Advances in Agronomy, Volume 98

V O LU M E N I N E T Y ADVANCES IN E I G H T AGRONOMY ADVANCES IN AGRONOMY Advisory Board PAUL M. BERTSCH RONA...

Author: Donald L. Sparks

99 downloads 3263 Views 7MB 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

DOWNLOAD PDF

V O LU M E

N I N E T Y

ADVANCES

IN

E I G H T

AGRONOMY

ADVANCES IN AGRONOMY Advisory Board

PAUL M. BERTSCH

RONALD L. PHILLIPS

University of Kentucky

University of Minnesota

KATE M. SCOW

LARRY P. WILDING

University of California, Davis

Texas A&M University

Emeritus Advisory Board Members

JOHN S. BOYER

KENNETH J. FREY

University of Delaware

Iowa State University

EUGENE J. KAMPRATH

MARTIN ALEXANDER

North Carolina State University

Cornell University

Prepared in cooperation with the American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America Book and Multimedia Publishing Committee DAVID D. BALTENSPERGER, CHAIR LISA K. AL-AMOODI

CRAIG A. ROBERTS

KENNETH A. BARBARICK

MARY C. SAVIN

HARI B. KRISHNAN

APRIL L. ULERY

SALLY D. LOGSDON

V O LU M E

N I N E T Y

ADVANCES

E I G H T

IN

AGRONOMY EDITED BY

DONALD L. SPARKS Department of Plant and Soil Sciences University of Delaware Newark, Delaware

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier 84 Theobald’s Road, London WC1X 8RR, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2008 Copyright # 2008 Elsevier Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made ISBN: 978-0-12-374355-8 ISSN: 0065-2113 (series) For information on all Academic Press publications visit our website at books.elsevier.com Printed and bound in USA 08 09 10 10 9 8 7 6 5 4 3 2 1

CONTENTS

Contributors Preface

1. Advances in Precision Conservation

ix xiii

1

Jorge A. Delgado and Joseph K. Berry 1. Introduction 2. Geospatial Technologies 3. Identifying Spatial Patterns and Relationships 4. Field Level Flows 5. Connection of Field with Off-Site Transport 6. Watershed Scale Considerations 7. Current Applications and Trends 8. Summary and Conclusions References

2. Reaction and Transport of Arsenic in Soils: Equilibrium and Kinetic Modeling

2 4 9 12 17 22 28 39 39

45

Hua Zhang and H. M. Selim 1. Introduction 2. Environmental Toxicity 3. Arsenic in Soils 4. Biogeochemistry 5. Transport in Soils 6. Modeling 7. Remediation of Contaminated Soils 8. Summary and a Look Ahead References

3. Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

46 47 48 52 73 81 101 104 105

117

Bijay-Singh, Y. H. Shan, S. E. Johnson-Beebout, Yadvinder-Singh, and R. J. Buresh 1. Introduction 2. Criteria for Evaluating Crop Residue Management Options

118 121 v

vi

Contents

3. 4. 5. 6.

Type and Abundance of Crop Residues Existing and Emerging Residue Management Options Evaluation of Options with Residues Managed During a Rice Crop Evaluation of Options with Residues Managed During a Non-Flooded Crop 7. Crop Residue and Bioenergy Options 8. Summary Acknowledgment References

4. Sampling and Measurement of Ammonia at Animal Facilities

123 125 135 160 181 183 185 186

201

Ji-Qin Ni and Albert J. Heber 1. Introduction 2. A General View of Ammonia Determination 3. Ammonia Sampling 4. Ammonia Concentration Measurement 5. Measurement Methods and Devices 6. Ammonia Concentration Data 7. Summary and Conclusions Acknowledgments References

5. Will Stem Rust Destroy the World’s Wheat Crop?

203 205 206 221 225 243 255 257 257

271

Ravi P. Singh, David P. Hodson, Julio Huerta-Espino, Yue Jin, Peter Njau, Ruth Wanyera, Sybil A. Herrera-Foessel, and Richard W. Ward 1. 2. 3. 4. 5.

Introduction Stem Rust Disease, Pathogen, and Epidemiology Breeding for Resistance Race UG99 and Why it is a Potential Threat to Wheat Production Breeding Strategies to Mitigate the Threat from UG99 and Achieve a Long-Term Control of Stem Rust 6. Conclusion and Future Outlook Acknowledgments References

272 274 277 281 288 305 306 306

6. Genetic Improvement of Forage Species to Reduce the Environmental Impact of Temperate Livestock Grazing Systems

311

M. T. Abberton, A. H. Marshall, M. W. Humphreys, J. H. Macduff, R. P. Collins, and C. L. Marley 1. Introduction 2. Reducing Diffuse Nitrogenous Pollution of Watercourses

312 315

Contents

3. Reducing P Pollution of Watercourses 4. Reducing Emissions to Air 5. Improving Soil Quality and Reducing Flood Damage 6. Enhancing Persistency and Resilience 7. Enhancing C Sequestration in Grasslands 8. Future Prospects Acknowledgments References

7. Mutagenesis and High-Throughput Functional Genomics in Cereal Crops: Current Status

vii

321 325 335 339 341 344 345 345

357

H. S. Balyan, N. Sreenivasulu, O. Riera-Lizarazu, P. Azhaguvel, and S. F. Kianian 1. Introduction 2. Insertional Mutagenesis 3. Non-Transgenic TILLING, DEALING, and DeleteageneTM Approaches 4. Phenomics Platform for Screening Mutagenized Population 5. Outlook Acknowledgments References Index

358 361 380 398 399 401 401 415

This page intentionally left blank

CONTRIBUTORS

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

M. T. Abberton (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom P. Azhaguvel (357) Texas A&M University Agricultural Research and Extension Center, 6500 Amarillo Blvd West, Amarillo, Texas 79106 H. S. Balyan (357) Department of Genetics and Plant Breeding, Ch. Charan Singh University, Meerut 250 004, India Joseph K. Berry (1) Berry and Associates, Spatial Information Systems, Fort Collins, Colorado 80525 Bijay-Singh (117) Department of Soils, Punjab Agricultural University, Ludhiana 141 004, Punjab, India R. J. Buresh (117) International Rice Research Institute, Los Ban˜os, Philippines R. P. Collins (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom Jorge A. Delgado (1) USDA-ARS, Soil Plant Nutrient Research Unit, Fort Collins, Colorado 80526 Albert J. Heber (201) Agricultural and Biological Engineering Department, Purdue University, West Lafayette, Indiana 47907 Sybil A. Herrera-Foessel (271) International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico David P. Hodson (271) International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico

ix

x

Contributors

Julio Huerta-Espino (271) INIFAP-CEVAMEX, 56230 Chapingo, Mexico M. W. Humphreys (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom Yue Jin (271) USDA-ARS, Cereal Disease Laboratory, St. Paul, Minnesota 55108 S. E. Johnson-Beebout (117) International Rice Research Institute, Los Ban˜os, Philippines S. F. Kianian (357) Department of Plant Sciences, North Dakota State University, Fargo, North Dakota 58105 J. H. Macduff (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom C. L. Marley (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom A. H. Marshall (311) Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom Ji-Qin Ni (201) Agricultural and Biological Engineering Department, Purdue University, West Lafayette, Indiana 47907 Peter Njau (271) Kenya Agricultural Research Institute, Njoro Plant Breeding Research Center (KARI-NPBRC), Njoro, Kenya O. Riera-Lizarazu (357) Department of Crop and Soil Science, Oregon State University, Corvallis, Oregon 97331 H. M. Selim (45) School of Plant, Environmental and Soil Sciences, Louisiana State University, Baton Rouge, Louisiana 70803 Y. H. Shan (117) College of Environmental Science and Engineering, Yangzhou University, Yangzhou 225009, China

Contributors

xi

Ravi P. Singh (271) International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico N. Sreenivasulu (357) Leibniz Institute of Plant Genetics and Crop Plant Research, Corrensstrasse-03, Gatersleben 06466, Germany Ruth Wanyera (271) Kenya Agricultural Research Institute, Njoro Plant Breeding Research Center (KARI-NPBRC), Njoro, Kenya Richard W. Ward (271) International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico Yadvinder-Singh (117) Department of Soils, Punjab Agricultural University, Ludhiana 141 004, Punjab, India Hua Zhang (45) School of Plant, Environmental and Soil Sciences, Louisiana State University, Baton Rouge, Louisiana 70803

This page intentionally left blank

PREFACE

Volume 98 contains seven comprehensive and timely reviews. Chapter 1 covers advances in precision conservation, including cutting-edge technologies and applications and trends. Chapter 2 deals with equilibrium and kinetic modeling of arsenic reactions and transport in soils and includes background material on the biogeochemistry of arsenic, a toxic element of concern worldwide. Chapter 3 covers crop residue management for lowland rice-based cropping systems in Asia with discussions on existing and emerging residue management options. Chapter 4 is a timely review on sampling and measurement of ammonia at animal facilities, including measurement methods and advances in data collection. Chapter 5 is concerned with stem rust and its effects on wheat production, including breeding efforts and strategies for reducing its impact. Chapter 6 covers genetic improvement of forage species with a goal of reducing the environmental impact of temperate livestock grazing systems. Topics dealing with reducing nitrogen and phosphorus pollution and air emissions are included. Chapter 7 is a comprehensive chapter on the current status of mutagenesis and high-throughput functional genomics in cereal grains. I am grateful for the authors’ outstanding reviews. DONALD L. SPARKS University of Delaware

xiii

This page intentionally left blank

C H A P T E R

O N E

Advances in Precision Conservation Jorge A. Delgado* and Joseph K. Berry† Contents 2 4 9 12

1. 2. 3. 4.

Introduction Geospatial Technologies Identifying Spatial Patterns and Relationships Field Level Flows 4.1. Variable erosion and transport (flows of gases, nutrients, and water) 4.2. Precision conservation for management of flows 5. Connection of Field with Off-Site Transport 5.1. Variable flows from field to nonfarm areas 5.2. Precision conservation buffers and riparian zones 6. Watershed Scale Considerations 6.1. Variable hydrology 6.2. Models and tools 6.3. Precision conservation at a watershed scale 7. Current Applications and Trends 8. Summary and Conclusions References

12 16 17 17 20 22 22 22 24 28 39 39

Population growth is expected to increase, and the world population is projected to reach 10 billion by 2050, which decreases the per capita arable land. More intensive agricultural production will have to meet the increasing food demands for this increasing population, especially because of an increasing demand for land area to be used for biofuels. These increases in intensive production agriculture will have to be accomplished amid the expected environmental changes attributed to Global Warming. During the next four decades, soil and water conservation scientists will encounter some of their greatest challenges to maintain sustainability of agricultural systems stressed by increasing food and biofuels demands and Global Warming. We propose that Precision Conservation will be needed to support parallel increases in soil and water conservation

* {

USDA-ARS, Soil Plant Nutrient Research Unit, Fort Collins, Colorado 80526 Berry and Associates, Spatial Information Systems, Fort Collins, Colorado 80525

Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00201-0

#

2008 Elsevier Inc. All rights reserved.

1

2

Jorge A. Delgado and Joseph K. Berry

practices that will contribute to sustainability of these very intensively-managed systems while contributing to a parallel increase in conservation of natural areas. The original definition of Precision Conservation is technologically based, requiring the integration of a set of spatial technologies such as global positioning systems (GPS), remote sensing (RS), and geographic information systems (GIS) and the ability to analyze spatial relationships within and among mapped data according to three broad categories: surface modeling, spatial data mining, and map analysis. In this paper, we are refining the definition as follows: Precision Conservation is technologically based, requiring the integration of one or more spatial technologies such as GPS, RS, and GIS and the ability to analyze spatial relationships within and among mapped data according to three broad categories: surface modeling, spatial data mining, and map analysis. We propose that Precision Conservation will be a key science that will contribute to the sustainability of intensive agricultural systems by helping us to analyze spatial and temporal relationships for a better understanding of agricultural and natural systems. These technologies will help us to connect the flows across the landscape, better enabling us to evaluate how we can implement the best viable management and conservation practices across intensive agricultural systems and natural areas to improve soil and water conservation.

1. Introduction Population growth is expected to increase, and the world population is projected to reach 10 billion by 2050, which will decrease the per capita arable land from 0.23 ha in 1995 to 0.14 ha by 2050 (Lal, 1995). More intensive agricultural production will have to meet the increasing food demands for this increasing population, especially because of an increasing demand for land area to be used for biofuels. These increases in intensive production agriculture will have to be accomplished amid the expected environmental changes attributed to Global Warming. Scientists are projecting future changes of weather patterns that include regions with higher evapotranspiration rates, lower precipitation in some areas, and higher precipitation in other areas, which may contribute to higher erosion rates (Hatfield and Prueger, 2004; Lal, 1995, 2000; Nearing et al., 2004; Pimentel et al., 1995). During the next four decades, soil and water conservation scientists will encounter some of their greatest challenges to maintain sustainability of agricultural systems stressed by increasing food and biofuel demands. Several scientists have reported on the potential impacts of global population increase, increase in greenhouse gases, and potential effects of climate change on soil and water quality and on soil erosion (Hatfield and Prueger, 2004; Lal, 1995, 2000; Nearing et al., 2004; Pimentel et al., 1995). There is a concern that if precipitation patterns continue to change, certain future

Advances in Precision Conservation

3

scenarios may cause conservation practices such as crop residue, no-till, and incorporation of manure to lose effectiveness very rapidly, resulting in dramatic increases in runoff, and higher impacts to soil and water quality (Hatfield and Prueger, 2004). It is also estimated that for every 25.4 mm increase in precipitation rate, erosibility increases by 1.7% (Nearing et al., 2004). Nearing et al. (2004) reported that the relationship between increases in rain, biomass production, and erosion is more complex. Although an increase in rain could increase biomass production, a decrease in biomass may also increase erosion rates. The more difficult area to evaluate was effects of climate change on land use and erosion rates, yet they concluded from their analysis that the average increase in erosibility will be 1.7% per 25.4 mm increase in precipitation. It is important to note that Meisinger and Delgado (2002) reported an average 10–30% of total N inputs in cropping systems are lost due to nitrate leaching. Thus, increases in precipitation and/or more intensive storms could potentially contribute to higher nitrate leaching rates as well. These assessments from Nearing et al. (2004) and Hatfield and Prueger (2004) clearly show the continuing need for soil and water conservation scientists and practitioners to continue looking for alternatives for managing future impacts to soil and water quality. Scientists and conservation practitioners will have to work together with farmers across all types of soils and weather to increase and sustain higher production to meet the demands of the increasing population, while managing for potential changes in weather patterns. This cooperation will also be necessary to develop cropping systems that produce enough to meet the increasing food and biofuel demands while maximizing soil and water conservation. The implementation of soil and water conservation will be necessary for the sustainability of these intensive efforts to maximize agricultural production. New technologies will help us to increase yields per hectare and these technologies will also be applied to understand and manage agricultural systems and to connect the flows from agricultural systems to natural areas in an effort to manage these regions for maximum yield and agroenvironmental sustainability. Precision Conservation was originally defined as a set of spatial technologies and procedures linked to mapped variables, which is used to implement conservation management practices that take into account spatial and temporal variability across natural and agricultural systems (Berry et al., 2003). Contrary to Precision Farming that was oriented to maximize yields in agricultural fields, Precision Conservation connects farm fields, grasslands, and range areas with the natural surrounding areas such as buffers, riparian zones, forest, and water bodies (Fig. 1). The goal of Precision Conservation is to use information about surface and underground flows to analyze the systems in order to make the best viable decisions for application of management practices that contribute to conservation of agricultural, rangeland, and natural areas.

4

Jorge A. Delgado and Joseph K. Berry

Precision conservation Precision Ag Wind erosion

Chemicals

Soil erosion Runoff Leaching

Leaching

Terrain

Leaching

Soils Yield Potassium

3-dimensional Flows Cycles

Coincidence

CIR image

2-dimensional Interconnected perspective

Isolated perspective

Figure 1 The site-specific approach can be expanded to a three-dimensional scale approach that assesses inflows and outflows from fields to watershed and region scales. (From Berry et al., 2003.)

Berry et al. (2003) acknowledged that there could be different degrees of Precision Conservation such as the use of nondigital, non-GIS maps and the use of survey methods that can help in the application of spatial conservation practices. However, the original definition of Precision Conservation is technologically based, requiring the integration of spatial technologies such as global positioning systems (GPS), remote sensing (RS), and geographic information systems (GIS) and the ability to analyze spatial relationships within and among mapped data according to three broad map analysis categories: spatial analysis, surface modeling, and spatial data mining (Fig. 2). Since Berry et al. (2003), several other papers related to the topic of Precision Conservation have been published describing how these new technologies can be applied for maximizing Precision Conservation.

2. Geospatial Technologies New GIS, GPS, RS, modeling, and computer program technologies are rapidly increasing our capacity to analyze large sets of information in space and time. Traditional statistics used for soil and water conservation studies and assessment of best management practices were initially nonspatial and analyzed a data set by fitting a numerical distribution (e.g., standard

Surface modeling

Point samples are spatially interpolated into a continuous surface

53.2 ppm

4.2 ppm

Field sample locations Phosphorus surface

Discrete data spikes

Min = 4.2 Max = 53.2 Avg = 13.4 SDev = 5.2

Spatial data mining 32c,62r

45c,18r

Map surfaces are clustered to identify data pattern groups

P 53.2

Relatively low responses in P, K, and N Relatively high responses in P, K, and N

11.0

Cluster 2 Cluster 1

N

K 412.0

177.0

27.9

32.9

N K P Geographic space

Data space

Clustered data zones

Figure 2 Surface modeling is used to derive map surfaces that utilize spatial data mining techniques to investigate the numerical relationships in mapped data.(From Berry et al., 2005.)

6

Jorge A. Delgado and Joseph K. Berry

normal curve) to generalize the central tendency of the data. The values used for soil and water conservation have traditionally used mean and standard deviation to describe the responses to a traditional conservation practice, informing its numerical distribution without any reference to the spatial distribution of the data sources. The basic assumption for this method of analysis was that these relationships among the data were randomly (or uniformly) distributed in geographic space. Many of the analysis techniques were considered less valid if the data exhibited spatial autocorrelation. New methods and advances in models use spatial technologies to analyze spatial relationships within and among mapped data for highly detailed insight into the field of Precision Conservation and the potential for sitespecific applications that can contribute to environmental sustainability (Berry, 1999, 2003a,b; Mueller et al., 2005; Qiu et al., 2007; Renschler and Lee, 2005; Schumacher et al., 2005). These new soil and water conservation analysis capabilities enabled by GIS can be grouped into two broad map analysis categories: Spatial Statistics, involving numerical relationships of surface modeling and spatial data mining and Spatial Analysis, involving geographical relationships, such as proximity and terrain configuration (Berry, 1999, 2003a,b). These new spatial techniques will contribute to an integrated evaluation of topography, hydrology, weather, management, and other physical and chemical parameters, providing new insight into sitespecific Precision Conservation for management of flow-interconnected agricultural and natural resources. Figure 3 outlines the fundamental differences between the traditional GIS mapping approach and the map analysis approach used in Precision Conservation. Most desktop mapping applications take a set of spatially collected data (e.g., parts per million, kilogram per hectare, etc.), then reduces the data set to a single value (total, average, median, etc.), and ‘‘paints’’ a fixed set of polygons with colors reflecting the scalar statistic of the field data occurring within each polygon. For example, the left side of Fig. 3 depicts the position and relative values for a set of field collected data; the right side shows the derived spatial distribution of the data for an individual reporting parcel. The average of the mapped data is shown as a superimposed plane ‘‘floating at average height of 22.0’’ and assumed to be the same everywhere within the polygon. But the data values themselves, as well as the derived spatial distribution, suggest that higher values occur in the northeast and lower values in the western portion. The first thing to notice in the figure is that the average exists hardly anywhere, forming just a thin band cutting across the parcel. Most of the mapped data is well above or below the average. That is what the standard deviation attempts to reveal—just how typical the computed typical value really is. If the dispersion statistic is relatively large, then the computed typical is not typical at all. The limitation inherent in previous computer

Map analysis Desktop mapping Field data Standard normal curve fit to the data

Spatially interpolated data

34.1% 34.1%

68.3% +/−1 standard deviation

Average = 22.0 StDev = 18.7

22.0

28.2

Discrete spatial object (generalized)

80 60 40 20 0 −20 −40 −60

High = 50

80 60 40 20 Average = 22.0

0 −20 −40 −60

N

Continuous spatial distribution (detailed)

Figure 3 Desktop mapping uses aggregated, nonspatial statistics to summarize spatial objects (points, lines, and polygons), whereas map analysis uses continuous spatial statistics to characterize gradients in geographic space (surfaces).

8

Jorge A. Delgado and Joseph K. Berry

applications arises from the fact that most desktop mapping applications ignore data dispersion and simply ‘‘paint’’ a color corresponding to the average regardless of numerical or spatial data patterns within a parcel. However, the central tendency assumption can be misleading. Assume the data is characterizing a toxic chemical in the soil that, at high levels, poses a serious health risk. The mean values for both the parcel on the left (22.0) and the right (28.2) are well under the ‘‘critical limit’’ of 50.0. Desktop mapping would paint both parcels a comfortable green tone, as their typical values are well below the level of concern. Even when considering the upper-tails of the standard deviations, the limit is not exceeded (22.0 + 18.7 = 40.7 and 28.2 + 19.8 = 48.0). So from a nonspatial perspective, the aggregated results indicate acceptable levels of the chemical in both parcels. However, the lower right portion of the figure portrays a radically different set of conditions. The left and right parcels are displayed as an increasing gradient from low levels (green) through areas that are above the critical limit (red tones). The high regions, when combined, represent a contiguous subarea of nearly 15% of the combined area that likely extends into adjacent parcels. The aggregated, nonspatial treatment of the spatial data fails to uncover the spatial pattern by assuming the average value is everywhere within the parcels. Similar surface modeling investigations can be used to compile point data into a continuous surface representation of data across the landscape to explain any variance. Point density mapping, spatial interpolation, and map generalization are examples of uses of surface modeling. Point density mapping can be used to evaluate the number of aggregate points within a specified distance (e.g., number of occurrences per hectare). Conservation practitioners and scientists will collect point-sampled data to derive maps of nutrient concentrations such as soil carbon. For example, we could use kriging for spatial interpolation of weight-average measurements within a localized area to assess carbon sequestration potential. An example of map generalization is the use of polynomial surface fitting to the entire data set. There are new techniques for spatial data mining that can be used to try to uncover relationships within and among multiple mapped data layers such as water tables, erosion potential, topography, soil texture, yields, vegetative cover, soil depths, and others (Berry, 1999, 2003a). Berry (2002) reported that these procedures, including coincidence summary, proximal alignment, statistical tests, percent difference, level-slicing, map similarity, and clustering can be used to assess similarities in data patterns. Another type of spatial data mining is the use of predictive models that use crop biomass cover (straw biomass production-dependent variable) and the soil nutrient values [soil texture, soil carbonates, topography, hydrology, water levels, and runoff (independent variables)], then quantify the data pattern. As thousands of map locations are analyzed, a predictable pattern

Advances in Precision Conservation

9

between crop biomass and the variables may appear. This crop residue production may be correlated to potential for reduction of erosion, of surface runoff, and other soil and water conservation outcomes. Scientists and practitioners can analyze the numerical relationships of spatial patterns inherent in mapped data using surface modeling and spatial data mining. These approaches can be used to explain variance by mapping and analyzing spatial distributions (Berry, 2002).

