishing long-term trends in the vegetation cover (ground cover) through remote-sensing techniques, and relating these to the area affected by soil degradation. 8. Standardizing the methodology of assessment of degradation and terminology (soil vs. land; degradation vs. desertification), and developing soil type and use-specific indicators of soil quality. 9. Developing channels of communication with policymakers on short- vs. long-term and on- vs. off-site effects of soil degradation, and creative relevant policy interventions to reversing the trends and restoring degraded soils. The strategy is to enhance awareness among leaders and policymakers. 10. Encouraging appropriate changes in education curricula for primary, middle, and secondary schools to enhance awareness among students during the formative stages of their development.
12.6 CONCLUSIONS Soils are among the most basic resources essential to the existence and well-being of all terrestrial life; therefore, preserving, restoring, and improving soil resources is important. In this regard, the data on state-of-the-soils is important for planning long-term and sustainable use of soil resources. It is thus important to create and
546
World Soil Resources and Food Security
strengthen a credible database on the extent and severity (degree) of soil degradation, establish the cause–effect relationship, standardize the methods of assessment of the areal extent and temporal changes and terminology, evaluate adverse impacts on agronomic production for a range of management scenarios, and identify indicators of soil quality in relation to the land use. The strategy is to establish long-term interdisciplinary and multi-institutional studies on the pilot scale for principal soils and major ecoregions. World soils have the capacity to feed current and future populations and provide other essential ecosystem services, providing that soil resources are judiciously used with site-specific RMPs that maintain and restore soil quality. Identifying longterm management options necessitates credible data on the state-of-the-soils and its impact on productivity and environment quality. The data on soil degradation must be validated against ground truth measurements, and agronomic information on productivity, nutritional quality, and economic well-being.
ACRONYMS AWC Bha C ECEC GLASOD MBC Mha NPP RMPs SOC SOM
Available water capacity Billion ha Carbon Effective cation exchange capacity Global Assessment of Human-Induced Soil Degradation Microbial biomass carbon Million ha Net primary production Recommended management practices Soil organic carbon Soil organic matter
REFERENCES Bai, Z.G., D.L. Dent, L. Olsson, and M.E. Schaepman. 2008. Proxy global assessment of land degradation. Soil Use & Management 24:223–234. FAO. 2010. FAOSTAT. Rome: FAO. http://faostat.fao.org/site/567/DesktopDefauly.aspx? pageID=567 (accessed December 15, 2010). Oldeman, L.R. 1998. Soil degradation: A threat to food security. ISRIC Report 98/01. Wageningen, the Netherlands: ISRIC. Oldeman, L.R., V.W.P. Van Engelen, and J.H.M. Pulles. 1990. The extent of human-induced soil degradation. In World map of the status of human-induced soil degradation, and explanitory note, 2nd ed., eds. L.R. Oldeman, R.D.A. Hakkleling, and W.G. Sombroeck. Wageningen, the Netherlands: GLASOD, ISRIC.