3. Identifying Spatial Patterns and Relationships For more than 8000 years, we have been using maps with features that identify special locations in the landscape to help us navigate. Precision Conservation is a new way to use advanced technologies to integrate thousands of data points and multiple layers of information contained in maps for management and conservation of the agricultural and natural areas. Specifically, Precision Conservation allows us to identify those management landscape combinations that produce or receive significant impact. Scientists have been using spatial information for soil and water conservation for decades. However, since the development of new computers and GIS technology in the early 1970s, mapped data have changed to digital representations that are linked to larger databases, thereby increasing the number of possible applications for Precision Conservation. These new developments and the capability to integrate thousands of points and multiple map layers of information to analyze spatial and temporal relationships are providing new answers for applications of Precision Conservation. There is even potential to use these map analyses to contribute to air quality conservation. We could use these new analyses to evaluate how conservation practices could be applied to reduce wind erosion from the most sensitive areas. Spatial emissions of trace gases such as nitrous oxide (N2O) and ammonia (NH3) volatilization could also be managed using Precision Conservation. There is potential to use these layers of information to develop Precision Conservation Management plans (Kitchen et al., 2005; Knight, 2005; Lerch et al., 2005). These advances in evolving technologies will continue to increase during the next four decades, which will facilitate and speed the collection and use of thousands of data points and multiple map layers. An example of these new technologies is the mote, a quarter-sized wireless smart sensor that fits anywhere. These smart sensors, initially developed by researchers at University of California at Berkeley and Intel, could have future applications in soil and water conservation. These sensors, called ‘‘smart dust’’ by their developers, Professors Kristofer Pister and Joseph Kahn of University

10

Jorge A. Delgado and Joseph K. Berry

of California at Berkeley, can be scattered, sending information from remote locations. These and other new developments may contribute to the collection of information that will be used to generate maps for use in analysis in the field of Precision Conservation (Berry et al., 2003). Map analysis procedures can be used to study landscape relationships among map features. These analyses can assess the relative position of features in the landscape and their connectivity to flows in the environment. We can use these map analyses to evaluate effective distances, indexes, optimal path connectivity, flows, biomass cover, soil texture, microterrain analysis, elevation, distances to water bodies, and other landscape characteristics. We could simulate the flows over an elevation map to estimate the erosion potential as described by Berry et al. (2003) by using an analysis that follows the downhill path over a terrain. Berry (2003b) described this type of map analysis as a method to account for all the areas sharing common paths (Fig. 4). The ability to model flows and interconnected cycles will benefit from the current evolutionary phase of GIS involving new geo-referencing approaches. In the 1970s, the research and early applications centered on Computer Mapping (display focus) that yielded to Spatial Data Management (data structure/management focus) in the next decade as we linked digital maps to attribute databases for geo-query (left side of Fig. 5). The 1990s centered on GIS Modeling (analysis focus) that laid the groundwork for whole new ways of assessing spatial patterns and relations, as well as for entirely new applications such as Precision Agriculture. Today, in its fourth decade, GIS is centered on Multimedia Mapping (mapping focus) which brings the technology full circle to its beginnings (Berry, 2007b). While advances in virtual reality and three-dimensional visualization can ‘‘knock your socks off,’’ they represent incremental progress in visualizing maps that exploit dramatic computer hardware/software advances. Radical innovation is being addressed by current geospatial research that is refocusing on data structure and analysis (Berry, 2007a). The bulk of the current state of geospatial analysis relies on ‘‘static coincidence modeling’’ using a stack of geo-registered map layers. However, the frontier of GIS research is shifting focus to ‘‘dynamic flows modeling’’ that tracks movement over space and time in three-dimensional geographic space. But a wholesale revamping of data structure is needed to make this leap. The impact of the next decade’s evolution will be huge and will shake the very core of GIS—the Cartesian coordinate system itself, a spatial referencing concept introduced by mathematician, Rene Descartes over 400 years ago. The current two-dimensional square for geographic referencing is fine for ‘‘static coincidence’’ analysis over relatively small land areas, but is woefully lacking for ‘‘dynamic three-dimensional flows.’’ It is likely that Descartes’ two-dimensional squares will be replaced by hexagons

Inclination of a fitted plane to a location and its eight surrounding elevation values

2418

2404

2393

2409

2395

2341

2383

2373

2354

Slope(47,64) = 33.23%

35% 30% 25% 20% 15% 10% 5% 1% 0%

Steep

Moderate Gentle flat

Slope map draped on elevation Slope map

Elevation surface

Flow(28,46) = 451 paths

537 Paths Heavy 256 Paths 123 Paths 64 Paths 32 Paths 16 Paths Moderate 8 Paths 4 Paths Light 2 Paths 1 Paths minimal

Total number of the steepest downhill paths flowing into each location Flow map draped on elevation Slope map

Figure 4 Maps of surface flow confluence and slope are calculated by considering relative elevation differences throughout a project area. (From Berry et al., 2005.)

12

Jorge A. Delgado and Joseph K. Berry

Future directions

Revisit analytics (2020s)

2D planar (X,Y data)

3D solid (X,Y,Z data)

Hexagon (6 sides)

Dodecahedron (12 pentagons)

Square (4 sides)

Hexahedron (6 squares)

4) Multimedia mapping (2000s) Revisit Geo-reference (2010s) Contemporary GIS 2) Spatial dB Mgt (1980s) 3) GIS modeling (1990s)

The early years 1) Computer mapping (1970s) Mapping focus Data/structure focus Analysis focus

Figure 5 Current GIS research focuses on revolutionary changes in geo-referencing, data structures, and analytical operations that will greatly advance dynamic flows and cycles modeling directly applicable to Precision Conservation.

(analogous to the pentagon patches forming a soccer ball) that better represent our curved earth’s surface. Current three-dimensional referencing using cubes will be replaced by nesting polyhedrons for a consistent and seamless representation of three-dimensional geographic space (Peterson, 2007). This change in referencing extends the current six sides of a cube for flow modeling to the 12 sides (facets) of a polyhedron (hexagonal polyhedron)—radically changing flow and cycle algorithms, as well as our historical perspective of mapping.

4. Field Level Flows 4.1. Variable erosion and transport (flows of gases, nutrients, and water) Quine and Zhang (2002) reported that eroded areas of the field with depleted nutrients had lower yields. This spatial relationship between erosion and crop yield is complex since other areas with high soil aggregation were also found to show lower yields (Quine and Zhang, 2002). Evaluation of variable erosion on yield production was more clear when long-term

Advances in Precision Conservation

13

simulations of the field were conducted. Quine and Zhang (2002) conducted a long-term evaluation of 40 years that clearly showed the future effects of spatial erosions, identifying that the more eroded areas of the field will also have lower yields. These model simulations clearly showed that there is a need to manage fields differently in order to reduce these higher site-specific erosion rates, which will eventually reduce yields in the most affected areas (Quine and Zhang, 2002). The data from Quine and Zhang (2002) show the underlying theory that informs the concepts of Precision Conservation on a field scale. Different spatial patterns of erosion that will affect yield productivity are clearly apparent. If all fields are managed with similar conservation practices, the higher erosion rates from the most affected areas may still continue to lower the yields as crop intensity increases. Precision Conservation proposes that there is a need in those affected areas for conservation managers to consider variable conservation as a means of increasing the sustainability of these systems (Berry et al., 2003, 2005; Mueller et al., 2005; Quine and Zhang, 2002; Schumacher et al., 2005). Schumacher et al. (2005) used a soil displacement of Cesium-137 and the Water and Tillage Erosion model to assess the erosion losses due to water and tillage across the field. They found that both methods were strongly correlated. The areas showing the higher slope were those areas showing the higher tillage and water erosion rates (Fig. 6). Spatial variability of nitrogen dynamics has previously been reported by several scientists. An example of the spatial variability of residual soil NO3-N was presented by Delgado (2001) and Delgado et al. (2001) in a study of the spatial variability for vegetable and small grain systems grown in similarly managed center pivot irrigated systems. The average residual soil NO3-N in the sandy loam zone of the center pivot system was higher than in the loamy sand. Delgado (2001) reported that for center pivot irrigated barley, canola, and potato grown on the loamy sand zone, the average residual soil NO3-N was 20, 44, and 109 kg N ha 1, respectively. The residual soil NO3-N for barley, canola, and potato grown on the sandy loam zone was 42, 51, and 136 kg N ha 1, respectively. Figure 7 shows similar results for residual soil NO3-N for center pivot irrigated corn grown on a sandy coarse soil of Northeastern Colorado (Delgado and Bausch, 2005). Residual soil NO3-N was negatively correlated with the percent sand content across the field. Spatial variability of NO3-N due to leaching has also been reported (Delgado, 2001; Wylie et al., 1995). Delgado (2001) reported that for center pivot irrigated barley, canola, and potato grown on a loamy sand zone, the average NO3-N leached was 32, 39, and 91 kg N ha 1, respectively. The amounts of NO3-N leached from the loamy sand zone were higher than the amounts leached from the sandy loam zone. The average NO3-N leached from center pivot irrigated barley, canola, and potato grown on the sandy loam zone was 29, 13, and 72 kg N ha 1, respectively. Figure 8 shows similar

14

Jorge A. Delgado and Joseph K. Berry

Tillage erosion

Water erosion

Tillage–water erosion

Total erosion (cesium-137 measurements)

−1 −1 Mg ha yr

−33

Soil loss net erosion

−22

Accelerated erosion

−11 0

Soil loss T value = −11 Mg ha−1 yr−1

11 22 33 120

Soil gain net deposition

Slope % map and cesium137 sampling sites Slope % 8 7 6 5 4 3 2 1 0 Elevation contour lines are overlaid on all maps elevation labels are shown only on total erosion map

Figure 6 Erosion patterns developed from tillage, water, tillage-water, and total erosion (137Cs) modeling of the research field are displayed. Cesium sampling sites are also displayed on a contour map of slope percentage for the field.(From Schumacher et al., 2005.)

results for NO3-N leaching for center pivot irrigated corn grown on a sandy coarse soil of Northeastern Colorado (Delgado and Bausch, 2005). Higher NO3-N leaching increased as the percentage sand increased across the field (Delgado and Bausch, 2005). Best management practices, modeling, and GIS

15

Advances in Precision Conservation

A 100

100

90

80

70

70

60

Sand (%)

90

80

60

50 50

200

200

150

150

100

100

50 0

50

kg NO3-N/ha 0–1.5m

B

0

Figure 7 Spatial distribution of sand content in the top 1.5 m of soil across different productivity zones (A). Spatial distribution of observed residual soil NO3-N in the top 1.5 m of soil across the different productivity zones during the 2000 growing season (B). (From Delgado and Bausch, 2005.)

can be used to evaluate the effects of management practices on spatially variable NO3-N transport and dynamics across regions (Hall et al., 2001). Spatial variability in emissions of trace gases such as N2O and in methane (CH4) uptake and sink were reported by Mosier et al. (1996). Mosier et al. (1996) reported that for a clay catena of the short grass steppe, the N2O emissions from the swale catena position were 2.5 mg N m 2 h 1 higher than the mid or top slope position of the catena, which averaged 1.4 and

16

Jorge A. Delgado and Joseph K. Berry

250

200

150

150

100 100 50

50

0

kg NO3-N/ha 0–1.5m

250

200

0

Figure 8 Spatial distribution of predicted NO3-N leaching from the root zone of corn (1.5 m depth) across the different productivity zones during the 2000 growing season. (From Delgado and Bausch, 2005.)

1.3 mg N m 2 h 1, respectively. Methane uptake rates were lower in the swale position than the mid and upper catena. Similar spatial observations in trace gas emissions have also been reported for Canada by Goddard (2005) and Pennock (2005).

4.2. Precision conservation for management of flows Schumacher et al. (2005) reported that spatial assessment of field erosion and the development of maps from the resultant data can be useful to identify highly sensitive areas of the fields. They recommended that these maps could then be used to develop site-specific conservation practices, including cover crops, organic matter additions, and no till for the site-specific areas that have higher rates of erosion. Berry et al. (2005) reported that creation of Precision Conservation Management Zones (PCMZ) might be a viable approach to enhance soil and water conservation practices. They reported that a combination of Site-Specific Management Zones (SSMZ) (Fleming et al., 1999; Khosla et al., 2002) and PCMZ could maximize economic returns, resource use efficiency, and soil and water conservation. Several studies have shown that with the implementation of SSMZ, grain yields have remained stable or increased, N use efficiencies have increased, and economic returns have been higher (Fleming et al., 1999; Khosla et al., 2002; Koch et al., 2003). Delgado et al. (2005) reported that SSMZ reduced NO3-N leaching compared to traditional management

Advances in Precision Conservation

17

practices. Remote sensing can also be used as a Precision Conservation technique to synchronize applied N with crop N uptake demands, which can increase N use efficiency by almost 50% while sustaining yields and reducing NO3-N leaching by 47% (Delgado and Bausch, 2005). The Bausch and Delgado (2003) method saved 102 kg N ha 1 year 1 with equivalent savings of about $55.00 ha 1 per season. Delgado and Mosier (1996) reported that controlled-release fertilizer and nitrification inhibitors can reduce N losses to the environment and the emissions of N2O. We suggest that a combination of management practices using nitrification inhibitors, controlled-release fertilizer, improved management of N applications that applied N considering N uptake demands (N budgets), split N applications, remote sensing, and management zones can reduce NO3-N leaching for those most sensitive areas. Cabot et al. (2006) used Precision Conservation technology for manure management to track location, timing, and rate of manure application. They reported that it is possible to apply manure more accurately across the landscape using Precision Conservation technology. Sharpley et al. (2007) indicated that this type of management can contribute to improved manure management and to reduced off-site transport (Sharpley et al., 2007). We know that landscape positions have been correlated with trace gas emissions (Goddard, 2005; Mosier et al., 1996; Pennock, 2005). There is potential to use Precision Conservation practices for these site-specific effects across the landscape to improve N management and reduce the spatial emissions of N2O from areas with higher emission rates (Goddard, 2005; Pennock, 2005). Delgado and Mosier (1996) reported that controlled-release fertilizer and nitrification inhibitors can reduce the rate of N2O emissions. We suggest that a combination of management practices using nitrification inhibitors, controlled-release fertilizer, and improved management of N applications that applied N fertilizer considering N uptake demands (N budgets), split N applications, and other localized practices can be used to reduce N2O emissions from the areas of landscape with higher emission rates.

5. Connection of Field with Off-Site Transport 5.1. Variable flows from field to nonfarm areas The connection between field and off-site transport was assessed by Feng and Sharratt (2007) using the wind erosion prediction system and GIS. They used this approach to scale the flows from field to region. They reported that, across the entire region in Washington State, wind erosion was higher in the areas with summer fallow rotations. These unprotected areas were more susceptible to wind erosion losses. The amount of wind erosion

18

Jorge A. Delgado and Joseph K. Berry

from the region was attributed to management, not to the crop land area. For example, summer fallow area represented only 28% of measured land area for Adams County, yet it contributed considerably (an average of 14,250 kg ha 1 in topsoil losses) to soil erosion within the county. In-depth analysis of soil type also found that Mollisols experienced higher wind erosion losses. Feng and Sharratt (2007) were able to identify the most problematic areas in the region, based on management and soil type. Berry et al. (2003, 2005) presented an example of how to use map analysis to assess the potential variable flows from field to surrounding natural areas. They used the new software to assess variable flow over the landscape and to create a potential erosion map superimposed over a topographic map. The map showed the locations where the flows originate and also showed the areas with greater confluence of water. The assessment identified the locations of the field that may have greater potential for concentrated runoff to natural areas. The Berry et al. (2003, 2005) example is straightforward, identifying the areas with the heaviest contribution to flows to adjacent areas and showing how to identify potential hot spots for surface runoff and sediment and agrochemical transport out of the field. These types of analyses can help producers cover these highly sensitive edge areas with Precision Conservation grasses, create buffers along the edge of the fields, or use other viable practices that may also take into account the potential temporal variability of the flows (Fig. 9). There is potential to use GIS software and models to evaluate nonpoint sources of pollutants in the vadose zone (Corwin et al., 1998; Hall et al., 2001). Hatch et al. (2001) reported that site-specific management must evaluate surface and underground flows since some watersheds may not be affected due to erosion. Additionally, the implementation of conservation practices may reduce erosions, but watersheds may have tile flows and the management practices that reduce erosion may increase infiltration and potential for greater NO3-N leaching. Precision Conservation is a threedimensional management scheme that accounts for both surface and underground flows (Berry et al., 2003, 2005). Variable transport of chemicals in shallow underground tile flows will be affected by the composition of the soil matrix, impermeable layers, slope, and others parameters (Vadas et al., 2007). There is potential to manage the sources and sinks for these variable tile flows (Vadas et al., 2007). Vadas et al. (2007) monitored N transport in drainage ditches by monitoring hydrology and groundwater N and P in 26 shallow 3 m wells for 27 months on a heavily ditched poultry farm in Maryland. They concluded that NO3-N leaching losses due to subsurface groundwater were probably occurring across the region. Vadas et al. (2007) reported that, for a poultry farm in Maryland, the groundwater flow to shallow ditches was only intermittent and often ceased during the drier periods across the region. They reported that the

Erosion potential

Slopemap

Reclassify

Overlay

Reclassify

3 steep 2 moderate 1 gentle

Flow/slope

Slope_classes Reclassify

33 heavy flow: steep 33 heavy flow: moderate 33 heavy flow: gentle 23 moderate flow: steep 22 moderate flow: moderate 21 moderate flow: gentle 13 light flow: steep 12 light flow: moderate 11 light flow: gentle

Flowmap

Erosion_potential High Moderate Low

Flow_classes 3 heavy 2 moderate 1 light

Effective erosion buffers Effective erosion potential distance

Erosion_potential

Far

Distance

Close

Erosion buffers Streams

Figure 9 Effective erosion buffers around a stream expand and contract depending on the erosion potential of the intervening terrain.(From Berry et al., 2005.)

20

Jorge A. Delgado and Joseph K. Berry

groundwater flow to deeper ditches was continuous through time across this region and a source of continuous NO3-N leaching transport. They concluded that the management of ditches, especially the deeper ditches that are continuously receiving the tile flow, presents a tremendous opportunity to reduce the NO3-N leaching losses. Hey et al. (2005) reported that there is potential to strategically locate nutrient management farms where waters with high NO3-N concentration will flow. Hey et al. (2005) envisioned that these site-specific nutrient harvesting farms will have wetlands and riparian buffers that will be used to clean the water. There is potential to employ these ecological engineering practices using Precision Conservation to reduce the transport of nutrients into the surrounding environment (Berry et al., 2003, 2005; Hey et al., 2005). Shuster et al. (2007) conducted a model simulation to evaluate the prospect of enhanced groundwater recharge via infiltration of urban storm water runoff. They evaluated the spatial distribution of expected recharge depth relative to the distribution of soils. Their results indicated strong possibilities for reducing storm water runoff by redirecting this runoff into enhanced recharge areas. Penn et al. (2007) reported that phosphorus sorbing materials (PSMs) can be used to decrease the potential for off-site transport of phosphorus in runoff water. They reported that structures called PSM traps can be installed using PSMs and that these structures can capture runoff phosphorus from large areas of land. The phosphorus removal structure captured 99% of the dissolved phosphorus that flowed through the structure in a 24-h runoff period. The efficiency of such structures installed in the future could be maximized by consulting temporal and spatial studies using models and GIS information that consider the flows and total amount of potential movement through the traps. There is also potential to use denitrification traps to remove NO3-N from underground flows or water flows (Hey et al., 2005; Hunter, 2001). We suggest that Precision Conservation techniques could be used to analyze map and data information to strategically locate these nutrient traps at positions that can maximize the effectiveness in removing phosphorus and nitrates via denitrification.

5.2. Precision conservation buffers and riparian zones Buffers, grass waterways, wetlands, and riparian areas can be good conservation tools with the potential to filter and improve water quality (Dosskey et al., 2002; Hey et al., 2005; Lowrance et al., 2000). These practices can help reduce the transport of chemicals, sediments, and denitrified NO3-N. Dosskey et al. (2005) reported that, to use riparian buffers effectively for Precision Conservation, we need to consider the site-specific characteristics of the flows.

Advances in Precision Conservation

21

Dosskey et al. (2007) reported that we can use soil survey for the identification of the best placement of buffers. They reported that vegetative buffers may have better performance for filtering runoff in some locations than others because of the soil physical and chemical properties of the locations where the buffers will be located. Dosskey et al. (2007) used RUSLE and the Vegetative Filter Strip Model (VFSMOD) to determine the best locations. They concluded that soil surveys may be used as screening tools to guide planners to locations where the buffers will probably have a greater impact on water. Lowrance et al. (2000) reported that the Riparian Ecosystem Management Model (REMM) can be used to evaluate buffers of different shapes and soil depths. These studies from Lowrance et al. (2000) and Dosskey et al. (2005, 2007) show that, in order to maximize soil and water conservation, we need Precision Conservation techniques in which multiple layers of information are evaluated to identify the best placement and shape to maximize buffer efficiency. Peterson and Vondracek (2006) reported that there are about 8340 sinkholes in the karst terrain of southeast Minnesota. They reported that vegetative buffers around these sinkholes will significantly contribute to improved water quality for the region. They used computer models to evaluate effectiveness of the buffer ranging from 2.5 to 30 m wide and found that 30 m wide buffers reduced pollution by 80%. Smith et al. (2006) reported that the ideal buffer width to maximize water quality benefits while minimizing land utilization is difficult to determine. They reported that increasing the buffer width from 9 to 30 m was effective in reducing shallow groundwater NO3-N along the stream bank to below 1 mg l 1. However, because of the severity of NO3-N problems associated with groundwater in the deeper samples, there were not detectable improvements in NO3-N in deeper samples taken after widening of the buffer. They concluded that standardized buffer width recommendations for a variety of landscapes are difficult to generate, but that wider buffers work well in those areas where the water table remains within 1 m of the surface. The previous discussion indicates that there is potential to use variable information across the watershed to identify the best positions of the buffers. There is also potential to use Precision Conservation techniques to employ buffers of different widths that account for the variability of flows (Dosskey et al., 2005). The width of the buffer and its effectiveness will be correlated with both surface flows and water table flows (Dosskey et al., 2005; Smith et al., 2006). This is another example that shows the need to consider surface and underground flows when using vegetative barriers to serve as filters of sediment and/or chemicals. Precision Conservation techniques and computer models can be used to conduct some of these assessments.

22

Jorge A. Delgado and Joseph K. Berry

6. Watershed Scale Considerations 6.1. Variable hydrology Qiu et al. (2007) reported that surface runoff is a major contaminant threat to water quality in the USA and proposed that, by incorporating the variable surface area (VSA) hydrology into watershed management practices, we can concentrate our efforts in key areas of the watershed that are the most sensitive. They reported that Hewlett and Hibbert (1967) are credited with the concept of VSAs. Qiu et al. (2007) suggested the need to more closely assess the key management alternatives that will contribute to managing variable source pollution and concluded that Precision Conservation is a good approach to managing this variability. Qiu et al. (2007) reported that managing variable source pollution emphasizes the interconnection between land and water and the different roles varying landscapes play in water resource protection. There is an opportunity to apply these new technologies to address macro- and microscale issues, such as watershed and regional water quality as well as subfield and subwatershed levels (Berry et al., 2003, 2005; Renschler and Lee, 2005). Qiu et al. (2007) reported that the identification of the hydrologically sensitive areas and critical management areas using variable source hydrology will provide the scientific basis for applying Precision Conservation techniques, when applicable. They also reported that it is critical when managing variable source hydrology to also simultaneously assess the area’s temporal variability to identify the most sensitive areas. It is important to consider both the variable source hydrology and the temporal variability to identify those areas that have higher pollution source potential or erosion potentials (Qiu et al., 2007). The same principle applies when trying to identify the areas that will receive the concentrated flows and the season when those concentrated flows will be occurring. It is important to know the sources, flows, and deposition areas to better manage the watershed. It is important to connect these variable flows, at both surface and underground levels to improve management across the watershed (Berry et al., 2003, 2005).

6.2. Models and tools New advances in computer software are allowing for faster integration of information layers used to assess the spatial and temporal flows across the watershed and to identify the best locations for Precision Conservation management practices (Berry et al., 2003, 2005; Dosskey et al., 2005, 2007; Qiu et al., 2007; Renschler and Lee, 2005; Secchi et al., 2007). The initial efforts to identify these spatial erosion impacts by accounting

Advances in Precision Conservation

23

for topography and other parameters were reported by Wheeler (1990), Mitasova et al. (1995), Desmet and Govers (1996), Siegel (1996), Mitas et al. (1997), and Wang et al. (2000). Wischmeier and Smith (1965) took initial steps in this direction by using the Universal Soil Loss Equation (USLE) to calculate average soil losses on slope sections (Wischmeier and Smith, 1965). Several scientists followed this initial effort by using USLE extensively on a watershed scale (Foster and Wischmeier, 1974; Williams and Berndt, 1972; Wilson, 1986). Now we have new models and algorithms that account for spatial erosion variabilities using GIS and Digital Elevation Models (DEMs) (Desmet and Govers, 1996). The use of this new software, integration of layers of information with GIS, remote sensing, and computer modeling was reported by Berry et al. (2003, 2005) as an approach for Precision Conservation, facilitating the identification of variable flows and connecting the flows from field to the watershed. The modeling approach to Precision Conservation was used by Secchi et al. (2007) to assess the effect of management practices across the watershed and how to generate more efficient use of the economical resources to reduce environmental impacts (Secchi et al., 2007). These new models and techniques used by Secchi et al. (2007), Renschler and Lee (2005), Qiu et al. (2007), Dosskey et al. (2005, 2007), and Bonilla et al. (2007) can be used to assess hot spots, identify most susceptible locations, and to implement best management practices for Precision Conservation. Some of the models used to evaluate watersheds are the Agricultural NonPoint Source Pollution (AGNPS) model (Young et al., 1987) and the Soil and Water Assessment Tool (SWAT) model (Arnold et al., 1993). FitzHugh and Mackay (2001) used the SWAT model and reported that data aggregation affected model behavior differently depending on whether the watershed was sediment source limited or transport limited. They concluded that it is important to characterize stream channel processes and to improve the selection of subwatershed size to match SWAT. The AGNPS model (Young et al., 1987) that divides the watershed into small discrete square cells representing variability in agricultural practices was used by Bhuyan et al. (2003) to assess erosion at the watershed scale. They used input parameters such as aspect/flow direction, slope, slope shape, slope length, soil erodibility factor (k-factor), C-factor, conservation practice factor (P-factor), soil texture, fertilizer availability, pesticide indicators, and other parameters. Their approach used sediment yields calculated from a modified USLE (Wischmeier and Smith, 1978) with runoff volume calculated by the SCS-CN method (SCS, 1968). The assessment of chemical movement, runoff, and erosion were calculated using the Agricultural Management Systems (CREAMS, Smith and Williams, 1980). They improved their assessment of topography by using databases that included inputs from the DEM. The inclusion of an RS approach improved the efficiency of the evaluation

24

Jorge A. Delgado and Joseph K. Berry

and reduced the time needed to evaluate the watershed. They concluded that this new RS-GIS modeling process (DEMs) was effective for the calculation of pollutant levels and chemical transport for small watersheds. Other scientists have used this approach of using multiple models and GIS to increase their ability to process several layers of information to assess transport and pollution levels. Renschler and Lee (2005) used three models and GIS to evaluate the effects of BMPs on both short (4–8 years) and long scales (100 years). They used the Water Erosion Prediction Project (WEPP) for hill slope and small watersheds. They also used the Geospatial interface for WEPP (GeoWEPP) in conjunction with GIS databases. They linked GeoWEPP to the SWAT model to assess larger watershed scales. Renschler and Lee (2005) concluded that this approach allows scientists to generate soil loss and sediment yield predictions within a watershed that can be used to detect hot spots for implementation of preferred management options such as spatially distributed BMPs. Bonilla et al. (2007) used the Precision Agricultural-Landscape Modeling System (PALMS) to estimate spatial water erosion in topographically complex landscapes. Bonilla et al. (2007) reported that PALMS can evaluate the effects of local soil properties and microtopography on changes in soil detachment and deposition across short distances. Bonilla et al. (2007) reported that PALMS also has the capabilities to quantify spatial and temporal erosion, deposition, sediment yield, evapotranspiration, soil evaporation, photosynthesis, plant and soil respiration, infiltration, drainage (with and without tiles), crop growth, yield, and other parameters. As we continue to develop capabilities to process multiple layers of information and to calibrate and validate new models that account for surface and underground flows from fields to natural areas, we will be able to improve the capabilities of Precision Conservation practices that minimize environmental impacts and maximize sustainability of increasingly intensive production systems.

6.3. Precision conservation at a watershed scale Precision Conservation principles are directed to use GIS, RS, and other models to handle large sets of information that consider spatial and temporal variability and allow the identification of variable and temporal flows in the environment. This identification informs decisions that can lead to the site-specific implementation of conservation practices that maximize conservation efforts. Secchi et al. (2007) addressed Precision Conservation with the use of computer models to assess the cost of clean water by assessing pollution reduction at a watershed scale. Using simulation models, they assessed the effect and cost of implementing and evaluating conservation practices designed to reduce phosphorus and nitrate levels.

Advances in Precision Conservation

25

Secchi et al. (2007) conducted a simulation suggesting that the cost to reduce nonpoint source pollution levels in Iowa could be very high. They suggested a method by which conservation practices designed to help reduce nutrient and sediment losses can be implemented using model simulations as guides. They identified conservation practices to adopt for their model evaluation based on the physical characteristics of the agricultural lands, in a way similar to the concept of Precision Conservation as described by Berry et al. (2003, 2005). They evaluated the following conservation practices: return of all cropland within 100 ft of a waterway (set aside), retirement of additional land from cropland based on the NRI index, terracing of all land with a slope greater than 7% in western Iowa and land with a slope greater than 5% in the remainder of Iowa, contouring, installation of grass waterways, conversion of significant areas into no-till, and impalement nutrient management planning enacted by reducing N inputs by 10%. Tomer et al. (2007) reported on the spatial patterns of sediment and phosphorus accumulation and flow in a riparian buffer in western Iowa. They proposed that we can use spatial vegetation that accounts for spatial patterns in flow as a mechanism to increase activities such as water use and uptake of nutrients in accordance with the spatial inputs of flows and temporal variability. Tomer et al. (2007) reported that half of the runoff was delivered from mid-April through mid-June. The soil–water phosphorus concentrations at depths of 1.5 m were higher in the riparian zone (where the switch grass was grown) than below the crop areas. The temporal variability when the nutrient flow was higher was during the time when the switch grass (a warm season grass) was not growing. Planting a grass in the lower areas of the buffer that transpires at a higher rate early in the season when sediment is accumulating the fastest contributes to better buffer performance (Tomer et al., 2007). This variable planting of varieties that account for spatial and temporal flows of sediments, nutrients, and water is a useful Precision Conservation technique within these riparian buffers. Strock et al. (2007) reported that appropriately managed ditches can provide an opportunity to manage N and reduce its losses by removing biologically available forms of N via physical and biogeochemical processes. Proper management of ditches, especially deeper ditches, may provide an opportunity to efficiently manage nutrient transport (Strock et al., 2007; Vadas et al., 2007). There are opportunities to use new models, GIS, RS, and GPS techniques to improve the spatial management of ditches across a region. Lowrance et al. (2007) reported on the effects of land use and management on nutrient transport. They found significant differences in runoff between two watersheds, mainly due to land use practices and the use of sediment ponds. They reported that one of the watersheds with sediment ponds using as little as 6.3% of the basin area had significantly cleaner water.

26

Jorge A. Delgado and Joseph K. Berry

The installation of plastic covers on 26% of the area of one of the subwatersheds contributed significantly to higher rates of erosion and off-site transport of sediment and total nitrogen. Lowrance et al. (2007) demonstrated in this study that placement of sediment ponds at strategic places in a watershed could help reduce both the off-site transport of sediment and total nitrogen transport. George et al. (2008) put GPS collars on cows and used supplemental feed to manage and monitor behavior (Fig. 10). George et al. (2008) found that they can use supplemental feed to manage cow behavior in a way that considers forest and grassland areas, temporal variability, and water bodies to enhance soil and water conservation. The results from the George et al. (2008) study show the effectiveness of new technologies to connect animal management with potential animal behavior that reduces environmental impacts. The results from George et al. (2008) are important because they show that management can help reduce the amount of time beef cows spend in riparian areas. These results are in agreement with the results of the studies from Bailey et al. (2001), which report that cattle spend more time and graze more forage within 600 m of supplement sites. Several researchers have shown that there is potential to use supplemental feed to manipulate animal behavior (Bailey, 2003; Bailey and Welling

GPS collar

Water source

Supplemental feed Sandy soils

Clay soils

Figure 10 Potential to use Precision Conservation for animal management and conservation of soil and water. (Adapted from George et al., 2008.)

Advances in Precision Conservation

27

1999, 2002; George et al., 2008; McDougald et al., 1989). George et al. (2008) presented their concept and stated the potential application for Precision Conservation, demonstrating that strategic placement of supplemental feed can be effective for soil and water conservation by reducing grazing in riparian patches and sensitive soil areas. The animal management industry can use precision supplemental feeding practices that take landscape and temporal variability into account through reference to spatial technologies such as GIS to implement conservation practices that result in improved environmental outcomes (George et al., 2008). They proposed that animal managers can contribute to Precision Conservation of soil and water by precisely placing nutrient supplements based on management decisions that consider economic returns, as well as environmental and site-specific factors. They also pointed out that there is even greater potential for continued animal management industry contribution to Precision Conservation because supplemental feeding can be moved as needed to prevent degradation of supplement sites. There are several ecological engineering principles that can be applied for Precision Conservation at a watershed scale. Nutrient farming can be used to reduce N losses to the environment (Hey et al., 2005). Hey et al. (2005) reported that, since about one-third of applied N enters drainage systems, there is potential to improve water quality and reduce losses and impacts of nutrients to rivers using drainage and water management to manage location of wetlands. They reported that we need to develop Nitrogen Trading by which nutrient farmers can use denitrification techniques and trade the reduced N with the surrounding environment. They reported that planners and nutrient managers need to evaluate the field management practices connected with streams, water channels, and nutrient farms (wetlands). For effective nutrient farming, we need to consider that denitrification of N in riparian zones can be an important mechanism for N removal from the system (Hey et al., 2005; Schade et al., 2001; Verchot et al., 1997). We propose that if the concept of Nitrogen Trading develops into viable alternative practices, Precision Conservation practices need to be considered to maximize the managed effectiveness of nutrient harvesting and to minimize transport of nutrients downstream. A spatial/temporal N loss evaluation tool such as NLEAP-GIS can be used to quickly identify the management scenario that shows the greatest potential to maximize the reduction in N losses at the field level and minimize N loss impacts to the environment. This temporal and spatial approach, combined with positive reductions in farm N inputs and other management changes that improve nitrogen use efficiencies, could be used to help identify opportunities to use the Nitrogen Trading Tool (Delgado et al. 2008). We suggest that Precision Conservation can be a key component in identifying Nitrogen Trading opportunities.

28

Jorge A. Delgado and Joseph K. Berry

7. Current Applications and Trends The Berry et al. (2003) publication about Precision Conservation generated enough interest that the Soil Science Society of America, Canadian Soil Science Society, Mexican Soil Science Society, and the Division of Soil Water and Management and Conservation celebrated a joint symposium entitled: ‘‘Precision Conservation in North America’’ at the November 1–4, 2004 annual meeting in Seattle, Washington. A special issue of selected papers was published in the Journal of Soil and Water Conservation (2005). Most recently, the Chinese Academy of Sciences conducted the first International Conference in Precision Conservation of Soil and Water October 22–24, 2007, at Shijiazhuang, China. Several speakers from Asian countries presented recent advances in GPS, GIS, RS, and other new equipment along with their potential applications for Precision Conservation of soil and water. Precision Conservation was also listed as one of the themes for the 9th International Conference in Precision Agriculture, which will be held in July of 2008 in Denver, Colorado. There is increased interest from both North American and international science communities in the use of new technologies for Precision Conservation of soil and water. Additionally, the USDA NRCS is committed to continued advancement in Precision Conservation and to the development of new tools to support Precision Conservation applications (Knight, 2005). Precision Conservation benefits producers by helping them to efficiently manage their operations (Knight, 2005). There are also benefits for taxpayers and for environmental conservation because Precision Conservation can be used to identify hot spots on the farm and throughout the watershed for a more efficient use of agricultural resources (Knight, 2005). Precision Conservation techniques can identify connections of flow from farm areas to the watershed to help us identify the best location to implement conservation practices that reduce environmental impacts while maximizing use of economical resources (Knight, 2005; Secchi et al., 2007). We have the potential to integrate multiple layers of information to assess spatial erosion variability at a field scale (Bonilla et al., 2007; Mueller et al., 2005; Schumacher et al., 2005). Spatial erosion variability reduces yields at site-specific locations within the field. If this variability in erosion across the field is not properly managed, the yields will continue to decline significantly after decades of uniform management (Quine and Zhang, 2002). There is potential to use Precision Conservation to integrate spatial and temporal information to better implement conservation management practices that account for this erosion variability (Mueller et al., 2005; Quine and Zhang, 2002; Schumacher et al., 2005).

Advances in Precision Conservation

29

Precision Conservation can integrate spatial and temporal information to identify and locate the best locations for riparian buffers, grass waterways, ditches, and wetlands within the watershed (Berry et al., 2003, 2005; Dosskey et al., 2002, 2005, 2007; Qiu et al., 2007; Renschler and Lee, 2005; Tomer et al., 2007). We can use these new technologies to connect multiple layers of information for improved rangeland management that integrates spatial variability of resources, including soil and water resources (George et al., 2008). Precision Conservation can also be used to integrate assessment of multiple layers of information that account for temporal variability of hydrology and flows of pollutants (Berry et al., 2003; Dosskey et al., 2007; George et al., 2008; Hey et al., 2005; Qiu et al., 2007; Renschler and Lee, 2005; Secchi et al., 2007). There is potential to use site-specific information of temporal and spatial flows to develop erosion maps that identify highly sensitive areas of the fields or to develop variable hydrology maps that represent variable movement of soil and chemicals across the watershed (Qiu et al., 2007; Schumacher et al., 2005). Table 1 highlights potential-related Precision Conservation practices. Users can integrate spatial and temporal information using GIS and/or models to analyze databases and to develop recommendations for implementation of these conservation practices. For example, we can use models to identify highly sensitive erosion areas of a field which we may then decide to set aside for hay production to reduce the erosion and movement of soil and chemicals out of the field. Alley cropping, conservation crop rotation, cover crops, field borders, riparian herbaceous cover, riparian forest buffers, filter strips, residue management, supplemental feed, sediment ponds, isolated hay production areas with permanent cover, nutrient traps, and buffers are some of the potential conservation practices that may result from an integration of spatial and temporal information about flows and the use of layers of information to develop more effective practices (Table 1). Other nutrient management practices such as remote sensing, site-specific management zones, and Precision Irrigation that contribute to reduced NO3-N leaching were not listed (Delgado and Bausch, 2005; Delgado et al., 2005; Sadler et al., 2005). The practices listed in Table 1 connect the flows from field to natural areas, and contribute to enhanced soil and water conservation. We propose that the application of these conservation practices could be more effective by using new technologies that integrate multiple layers of information spatially and temporally, thereby identifying hot spots. We also propose that the efficiency of these practices could be increased by using variable designs that integrate variable widths, use variable species and varieties, and apply the practices precisely at the hot spots in the field or watershed.

30

Table 1 Potential conservation practices that could be used to manage spatial and temporal variability across the landscape to increase precision conservation of soil and watera

Location

Field

Conservation practice

Alley cropping (CODE 311)

Definition

Precision conservation potential

Trees or shrubs are planted in sets of single or multiple rows with agronomic, horticultural crops or forages produced in the alleys between the sets of woody plants that produce additional products.

There is the potential to use new models, remote sensing, and computer software to integrate spatial and temporal information about flows to develop information that can be used to improve site-specific management decisions for alley cropping locations. Spatial soil properties and underground water tables and flows can also be considered. There is also potential to plant single or multiple drills of trees or shrubs, taking these spatial soil properties into consideration. The varieties of trees or shrubs planted in rows could also be changed based on soil property data that may account for different water tables, salinity levels, and site-specific chemical and physical properties. The objective for this application is to determine the number of rows and appropriate species (considering the species’ water use and nutrient uptake) to match the temporal variability of water flows, water tables,

Field

Conservation crop rotation (CODE 328)

Growing crops in a recurring sequence on the same field.

Field

Cover crop (CODE 340)

Crops including grasses, legumes and forbs for seasonal cover and other conservation purposes.

salinity, and precipitation. There is also potential to use this practice to mine and recover NO3-N leached below the roots of shallower crops in areas where NO3-N leaching presents a problem across the field (Allen et al., 2004; Delgado, 1998, 2001; Rowe et al., 1999; Tomer et al., 2007). There is potential to use different crops to reduce soil erosion considering variation in soil types and variable field erosion. In case where bales of straw are removed from the fields, some areas of the field with lower soil organic matter and/or higher erosion potential could be managed differently with full incorporation of crop residue. In highly saline areas of the fields, a more salt-tolerant conservation crop could be planted to manage saline seeps. There is the potential to use deeply rooted crops in some areas of the field that would filter underground water when planted with shallowly rooted crops (Delgado, 1998, 2001; Schumacher et al., 2005). There are potential to use cover crops in the most sensitive areas of the field and natural areas to reduce soil erosion. Cover crops are highly beneficial in most cases. Cover

31 (continued)

Table 1

(continued)

32 Location

Outside field/ natural area

Conservation practice

Field border (CODE 386)

Definition

A strip of permanent vegetation established at the edge or around the perimeter of a field.

Precision conservation potential

crop use can be differentiated according to the soil type and N cycling. For fields with spatial variability of soil type, cover crops such as winter cover rye and winter wheat (both effective cover crop scavengers of leached NO3-N) may be planted in areas that have high leaching potential. In the areas of the fields with the finer soil texture and lower leaching potential, leguminous cover crops can be planted to increase nitrogen input into the system (Delgado, 1998, 2001). There is the potential to use new models, remote sensing, and computer software to integrate spatial and temporal information about flows to develop information that can be used to identify the location of concentrated flows of water, nutrients, and sediment from the field. This will allow the identification of hot spots and temporal and spatial patterns at the field border. This information can be used to decide how wide to make the vegetative field border, what species to plant, how deep the root systems should be, and how

Outside field/ natural area

Riparian herbaceous cover (CODE 390)

Grasses, grass-like plants, and forbs that are tolerant of intermittent flooding or saturated soils and that are established or managed in the transitional zone between terrestrial and aquatic habitats.

tall and thick the vegetation should be at the surface. Spatial and temporal analysis of the flows coming out of the field could be used to improve the design of field borders that reduce the off-site transport of soil particles, organic matter, chemicals, and water. These field edge practices can maximize farmers’ soil and water conservation, as well as their use of energy and resources. Precision Conservation can determine the best plant type for field borders, whether grass, legumes, or shrubs, considering the potential of each to reduce off-site transport of soil, soil organic matter, and nutrients due to water and wind erosion. These barriers could be precisely designed to eliminate flows from end rows, headlands, and other areas of concentrated flow (Berry et al., 2003; Dosskey et al., 2005; Tomer et al., 2007). There is potential to develop riparian herbaceous cover that accounts for temporal and seasonal site-specific hydrology. There is potential to use models, remote sensing, and computer software to integrate spatial and temporal

33

(continued)

Table 1

(continued)

34 Location

Conservation practice

Definition

Precision conservation potential

information and to develop management decisions about the best location(s) for the riparian herbaceous cover. There is also potential to plant variable species across the riparian and herbaceous cover zones to try to synchronize the vegetation growth and water and nutrient use with periods of maximum water flows across the riparian buffers. These riparian zones can be applied to areas adjacent to perennial and intermittent watercourses or water bodies, accounting for the spatial and temporal hydrology. Site-specific hydrology, including water table data and the potential for concentrated flows in extreme cases, can be factored into these decisions to establish a site-specific riparian herbaceous cover that maximizes water quality, using variable widths and species. There is also potential to use the multiple layers of site-specific information to design the best viable shape of the riparian herbaceous cover to account for variable flows and to identify the best placement to maximize buffer affectivity for soil and

Outside field/ natural area

Riparian forest buffer (CODE 391)

An area comprised predominantly of trees and/or shrubs located adjacent to and upgradient from watercourses or water bodies.

Outside field/ natural area

Filter strip (CODE 393)

A strip or area of herbaceous vegetation situated between cropland, grazing land, or disturbed land (including forestland) and environmentally sensitive areas.

water conservation (Dosskey et al., 2002, 2005, 2005; Hey et al., 2005; Tomer et al., 2007). There is potential to use spatial and temporal information to develop riparian forest buffers that improve and protect water quality by reducing the amount of sediment, nutrient, and surface flows and shallow groundwater chemical movement. There is potential to use the variable hydrology and flow information to identify both the best viable shape for riparian forest buffers to account for variable flows and the best buffer locations for effective management of surface and underground flows (Hey et al., 2005). There is potential to develop filter strips to improve and protect water quality by reducing the amount of sediment and nutrient runoff and movement in surface runoff and shallow groundwater. Sitespecific spatial and temporal information can be used to determine the best locations for filter strips in areas below cropland, grazing land, or disturbed land (including forest land). Filter strips can also be strategically located in areas where

35 (continued)

Table 1

(continued)

36 Location

Field

Conservation practice

Seasonal residue management (CODE 344)

Animal Supplemental systems feed

Definition

Managing the amount, orientation, and distribution of crop and other plant residues on the soil surface during a specified period of the year, while planting annual crops on a clean-tilled seedbed, or while growing biennial or perennial seed crops.

Use of supplemental feed to manage cow behavior in a way that considers forest and grassland areas, temporal variability, and

Precision conservation potential

sediment, particulate matter, and/or dissolved concentrated contaminants may be leaving and entering environmentally sensitive areas. These filter strip areas need to be comprised of permanent vegetation, and fully established prior to the first irrigation. This site-specific information can also be used to design the best viable shape for the filter strips (Tomer et al., 2007). There is potential to spatially manage residue to reduce erosion from the most sensitive areas of the field. There is potential to concentrate residue in those areas that are more susceptible to erosion. In case where bales of straw are removed from the fields, some areas of the field with lower soil organic matter and/or higher erosion potential could be managed differently with full incorporation of crop residue (Schumacher et al., 2005). There is potential to use practice to maximize carbon sequestration. There is potential to use Precision Conservation techniques to strategically place supplemental feed to manipulate

water bodies to enhance soil and water conservation.

Outside field/ natural area

Sediment ponds

Use of sediment ponds to reduce the movement of soil and chemicals.

Field

Set aside hay areas with permanent cover

Use of set aside hay areas with permanent cover.

animal behavior. Strategic placement of supplemental feed can be effective for soil and water conservation by reducing grazing in riparian patches and sensitive soil areas. Animal managers can take landscape and temporal variability into account through reference to spatial technologies such as geographic information systems (GIS) to implement supplemental feeding-based conservation practices that result in improved environmental outcomes. Supplemental feeding sites can be moved as needed to prevent site degradation (George et al., 2008). There is potential to use sediment ponds to reduce the movement of soil and nutrients from fields and from subwatersheds. There is potential to strategically place these ponds taking variable hydrology and flows into account (Lowrance et al., 2007). Spatial assessment of field erosion and development of maps can be used to identify highly sensitive areas of fields. There is potential to manage the most erosion-sensitive areas by setting aside areas for hay production to reduce the erosion and movement of soil and

37

(continued)

Table 1

(continued)

Location

Field/ natural area

a

Conservation practice

Nutrient traps

Definition

Installation of nutrient traps or denitrification traps to remove nutrients from field outflows.

Precision conservation potential

chemicals out of the field (Schumacher et al., 2005). Spatial assessment of field erosion and variable hydrology can be used for development of maps to identify areas with higher flows of phosphorus and nitrates in fields and/or field borders and natural areas. There is the potential to use phosphorus sorbing materials (PSMs) to decrease the potential for off-site transport of phosphorus in runoff water. There is also potential to use denitrification traps to reduce NO3-N concentrations in runoff water or underground water flows (Hunter, 2001; Penn et al., 2007).

Some of these are practices recommended by USDA-NRCS. We suggest that there is potential to apply the concepts of Precision Conservation to these USDANRCS-recommended practices and to other practices included in this report. We suggest that these practices can be implemented site specifically in fields and natural areas by using layers of information that identify hot spots across the landscape. We also suggest that there is potential to use models, map and data analysis software, and Precision Conservation techniques to modify these practices, taking the site-specific spatial and temporal information about flows into consideration. There is also potential to implement these practices using different device shapes and/or species to better account for variable spatial and temporal hydrology and flows. We suggest that there is significant potential to develop new practices for Precision Conservation of soil and water, such as the integration of animal behavior management with soil and water conservation.

Advances in Precision Conservation

39

8. Summary and Conclusions There are multiple examples of advances in Precision Conservation published during the last 4 years showing how new spatial technologies and the integration of GPS, GIS, RS, and models can be applied to improve management decisions that contribute to Precision Conservation of soil and water. Precision Conservation can more precisely identify where to locate riparian buffers, sediment ponds, nutrient management farms, and other ecological engineering practices to most effectively reduce environmental impacts from hot spots across the watershed. These technologies can be used to simultaneously consider variable hydrology and temporal flows to identify the best locations for the implementation of conservation practices at the watershed and subwatershed levels. These technologies can also be used to design better buffers to manage flows at field borders, to identify the best locations for phosphorus recovery devices, and to locate potential denitrification trap sites. These new approaches can contribute to better management of variable surface and underground flows across grass waterways, buffers, riparian buffers, ditches, wetlands, and watersheds. New advances even show that there is potential to integrate management of rangeland animal behavior with management practices that account for spatial and temporal variability to enhance Precision Conservation of soil and water resources. With continued increases in population growth and increased demands of land resources for food and biofuel production, maximizing agricultural production is increasingly necessary. Precision Conservation can be used to synchronize best management practices that maximize yields while reducing unnecessary inputs and losses of sediment and other chemicals to the environment. We propose that, as new technological advances continue to emerge, adaptations of Precision Conservation by land owners, managers, farmers, and extension personnel will be widely implemented. These new technologies can contribute to higher efficiency of resource management, economical returns, and environmental sustainability (Berry et al., 2003; Knight, 2005; Secchi et al., 2007). As new advances in computer models, remote RS, GIS, and GPS continue, these technologies will become increasingly accessible for the conservation of agricultural and natural resources. Precision Conservation will play a significant role in maximizing and sustaining agricultural yields while contributing to global sustainability in the 21st century.

REFERENCES Allen, S. C., Shibu, J., Nair, P. K. R., Brecke, B. J., Nkedi-Kizza, P., and Ramsey, C. L. (2004). Safety-net role of tree roots: Evidence from a pecan (Carya illinoensis K. Koch)-cotton

40

Jorge A. Delgado and Joseph K. Berry

(Gossypium hirsutum L.) alley cropping system in the southern United States. For. Ecol. Manag. 192(2–3), 395–407. Arnold, J. G., Allen, P. M., and Bernhardt, G. (1993). A comprehensive surfacegroundwater flow model. J. Hydrol. 142, 47–69. Bailey, D. W. (2003). Managementstrategies for optimal grazing distribution and use of arid rangelands. In ‘‘Proceedings of the Forage Strategies for Arid Climates Symp,’’ Journal Series No. 2003–31, pp. E147–E153. Montana Agricultural Experiment Station, Bozeman, Montana. Bailey, D. W., and Welling, G. R. (1999). Modification of cattle grazing distribution with dehydrated molasses supplement. J. Range Manage. 52, 575–582. Bailey, D. W., and Welling, G. R. (2002). Comparisonof low-moisture molasses blocks and loose dry mineral mixes as delivery systems for supplementing trace minerals to rangeland cattle. In ‘‘Proceedings of the Annual Meeting on Society Range Management,’’ 55th, p.113. Kansas City, Missouri. Bailey, D. W., Welling, G. R., and Miller, E. T. (2001). Cattle use of foothills rangeland near dehydrated molasses supplement. J. Range Manage. 54, 338–347. Bausch, W. C., and Delgado, J. A. (2003). Ground-based sensing of plant nitrogen status in irrigated corn to improve nitrogen management. In ‘‘Digital Imaging and Spectral Techniques: Applications to Precision Agriculture and Crop Physiology’’ (T. VanToai, D. Major, M. McDonald, J. Schepers, and L. Tarpley, Eds.), pp. 151–163. ASA Spec. Publ. 66. ASA, CSSA, and SSSA, Madison, WI. Berry, J. K. (1999). GIS technology in environmental management: A brief history, trends and probable future. In ‘‘Handbook of Global Environmental Policy and Administration’’ (D. L. Soden and B. S. Steel, Eds.), pp. 49–76. Marcel Dekker Publishers, New York. Berry, J. K. (2002). Quantitative methods for analyzing map similarity and zoning. In ‘‘Proceedings of GeoTech Conference on Geographic Information Systems,’’ 8–11 April. Toronto, Ontario, Canada. Berry, J. K. (2003a). ‘‘Map Analysis: Procedures and Applications in GIS Modeling.’’ BASIS Press, Fort Collins, CO. Berry, J. K. (2003b). Analyzing spatial content. In ‘‘Analyzing Precision Ag Data: A Handson Case Study in Spatial Analysis and Data Mining’’ ( J. K. Berry, Ed.), pp. 55–63. BASIS Press, Fort Collins, CO. Berry, J. K. (2007a). GIS innovation drives its evolution. GeoWorld 20(8), 14–15. Berry, J. K. (2007b). Geo-referencing is the cornerstone of GIS. GeoWorld 20(4), 14–15. Berry, J. K., Delgado, J. A., Khosla, R., and Pierce, F. J. (2003). Precision conservation for environmental sustainability. J. Soil Water Conserv. 58, 332–339. Berry, J. K., Delgado, J. A., Pierce, F. J., and Khosla, R. (2005). Applying spatial analysis for precision conservation across the landscape. J. Soil Water Conserv. 60, 363–370. Bhuyan, S. J., Marzen, L. J., Koelliker, J. K., Harrington, J. A., Jr., and Barnes, P. L. (2003). Assessment of runoff and sediment yield using remote sensing, GIS, and AGNPS. J. Soil Water Conserv. 57, 351–364. Bonilla, C. A., Norman, J. A., and Molling, C. C. (2007). Water erosion estimation in topographically complex landscapes: Model description and first verifications. Soil Sci. Soc. Am. J. 71, 1524–1537. Cabot, P. E., Pierce, F. J., Nowak, P., and Karthikeyan, K. G. (2006). Monitoring and predicting manure application rates using precision conservation technology. J. Soil Water Conserv. 61, 282–292. Corwin, D. L., Loague, K., and Ellsworth, T. R. (1998). GIS-based modeling of non-point source pollutants in the vadose zone. J. Soil Water Conserv. 53, 34–38. Delgado, J. A. (1998). Sequential NLEAP simulations to examine effect of early and late planted winter cover crops on nitrogen dynamics. J. Soil Water Conserv. 53, 241–244.

Advances in Precision Conservation

41

Delgado, J. A. (2001). Use of simulations for evaluation of best management practices on irrigated cropping systems. In ‘‘Modeling Carbon and Nitrogen Dynamics for Soil Management’’ (M. J. Shaffer, L. Ma, and S. Hansen, Eds.), pp. 355–381. Lewis Publishers, Boca Raton, FL. Delgado, J. A., and Bausch, W. (2005). Potential use of precision conservation techniques to reduce nitrate leaching in irrigated crops. J. Soil Water Conserv. 60, 379–387. Delgado, J. A., and Mosier, A. R. (1996). Mitigation alternatives to decrease nitrous oxides emissions and urea-nitrogen loss and their effect on methane flux. J. Environ. Qual. 25, 1105–1111. Delgado, J. A., Ristau, R. J., Dillon, M. A., Duke, H. R., Stuebe, A., Follett, R. F., Shaffer, M. J., Riggenbach, R. R., Sparks, R. T., Thompson, A., Kawanabe, L. M., Kunugi, A., et al. (2001). Use of innovative tools to increase nitrogen use efficiency and protect environmental quality in crop rotations. Commun. Soil Sci. Plant Anal. 32, 1321–1354. Delgado, J. A., Khosla, R., Bausch, W. C., Westfall, D. G., and Inman, D. (2005). Nitrogen fertilizer management based on site-specific management zones reduce potential for NO3-N leaching. J. Soil Water Conserv. 60, 402–410. Delgado, J. A., Shaffer, M. J., Lal, H., Mckinney, S. P., Gross, C. M., and Cover, H. (2008). Assessment of nitrogen losses to the environment with a Nitrogen Trading Tool (NTT). Comput. Electron. Agric. (In Print). Desmet, P. J. J., and Govers, G. (1996). A GIS procedure for automatically calculating the USLE LS factor on topographic complex landscape units. J. Soil Water Conserv. 51, 427–433. Dosskey, M. G., Helmers, M. J., Eisenhauer, D. E., Franti, T. G., and Hoagland, K. D. (2002). Assessment of concentrated flow through riparian buffers. J. Soil Water Conserv. 57, 336–344. Dosskey, M. G., Eisenhauer, D. E., and Helmers, M. J. (2005). Establishing conservation buffers using precision information. J. Soil Water Conserv. 62, 349–354. Dosskey, M. G., Helmers, M. J., and Eisenhauer, D. E. (2007). An approach for using soil surveys to guide the placement of water quality buffers. J. Soil Water Conserv. 61, 344–354. Feng, G., and Sharratt, B. (2007). Scaling from field to region for wind erosion prediction using the wind erosion prediction system and geographical information system. J. Soil Water Conserv. 62, 321–328. FitzHugh, T. W., and Mackay, D. S. (2001). Impact of subwatershed partitioning on modeled source- and transport-limited sediment yields in an agricultural nonpoint source pollution model. J. Soil Water Conserv. 56, 137–143. Fleming, K. L., Westfall, D. G., Wiens, D. W., Rothe, L. E., Cipra, J. E., and Heermann, D. F. (1999). Evaluating farmer developed management zone maps for precision farming. In ‘‘Proceedings of the International Conference on Precision Agriculture,’’ 4th, 19–22 July 1998 (P. C. Robert, R. H. Rust, and W. E. Larson, Eds.), pp. 335–343. ASA, Madison, WI. Foster, G. R., and Wischmeier, W. H. (1974). Evaluating irregular slopes for soil loss prediction. J. Trans. ASAE 17, 305–309. George, M. R., McDougald, N. K., Jensen, W. A., Larsen, R. E., Cao, D. C., and Harris, N. R. (2008). Effectiveness of nutrient supplement placement for changing beef cow distribution. J. Soil Water Conserv. 63, 11–17. Goddard, T. W. (2005). An overview of Precision Conservation in Canada. J. Soil Water Conserv. 62, 456–461. Hall, M. D., Shaffer, M. J., Waskom, R. M., and Delgado, J. A. (2001). Regional nitrate leaching variability: What makes a difference in Northeastern Colorado. J. Am. Water Resour. Assoc. 37, 139–150.

42

Jorge A. Delgado and Joseph K. Berry

Hatch, L. K., Mallawatantri, A., Wheeler, D., Gleason, A., Mulla, D., Perry, J., Easter, K. W., Smith, R., Gerlach, L., and Brezonik, P. (2001). Land management at the major watershed-agroecoregion intersection. J. Soil Water Conserv. 56, 44–51. Hatfield, J. L., and Prueger, J. H. (2004). Impacts of changing precipitation patterns on water quality. J. Soil Water Conserv. 59, 51–58. Hewlett, J. D., and Hibbert, A. R. (1967). Factors affecting the response of small watersheds to precipitation in humid regions. In ‘‘Forest Hydrology’’ (W. E. Sopper and H. W. Lull, Eds.), pp. 275–290. Pergamon Press, Oxford. Hey, D. L., Urban, L. S., and Kostel, J. A. (2005). Nutrient farming: The business of environmental management. Ecol. Eng. 24, 279–287. Hunter, W. (2001). Remediation of drinking water for rural populations. In ‘‘Nitrogen in the Environment’’ (R. F. Follett and J. L. Hatfield, Eds.), pp. 433–453. CRC Press, New York, NY. Khosla, R., Fleming, K., Delgado, J., Shaver, T., and Westfall, D. (2002). Use of site-specific management zones to improve nitrogen management for precision agriculture. J. Soil Water Conserv. 57, 513–518. Kitchen, N. R., Sudduth, K. A., Myers, D. B., Massey, R. E., Sadler, E. J., Lerch, R. N., Hummel, J. W., and Palm, H. L. (2005). Development of a conservation-oriented precision agriculture system: II. Crop production assessment and plan implementation. J. Soil Water Conserv. 62, 421–430. Knight, B. L. (2005). Precision conservation. J. Soil Water Conserv. 60, 137A. Koch, B., Khosla, R., Frasier, M., and Westfall, D. G. (2003). Economic feasibility of variable-rate nitrogen application in site specific management. In ‘‘Proceedings of the Western Nutrient Management Conference,’’ 6–7 March 2003, Vol 5. pp. 107–112. Salt Lake City, UT. Lal, R. (1995). Global soil erosion by water and carbon dynamics. In ‘‘Soils and Global Change’’ (R. Lal, J. Kimble, E. Levine, and B. A. Stewart, Eds.), pp. 131–140. Lewis Publishers, Boca Raton, FL. Lal, R. (2000). A modest proposal for the year 2001: We can control greenhouse gases and feed the world . . . with proper soil management. J. Soil Water Conserv. 55, 429–433. Lerch, R. N., Kitchen, N. R., Kremer, R. J., Donald, W. W., Alberts, E. E., Sadler, E. J., Sudduth, K. A., Myers, D. B., and Ghidey, F. (2005). Development of a conservationoriented precision agriculture system: I. Water and soil quality assessment. J. Soil Water Conserv. 62, 411–421. Lowrance, R., Altier, L. S., Williams, R. G., Inamdar, S. P., Sheridan, J. M., Bosch, D. D., Hubbard, R. K., and Thomas, D. L. (2000). REMM: The riparian ecosystem management model. J. Soil Water Conserv. 55, 27–34. Lowrance, R., Sheridan, J. M., Williams, R. G., Bosch, D. D., Sullivan, D. G., Blanchett, D. R., Hargett, L. M., and Clegg, C. M. (2007). Water quality and hydrology in farm-scale coastal plain watersheds: Effects of agriculture, impoundments, and riparian zones. J. Soil Water Conserv. 62, 65–76. McDougald, N. K., Frost, W. E., and Jones, D. E. (1989). Use of supplemental feeding locations to manage cattle use of riparian areas of hardwood rangelands. In ‘‘Proceedings of the California Riparian Systems Conference: Protection, Management, and Restoration for the 1990s,’’ General Tech. Rep. PSW-110 (D. L. Abell, Ed.). Pacific Southwest Forest and Range Experiment Station, USDA Forest Service, Berkeley, CA. Meisinger, J. J., and Delgado, J. A. (2002). Principles for managing nitrogen leaching. J. Soil Water Conserv. 57, 485–498. Mitas, L., Brown, W. M., and Mitasova, H. (1997). Role of dynamic cartography in simulations of landscape processes based on multivariate fields. Comput. Geosci. 23, 437–446.

Advances in Precision Conservation

43

Mitasova, H., Mitas, L., Brown, W. M., and Gerdes, D. P. (1995). Modeling spatially and temporally distributed phenomena: New methods and tools for GRASS GIS. Int. J. Geogr. Inf. Sci. 9, 433–446. Mosier, A. R., Parton, W. J., Valentine, D. W., Ojima, D. S., Schimel, D. S., and Delgado, J. A. (1996). CH4 and N2O fluxes in the Colorado shortgrass steppe: 1. Impacts of landscape and nitrogen addition. Global Biogeochem. Cycles 10, 387–399. Mueller, T. G., Cetin, H., Fleming, R. A., Dillon, C. R., Karathanasis, A. D., and Shearer, S. A. (2005). Erosion probability maps: Calibrating precision agriculture data with soil surveys using logistic regression. J. Soil Water Conserv. 62, 462–468. Nearing, M. A., Pruski, F. F., and O’Neal, M. R. (2004). Expected climate change impacts on soil erosion rates: A review. J. Soil Water Conserv. 59, 43–50. Penn, C. J., Bryant, R. B., Kleiman, P. J. A., and Allen, A. L. (2007). Removing dissolved phosphorous from drainage ditch water with phosphorous sorbing materials. J. Soil Water Conserv. 6, 269–276. Pennock, D. J. (2005). Precision conservation for co-management of carbon and nitrogen on the Canadian prairies. J. Soil Water Conserv. 62, 396–401. Peterson, P. (2007). Personal communication on extending map algebra on multidimensional tessellations using the Rhombic Dodecahedron and A3 Lattice as expressed in the PYXIS digital earth reference model. (Website at www.pyxis.com.au). Peterson, A., and Vondracek, B. (2006). Water quality in relation to vegetative buffers around sinkholes in karst terrain. J. Soil Water Conserv. 61, 380–390. Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., Crist, S., Shpritz, L., Fitton, L., Saffouri, R., and Blair, R. (1995). Environmental and economic cost of soil erosion and conservation benefits. Science 267, 1117–1123. Qiu, Z., Walter, M. T., and Hall, C. (2007). Managing variable source pollution in agricultural watersheds. J. Soil Water Conserv. 62, 115–122. Quine, T. A., and Zhang, Y. (2002). An investigation of spatial variation in soil erosion, soil properties, and crop production within an agricultural field in Devon, United Kingdom. J. Soil Water Conserv. 57, 55–64. Renschler, C. S., and Lee, T. (2005). Spatially distributed assessment of short- and long-term impacts of multiple best management practices in agricultural watersheds. J. Soil Water Conserv. 62, 446–456. Rowe, E. C., Hairiah, K., Giller, K. E., van Noordwijk, M., and Cadisch, G. (1999). Testing the safety-net role of hedgerow tree roots by 15N placement at different soil depths. Agrofor. Syst. 4, 81–93. Sadler, E. J., Evans, R. G., Stone, K. C., and Camp, C. R. (2005). Opportunities for conservation with precision irrigation. J. Soil Water Conserv. 62, 371–379. Schade, J. D., Fischer, S. G., Grimm, N. B., and Seddon, J. A. (2001). The influence of a riparian shrub on nitrogen cycling in a Sonoran Desert stream. Ecology 82, 3363–3376. Schumacher, J. A., Kaspar, T. C., Ritchie, J. C., Schumacher, T. E., Karlen, D. L., Ventris, E. R., McCarty, G. M., Colvin, T. S., Jaynes, D. B., Lindstrom, M. J., and Fenton, T. E. (2005). Identifying spatial patterns of erosion for use in precision conservation. J. Soil Water Conserv. 62, 355–362. Secchi, S., Gassman, P. W., Jha, M., Kurkalova, L., Feng, H. H., Campbell, T., and Kling, C. L. (2007). The cost of cleaner water: Assessing agricultural pollution reduction at the watershed scale. J. Soil Water Conserv. 62, 10–21. Sharpley, A. N., Herron, S., and Daniel, T. (2007). Overcoming the challenges of phosphorous-based management in poultry farming. J. Soil Water Conserv. 62, 375–389. Shuster, W. D., Gehring, R., and Gerken, J. (2007). Prospects for enhanced groundwater recharge via infiltration of urban storm water runoff: A case study. J. Soil Water Conserv. 62, 129–137.

44

Jorge A. Delgado and Joseph K. Berry

Siegel, S. B. (1996). ‘‘Evaluation of Land Value Study.’’ Stud. Rep. CAA-SR-96–5. U.S. Army Concepts Analysis Agency, U.S. Gov. Print. Office, Washington, DC. Smith, R. E., and Williams, J. R. (1980). Simulation of the surface hydrology. In ‘‘CREAMS: A Field Scale Model for Chemical Runoff, and Erosion from Agricultural Management Systems’’ (W. Knisel, Ed.). Conserv. Res. Rep. 26, USDA-ARS, Washington, DC. Smith, T. A., Osmond, D. L., and Gilliam, J. W. (2006). Riparian buffer and nitrate removal in a lagoon-effluent irrigated agricultural area. J. Soil Water Conserv. 61, 273–281. Soil Conservation Service (SCS). (1968). ‘‘Hydrology.’’ Supplement A, Sec. 4. Eng. Handb. USDA-SCS, Washington, DC. Strock, J. S., Dell, C. J., and Schmidt, J. P. (2007). Managing natural processes in drainage ditches for nonpoint source nitrogen control. J. Soil Water Conserv. 62, 188–196. Tomer, M. D., Moorman, T. B., Kovar, J. L., James, D. E., and Burkart, M. R. (2007). Spatial patterns of sediment and phosphorous in a riparian buffer in western Iowa. J. Soil Water Conserv. 62, 329–338. Vadas, P. A., Srinivasan, M. S., Kleinman, P. J. A., Schmidth, J. P., and Allen, A. L. (2007). Hydrology and groundwater nutrient concentrations in a ditch-drained agroecosystem. J. Soil Water Conserv. 62, 178–188. Verchot, L. V., Franklin, E. C., and Gilliam, J. W. (1997). Nitrogen cycling in Piedmont vegetated filter zones: I. Surface soil processes. J. Environ. Qual. 26, 327–336. Wang, G., Gertner, G., Parysow, P., and Anderson, A. B. (2000). Spatial prediction and uncertainty analysis of topographic factors for the revised universal soil loss equation (RUSLE). J. Soil Water Conserv. 55, 374–384. Wheeler, P. H. (1990). An innovative county soil erosion control ordinance. J. Soil Water Conserv. 45, 374–378. Williams, J. R., and Berndt, H. D. (1972). Sediment yield computed with universal equation: Proc. of the ASCE. J. Hydraul. Div. 98, 2087–2098. Wilson, J. P. (1986). Estimating the topographic factor in the universal soil loss equation for watersheds. J. Soil Water Conserv. 41, 179–184. Wischmeier, W. H., and Smith, D. D. (1965). ‘‘Predicting Rainfall Erosion Losses from Cropland East of the Rocky Mountains.’’ USDA-ARS. Agric. Handb. 282. U.S. Gov. Print. Office, Washington, DC. Wischmeier, W. H., and Smith, D. D. (1978). ‘‘Predicting Rainfall Erosion Losses.’’ Agric. Handb. 537. USDA-ARS, U.S. Gov. Print. Office, Washington, DC. Wylie, B. K., Shaffer, M. J., Brodahl, M. K., Dubois, D., and Wagner, D. G. (1995). Predicting spatial distributions of nitrate leaching in northeastern Colorado. J. Soil Water Conserv. 49, 288–293. Young, R. A., Onstad, C. A., Bosch, D. D., and Anderson, W. P. (1987). ‘‘AGNPS: Agricultural Nonpoint Source Pollution Model: A Large Watershed Analysis Tool.’’ Conserv. Res. Rep. 35. USDA-ARS, U.S. Gov. Print. Office, Washington, DC.

C H A P T E R

T W O

Reaction and Transport of Arsenic in Soils: Equilibrium and Kinetic Modeling Hua Zhang and H. M. Selim Contents 46 47 48 48 49 51 52 52 56 57 60 62 64 66 69 73 73 78 81 81 84 86 90 97 99 101 104 105

1. Introduction 2. Environmental Toxicity 3. Arsenic in Soils 3.1. Background concentrations 3.2. Anthropogenic sources 3.3. Speciation 4. Biogeochemistry 4.1. Retention mechanisms 4.2. pH dependency 4.3. Effect of solution composition 4.4. Sorption kinetics 4.5. Desorption 4.6. Reaction with sulfides 4.7. Heterogeneous oxidation 4.8. Microbial-mediated reduction and oxidation 5. Transport in Soils 5.1. Transport mechanisms 5.2. Mobility under field conditions 6. Modeling 6.1. Equilibrium thermodynamic models 6.2. Empirical equilibrium models 6.3. Surface complexation models 6.4. Kinetic models 6.5. Transport models 6.6. Field application 7. Remediation of Contaminated Soils 8. Summary and a Look Ahead References School of Plant, Environmental and Soil Sciences, Louisiana State University, Baton Rouge, Louisiana 70803 Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00202-2

#

2008 Elsevier Inc. All rights reserved.

45

46

Hua Zhang and H. M. Selim

Arsenic contamination of the soil and groundwater poses great risk to human and animal health. There is a growing public interest in developing risk assessment framework, environment regulations, and remedial strategies for protecting ecosystems and human from arsenic poisoning. Although extensive research efforts have been made over the past four decades, the prediction of the fate and transport of arsenic in soils are often inaccurate due to the complex biogeochemical reactions of various arsenic species in soil and water environments. In-depth knowledge of factors that influence the behavior of arsenic in aqueous and solid phases are critical in making accurate determinations of the mobility, bioavailability, and toxicity of arsenic in the soil root zone. In this contribution, we present a review of the state of knowledge on reactions and transport of arsenic in soils with emphasis on modeling of the physical, chemical, and biological interactions of arsenic in soil environment. Specifically, we present an overview of (i) biogeochemical mechanisms of arsenic adsorption–desorption, oxidation– reduction, and precipitation–dissolution; (ii) reactive transport mechanisms of arsenic in the natural environment as affected by factors including arsenic species, redox potential, solution chemistry, flow regime, and colloid-facilitated transport; and (iii) equilibrium and kinetic modeling approaches to simulating the geochemical reactions and transport mechanisms of arsenic in porous media. A range of remedial technologies have been reviewed and their effectiveness and feasibility in the removal or in situ stabilization of arsenic in contaminated soils are discussed. Future research needs are also outlined.

1. Introduction Arsenic (As) is a highly toxic element widely present in soils, plants, and water at trace levels. Increasing amounts of arsenic are being introduced into soil and water environments as a result of natural and anthropogenic processes. The U.S. Environmental Protection Agency (USEPA) classified arsenic as a human carcinogen contaminant and lowered the maximum contaminant level (MCL) in drinking water from 50 ppb to 10 ppb (USEPA, 2001). Arsenic concentrations in groundwater above the environmental standard have been observed in many countries including Bangladesh, India, Vietnam, China, and United States, and were attributed to geologic or anthropogenic sources (Nordstrom, 2002; Smedley and Kinniburgh, 2002). In addition, increased use of arsenic-containing compounds, such as pesticides, herbicides, wood preservatives, and livestock feed additives, have introduced large amounts of arsenic into the soil system (Smith et al., 1998). Contamination of ground and surface water by arsenic from soils and aquifers pose significant threat to human health (WHO, 2004). The occurrence of arsenic in groundwater of Bangladesh, resulting from the dissolution of arsenic containing aquifer material, caused massive poisoning of the people living in that area (BGS, 2001). In United States,

Reaction and Transport of Arsenic in Soils

47

arsenic is a contaminant of concern (COC) at 568 Superfund sites, making it the second most common inorganic contaminant on the National Priority List (USEPA, 2002). The thorough understanding of the fate of arsenic in soil environment is urgently required for environment risk assessment and remediation plan. Recently, extensive research efforts were devoted to unraveling the complex geochemical reactions of arsenic in natural environment. This is reflected by the large volume of literatures published in this area over the past three decades. Numerous laboratory and field studies demonstrated that complex chemical (e.g., adsorption–desorption, oxidation–reduction, precipitation–dissolution), physical (e.g., advection and dispersion), and biological (e.g., biotransformation) processes are involved in regulating the behavior of arsenic in soil, sediment, aquifer, and the surface water environment (Smedley and Kinniburgh, 2002). While our understanding of arsenic cycle was greatly improved in the last several decades, the influence of various environmental factors and their combinations on the retention and transport of arsenic have not been fully explored. Prediction of the mobility of arsenic at contaminated sites was impeded by the lack of knowledge in the hydrogeochemical processes governing arsenic speciation, retention, release, and transport in subsurface. The heterogeneous natural of the geological materials multiplied the complexity of predicting the fate of arsenic existed or released in natural environment. This literature review highlights recent scientific advances toward the understanding of the fate and behavior of arsenic in the soil environment. We present overview of laboratory and field observations and discuss equilibrium and kinetic approaches for describing arsenic retention and transport soils. In addition, we briefly discussed various techniques employed in the remediation of arsenic contaminated sites and identified future research needs.

2. Environmental Toxicity The toxicity of arsenic depends on its chemical form. Organic arsenic compounds are much less toxic than inorganic arsenic. Among the inorganic arsenic, arsine gas (AsH3) is the most toxic form. However, arsine gas rarely exists in the natural environment. Two dominant arsenic forms in the environment, arsenate and arsenite, are highly toxic. As a molecular analog of phosphate, arsenate blocks oxidative phosphorylation, short-circuiting life’s main energy generation system. Arsenite is even more toxic by binding to sulfhydryl groups, impairing the function of many proteins (Oremland and Stolz, 2003). The targets of arsenic toxicity are the respiratory system, the circulatory system, the skin, the nervous system, and the reproductive system, among

48

Hua Zhang and H. M. Selim

others. Acute arsenic poisoning affects the central nervous system, blood vessels, kidney, and can cause death in 1–3 days (Reigart and Roberts, 1999). Drinking water rich in arsenic over a long period leads to arsenic poisoning or arsenicosis. The health effects of arsenicosis include skin problems (such as color changes on the skin, and hard patches on the palms and soles of the feet), skin cancer, cancers of the bladder, kidney and lung, diseases of the blood vessels of the legs and feet, and possibly also diabetes, high blood pressure and reproductive disorders (WHO, 2004). The USEPA classified arsenic as Group A (human carcinogen) contaminant. Several incidence of arsenic poisoning have been reported in Bangladesh (Nickson et al., 1998), India (Acharyya et al., 1999), Vietnam (Berg et al., 2001), China (Smedley and Kinniburgh, 2002), Taiwan (Chen et al., 1994), and United States (USGS, 2004). The culprit of arsenic poisoning in those range from arsenic in drinking water from geological sources, arsenic released from industrial sources and mining activities, or arsenic in contaminated food (Mandal and Suzuki, 2002). Arsenic enters human body via respiration of arsenic in dust and fumes and ingestion of arsenic in water, soil, and food (Mandal and Suzuki, 2002). The air exposure of arsenic is generally low. However, combustion of arsenic-containing coal may result in locally high arsenic levels in some areas. In general, drinking arsenic contaminated water is the major route of arsenic poisoning around the world. Millions of people suffer from the toxic effects due to drinking of arsenic-rich groundwater (Smedley and Kinniburgh, 2002). Soil ingestion is another important pathway of arsenic poisoning, especially for children (Rodriguez et al., 1999). Arsenic may accumulate in crops, vegetables, and fruits grown on contaminated soil (Meharg and Hartley-Whitaker, 2002). Consumption of arsenic polluted food is thus another serious threat to human health.

3. Arsenic in Soils 3.1. Background concentrations The main sources of arsenic in soils come from arsenic containing parent material. The mean value of arsenic abundance in crystal rocks is around 2 mg kg1 with considerable variance between different types of rocks. Arsenic minerals commonly occurred in the environments are arsenopyrite (FeAsS), orpiment, realgar, arsenides, enargite, colbaltite, and proustite. The weathering of those arsenic containing minerals brings dissolved arsenic into soil environment and subsequently adsorbed or precipitated on mineral surfaces. Background concentrations of arsenic in soils vary among soil types, depending on the parent materials from which the soil is derived. According

Reaction and Transport of Arsenic in Soils

49

to the National Geochemical Survey conducted by The U.S. Geological Survey (USGS), arsenic concentrations of most soils in the United States are well below 10 mg kg1 (USGS, 2004). Researchers reported that background arsenic concentrations of soils in Australian and New Zealand is 0.2–30 mg kg1, and they suggested environmental investigation for the concentrations greater than 20 mg kg1 (Barzi et al., 1996). The study of Bradford et al. (1996) showed that the geometric mean of arsenic concentration in 50 soils from California was 2.8 mg kg1, with a range of 0.6–11.0 mg kg1. Similarly, Chen et al. (2002) reported that the mean arsenic concentration of Florida soils was 0.42 mg kg1, ranging from 0.01 to 50.6 mg kg1, with considerable difference between soil types. The survey of soils in Mississippi found the mean arsenic concentration was 8.25 mg kg1, with a range of 0.26–24.43 mg kg1 (Pettry and Switzer, 2001). It is observed that soil arsenic concentration may correlate with clay content, pH, cation exchange capacity, organic matter content, and most significantly Fe and Al concentration (Bradford et al., 1996; Chen et al., 2002; Ori et al., 1993; Pettry and Switzer, 2001).

3.2. Anthropogenic sources Arsenic is frequently associated with various types of mineral deposits, especially sulfide ore (Foster et al., 1968ab; Paktunc et al., 2003, 2004). The mining process of Pb, Cu, Zn, Co, Ni, and Au often produces tailing of high residual arsenic concentrations due to the presence of arsenic minerals in the ores, such as FeAsS, arsenolite (As2O3), olivenite (Cu2OHAsO4), mimetite (Pb5Cl(AsO4)3, and cobaltite (CoAsS). Soil arsenic concentrations near the mining dump site are reported as high as 30,000 mg kg1, though the levels rapidly decreased with distance away from dump sites (O’Neill, 1995). Large amounts of arsenic containing coal are combusted in power plants worldwide. Combustion of coals adds arsenic containing fly ash into the atmosphere, which eventually accumulate in soils and water (Ishak et al., 2002; Qafoku et al., 1999; Sakulpitakphon et al., 2003). Generally, arsenic may present in coals at concentrations ranging from 2 to 84 mg kg1, depending on the geological background. However, high concentrations of arsenic (1500 mg kg1) within the brown coal from the former Czechoslovakia were reported. Even higher concentration of arsenic (100–9000 mg kg1) within the coal was reported in Guizhou, China as a result of epigenetic mineralization (Liu et al., 2002). Approximately 90% of industrial arsenic in the United States is currently used as a wood preservative. Arsenic-containing compounds, such as chromated copper chromate (CCA), ammoniacal copper arsenate (ACA), and ammoniacal copper zinc arsenate (ACAA), have been extensively used as wood preservatives in order to reduce bacterial, fungal, and insect decay in woods. CCA has been the dominant chemical used to treat wood for decks and other

50

Hua Zhang and H. M. Selim

outdoor uses, constituting 75% of the pressure treatment wood market by volume. The treated woods commonly contain 1000–5000 mg kg1 of arsenic. Arsenic in those wood preservatives could diffuse into adjacent soil and leach into groundwater. Chirenje et al. (2003) showed that mean soil arsenic concentration as high as 23 mg kg1 close to CCA-treated wood structures compared with less than 3 mg kg1 at distance about 1.5 m away. Similarly, Rahman et al. (2004) concluded that arsenic diffused from CCA-treated wood to adjacent garden soil and found that vegetable crops grown in these gardens can accumulate significant concentrations of arsenic. EPA granted the cancellation and used termination requests affecting virtually all residential uses of CCA-treated wood after Dec. 31, 2003 (USEPA, 2003). Compounds containing arsenic had been extensively used as pesticides, insecticides, herbicides, soil sterilants, silvicides, and desiccants in cotton, orchards, silviculture, and turf since late 1800s. Commonly used arsenical pesticide includes inorganic compounds such as lead arsenate (PbAsO4), calcium arsenate (CaAsO4), magnesium arsenate (MgAsO4), zinc arsenate (ZnAsO4), zinc arsenite [Zn(AsO2)2], and Paris green [Cu(CH3COO)2 3Cu(AsO2)2], as well as organic compounds such as DMAA [dimethylarsonic acid, (CH3)2AsO2H], MSMA [monosodium methane arsenate, CH3AsO3HNa], MAMA [Monoammonium methane arsonate, CH3AsO2NH4OH], and MAA [Methylarsonic acid, CH3AsO2(OH)2] (Reigart and Roberts, 1999). Woolson et al. (1971) reported that surface soil in orchards with history of arsenic pesticide application averaged 165 mg As kg1 versus 13 mg As kg1 in soils without such applications. Before 1968, CaAsO4 was typically used in Louisiana as for cotton defoliation which subsequently resulted in arsenic in soils (Bednar et al., 2002; Ori et al., 1993). As a result of government regulations, several arsenical pesticides have been prohibited in the United States. Nevertheless, considerable amounts of arsenic are retained in by the soil matrix. In Denver, Colorado, some half a century after application of an arsenical herbicide (As2O3 þ PbAsO4) on turf, arsenic concentrations in residential soils at concentrations up to 1440 mg kg1 were measured (Folkes et al., 2001). In the US poultry industry, organic arsenic compounds, such as roxarsone (3-nitro-4-hydroxyphenylarsonic acid), are common feeds for broilers to control coccidial intestinal parasites and improve feed efficiency. Little of these organoarsenicals is retained in the meat and most of the arsenic is rapidly excreted. Thus, poultry litter containing 10–40 mg kg1 of arsenic has been largely recycled as organic amendment to the agriculture fields. It is reported that 20,000–50,000 kg of arsenic is annually introduced into the environment by poultry farmers along the eastern shore of the United States (e.g., Delaware, Maryland, and Virginia) (Arai et al., 2003). Han et al. (2004) found that after 25 years of application of arsenic containing poultry litter, arsenic concentrations were 8.4 and 2.7 mg kg1 in amended and nonamended soils, respectively.

Reaction and Transport of Arsenic in Soils

51

From 1906 to 1962, dipping solutions containing 1400–2200 mg kg1 of arsenic were widely used among thousands of cattle-dipping vats throughout southern United States in an effort to eradicate ‘‘southern cattle fever,’’ a transmitting disease. These practices have resulted in the contamination of soils and groundwater due to arsenic leaching and disposal from cattle-dipping vats (Thomas and Rhue, 1997). Mclaren et al. (1998) reported that surface soils surrounding cattle-dipping vats in Australia with a history of arsenicals usage were contaminated with arsenic ranged up to 3542 mg kg1.

3.3. Speciation Arsenic can exist in many organic and inorganic forms, depending on the original sources and dominant reactions in soils. Arsenic can form organic compounds by methylation as a biological process, producing both trivalent and pentavalent organoarsenic compounds (O’Neill, 1995; Smith et al., 1998). Environmentally significant arsenic compounds are arsenate and arsenite, because they are soluble in water and toxic. Distribution of arsenate and arsenite in the solution and solid phases is largely determined by adsorption and redox reactions in soils, which will be discussed in the following sections. Arsenic in sediments and soils is bound with solid phases at different strengths. The total arsenic concentration is not necessarily a good indicator of its potential mobility and bioavailability. Several sequential chemical extraction methods have been proposed to apportion soil arsenic into various pools based on the types of extractant used. Sequential extraction methods have been extensively used to characterize the mobility and bioavailability of arsenic in sediments and soils and regarded as operational (Han et al., 2004; Keon et al., 2001; Matera et al., 2003; Mclaren et al., 1998; Rodriguez et al., 2003; Woolson et al., 1973). It should be noted that most extraction procedures were originated from that of Tessier et al. (1979), where the following chemical fractions or heavy metals pools were considered; ion-exchangeable, surficially adsorbed, precipitated, organic chelated, and occluded. Although selective chemical extraction provides empirical determination of the dominant constituents responsible for As retention soils, the procedure does not provide specific information on interfacial reactions between As and various soil constituents. Spectroscopic techniques have been applied to determine the sorption reactions occurring in heterogeneous geological material (Foster, 2003). Spectroscopic analyses, including Fourier transformed infrared (FTIR), Raman spectroscopy, X-ray photoelectron spectroscopy, extended X-ray absorption fine structure (EXAFS), X-ray absorption near-edge spectroscopy (XANES), were able to determine the electronic energy levels of atoms or molecules in the system.

52

Hua Zhang and H. M. Selim

Therefore, they are useful in determining the structure of a particular solid phase at molecular level. Extensive researches have been conducted in this area and the reader is referred to Foster (2003) among others for a comprehensive review.

4. Biogeochemistry The biogeochemistry of arsenic in heterogeneous soil systems is rather complex, comprising a large array of chemical and microbiological reactions, for example, adsorption–desorption, reduction–oxidation, dissolution– precipitation, acid–base reactions, and biomethylation. Those reactions are affected by a series of environmental conditions such as pH, Eh, soil constituents, electrolytes, microbial activity, temperature, and residence time. A schematic diagram of the biogeochemical reactions of arsenic in soils is presented in Fig. 1. In this section, we briefly discuss biogeochemical reactions in Fig. 1 and focus on the impacts of those reactions on the retention and transport of arsenic in soils. Details of chemical and microbiological reactions of arsenic have been reviewed by others including Smith et al. (1998), Smedley and Kinniburgh (2002), Mahimairaja et al. (2005).

4.1. Retention mechanisms Several processes including ion exchange, surface complexation, precipitation, and surface precipitation contribute to the removal of arsenic from aquatic solution by the soil matrix. Retention of arsenic depends on arsenic

Agricultural input

As (V) Minerals Precipitation Demethylation

Organic As Demethylation

Methylation

Arsine

Agricultural industrial input

Dissolution

Secondary precipitation

Adsorption

Sorbed As (V)

As (V) Methylation Oxidation Demethylation Methylation

As (III)

Reduction Dissolution

Atmosphere

Desorption Reduction Adsorption

Oxidation

Reduction

Sorbed As (III)

Desorption Dissolution Precipitation

As Sulfides

Arsenide Geological industrial input

Figure 1

Schematic diagram of arsenic biogeochemistry in soils.

Reaction and Transport of Arsenic in Soils

53

concentration in solution, pH, reaction kinetics, arsenic species, competing ligands, as well as soil mineralogical composition. Adsorption processes, especially sorption onto metal oxide surfaces, governs the arsenic distribution in aerobic soils (Livesey and Huang, 1981; Masscheleyn et al., 1991). Sorption of arsenate and arsenite on soils was found to be significantly correlated with extractable Al and Fe contents. For example, Jacobs et al. (1970) found that AsO4 adsorption increased with increasing content of Fe oxide and the removal of Fe and Al oxides eliminated or appreciably reduced AsO4 adsorption in soils. Wauchope (1975) showed that besides Fe and Al content, soil clay content can also affect arsenate and methylarsonate adsorption. Numerous studies have investigated AsO4 and AsO3 adsorption on Fe and Al oxides (e.g., Anderson et al., 1976; Dixit and Hering, 2003; Pierce and Moore, 1980, 1982; Raven et al., 1998). Arsenate and arsenite anions are strongly adsorbed on metal oxides and hydroxides. The dominant process is inner-sphere surface complexation via ligand exchange of As for OH2 and OH in the coordination spheres of surface structural metal atoms (Sun and Doner, 1996; Waychunas et al., 1993). Spectroscopic analysis (Fendorf et al., 1997; Manning et al., 1998) revealed that inner-sphere surface complexes formed between arsenic anions [AsO4 and AsO3] and metal oxides can be mononuclear monodentate, mononuclear bidentate, and binuclear bidentate (Fig. 2). In general, formation of monodentate complex is favored at low As surface coverage while bidentate–binuclear complex occupied highest proportion of mineral surface at highest As coverage (Fendorf et al., 1997). Grossl et al. (1997) measured chemical relaxation via conductivity detection during pressure-jump relaxation experiment and concluded that arsenate adsorption on goethite is a twostep process resulting in the formation of inner-sphere bidentate complex. EXAFS studies of Manning et al. (1998) showed that AsO3 formed bidentate binuclear bridging complexes on goethite surface. Goldberg and Johnston (2001) using FTIR investigated arsenate and arsenite adsorption on amorphous iron and aluminum oxides. Their conclusion is that arsenate form inner-sphere surface complexes on both amorphous Fe and Al oxides, while arsenite form inner-sphere and outer-sphere surface complexes on amorphous Fe oxides and outer-sphere surface complexes on Al oxides. Indirect macroscopic evidences such as shift of point-zero-charge (PZC), ionic strength effect, and OH release stoichiometry have also verified the formation of inner-sphere complex of As on iron oxide surfaces. Shift in PZC with increasing anion concentration can be seen as evidence of strong specific anion adsorption and inner-sphere surface complex formation. Adsorption of anions forming inner-sphere complexes normally shows little or no ionic strength dependence (Hingston et al., 1967). Electrophoretic mobility (EM) measures of Anderson et al. (1976) indicated PZC change to lower pH values as arsenate adsorption increased on amorphous

54

Hua Zhang and H. M. Selim

Bidentate comlpexes Mononuclear Monodentate complex

Fe O

O

Fe As

Fe O Fe

O

O As

Fe O

O

O

O

As-Fe 2.85 Å

O Binuclear

Fe

O As-Fe 3.59 Å

Fe

O

O As

Fe

O

O

O As-Fe 3.24 Å

Figure 2 Binding mechanisms of arsenic in soils. (Reproduced with permission from Fendorf et al.,1997.)

Al hydroxide surface. The titration curve produced by Jain et al. (1999) demonstrated that adsorption of arsenate resulted in PZC reduction from 8.5 of pure ferrihydrite to 6.1, adsorption of arsenite resulted in a slightly less reduction of 1.5. Goldberg and Johnston (2001) found PZCs of amorphous iron oxide significantly shifted to increasingly lower pH with increasing AsO4 or AsO3 concentration, but the same results were not observed on amorphous Al oxide. They also showed that ionic strength has little or no effect on AsO4 and AsO3 adsorption on amorphous Al and Fe oxides. Because of their negatively charged surfaces, clay minerals generally have low arsenic adsorption capacity (Frost and Griffin, 1977; Lin and Puls, 2000; Goldberg, 2002; Xu et al., 1988). The proposed adsorption sites of arsenic anions on layer silicates are positively charged AlOH2þ functional groups exposed at crystal edges (Manning and Goldberg, 1996a). Isomorphous substitution of Al by Fe in some clays may contribute to arsenic sorption. Generally, clay minerals with high surface area exhibit strong arsenic sorption.

Reaction and Transport of Arsenic in Soils

55

Other soil components such as calcite, sulfide, and organoclay matter may also contribute to As adsorption. Goldberg and Glaubig (1988) quantified arsenate adsorption on calcite and they concluded that at high pH range (>9), soil carbonates may play an important role in arsenate adsorption. Bostick and Fendorf (2003) studied arsenite sorption on iron sulfides and found that at low surface coverage, arsenite retention by sulfides conformed to a Langmuir isotherm. Moreover, the existence of soil organic matter (SOM) can reduce As adsorption due to competition for adsorption sites on mineral surfaces (Grafe et al., 2001) or form aqueous complex with As (Redman et al., 2002). However, Saada et al. (2003) showed evidence that humic acid (HA) nitrogen groups can form organoclay with clay mineral that resulted in enhanced As adsorption capacity. Precipitation is generally considered to contribute a small portion of arsenic retention except in highly contaminated soils (e.g., areas surrounding acid mines). If present at exceedingly high concentrations, direct precipitation, or coprecipitation of arsenic with solid phase Al, Fe, Mn, Mg, and Ca might occur. For example, in As-rich mine tailings piles, precipitates, such as scorodite, parasymplesite [Fe2(AsO4)38H2O], or rauenthalite [Ca3(AsO4)210H2O], may form, often as surface coatings on other mineral grains ( Walker et al., 2006). Alternatively, arsenic anions can substitute anions in secondary minerals including jarosite [KFe3(SO4)2(OH)6], gypsum [CaSO42H2O], calcite [CaCO3], and ettringite [Ca6Al2(SO4)3 (OH)1226H2O] (Myneni et al. 1997). Voigt et al. (1996) observed natural precipitation of mineral hoernesite [Mg3(AsO4)28H2O] in a contaminated soil. Juillot et al. (1999) reported the precipitation of 1:1 Ca arsenate such as weilite [CaHAsO4], haidingerite [CaHAsO4H2O], and pharmacolite [CaHAsO42H2O] and in a minor amounts, Ca–Mg arsenate such as picropharmacolite [(Ca,Mg)3(AsO4)26H2O] in a contaminated industrial site. After equilibrate slurries with varying Ca/As ratios for 4 years, Bothe and Brown (1999) observed the formation of Ca4(OH)2(AsO4)24H2O, Ca5(AsO4)3OH (arsenate apatite), and Ca3(AsO4)232/3H2O using Scanning Electron Microscope (SEM) and X-ray diffraction (XRD). Foster et al. (1998) showed the formation of scorodite [FeAsO42H2O] in mine waste using extended X-ray adsorption fine-structure spectroscopy (EXAFS). On the basis of results from EXAFS spectroscopic analyses, Grafe et al. (2004) reported the formation of adamite like [Zn2(AsO4)OH)] and korinigite precipitation on goethite when AsO4 and Zn were simultaneously introduced in high surface density ratio. Violante et al. (2006) reported the formation of poorly crystalline aluminum arsenate precipitates (a boehmite-like mineral) as influenced by pH, As/Al ratio, and aging. They suggested that arsenate appeared to be occluded within networks of short range ordered materials. Recent studies suggested that surface precipitation, that is, threedimensional growth of a particular surface phase, may occur for arsenate. Surface precipitation assumes anions absorbed on mineral surfaces to attract

56

Hua Zhang and H. M. Selim

dissolved Fe or Al. The adsorbed Fe or Al in turn adsorbs more anions, resulting in a multilayer adsorption. Contradictions were reported based on kinetic and spectroscopic studies. Waychunas et al. (1993) ruled out the possibility of arsenate surface precipitation on ferrihydrite with their EXAFS data. However, the possibility of Al–AsO4 surface precipitate formation was not excluded based on spectroscopic studies of arsenate adsorption on Al oxides (Arai and Sparks, 2002). The possibility of surface precipitation of AsO4 on ferrihydrite was suggested by Zhao and Stanforth (2001) through kinetic studies. More recently, Jia et al. (2006) demonstrated the formation of poorly crystalline ferric arsenate surface precipitates (a scorodite-like mineral) in undersaturated condition at low pH. Pedersen et al. (2006) reported that the arsenate was initially associated with the surface of more reactive iron oxides (ferrihydrite and lepidocrocite) but incorporated into the crystalline structure during Fe2þ catalyzed transformation into iron oxides into recrystallization products (goethite and magnetite).

4.2. pH dependency The effect of solution pH on arsenic sorption processes have been investigated on several minerals and soils (Dixit and Hering, 2003; Frost and Griffin, 1977; Goldberg and Glaubig, 1988; Manning and Goldberg, 2002; Manning and Goldberg, 1996; Manning and Goldberg, 1997a; Pierce and Moore, 1982; Smith et al., 1999; Xu et al., 1988). Solution pH controls two fundamental factors; mineral surface potential and arsenic speciation, which in turn impacts arsenic adsorption on mineral surfaces. Metal oxides and other minerals possess pH-dependent variable surface charges. At pH below PZC, adsorption of Hþ is in excess of that of OH, the surface becomes positively charged. The magnitude of negative potential on mineral surfaces increases with increasing pH. On the other side, there exists pH-pKa dependence of both arsenate (pKa1 ¼ 2.3, pKa2 ¼ 6.8, and pKa3 ¼ 11.6) and arsenite (pKa1 ¼ 9.2, and pKa2 ¼ 12.7). At neutral pHs, 2 arsenate exists as negatively charged H2 AsO 4 and HAsO4 , and the negative potential tend to increase as pH increases. In contrast to arsenate, arsenite mostly exists as zero charged H3 AsO03 below pH 9.2. The interaction between PZC of soil minerals and pKa of arsenate or arsenite determine the adsorption envelope, that is, amount of adsorption as a function of pH (Hingston et al., 1967). Arsenate adsorption on oxides and clays appears to be pH dependent with adsorption decreasing with increasing pH. In contrast, arsenite adsorption on soil minerals exhibits parabolic behavior with an adsorption maximum between 8 and 10. Under most conditions, AsO4 adsorbed more strongly than AsO3 on soil components (Manning and Goldberg, 1997a; Smith et al., 1999). However, arsenite is more strongly bound to metal

Reaction and Transport of Arsenic in Soils

57

oxides under highly alkaline condition (Goldberg, 2002; Jain and Loeppert, 2000; Manning and Goldberg, 1997a; Raven et al., 1998). These two arsenic oxyanions may compete with each other for adsorption sites. Jain and Loeppert (2000) reported that when both AsO4 and AsO3 concentrations were below 2.08 mmol As kg1Fe, the effect of AsO4 on AsO3 sorption on ferrihydrite was more pronounced than vice versa. At concentrations higher than 3.47 mmol As kg1 Fe, AsO4 did not influence AsO3 adsorption, but AsO3 significantly reduced AsO4 adsorption. Goldberg (2002) concluded that at pH lower than 8, arsenate has higher adsorption capacity on minerals and soils than arsenite and competition between arsenate and arsenite is relatively small. Dixit and Hering (2003) concluded that at lower pH (<7), arsenate sorption is more favorable than arsenite on amorphous Fe oxide and goethite and the opposite holds true at pH > 7 (Fig. 3).

4.3. Effect of solution composition Arsenic adsorption in soils can be affected by the presence of ligands that can compete for adsorption sites on mineral surfaces (such as phosphate, silicate, carbonate, and organic acid). Several studies indicate that, because of their similar chemical properties, phosphate [PO4] in soils competes with arsenate [AsO4] for available adsorption sites. Both AsO4 and PO4 are specifically sorbed on mineral surfaces by forming similar types of inner-sphere surface complexes through ligand exchange. Several studies indicated the existence of phosphate substantially suppressed the sorption of arsenate on minerals and soils (Darland and Inskeep, 1997a; Dixit and Hering, 2003; Livesey and Huang, 1981; Melamed et al., 1995; Violante and Pigna, 2002; Roy et al., 1986ab; Smith et al., 2002; Violante and Williams et al., 2003). Competitive sorption between phosphate and arsenate generally depends on the surface properties of the adsorbent, concentrations of AsO4 and PO4, pH, sequence of addition, and residence time ( Violante and Pigna, 2002). Arsenate and phosphate are specifically adsorbed on a similar set of surface sites, although evidence showed some sites are only available for either AsO4 or PO4. On the basis of competitive adsorption of anions on goethite and gibbsite, Hingston et al. (1971) proposed two types of adsorption sites on mineral surface; the first type is available for both anions where competition takes place while the second type of adsorption sites is specifically available for either anions. Violante and Pigna (2002) demonstrated that minerals rich in aluminium (Al) have a greater affinity of phosphate than arsenate, whereas metal oxides and phyllosilicates rich in Fe were more effective in adsorbing AsO4 than PO4. In general, the adsorption of both phosphate and arsenate on Fe/Al oxides decreases with increasing pH (Manning and Goldberg, 1996). Furthermore, Jain and Loeppert (2000) reported that the effect of phosphate on arsenate adsorption on ferrihydrite was greatest at high pH

58

Hua Zhang and H. M. Selim

A 100

%As adsorbed

80 60 40 20 0

4

5

6

7 pH

8

9

10

4

5

6

7 pH

8

9

10

B 100

%As adsorbed

80 60 40 20 0

Figure 3 Comparison of AsO4 and AsO3 sorption edges on (A) hydrous ferric oxide (HFO) and (B) goethite in the presence (solid symbols) and absence (open symbols) phosphate: AsO4 (squares) and AsO3 (circles). Total arsenic concentrations are 10 and 25 mM for HFO and goethite, respectively. Total phosphate concentration is 100 mM. Experimental conditions: 0.01 M NaClO4, 0.03 g L1 HFO, or 0.5 g L1 goethite. (Reproduced with permission from Dixit and Hering, 2003.)

than at low pHs. Zhao and Stanforth (2001) confirmed that arsenate and phosphate equally adsorbed on goethite when added simultaneously. When added sequentially, the desorption process was kinetically controlled, with a fraction of both AsO4 and PO4 remained nonexchangeable. Carbonate anions is commonly present at high concentration in soil solution and groundwater. Appelo et al. (2002) using surface complexation modeling calculated carbonate and ferrous ion sorption on ferrihydrite and their displacing effect on sorbed arsenate and arsenite. Their calculation demonstrated that sorption of particularly carbonate at common soil concentrations reduced the sorption capacity of arsenic on ferrihydrite significantly. In contrast, Arai et al. (2004) observed carbonate enhanced AsO4

Reaction and Transport of Arsenic in Soils

59

sorption on hematite surface. But when pH was held constant, dissolved carbonate effect was negligible. They suggested that the effect of dissolved carbonate on AsO4 adsorption were influenced by the reaction conditions. In natural systems, the presence of dissolved organic carbon (DOC) may compete with arsenic for adsorption sites on mineral surfaces and inhibit arsenic adsorption. Grafe et al. (2001) showed that the presence of HA and fulvic acid (FA) reduced AsO4 adsorption on goethite surfaces whereas citric acid (CA) indicated no effect. AsO3 adsorption was inhibited by all three organic acids in the order of CA > FA > HA. Redman et al. (2002) observed natural organic matter (NOM) dramatically delayed the sorption kinetic and diminished the sorption maximum of both arsenate and arsenite on hematite. The introduction of NOM displaced sorbed AsO4 and AsO3 from hematite surfaces, on the other side, arsenic similarly displaced NOM from hematite surfaces. They also observed the formation of aqueous complexes between NOM and arsenate or arsenite. Silicic acid is ubiquitous in soil and water environment and strongly adsorb to Fe oxides through ligand exchange. Waltham and Eick (2002) found that the presence of 1.0 mM silicic acid reduced 40% of arsenite adsorption on goethite, whereas it decreased the rate of arsenate adsorption but not the total sorption or capacity. We should mention that several studies have verified that other anions naturally occurring in the soil solution, such as Cl, NO3, and SO42, have little effect on arsenic retention and transport in soils (Livesey and Huang, 1981; Peryea and Kammereck, 1997; Qafoku et al., 1999; Smith et al., 2002). Few studies considered the effect of ionic strength on As behavior in soils. Adsorption of As indifferent to changes in ionic strength was seen as macroscopic evidence for inner-sphere surface complexation. Goldberg and Johnston (2001), Manning and Goldberg (1996a) showed that ionic strength have little effect on arsenite adsorption on Al and Fe oxides at the range of 0.02–0.1M. They observed that ionic strength have greater effect on arsenite than arsenate, indicating arsenite is more weakly bound. However, arsenic adsorption increases with increasing ionic strength was often observed in soils (Manning and Goldberg, 1997; Smith et al., 1999; Williams et al., 2003). Increasing ionic strength lessens electrostatic repulsion between the negatively charged surface and oxyanions and may result in increased adsorption of As onto mineral surface. The type of cations present in the soil system impacts on arsenic adsorption. Some cations may form ion pairs with arsenic oxyanions, resulting in decreasing arsenic activity in the soil solution and decreasing retention on soil surfaces (Myneni et al., 1998). Another effect is that the sorption of some cations on mineral surface may decrease negative charges on mineral surfaces and increase adsorption of arsenic oxyanions. Smith et al. (2002) observed that sorption of both AsO4 and AsO3 was enhanced by the presence of Ca2þ in the solution when compared with Naþ.

60

Hua Zhang and H. M. Selim

4.4. Sorption kinetics Traditionally, arsenic sorption studies have been carried out based on batch equilibration experiments conducted within a short period of reaction time. Few studies investigated the effect of long residence time on adsorption of arsenic in soils. However, a slow but significant reaction phase may exist due to diffusion into interparticle and intraparticle spaces, sites of different reactivity, or surface precipitation. The kinetics of arsenic adsorption– desorption must be understood for accurate predictions of the fate of arsenic in the soil environment (Sparks, 1998). Studies demonstrated that adsorption of arsenic on mineral surfaces is perhaps a two-phase reaction with a large amount of arsenic rapidly taken up by the adsorbent initially time and followed by a long plateau phase that can extend to years. Observed retention kinetics of arsenic is likely due to the heterogeneity of the soil surface where multiple chemical and physical processes may take place. Chemical reaction rates of surface complexation between anions and metal oxides are considered rapid. Using a pressure jump relaxation technique, Grossl et al. (1997) calculated a kinetic rate constant of 106.3 s1 for the formation of monodentate inner-sphere surface complex on goethite surfaces. In addition, a forward rate constant of 15 s1 was associated with the succeeding reaction for the formation of bidentate mononuclear surface complex. Because of their rapid reaction rates, surface complexation is not a rate-limiting step of AsO4 adsorption in soils. However, different types of surface complexes (e.g., monodentate, bidentate, mononuclear, binuclear) can be formed on oxide surfaces at high or low surface coverage. This heterogeneity of sorption sites may contribute to observed adsorption kinetics where sorption takes place preferentially on high affinity sites and followed subsequently by slow sorption on sites of low sorption affinity. The development of surface precipitates is a slow process involving multiple reaction steps and may explain in part the slow AsO4 kinetics in soils. Zhao and Stanforth (2001) suggested the slow buildup of surface precipitates as the mechanisms of irreversible AsO4 and PO4 retention on goethite. More recently, the XRD and Raman spectroscopy results of Jia et al. (2006) confirmed the formation of poorly crystalline ferric AsO4 surface precipitates on ferrihydrite under high As/Fe molar ratio, low pH, and extended reaction time. Diffusion of AsO4 to reaction sites within the soil matrix was proposed as an explanation to the time-dependent adsorption by Fuller et al. (1993), Raven et al. (1998), among others. A two-phase process was generally assumed for diffusion-controlled adsorption, with the reaction occurring instantly on liquid–mineral interfaces during first phase whereas slow penetration or intraparticle diffusion is responsible for the second phase. Pore space diffusion model has been employed by Fuller et al. (1993) and Raven

61

Reaction and Transport of Arsenic in Soils

As(V) sorbed (mg kg−1)

600 Windsor 450

300

150

0 0

15 30 45 60 As(V) in solution (mg L−1)

75

Figure 4 Time-dependent sorption isotherm of arsenate on Windsor soil. Symbols are for different reaction times of 24, 72, 168, 336, and 504 h (from bottom to top). Solid curves depict results of curve fitting with the Freundlich equation. (Reproduced with permission from Zhang and Selim, 2005.)

et al. (1998) to describe the slow sorption of AsO4 on ferrihydrite. For heterogeneous soil system, the complex network of macro- and micropores may further limit the access of solute to the adsorption sites and cause the time-dependent adsorption. Because of the intrinsic chemical and physical heterogeneity of soils, it is difficult to describe and to accurately predict the kinetics of arsenic adsorption on the soil matrix. Most studies have demonstrated that the residence time significantly affected arsenic retention by soils (Fig. 4). Similar to observations on mineral surfaces, it is widely reported that arsenic sorption is biphasic with an initial rapid reaction followed by a slow (kinetic) phase. For example, Livesey and Huang (1981) concluded that AsO4 adsorption on soils were rapid initially with limited adsorption occurring after 24 h. Elkhatib et al. (1984a) studied kinetic of AsO3 adsorption on surface and subsurface soils and described their results with an Elovich equation and a modified Freundlich equation. They concluded that the initial reaction was rapid with more than 50% of AsO3 sorbed on soils in the first 0.5 h. Their regression results showed that reaction rate can be related to soil clay content or Fe oxide content. Similarly, Carbonell-Barrachina et al. (1996) found that AsO3 sorption on soils was initially rapid and sorption rate decreased with time. Their results showed that sorption processes continued during 50 h of reaction. Corwin et al. (1999) found that Langmuir adsorption coefficient (KL) for arsenate adsorption was 1.70, 2.92, and 3.98 L kg1 after 1, 7, 14 days, respectively. Smith et al. (1999) showed that AsO4 retention by soils was initially rapid, attained apparent equilibrium in less than 1 h, followed by a steady and slow rate for the 72 h investigated. Williams et al. (2003) conducted long-term AsO4 adsorption experiment on an iron oxide containing subsurface soil. They concluded that AsO4

62

Hua Zhang and H. M. Selim

sorption exhibit biphasic pattern with a rapid period before 48 h, followed with a slow process for several weeks. The Kd value after 3 weeks was three to four times that after 1 week. Darland and Inskeep (1997b) demonstrated that adsorption of AsO4 on iron oxides continued for at least 96 h. Brouwere et al. (2004) showed that Kd for AsO4 increased on average 1.8 fold between day 2 and 7. Manning and Suarez (2000) showed that the rate of AsO3 adsorption on soils was closely dependent on soil properties including extractable metals, soil texture, specific surface area, and pH.

4.5. Desorption Recent research has shown that desorption of arsenic is highly hysteric and sorbed arsenic is not easily removable from the soil matrix. Observed desorption hysteresis might be due to kinetic retention behavior, such as slow diffusion, and irreversible retention. In contrast to adsorption studies, relatively few work have been done to investigate desorption or release of arsenic from minerals or soils. The effects of residence time on desorption of arsenic from soil minerals or soils are not clear. Lin and Puls (2000) found that desorption of AsO3 and AsO4 from clay minerals was significantly decreased with increasing aging time. They explained this phenomenon by assuming diffusion of arsenic into internal sorption sites, which are not readily accessible by the bulk solution. O’Reilly (2001) showed that a significant amount of AsO4 bound to goethite (>60%) is not readily desorbable by PO43 after 5 months of reaction. However, they found residence time (0.7–4846 h) have little effect on AsO4 desorption from goethite in the presence of 6 mM PO43 at pH 4 and 6. In contrast, Arai and Sparks (2002) found that AsO4 desorption from aluminum oxide surfaces decreased with increasing reaction time (3 days to 1 year). Furthermore, their EXAFS studies provided microscopic evidence of rearrangement of surface complexes and surface precipitation. These different results may due to the different desorption solution used. O’Reilly (2001) used high concentration of phosphate solution while Arai and Sparks (2002) used a complex solution of 0.096M NaCl, 1 mM sodium sulfate, and 2 mM organic buffer. Pigna et al. (2006) studied desorption of AsO4 from Fe and Al oxides by PO4 as a function of residence time and surface coverage. Their data indicate that more AsO4 was desorbed from Al oxides than from Fe oxides. Increased residence time from 24 h to 360 h significantly decreased AsO4 desorption due to rearrangement of AsO4 on mineral surfaces from desorbable into more resistant forms. Desorption or release of arsenic from soil is a complex process since sorbed arsenic species may bound to solid matrices in various energy states. Jacobs et al. (1970) examined the extractability of AsO4 sorbed by soil decreased with increasing equilibration time. The time required to reach equilibrium of arsenic in soils varied from 1 to 6 months depending on soil

63

Reaction and Transport of Arsenic in Soils

texture and arsenic level. Woolson et al. (1973) showed that AsO4 solubility in soils decreased with time and reached equilibrium in 4–6 weeks. Elkhatib et al. (1984b) concluded that arsenite desorption soils was hysteretic with only a small fraction of the sorbed AsO3 released after five desorption steps with deionized (DI) water. Carbonell-Barrachina et al. (1996) stated that AsO3 sorption was a reversible process and their data showed about 50% of the sorbed arsenic can be released from soils after five desorption steps in 36 h. Lombi et al. (1999) evaluated the kinetics and reversibility of AsO3 and AsO4 sorption by Fe oxide-coated sand and several soils. They demonstrated that a significant portion of arsenic was converted to less mobile as a subsequent to decreased As in the easily extractable form, and increasing As in the more recalcitrant form. Desorption or release results of Zhang and Selim (2005), which are presented as isotherms in the traditional manner in Fig. 5, demonstrated distinct discrepancies between adsorption and successive desorption isotherms and indicate considerable hysteresis for AsO4 release. This observed hysteresis was attributed to kinetic retention behavior, such as slow diffusion and irreversible retention. Desorption curves also demonstrated that percentage of desorption increased with AsO4 surface coverage, indicating the high sorption affinity at low surface coverage. Under strongly acidic and reducing conditions, iron oxides may be dissolved either biotically or abiotically. It is believed that arsenic associated (adsorbed or precipitated) with oxides will be released into aqueous solution during iron oxide dissolution (Smedley and Kinniburgh, 2002). Pedersen et al. (2006) observed the release of arsenic during the reduction of ferrihydrite, lepidocrocite, and goethite. For ferrihydrite and goethite, they attributed the release of arsenic as a consequence of reduced surface area of iron 600 As(V) sorbed (mg kg−1)

Windsor 450

300 Desorption Adsorption

150

0 0

30 10 20 As(V) in solution (mg L−1)

40

Figure 5 Isotherms of arsenate desorption from different soils based on successive dilution after the last adsorption step for different initial concentrations (Co) of 20, 40, 80, and 100 mg l1. The solid and dashed curves depict results of curve fitting with the Freundlich equation for 504-h adsorption and desorption isotherms, respectively. (Reproduced with permission from Zhang and Selim, 2005.)

64

Hua Zhang and H. M. Selim

oxides during dissolution. In contrast, they found arsenic was released from lepidocrocite before the release of Fe. A study by Herbel and Fendorf (2005) investigated the mobilization of arsenic under dynamic flow conditions in ferric hydroxide-coated sands inoculated with arsenate reducing bacteria (Surfurosprillum barnesii strain SES-3). They suggested that the release of arsenic into the aqueous phase is associated with the mineralogical transformation of iron oxides due to microbial reduction.

4.6. Reaction with sulfides Chemical weathering is a major process controlling the geochemical cycling of arsenic in the environment. Arsenic in the parent material is originally present in the form of chemically reduced minerals such as realgar (AsS), orpiment (As2S3), FsAsS, and amorphous As2S3 or AsS (Lengke and Tempel, 2005; Oremland and Stolz, 2003). The weathering process can oxidize arsenic in sulfide minerals to arsenite or arsenate minerals and subsequently release arsenic anions into aqueous phase. The reaction occurring at the interface between sulfide minerals and aqueous solution depends on environmental factors such as pH, redox potential (pe), carbonates, temperature, water flow, and morphology of the minerals. In general, the dissolution arsenic from sulfide minerals is a slow process that can continue for several years. Considerable amount of arsenic will be released into aqueous solution during this process. Lengke and Tempel (2005) studied the oxidative dissolution of amorphous As2S3 and AsS, orpiment, and realgar using mixed flow reactors. The proposed overall reactions can be expressed as:

As2 S3 ðsÞ þ 7O2 ðaqÞ þ 6H2 O ! 2HAsO2 4 ðaqÞ þ þ 3SO2 4 ðaqÞ þ 10H ðaqÞ

AsSðsÞ þ 2:75O2 ðaqÞ þ 2:5H2 O ! HAsO2 4 ðaqÞ þ þ SO2 4 ðaqÞ þ 4H ðaqÞ

ð1Þ ð2Þ

They found that the oxidation rates of arsenic sulfides increase with increasing pH and dissolved oxygen (DO) concentration. Their results demonstrated that at pH value of 7–8, the oxidation rate of arsenic sulfide solids is in the decreasing order of As2S3(am) AsS (am)>orpiment realgar. Arsenite was identified as the dominant senic species released from oxidation process. Calculated values of activation energies suggested that oxidation kinetics of sulfides minerals are controlled by surface reactions, that is, transfer of electron from the sulfide mineral to the oxidant (DO). The oxidative dissolution of FeAsS is commonly observed in many mining sites resulting in the release of high concentrations of arsenic in the

Reaction and Transport of Arsenic in Soils

65

effluent. Because As, F, and S can coexist in multiple oxidation states, the dissolution mechanism of FeAsS is rather complex where multiple reactions occur simultaneously or sequentially (Nesbitt et al., 1995). Using X-ray photoelectron spectroscopy (XPS), Nesbitt et al. (1995) reported that As1 was predominant on surfaces of unoxidized FeAsS. Significant oxidation to As5þ and As3þ was observed upon reaction with air saturated water. Using mixed flow reactor, Walker et al. (2006) determined the rate of FeAsS oxidation by DO at pH of 6.3–6.7 and suggested these reaction mechanisms:

4FeAsSðsÞ þ 11O2 ðaqÞ þ 6H2 O ! 4H3 AsO3 ðaqÞ 2þ þ 4SO2 4 ðaqÞ þ 4Fe ðaqÞ

ð3Þ

4Fe2þ ðaqÞ þ O2 ðaqÞ þ 10H2 O ! 4FeðOHÞ3 ðsÞ þ 8Hþ ðaqÞ ð4Þ þ 2H3 AsO3 ðaqÞ þ O2 ðaqÞ ! 2HAsO2 4 ðaqÞ þ 4H ðaqÞ

ð5Þ

Their results indicate that the oxidation rate of FeAsS (10–10.14 mol m2 was independent of DO in the range of 0.3–17 mg l1. However, increasing level of DO resulted in the increasing ratio of AsO4/AsO3 in the effluent. They suggested that the rate limiting reaction step is the slow reduction (electron release) of water at anodic sites on FeAsS surface. Yu et al. (2004) investigated the dissolution of FeAsS in Fe2(SO4)3 solution under acidic condition (pH = 1.8). Their results demonstrated that the dissolution rate of FeAsS increased with increasing concentration of ferric iron [Fe(III)] and temperature. The dissolution rate of FeAsS was higher in ferric chlorite (FeCl3) solution than in ferric sulfate solution. In addition, the dominant species of arsenic releasing was arsenite in Fe2(SO4)3 solution, whereas arsenate was the major form of arsenic in FeCl3 solution. It should be noted that precipitation of iron oxides commonly occurs as a result of FeAsS oxidation. Thus, adsorption of released arsenite and arsenate on iron oxides attenuated arsenic concentration in solution. In anoxic sulfidic environment, reaction of arsenic with sulfide minerals controls the geochemical cycling of arsenic. Using X-ray adsorption nearedge structure (XANES), Reynolds et al. (1999) identified the precipitation of FeAsS in Santa and Palouse soils after 14 days of flooding and those FeAsS precipitation was largely destroyed upon reaeration. Under acidic and reduced conditions, Wilkin and Ford (2002) reported disordered orpiment or alacranite precipitations formed after the reaction between soluble arsenic and H2S. Using As K-edge X-ray absorption spectroscopy (XAS), Farquhar et al. (2002) investigated the mechanisms of the interaction between AsO3 and AsO4 in aqueous solution (pH 5.5–6.5) with the surfaces of crystalline mackinawite (tetragonal FeS) and pyrite (FeS2). They suggested the formation s1)

66

Hua Zhang and H. M. Selim

of outer-sphere complexation at low arsenic concentrations and the formation of poorly crystalline arsenic sulfide at high arsenic concentrations. Bostick and Fendorf (2003) studied arsenite sorption on troilite (FeS) and FeS2 and they identified surface precipitation of FeAsS using XAS. They observed that sorption increased with pH. This was explained on the basis of the formation of Fe(OH)3 associated with FeAsS precipitation, because the formation of Fe(OH)3 is favored with increasing pH. Wolthers et al. (2005) reported that the formation of outer-sphere complex between AsO4 and AsO3 and surface of sulfide minerals is a fast process. Thus, sorption isotherms can be described with a Freundlich equation where the distribution coefficient (Kd) for arsenate always reported larger than that for arsenite.

4.7. Heterogeneous oxidation Oxidation–reduction reactions play an important role in determining arsenic solubility, mobility, bioavailability, and toxicity in soils. Under natural environmental conditions, arsenate [AsO4] and arsenite[AsO3] are the most abundant forms of arsenic (Masscheleyn et al., 1991). In soils and water systems, AsO4 is dominant under aerobic condition and AsO3 under anoxic or anaerobic condition. The distribution and transformation between arsenate [AsO4] and arsenite [AsO3] is largely controlled by the redox condition of the soil environment. Besides soil redox potential (Eh/pe), other factors, such as pH, Fe and Mn oxides, sulfides, organic matter, and microbial activity, also impact reduction and oxidation of arsenic ( Jones et al., 1997). The oxidation–reduction reactions of arsenic can be chemical or biological and dependent on the substances present in the environment. Although AsO3 is not thermodynamically stable under oxidized conditions, high concentrations of AsO3 are frequently observed in water environments in the presence of DO. Nonequilibrium conditions may be due to slow kinetics of arsenite oxidation by oxygen in homogeneous solutions (with a half time of 1 year) (Scott and Morgan, 1995). However, the existence of Mn oxides in soils and water environments can promote heterogeneous oxidation of AsO3 to AsO4 at considerably high reaction rates (Oscarson et al., 1981ab; Scott and Morgan, 1995). There is considerable evidence that heterogeneous oxidation on the surfaces of Mn oxides play an important role in the transformation of AsO4 and AsO3. Oscarson et al. (1981a) showed that Mn oxides can effectively oxidize AsO3 into AsO4, accompanied by the reduction of Mn(IV) to Mn(II). The oxidation rates were unaffected with the insulation from oxygen or the inhibition of biological activity (Oscarson et al., 1981a). The depletion of AsO3 through oxidation on surfaces of birnessite, crytptomelane, and pyrolusite followed first-order kinetics with rate constants dependent on crystallinity and surface area of the minerals (Oscarson et al.,

67

Reaction and Transport of Arsenic in Soils

1983a). Through the masking of electron-accepting sites, surface coating of Fe and Al oxides or calcium carbonate decreased the rate constants of AsO3 oxidation by Mn oxides (Oscarson et al., 1983b). Their spectroscopic study showed that birnessite was an active oxidant of AsO3 both in solution and on the goethite surface (Sun and Doner, 1998). Scott and Morgan (1995) demonstrated that AsO3 oxidation by synthetic birnessite is rapid, with a timescale of minutes. They explained this reaction with a four-step multiprocess surface mechanism, (1) adsorption of AsO3; (2) electron transfer from AsO3 to Mn(IV); (3) release of AsO4; (4) release of Mn(II), with the adsorption as the slowest step (Fig. 6). Furthermore, surface spectroscopic studies suggest that the electron transfer process involves two steps with the formation of Mn(III) as an intermediate product (Nesbitt et al., 1998). The reaction stoichiometry can be expressed as:

2MnO2 ðsÞ þ H3 AsO3 ¼ 2MnOOH ðsÞ þ H3 AsO4

ð6Þ

2MnOOH ðsÞ þ H3 AsO3 ¼ 2MnO þ H3 AsO4 þ H2 O

ð7Þ

where MnOOH* is an intermediate reaction product. The oxidation of AsO3 by Mn(III) oxide manganite (g-MnOOH) was shown to be occurring on the timescale of hours (Chiu and Hering, 2000). In addition, chemical and miscroscopy techniques demonstrated the formation of manganese arsenate (Krautite) precipitate at the surface of H-birnessite with high crystallinity (Tournassat et al., 2002). The reaction can be written as follows:

Mn2þ þ H2 AsO4 þ H2 O ¼ MnHAsO4 H2 O þ Hþ

ð8Þ

Although thermodynamically favorable, oxidation rate of AsO3 on surfaces of Fe(III) oxides is rather slow. XPS evidences showed no redox reaction between Fe(III) oxide and AsO3 within 72 h (Oscarson et al., 1981b). EXAFS results of AsO3 sorption showed no evidence of heterogeneous oxidation to AsO4 on goethite surface at pH 6.4 to 8.6 (Manning and Goldberg, 1997b). However, XANES and IR spectral demonstrated that 20% of AsO3 sorbed on goethite surface was oxidized into AsO4 at pH 5.0 after 20 days of incubation, while no AsO4 was detected at pH 8.0 (Sun and Doner, 1998). There is considerable evidence that Fe(II) can catalyze the oxidation of AsO3 by O2 in the presence of iron oxides (Sung and Morgan, 1980, 1981). It was reported that the addition of single oxidants [O2, H2O2, dissolved Fe(III), or iron(III) (hydr)oxides] did not oxidize AsO3. However, partial or complete oxidization of AsO3 was observed in parallel to the oxidation of Fe(II) by O2 or H2O2. A reaction scheme is proposed in which the reaction of Fe(II) with H2O2 forms free radical intermediates, possibly OH radicals at low pH or an Fe(IV) species at higher pH (>5.24), followed by the

68

Hua Zhang and H. M. Selim

Mn

O

Mn

Mn

OH

O

O

O

Mn

OH

O

O

OH Mn

O

O Mn

Mn

OH

Mn

O O

Mn

O

O OH

Mn

Mn

O

As OH

O O

Mn

OH

Adsorption of As(III) A B Electron transfer

Equivalent surfaces Mn

O

Mn

C

E

Mn

Mn

Mn

OH

Release of Mn(II)

Mn

OH

OH O

O

OH O

O

O Mn2+

OH

O Mn

OH

OH

O Mn

Mn

Mn

OH O

OH O

Mn

As

O

OH OH

Release of As(V) D

Mn

O

O

O Mn

OH

OH

O Mn

Mn

Mn

OH OH

OH O

Mn

−O

As

O

OH OH

Figure 6 (A) Schematic representation of the cross section of the surface layer of a Mn(IV) oxide and (B) the resulting surface structure following arsenite adsorption, (C) electron transfer, (D) arsenate release, and (E) Mn2þ release. (Reproduced with permission from Scott and Morgan, 1995.)

oxidation of AsO3 to As(IV) by intermediate Fe(IV), then As(IV) is oxidized to AsO4 by molecular oxygen. The oxidation of AsO3 occurs on a timescale of hours and incomplete oxidation was observed even though Fe(II) was in excess (Hug and Leupin, 2003). A reaction transport model was developed to simulate the complex kinetics of Fe(II) catalyzed AsO3 oxidation in sediments (Bisceglia et al., 2005).

Reaction and Transport of Arsenic in Soils

69

It is generally accepted that phyllosilicate clay minerals has lower affinity for anions than iron and aluminum oxides and have limited or no capability of oxidizing AsO3 to AsO4. No oxidation of AsO3 in solution was observed after 48 h reaction with suspensions of illite, montmorillonite, kaolinite, vermiculite, ferruginous smectite, microcline, orthoclase, or calcite (Oscarson et al., 1981b). The heterogeneous oxidation of AsO3 to AsO4 in the presence of kaolinite and illite was attributed to MnO2 impurity in clay minerals (Manning and Goldberg, 1997b). Lin and Puls (2000) demonstrated that oxidation of AsO3 to AsO4 occurred at the clay surface, whereas no reduction of AsO4 to AsO3 was observed. The kinetics of AsO3 oxidation in aerated soil and sediment is generally controlled by their reactions on mineral surfaces. Kinetic studies of AsO3 sorption and oxidation by aquifer materials demonstrated that AsO3 removal rate increased with increasing Mn oxides contents (Amirbahman et al., 2006). Production of solid AsO4 species occurred simultaneously with AsO3 adsorption on solid phases (Fig. 7). Natural gradient tracer study with AsO3 demonstrated that arsenic transport in oxic zones was substantially retardation through the oxidation to AsO4 by Mn oxides in the aquifer (Stadler et al., 2001). Using speciation with XANES spectroscopy, Manning (2005) recently reported that, in a batch reaction, AsO3 was either partially or completely oxidized to AsO4 on soil surface. Fitting XANES spectra as linear combinations of several well-characterized AsO3- and AsO4-treated model compounds demonstrated that the reaction products are AsO3 on Fe oxides and AsO4 on Fe and Al oxides as well as clay minerals.

4.8. Microbial-mediated reduction and oxidation Numerous bacteria, fungi, and algae organisms that are capable of reducing or oxidizing arsenic have been identified and isolated. Under anaerobic condition, reduction of AsO4 to AsO3 is generally mediated through two principal biological mechanisms: (1) dissimilatory reduction (respiration) of microbes such as Escherichia coli, Staphylococcus aureus, and Staphylococcus xylsis where AsO4 is utilized as a terminal electron acceptor; (2) detoxification activity controlled by ars genes that encode AsO4 reduction via an AsO4 reductase, followed by AsO3 release from the cell with an efflux pump (Jones et al., 1997; Langner and Inskeep, 2002; Tamaki and Frankenberger, 1992). Bacteria capable of oxidizing arsenite into arsenate was observed many years ago, first in the cattle dips, later in raw sewage and contaminated soils, as well as in mine deposit. The majority of arsenic oxidizing bacteria, such as Alcaligenes, operates through detoxification mechanism utilizing arsenite oxidase. However, bacteria that are capable of chemolithoautotrophic growth using the energy from arsenite oxidation have been documented (Santini et al., 2000).

A

B

30

60

50

Solid As(V)

Solid As(V) 20

40

30 10

20

aq. As(III)

aq. As(III)

Concentration (mm)

10 Solid As(III)

Solid As(III) 0

0 0

50

100

150

C 100

50

0

D

100

150

500 3

80

2

400 aq. as (III)

Aq. As(III) Solid As(V)

aq.As(V)

1

60

300 0 0

40

200

20

100

50 100

150

200

250

Solid As(V)

Solid As(III)

Solid As(III) 0

0 0

50

100

150

0

50

100

150

200

250

Time (h)

Figure 7 Oxidation kinetics of AsO3 by F168–15 material. pH=5.2, and initial [AsO3]=28.3 mM (A), 59.2 mM (B), 88.7 mM (C), and 483.0 mM (D). All lines are model fits to the experimental data. (Reproduced with permission from Amirbahman et al., 2006.)

Reaction and Transport of Arsenic in Soils

71

In the soil environment, Macur et al. (2004) showed that bacteria capable of either oxidizing AsO3 or reducing AsO4 coexist and are ubiquitously present. Their unsaturated column study showed that AsO3 was readily oxidized into AsO4, whereas no apparent AsO4 reducing was observed. Langner and Inskeep (2000) investigated microbial reduction of arsenate in the presence of ferrihydrite. They found that an AsO4 reducing, glucosefermenting microorganism was able to rapidly reduce the aqueous AsO4 to AsO3 but not able to reduce AsO4 aorbed on ferrihydrite surface. Although the aqueous AsO4 was highly reduced, the desorption rate of AsO4 from ferrihydrite is too slow to cause increase of arsenic solubility. Moreover, Langner et al. (2001) observed the rapid oxidation (first-order rate constant of 1.2 min1) of AsO3 to AsO4 in stream waters from geothermal springs with the presence of live organisms and high Fe/Al content. Microbial-mediated arsenic reduction through either dissimilatory reduction or detoxification pathways generally demonstrates first-order reaction kinetics with half-life of hours to days, depending on the environmental conditions (Langner et al., 2002). Because of the slow kinetics of biological processes, both AsO4 and AsO3 are often found in the soil environment regardless of the redox conditions. For example, Masscheleyn et al. (1991) documented the persistence of AsO4 under reducing condition and AsO3 under oxidizing condition, which was attributed to the slow reaction kinetics. Onken and Hossner (1996) identified both convergence from arsenite to arsenate and arsenate to arsenite in the soil solution under flooded conditions. Many environmental factors impact the microbial-mediated arsenic reduction and oxidation. McGeehan and Naylor (1994) have shown that the reduction of AsO4 to AsO3 was highly dependent on the sorption process of arsenic. Manning and Suarez (2000) observed that heterogeneous oxidation of AsO3 to AsO4 was controlled by soil properties including pH, content of Al, Fe and Mn oxides. Takahashi et al. (2004) found that arsenic quickly released from flooded paddy soils as a result of reductive dissolution of Fe hydro(oxide) accompanied by reduction from AsO4 to AsO3. Reduction of AsO4 to AsO3 was observed in a natural gradient tracer study in the anoxic zone of a sandy aquifer on Cape Code. Microbial analysis with the sediment material revealed the presence of AsO4 reducing microorganisms (Ho¨hn et al., 2006). Other substances occurring may influence redox chemistry of arsenic in soils. For example, Reynolds et al. (1999) reported that the addition of H2PO4 enhanced arsenic reduction rate in two soils. Senn and Hemond (2002) demonstrated that the existence of nitrate under anaerobic condition can oxidize AsO3 into AsO4. The accumulation of nitrate also produced more As-sorbing Fe(III) oxides. As a result, the presence of nitrate reduced the toxicity of arsenic. Methylation of inorganic arsenic species by aerobic and anaerobic microorganisms produces monomethylarsonic acid (MMA), dimethylarsinic acid

72

Hua Zhang and H. M. Selim

(DMA), and trimethylarsine oxide (TMAO). The biological arsenic methylation process in soils can be influenced by abiotic factors, such as pH and temperature. It was demonstrated that several aerobic and anaerobic microorganisms, that is, methanogenic and sulfate-reducing bacteria are accountable for arsenic methylation (Cullen and Reimer, 1989). It is commonly accepted that the biomethylation of arsenic goes through the challenger pathway, where arsenic undergoes an alternating sequence of reduction and oxidative methylation reactions as schematically represented in Fig. 8 (Dombrowski et al., 2005). Once believed to be a detoxification mechanism for arsenic, the methylated species may actually be more toxic and reactive along the methylation pathway. Relatively, few studies have been carried out on the adsorption of methyl-arsenic on minerals and soils. In general, adsorption of arsenic decreases with the methylation. Cox and Ghosh (1994) found that adsorption of MMA(V) and DMA(V) on alumina and hydrous ferric oxide (HFO) was insensitive to changes in ionic strength, indicating that these arsenic species form inner-sphere surface complexes with Fe and Al oxides. In addition, their studies demonstrated that maximum adsorption of MMA(V) and DMA(V) occurs at low pH and the amount of adsorption decreases with increasing pH. Lafferty and Loeppert (2005) observed that adsorption of MMA(III) and DMA(III) on iron oxides was insignificant with pH ranging from 3 to 11.

CH3 + As

−O

OH Arsenite

MMA(III)

OH

HO As

Adding methyl cation

Reduction

n hy lat io et

n tio hy la et M

HO As

CH3

CH3+

CH3

CH3 DMA(III)

••

OH

CH3

CH3

••

As(III) HO As

Reduction

n tio hy la et

CH3

••

−O

OH

CH3

OH

M

Reduction

CH3 + As

+ −O As OH

OH Reduction

TMA(V)

M

OH + −O As OH

DMA(V)

CH3

••

As(V)

MMA(V)

Reduction

Arsenate

As

CH3

CH3 TMA(III)

Figure 8 Schematic of the Challenger pathway for biomethylation, with the alternating sequence of reduction and oxidative methylation of arsenic. (Reproduced with permission from Dombrowski et al., 2005.)

Reaction and Transport of Arsenic in Soils

73

5. Transport in Soils 5.1. Transport mechanisms The transport of arsenic in heterogeneous soils is largely controlled by adsorption–desorption processes on solid matrix surfaces. Non-nonlinear or concentration dependent as well as kinetic adsorption–desorption behavior is often regarded as responsible for observed nonequilibrium transport of arsenic in soils. Melamed et al. (1995) observed asymmetrical AsO4 breakthrough curves (BTCs) from columns of an Oxisol and suggested that physical and chemical nonequilibrium as the dominant processes for arsenic movement in soils. Kuhlmeier (1997a) investigated the transport of As in columns of silty and sandy soils sands and quantified time- and concentration-dependent Kd values. In a separate column experiment with a contaminated clayey soil, Kuhlmeier (1997b) suggested that the slow release of arsenic as a result of kinetically controlled or rate limited mass transfer in soils. Darland and Inskeep (1997a,b) found that AsO4 transport exhibited significant retardation, tailing, and poor recovery. Nonequilibrium of AsO4 transport in a subsurface soil with high arsenic sorption capacity was also demonstrated by Williams et al. (2003). Because of the nonequilibrium behavior of arsenic transport in soils, water flow velocity (or residence time) significantly impacts its mobility in soils. Puls and Powell (1992) observed that the distribution factors (Kd) determined from column transport experiments decreased from 3.0 L kg1 to 1.4 L kg1 when the flow rate (q) was doubled. Darland and Inskeep (1997a) demonstrated that increasing pore volume velocity from 0.2 to 90 cm h1 resulted in some tenfolds increase of AsO4 leaching or recovery from saturated sandy columns (Fig. 9). Williams et al. (2003) reported enhanced AsO4 mobility and decreased retardation of BTCs in a subsurface soil by increasing the pore water velocity. Nikolaidis et al. (2004) studied the mobility of arsenic in contaminant lake sediments and found that a significant portion of arsenic was leached due to increased flow velocity. Furthermore, Radu et al. (2005) studied the effect of pore water velocity (0.23 and 2.3 cm/min) on the transport of AsO3 in saturated columns of sand. Their results demonstrated that the increasing pore water velocity increased the mobility of AsO3 in goethite-coated sand columns. A number of column studies have been conducted to evaluate the impact of various conditions on the transport of arsenic in saturated soils. Hiltbold et al. (1974) studied the transport of MSMA in several soils. The BTCs from surface soils exhibited relatively little retardation, whereas extensive retardation was observed in BTCs from the subsoils. The Kd values calculated from batch and column experiments showed distinct

74

Hua Zhang and H. M. Selim

Relative concentration (C/C0)

0.6 As recovery in effluent @ PWV 10 PVs 0.2 cm/h 7.24% 1 cm/h 35.6%

0.5 0.4

10 cm/h 90 cm/h

0.3

53.3% 74.3%

0.2 0.1 0.0 0

2

4

6

8

10

Pore volumes

Figure 9 Breakthrough curves of AsO4 at four different pore water velocities (0.2, 1, 10, and 90 cm h1). Column conditions: background 0.01 M KCl; AsO4 pulse 1 pore volume at 133 mM. (Reproduced with permission from Darland and Inskeep, 1997a.)

Relative concentration (C/C0)

1.0

pH 4.5, 0.53 cm min−1 pH 4.5, 0.53 cm min−1, 0.25 mM PO4

0.8

pH 9, 0.53 cm min−1 pH 4.5, 1.6 cm min−1

0.6

0.4

0.2

0.0 0

500

1000

1500

2000

Pore volume (V/V0)

Figure 10 Breakthrough curves of AsO4 under varying pH, phosphate, and pore water velocity. Column conditions: background 0.01 M NaNO3; AsO4 initial concentration at 1 mg l1. (Reproduced with permission from Williams, 2001.)

discrepancies which was attributed to the rather limited residence time of arsenic in the column (high flow rate). Williams et al. (2003) found that increasing ionic strengths from 0.01 M to 0.1 M extended AsO4 adsorption (Fig. 10). They explained that increases in ionic strength lessen the electrostatic repulsion between negatively charged surfaces and oxyanion which results in an increased adsorption of arsenic onto mineral surfaces.

Reaction and Transport of Arsenic in Soils

75

Several miscible displacement experiments provided evidence that the presence of PO4 greatly enhanced arsenic mobility in soils. Woolson et al. (1973) found that 77% of arsenic in a sandy soil column was leached due to the application of 0.05 M KH2 PO4 solution. The significant enhanced transport of AsO4 through columns of an aggregated Oxisol by the increasing addition of phosphate was demonstrated by the left shifted BTCs observed by Melamed et al. (1995). Similarly, Peryea and Kammereck (1997) found that the application of phosphate significantly mobilized arsenate and resulted in the leaching of 50% of soil arsenic after 10 pore volumes. Qafoku et al. (1999) studied the effect of competitive anions (phosphate and sulfate) on arsenic transport through a packed Cecil soil column amended with fly ash. They found that with calcium phosphate as the displacing solution, arsenate concentration in the leachate increased an order of magnitude when compared to calcium sulfate as the displacing solution. The displacement of AsO4 by phosphate under dynamic flow conditions were also reported by other researchers including Darland and Inskeep (1997b), Williams et al. (2003), among others. Column experiments were employed to study the effect of carbonate on arsenic transport and found that the presence of 0.1 mM CO3 resulted in a limited increase of AsO4 transport in columns of subsurface soil. Radu et al. (2005) studied the effects of high aqueous carbonate concentrations on the transport of arsenic in a synthetic iron oxide-coated sand columns. They found that increasing carbonate concentrations had relatively little effect on the adsorption and transport of AsO3 and AsO4. A number of experiments demonstrated that AsO4 was generally more mobile under high pH condition which is due to decreased adsorption. For example, Mariner et al. (1996) reported that Kd values measured on excavated core samples of contaminated aquifer material decreased at least tenfold as the pH increases from 8.5 to 11. Darland and Inskeep (1997b) found that increasing the pH from 4.5 to 8.5 caused a retardation or a delay of the AsO4 BTC. A more symmetrical BTC was observed at pH 8.5 due to increased AsO4 mobility. Similar effect of pH was observed by Williams et al. (2003) on a subsurface soil having high iron content. However, Radu et al. (2005) reported early breakthrough for AsO3 at pH 4.5 than at pH 9, as a result of increased adsorption of AsO3 under high pH. The oxidation reduction potential of the geological materials determines chemical speciation of arsenic, hence substantially impacts the extent of arsenic retention and transport. A series of small-scale natural gradient tracer experiment with AsO3 and AsO4 were conducted in oxic, suboxic, and anoxic zones of a shallow sandy aquifer at Cape Cod, Massachusetts by Stadler et al. (2001)and Ho¨hn et al. (2006). Stadler et al. (2001) observed that a significant portion of applied AsO3 was oxidized to AsO4 2.2 m downstream from the injection well in both oxic and suboxic zones. There was relatively little retardation of total arsenic breakthrough relative to

76

Hua Zhang and H. M. Selim

conservative Br tracer in suboxic zone. However, substantial retardation (R ¼ 1.3) and irreversible retention (one-third of the applied arsenic) was observed in oxic zone. The transformation from AsO3 to AsO4 in the aquifer was explained by abiotic oxidation on the surface of Mn oxides. Ho¨hn et al. (2006) observed that applied AsO4 was reduced to AsO3 some 1 m downstream of the injection well. This process was attributed to AsO4 reducing microorganisms in the anoxic zone. A spatial snapshots of AsO4 and AsO3 30, 45, 63, and 104 days after injection are illustrated in Fig. 11 using two-dimensional longitudinal transects. Their results demonstrated that the reduction of AsO3 was occurred after nitrate concentration had been attenuated by reaction with Fe(II) and the transport of AsO3 was faster than AsO4 under reducing chemical conditions. Biotransformation between AsO3 and AsO4 as a result of microbial activity might be an important factor contributing to the retention and transport of arsenic in aquifer and vadose zone. Herbel and Fendorf (2006) investigated the mobilization of adsorbed arsenic [AsO4 or AsO3] in saturated columns of ferric hydroxide-coated sands. The sand was inoculated with Surfurosprillum barnesii, an Fe(III) and AsO4 respiring bacterium, or Bacillus benzoevorans, an AsO4 respiring bacterium incapable of Fe(III) reduction. Extensive release of arsenic, predominantly in the form of AsO3, was observed and the desorption was promoted by mildly reducing conditions where limited Fe(II) was present. Another important factor affecting in arsenic mobility in soils is that of colloid facilitated contaminant transport in porous media which was recognized since the early 1990s. Traditionally, contaminants are assumed to be either immobilized due to sorption by the solid phase or mobile in the aqueous phase. However, there is considerable evidence that nonaqueous mobile colloid could transport low solubility contaminants for considerable distances (Kretzschmar et al., 1999). Puls and Powell (1992) conducted column experiments with aquifer material to investigate the possible effect of colloidal iron oxide on facilitating AsO4 transport. They observed substantial mobilization of colloid associated arsenate when flushing the column with DI water. Ishak et al. (2002) investigated arsenic leaching from fly ash through an loamy sand column showed that arsenic levels the leachates correlated with effluent turbidity, which support the supposition that arsenic movement was generally associated with mobilized colloids. Recently, Ghosh et al. (2006) observed that arsenic release from granular ferric hydroxide residuals under mature landfill conditions was associated with suspended particulate matter and mediated by microbial reduction of ferric hydroxides. Zhang and Selim (2007a) evaluated colloid mobilization by changing the background solution from 0.01 M NaCl to DI water and its effect on arsenic(III) transport in soils. Their results revealed that colloid facilitated transport contributed little to arsenic movement under steady flow and constant ionic strength. However, enhanced transport of arsenic

A M2

M3

B M9 M2

M5

M3

M5

M9

2.5 2

5 4

0.02

0.06

2

0.06

0.06

2

30th day

1.5

0.02

0.02

As(V) [mM]

1

As(III) [mM]

2.5 2

2.4

1.2

2 1.5

0.2

0.6

0.4

1.2 0.4

45th day

1 2.5 0.4

0.8

1.6 0.4

1.5

0.

2

0.8 1.2

2

0.2

63th day 1

MSL [m]

2.5 0.4

0.6

2

0.4

1 1.4

0.25

0.25

1

14 1

1

1.4 0.1

1.5 0.1

104th day 1 0

1

2 3 Distance [m]

4

5 0

1

2 3 Distance [m]

4

5

Figure 11 Longitudinal profiles showing concentrations of (A) AsO4 and (B) AsO3 30, 45, 63, and 104 days after starting injection. Mean flow velocity is 0.3–0.4 m d1. (Reproduced with permission from Ho¨hn et al., 2006.)

78

Hua Zhang and H. M. Selim

by colloid generation was observed as a result of changing ionic strength and flow interruption (Fig. 12). Preferential flow occurring via distinct flow pathways in clay or sand lenses, cavities, fissures, and other macropores is another factor influencing the movement of arsenic through soils. Few studies have been conducted to investigate arsenic transport through soils when preferential flow conditions are dominant. Simulations conducted by Corwin et al. (1999) demonstrated that, a lysimeter column study, without accounting for preferential flow, 100% of the applied As was isolated in the top 0.75 m over a 2.5-year period. However, when preferential flow was considered and a representative bypass coefficient was used, only 0.59% of applied As moved beyond 1.5 m, which compares favorably with that measured. Hopp et al. (2006) attributed the transport of arsenic in the vadose zone of a former wood preserving site to strong preferential flow patterns which was supported by a dye experiment.

5.2. Mobility under field conditions Elevated concentrations of arsenic in soils and aquifers have caused concern over potential contamination of surface and groundwater from arsenic release due to its leaching. Downward movement of arsenic has been observed in soils contaminated with arsenic pesticides. Isensee et al. (1973) investigated arsenate residual in Metapeake silt loam 14 years after massive application of arsenical herbicides. They found that large amount of arsenic remained in the soil profile and the concentration decreased with depth, indicative of slow As leaching. Hiltbold et al. (1974) studied MSMA transport in surface and subsurface Dothan and Decatur loamy soils and reported that leaching of MSMA was not observed. This is possibly because of the relatively short period after herbicide application (6 years) and high As adsorption of the subsoils. Peryea and Creger (1994) investigated the vertical distribution of arsenic in an orchard soil contaminated by historical application of PbAsO4 insecticides. Their results demonstrated that arsenic was relatively depleted in the surface soil but enriched in subsoil and a significant amounts of As (0.07–0.63 mmol/kg soil) were leached to the depth of 120 cm. Mariner et al. (1996) studied the mobilization of arsenic from buried sodium arsenite waste pits to adjacent waterway through a high-pH groundwater plume. They found that low soil permeability significantly affect transport of As in soils. Corwin et al. (1999) conducted a mesoscale (0.6 m in diameter and 1.83 m in height) weighing lysimeter experiment to assess the mobility of As through the root zone of a Rocky Mountain Arsenal soil. Their results suggested that even though the movement of As is significantly retarded due to adsorptive processes, preferential flow and chemical perturbation might mobilize arsenic above environmental standards. Robinson et al. (2007) investigated the arsenic concentration in soils, stream sediments,

79

Reaction and Transport of Arsenic in Soils

8 Olivier DIW leaching Begins

Arsenic concentration (mg L−1)

6

Total Dissolved 4

2

0

10

0

20

30

40

5 Windsor

Arsenic concentration (mg L−1)

4

3 DIW leaching Begins 2

1 Total 0

Dissolved

0

10

20 30 Pore volumes (V/Vo)

40

50

Figure 12 Breakthrough curves of total and dissolved arsenic for the Olivier and Windsor soil columns. Arrows indicate pore volumes when flow interruptions or leaching with deionized (DI) water occurred. Initial AsO3 concentration Co ¼ 10 mg l1. Pore water velocities are 0.83 and 1.03 cm h1 for Olivier and Windsor columns, respectively. (Reprinted with permission from Zhang and Selim, 2007a.)

80

Hua Zhang and H. M. Selim

and groundwater near old orchards with historical application of arsenic pesticides in Virginia and West Virginia. Their results indicated that arsenic in pesticide contaminated soils had limited mobility to the groundwater. Soils around historical cattle-dipping vats commonly contain high concentrations of arsenic. There were fears that arsenic in such contaminated soils may be leached to groundwater and cause contamination of the drinking water resources. Mclaren et al. (1998) investigated soils surrounding cattle dips and showed that considerable movement of arsenic down through the soils had occurred. Arsenic in the subsurface soil (20–40 cm) near cattle-dip sites in Australia can be as high as 2282 mg As kg1. However, the migration of arsenic has been found to be slow, controlled by chemical and physical properties of soils. Kimber et al. (2002) analyzed the shallow groundwater around 28 cattle-dipping vats in Australia with piezometers. The highest arsenic concentration (5.69 mg l1) in the groundwater was found adjacent to a contaminated site with soils of sandy texture. The concentration declined to approximately background levels with 20 m distance from the contaminated site. There are many industrial sites contaminated by CCA which may pose a serious threat to the groundwater. Studies demonstrated that the arsenic contained in CCA leached from the wood surface to adjacent soil (Chirenje et al., 2003). Moreover, it is also possible that arsenic may be partially released into aquatic environment causing human health concern. Andersen et al. (1996) sampled soil material and soil solution at a wood impregnation site and provided evidence of both a strong accumulation and a high mobility of As and Cr. Allinson et al. (2000) investigated the release of CCA constituents from undisturbed soil monolith lysimeters of a sandy loam soil. They observed that up to 13% of the applied arsenic was detected in the leachate at 15 cm depth and breakthrough was observed 25 days after CCA application. Hingston et al. (2001) conducted literature review on the leaching studies of CCA-treated wood and they concluded there is insufficient data to quantitatively predict the leaching rate of elements under various environment conditions (pH, salinity, and temperature). More recently, Khan et al. (2006a,b) investigated the leaching of arsenic from CCA-treated wood during service as well as disposal with lysimeter tests. They found cumulatively around 2000 mg arsenic was leached out from landfills in a 1-year period, with inorganic AsO4 and AsO3 as the major species. Hopp et al. (2006) monitored the spatial and temporal variability of As and Cr in the soil pore water at a former wood-preserving site. They found that the high spatial variability (up to three orders of magnitude) of arsenic levels in soil pore water. They suggested that spatial variability is due to heterogeneous and uneven leaching patterns caused by water repellency of the surface soil and preferential flow. In contrast As concentrations in the soil water, only low As concentrations (<12 mg l1), were detected in the groundwater at 4.5 m depth.

81

Reaction and Transport of Arsenic in Soils

6. Modeling 6.1. Equilibrium thermodynamic models Geochemical models based on thermodynamic equilibria are frequently used for the prediction of arsenic reactions in solid and aqueous phases as well as its mobility in the environment. The standard Gibbs free energies for the element reactions are required for the calculation of the equilibrium state of multicomponent systems,

DG ¼ RT lnKeq

ð9Þ

where Keq is the equilibrium reaction constant, and DG is the standard Gibbs energy for the formation of unit mole of a compound in its standard state. The standard Gibbs energy of some important arsenic species are given in Nordstrom and Archer (2003). On the basis of thermodynamic laws, equilibrium or rate constants of chemical reactions are functions of temperature. The effect of temperature on certain reaction constant k can be described with Arrhenius equation:

Keq ¼ AeEa =RT

ð10Þ

where A is constant pre-exponential factor and Ea is the energy of activation, which is commonly from the slope of a plot of ln Keq versus 1/T. The calculation of equilibrium constants Keq are based on the activities (ai) of ions in multicomponent solution. The activity coefficients of charged ionic species can be calculated from solution ionic strength (I ) using extended Debye-Hu¨ckel equation:

log g ¼

Az2 I1=2 þ bI 1 þ BaI1=2

ð11Þ

where gi is the single-ion activity coefficient of ionic species i, constant A ¼ 0.512 in water at 25 C, B ¼ 0.33 in water at 25 C, a and b are adjustable parameters corresponding to the size of the ion. The commonly used Davies equation is modified from Debye-Hu¨ckel limiting law with the form of

log gi ¼

Az2i

I1=2 0:3I 1 þ I1=2

ð12Þ

Both arsenic acid (H3AsO4) and arsenious acid (H3AsO3) are weak triprotic acids. The distribution of acid and base forms of the anions as a function of pH can be calculated with the Henderson-Hasselbalch equation:

pH ¼ pKa þ log

ðA Þ ðHAÞ

ð13Þ

82 Table 1

Hua Zhang and H. M. Selim

Dissociation constants for various arsenic species

Name

Formulae

pKa1

pKa2

pKa3

Arsenic acid [AsO4] Methylarsonic acid [MMAsO4] Dimethylarsonic acid [DMAsO4] Arsenious acid [AsO3]

H3AsO4 H2AsO3(CH3)

2.20 4.19

6.97 8.77

11.53 –

HAsO2(CH3)2

6.14

–

–

H3AsO3

9.22

12.13

13.4

Lafferty and Loeppert, 2005.

where A is base form and HA is acid form. Using the dissociation constants ( pKa) given in Table 1, the distribution of various protonated forms of arsenic species under certain pH value can be calculated using Eq. (12) (Goldberg, 2002). At near-neutral pH, reductive transformation between arsenate and arsenite anions can be described with equation:

1 3 þ 1 1 H2 AsO 4 þ e þ H , H3 AsO3 þ H2 O 2 2 2 2

ð14Þ

The equilibrium thermodynamic relationship between pH, electron activity ( pe), and the equilibrium speciation of the anions can be described with Nernst equation: (Bohn,1976)

H2 AsO 3 1 4 pe ¼ pe pH þ log 2 2 ðH3 AsO3 Þ

ð15Þ

where standard pe ¼ DG =2:303RT . Figure 13 presents a calculated pH-pe diagram describing the equilibrium distribution of dominant arsenic species in solution as a function of pH and pe. Equilibrium thermodynamic calculation based on the standard reaction constants for AsO4/AsO3 redox couple are commonly used to describe the equilibrium distribution of arsenic species under various Eh and pH conditions. The simulation using thermodynamic data showed that AsO4 species are more abundant at pe þ pH > 9, while AsO3 forms dominate with pe þ pH < 7 (Sadiq, 1997; Sadiq et al., 1982). However, it is well known that a single measure of redox potential ( pe) does not provide sufficient information for simultaneously determining the ratio of oxidized species to reduced species for all redox couples. Equilibrium arsenic concentration in mixed solid solution system is determined by the solubility of minerals, which is commonly described

83

Reaction and Transport of Arsenic in Soils

24 20 1.0

16

H3AsO40 H2AsO4−

pe

8

0.5 HAsO42−

4 H3AsO30

0.0

AsO43−

0

Eh, in volts

12

−4 −8

−0.5

H2AsO3−

−12 −16

0

2

4

6

8

10

12

14

pH

Figure 13 pe-pH diagram for predominant aqueous species of arsenic at equilibrium and 298.15 K and 1 atmosphere pressure. (Reproduced with permission from Nordstrom and Archer, 2003.)

with solubility product constant (Ksp, mol l1). Consider the dissolution of scorodite in neutral solution which can be expressed in the form of:

FeAsO4 2H2 OðsÞ þ Hþ ðaqÞ , H2 AsO 4 ðaqÞ þ FeðOHÞ2þ ðaqÞ þ H2 O

ð16Þ

At equilibrium, this reaction can be described by

2þ Fe ð OH Þ H2 AsO 4 Ksp ¼ ðFeAsO4 2H2 OÞs ðHþ Þ

ð17Þ

To determine if the solution phase is in equilibrium with the scorodite mineral, the ion activity product (IAP) is compared with the solubility product (Ksp) using the saturation index (SI )

SI ¼ log

IAP Ksp

ð18Þ

Solution is at thermodynamic equilibrium with respect to the mineral when SI ¼ 0. When SI > 0, solution is supersaturated and scorodite mineral

84

Hua Zhang and H. M. Selim

should precipitate. On the contrary, if SI < 0, solution is undersaturated and scorodite mineral should dissolve. The term of scorodite (FeAsO42H2O) is usually not presented in Eq. (16) because the activity of pure solid phases in standard state is defined as being equal to one. However, most naturally occurring arsenic minerals are deviated from pure solid state due largely to isomorphous ion substitution in the crystalline lattice. This results in the mixture of two or more solid mineral phases, which is called as solid solution. Assuming ideal ion substitution, the free energy of a binary solid solution may be represented by (Davis et al., 1996)

DGi;j ¼ xDGi þ ð1 xÞDGj þ DGmix

ð19Þ

where DGi;j is the standard free energy of the solid solution, x is the mole fraction of mineral component i, and DGi and DGj are the free energies of mineral components i and j. Assuming random mixing, the free energy of mixing (DGmix ) is given by

DGmix ¼ nRT ½x ln x þ ð1 xÞlnð1 xÞ

ð20Þ

Using Eqs. (18) and (19), Davis et al. (1996) calculated the free energy of anhydrous arsenate–phosphate solid solutions of Cu, Fe, Pb, and Zn, and the resulting DGi;j was used to calculate the limiting solubilities. Thermodynamic models have been employed for the simulation of equilibrium distribution of arsenic species regarding arsenic sulfide minerals. For example, Craw et al. (2003) calculated the Eh-pH diagram for As-Fe-S-O system, which showed that FeAsS was relatively stable in the surficial environment. The prediction capacity and quality of thermodynamic models are highly dependent on the data input on stability constants or solubility constants. As pointed out by Nordstrom and Archer (2003), many constants in the published thermodynamic databases are derived from various sources and may suffer from internal inconsistencies.

6.2. Empirical equilibrium models Equilibrium batch experiment, carried out by equilibrating arsenic solution with mineral or soil solid for a certain amount of time (usually 24 h) at constant temperature, is employed to quantify the adsorption capacity and affinity of arsenic in soils. The relationship between the equilibrium concentration in the aquatic solution and the amount adsorbed on the solid surface are commonly described with adsorption isotherms. The L-type and H-type isotherms are usually employed to describe the arsenic adsorption on mineral and soil surfaces. Both types are highly nonlinear (concentration

85

Reaction and Transport of Arsenic in Soils

dependent) and indicative of high affinity chemical adsorption, which are commonly described with either an equilibrium model of the Freundlichor Langmuir-type. The Freundlich equation is an empirical adsorption model that can be expressed as

S ¼ KF C N

ð21Þ

where S represents the (total) amount of adsorption (mg kg1), KF is the distribution or partition coefficient (mg kg1 (mg l1)N ), and N is the dimensionless reaction order commonly less than 1. The Langmuir equation has the advantage of providing a sorption maximum Smax (mg kg1) that can be correlated to soil sorption properties. It has the form:

S ¼ Smax

KL C 1 þ KL C

ð22Þ

where KL (L mg1) is a Langmuir coefficient related to the binding strength. Langmuir and Freundlich models can be incorporated into solute transport models to predict the transport of arsenic in soils. The nonlinear adsorption behavior of arsenic was numerically described with Langmuir equation (Darland and Inskeep, 1997a; Livesey and Huang, 1981; Manning and Goldberg, 1997; Pierce and Moore, 1980) and Freundlich equation (Elkhatib et al., 1984a; Puls and Powell, 1992; Smith et al., 1999; Williams et al., 2003). The value of model parameters (KF, N, KL, and Smax) varies considerably among various studies, and depends on several factors such as the soil mineralogy, reaction time, solution pH, initial arsenic concentration, and oxidation state. For arsenate adsorption on a soil from Rocky Mountain area (Corwin et al., 1999), the Langmuir coefficient (KL ) were 1.74, 2.92, and 6.13 L kg1 for temperature of 8, 25, and 40 C, respectively, whereas no significant difference were found between adsorption maxima (66 mg kg1) for three temperatures. The presence of ligands can compete with arsenic for adsorption sites on mineral surfaces. Sheindorf et al. (1981) introduced a modification of Freundlich Eq. (1) in order to account for adsorption when more than one competing ions is present in the solution. Specifically, the SheindorfRebhun-Sheintuch (SRS) model was developed to describe competitive equilibrium sorption for multicomponent systems. Here it was assumed that for a single component, sorption isotherms follow the Freundlich equation. The derivation of SRS equation was also based on the assumption of an exponential distribution of adsorption energies for each component. A general form of the SRS equation may be written as

Si ¼ KFi Ci

l X j¼1

!Ni 1

aij Cj

ð23Þ

86

Hua Zhang and H. M. Selim

where l is the total number of components, i and j represent components i and j aij is a dimensionless competition coefficient that describes the inhibition by component j to the adsorption of component i where aij ¼ 1 when i ¼ j. In the absence of competitive sorption, that is, aij ¼ 0 for i 6¼ j, Eq. (22) yields the Freundlich Eq. (20) for a single component. Equation (22) was successfully employed by Roy et al. (1986a,b) to describe the competitive adsorption isotherms of AsO4 and PO4 in several soils. Barrow et al. (2005) have developed a scheme to numerically solve the nonlinear set of equations through iterative improvements and estimate the competitive coefficient with nonlinear optimization. Zhang and Selim (2007b) extended the equation to kinetic form and incorporated it in the solute transport model for the simulation of competitive transport of AsO4 and PO4 in soils.

6.3. Surface complexation models Empirical laws, such as the Langmuir or Freundlich equations, do not provide specific information about the retention mechanisms. As a matter of fact, a number of reactions including ion exchange, outer-sphere surface complexation, inner-sphere surface complexation, and/or surface precipitation can be embraced in simple equilibrium models. Unlike empirical models, surface complexation models (SCMs) are chemical models that give a general molecular description of adsorption phenomena using an equilibrium approach. The models treat arsenic adsorption as inner-sphere surface complexation reactions through ligand exchange mechanism. Because surface sites are explicitly defined as the number of functional groups SOH on mineral surfaces, those model predict a Langmuir-type adsorption isotherm. SCMs are commonly employed to describe the adsorption envelopes, that is, adsorption of arsenic anions as a function of solution pH. Various types of SCM have been proposed based on different assumption of the distribution of surface electrostatic potential. The constant capacitance model (CCM) assumes linear relationship between surface charge (s, mol L1) and surface potential (c, volts), which can be described with (Goldberg, 1992):

s¼

CAs MV c F

ð24Þ

where As is the specific surface area of absorbent (m2 g1), MV is the suspension density of absorbent (g L1), capacitance density C ¼ 1.06 F m2, and Faraday constant F ¼ 9.65 10–4 coulombs mol1. Table 2 gives the reactions used to calculate monodentate and bidentate surface complexation of arsenate on the goethite surface:

Table 2 Surface complexation reactions and constant capacitance model (CCM) intrinsic surface complexation constants for arsenate adsorption on goethite Reaction

Surface hydrolysis reactions ð1Þ XOH þ Hþ ¼ XOHþa 2

ð2Þ XOH ¼ HO þ Hþ

Equilibrium expressions and constants

½XOHþ Fc 2 ¼ 107:31b exp RT ½XOH½Hþ ½XO ½Hþ Fc Ka2 ðintÞ ¼ exp ¼ 108:81b ½XOH RT

Ka1 ðintÞ ¼

Formation of inner-sphere monodentate oxyanion/goethite surface complexes ð3Þ XOH þ H3 AsO4 ¼ XH2 AsO4 þ H2 O þ ð4Þ XOH þ H3 AsO4 ¼ XHAsO 4 þ H2 O þ H þ ð5Þ XOH þ H3 AsO4 ¼ XHAsO2 4 þ H2 O þ 2H

½XH2 AsO4 ¼ 1010 ½XOH½H3 AsO4 þ XHAsO Fc 4 ½H 2 exp KAs ðintÞ ¼ ¼ 105:1 ½XOH½H3 AsO4 RT þ2 2 XAsO ½H Fc 4 3 exp KAs ðintÞ ¼ ¼ 100:55 RT ½XOH½H3 AsO4 1 KAs ðintÞ ¼

Formation of inner-sphere bidentate oxyanion/goethite surface complexes ð8Þ 2XOH þ H3 AsO4 ¼ X2 HAsO4 þ 2H2 O þ ð9Þ 2XOH þ H3 AsO4 ¼ X2 AsO 4 þ 2H2 O þ H

4 KAs ðintÞ ¼

½X2 HAsO4 ¼ 1017 ½XOH2 ½H3 AsO4

5 KAs ðintÞ

þ X2 AsO 4 ½H

Fc exp ¼ 2 RT ½XOH ½H3 AsO4

¼ 1011:4

88

Hua Zhang and H. M. Selim

The intrinsic equilibrium reaction constants

½Kþ ðintÞ; K ðintÞ; K1sAsO4 ðintÞ; K2sAsO4 ðintÞ; K3sAsO4 ðintÞ; K1sAsO3 ðintÞ; and K2sAsO3 ðintÞ are commonly obtained through fitting the batch reaction data to the model. The mass balance for the surface functional group is: ½SOHT ¼ ½SOH þ bSOHþ 2 c þ bSO c þ ½SH2 AsO4 2 þ bSHAsO 4 c þ bSAsO4 c þ ½SH2 AsO3 þ bSHAsO3 c

ð25Þ and the charge balance expression is: 2 s ¼ bSOHþ 2 c bSO c bSHAsO4 c 2bSAsO4 c bSHAsO3 c

ð26Þ Constant capacitance model was fitted to the envelope of arsenic adsorption on several pure minerals to obtain the surface complexation constant which can be used to predict the adsorption behavior of arsenic. The model has been successfully employed for the description of adsorption envelope of arsenate adsorption on Al and Fe oxides (Goldberg, 1986), arsenate on calcareous soils (Goldberg and Glaubig, 1988), arsenite on amorphous Al and Fe oxides (Goldberg and Johnston, 2001). Furthermore, the reactions and corresponding reaction constants for competing ligands (e.g., arsenite, phosphate) can be added to the reaction system for the simulation of multicomponent adsorption. The CCM model successfully described the competition of arsenate with phosphate and molybdate on oxides minerals (Manning and Goldberg, 1996a), arsenate with phosphate on clay minerals (Manning and Goldberg, 1996b). In addition, Goldberg (2002) simulated the competition between arsenate and arsenite on Fe and Al oxides as well as clay minerals. However, CCM model predictions can only qualitatively describe the shape of adsorption curves (Goldberg, 2002; Goldberg and Glaubig, 1988; Manning and Goldberg, 1996, 1997; Williams et al., 2003). The diffuse double layer (DDL) model developed by Dzombak and Morel (1990) was frequently used to describe the arsenic sorption on oxides (Dixit and Hering, 2003). Like the CCM model, this model assumes that all surface complexes are inner-sphere surface complexes. However, the DDL model considers a diffuse layer where the relationship between surface

89

Reaction and Transport of Arsenic in Soils

charge and surface potential can be described with the Gouy-Chapman equation:

s ¼ 0:1174I 1=2 sinh

zFc 2RT

ð27Þ

where I is ionic strength, z is the valence of the symmetrical electrolyte and commonly takes as unity. Similar to CCM model, a set of equilibrium constants defined for the surface reactions are required as the model parameter. The intrinsic surface complexation constants for HFO, goethite, and magnetite provided by Dixit and Hering (2003) are presented in Table 3. SCMs have been incorporated in geochemical modeling packages such as MINTEQA2 (Allison et al., 1991) and PHREEQC2 (Parkburst and Appelo, 1999). A database compiled by Dzombak and Morel (1990) for the surface reactions of cations and anions on HFO is included in those models and frequently employed to predict the retention and transport of arsenic in soil and water environment. For example, adsorption of arsenic anions during its transport in Benthic sediment has been simulated with the DDL model by Smith and Jaffe (1998) using MINTEQA2. Using DDL model in PHREEQC2, Appelo et al. (2002) simulated the competitive effect of ferrous iron and carbonate on arsenic sorption on ferrihydrite. Table 3 Intrinsic surface complexation constants for hydrous ferric oxide (HFO), goethite, and magnetite (Bixit and Herring, 2003) Intrinsic surface complexation constants

HFO

Goethite

Magnetite

FeOH þ Hþ ¼ FeOHþ 2

7.29

7.47

4.60

8.93

9.51

8.20

29.88

31.00

24.43

26.81

18.10

20.22

¼

þ H2 O Arsenate adsorption constants þ

FeOH þ AsO3 3 þ 3H

38.76

39.93

38.41

¼ FeH2 AsO3 þ H2 O þ

FeOH þ AsO3 3 þ 2H

31.87

32.40

33.02

FeOH ¼ FeO þ Hþ Arsenate adsorption constants þ

FeOH þ AsO3 4 þ 3H ¼ FeH2 AsO4 þ H2 O þ

FeOH þ AsO3 4 þ 2H ¼ FeHAsO 4 þ H2 O þ

FeOH þ AsO3 4 þH FeAsO2 4

¼

FeHAsO 3

þ H2 O

90

Hua Zhang and H. M. Selim

6.4. Kinetic models 6.4.1. Kinetic Retention In general, the rate of sorption is a dependent on the following processes: (1) the transport of arsenic anions from bulk solution to the reaction sites on mineral surfaces; and (2) the chemical reaction at the surfaces. Specifically, transport processes include: (i) diffusion in the aqueous solution; (ii) film diffusion at the solid/liquid interface; (iii) intraparticle diffusion in micropores and along pore wall surfaces; and (iv) interparticle diffusion inside solid particles. The chemical processes may include reactions such as ion exchange, formation of inner-sphere surface complexes, precipitation into distinct solid phases, or surface precipitation on minerals (Sparks, 1998). Because of the complexity of the sorption process, it is impractical to derive the mechanism-based reaction rates. Instead, a wide variety of empirical kinetic rate expressions was developed in the last three decades. Those kinetic equations have been widely applied to describe the results from sorption and desorption kinetic experiments. The pseudo first-order equation assumes a fixed amount of adsorption at equilibrium (Seq). It has the form of:

dS ¼ k Seq S dt

ð28Þ

where k is the sorption rate (h1). For the initial condition S ¼ 0 at t ¼ 0, the integrated form of the pseudo first-order equation for adsorption is:

S ¼ 1 ekt Seq

ð29Þ

Similarly, the pseudo second-order equation has the form of:

2 dS ¼ k Seq S dt

ð30Þ

Again for the initial condition S ¼ 0 at t ¼ 0, the integrated form of pseudo second-order equation for adsorption is:

1 1 þ kt ¼ Seq S Seq

ð31Þ

Another kinetic adsorption equation is that of the fraction power equation where,

dS ¼ kt n1 dt

ð32Þ

91

Reaction and Transport of Arsenic in Soils

and n is a parameter between 0 and 1. Integration (when S ¼ 0 at t ¼ 0 the familiar adsorption equation,

S ¼ kt n

ð33Þ

which represents nonlinear kinetic adsorption. This equation can be extended to describe desorption as follows. When S ¼ Sc at t ¼ tc, desorption can be expressed as

S ¼ Sc k t n tcn

ð34Þ

Retention of solute by soil is commonly described with the empirical Elovich equation of the form:

dS ¼ aebS dt

ð35Þ

where a is the initial adsorption rate and b is a constant. Assuming abt>>1 and initial condition S ¼ 0 at t ¼ 0, the above rate equations yields this model expressing a linear relationship between S and ln t:

1 1 S ¼ lnðabÞ þ lnðtÞ b b

ð36Þ

This model can also be employed to describe desorption kinetics of solute from the surfaces of adsorbent by applying the condition: S ¼ Sc at t ¼ tc, which yields

1 1

t S ¼ lnðabGÞ þ ln 1 þ b b G

ð37Þ

where G ¼ ebSc =ab tc . Even though adherence to the Elovich model has been proposed as evidence of diffusion-controlled retention mechanism, the use of the model alone to describe kinetic data should not be used for determining retention mechanisms (Fuller et al., 1993). The parabolic diffusion model is based on the assumption of diffusioncontrolled rate-limited process in media with homogeneous particle size. The parabolic diffusion equation was derived from the Fick’s second law of Diffusion in radial coordinate system.

3=2 S 4 Dt 1=2 Dt 1 Dt ¼ 1=2 2 2 1=2 2 Seq p r r r 3p

ð38Þ

For relatively short durations in kinetic studies, the third and subsequent terms may be ignored, thus

1 S 4 D 1=2 1 D ¼ 1=2 2 2 1=2 t Seq p r r t

ð39Þ

92

Hua Zhang and H. M. Selim

The above kinetic models have been employed in describing the timedependent sorption of arsenic on minerals and soils. The Elovich and fraction power equations were successfully used by Elkhatib et al. (1984a,b) to describe the sorption and desorption of arsenite by several soils. Prediction results showed that the fraction power equation had higher coefficients of determination (r2 ) than the Elovich equation. Furthermore, the sorption and desorption rate constants k for the fraction power equation were correlated with the Eh and soil content of Fe oxides. Similarly, CarbonellBarrachina et al. (1996) compared Elovich and power equation for arsenite sorption–desorption kinetics and their results showed the Elovich was superior to the power equation. Pigna et al. (2006) performed kinetic studies on the sorption of AsO4 on Fe and Al oxides and desorption by PO4 and were best described by the Elovich kinetic model. Raven et al. (1998) described arsenite and arsenate sorption kinetic results on ferrihydrite using first order, second order, power function, Elovich, and parabolic diffusion equations. The parabolic diffusion equation provided best fit to kinetic data and they suggested that the retention of arsenic on ferrihydrite was diffusion controlled. The time-dependent sorption data demonstrated that a fast reaction is followed by a slow reaction, suggesting that different sorption mechanisms involved in the process. Fuller et al. (1993) developed a pore-space diffusion model with Freundlich equilibrium sorption to describe the timedependent sorption of arsenate on ferrihydrite. Their model has the form of

2 Sa @ C 2@C N 1 @C ¼D e þ KF C þ @t @r 2 r@r N

ð40Þ

where e is the initial porosity, Sa is the reactive surface area, KF and N are Freundlich parameters (Eq. 20). The numerical solution of this model successfully described AsO4 adsorption kinetics on ferrihydrite. The reversible nonequilibrium sorption models have been proposed by many researches to describe the sorption–desorption kinetics involving multiple chemical and physical reaction processes. The reversible first-order kinetic sorption equation has the form of:

@S y ¼ kf C kb S @t r

ð41Þ

where kf (h1) and kb (h1) are the forward and backward reaction rates, respectively. Under equilibrium conditions, that is, @S=@t ¼ 0, Eq. (41) yields linear adsorption equation of S ¼ Kd C. The reversible nth-order (Freundlich-type) kinetic sorption equation is in the form of

@S y ¼ kf C b kb S @t r

ð42Þ

Reaction and Transport of Arsenic in Soils

93

where kf and kb are the forward and backward reaction rate coefficients (h1), respectively, b is a nonlinear parameter usually less than 1, t is reaction time (h), r is the soil bulk density (g cm3), and y is the volumetric water content (cm3 cm3). Under equilibrium conditions, that is, @S=@t ¼ 0; Eq. (41) yields Freundlich Eq. (20) assuming KF ¼ kf =kb y=r and N ¼ b. The Langmuir kinetic equation is another reversible type where a concentration maxima (Smax) is assumed,

@S y ¼ kf ðSmax SÞC kb S @t r

ð43Þ

Under equilibrium conditions, @S=@t ¼ 0, Eq. (43) yields the Langmuir equilibrium Eq. (21) where KL ¼ kf =kb y=r. The Freundlich- and Langmuir-type reversible kinetic sorption equations are commonly solved using numerical algorithms such as the fourthorder Runge-Kutta methods. The numerical solutions give us the flexibility of simulating a wide range of initial conditions. The nonequilibrium kinetic equations are incorporated into the dispersion–advection transport model for the simulating dynamics of solute concentration across space and time. For example, the batch kinetic sorption data of Darland and Inskeep (1997a) were described using first-order and nth order reversible adsorption. The first-order forward and backward rate constants (kf and kb) were 2.65 10–1 and 8.75 10–3 h1 for arsenate sorption on acid-washed sand. Because of the complex sorption processes, the simple chemical kinetic models may not be appropriate for describing sorption kinetics in heterogeneous soils where a range of particle sizes and multiple types of reaction sites exists. Recent approaches based on soil heterogeneity and kinetics of adsorption–desorption have been proposed for the purpose of describing the time-dependent sorption of heavy metals in the soil environment. The multireaction model (MRM) kinetic approach presented here considers several interactions of heavy metals with soil matrix surfaces (Amacher et al., 1988; Selim, 1992). Specifically, the model assumes that a fraction of the total sorption sites is kinetic in nature whereas the remaining fractions interact rapidly or instantaneously with solute in the soil solution. The model accounts for reversible as well as irreversible sorption of the concurrent and consecutive type (Fig. 14). The model can be presented in the following formulations:

Se ¼ Ke C n

ð44Þ

@Sk y ¼ k1 C m ðk2 þ k3 ÞSk r @t

ð45Þ

94

Hua Zhang and H. M. Selim

Se

Multi-reaction model

Ke k1 C

k2

S1

k3

S2

ks Ss

Figure 14 A schematic diagram of the multireaction model (MRM) with equilibrium, kinetic, and irreversible adsorption sites. Here C is concentration in solution, Se is the amount sorbed on the equilibrium sites, S1 is the amount sorbed on the kinetic sites, S2 is the amount retained on consecutive irreversible sites, and Ss is amount retained on concurrent irreversible sites, where Ke, k1, k2, k3, and ks are the respective rates of reactions. (Reprinted with permission from Zhang and Selim, 2006.)

@Si ¼ k3 Si @t

ð46Þ

@Ss y ¼ ks C r @t

ð47Þ

where Se is the amount retained on equilibrium sites (mgl kg1), Sk is the amount retained on kinetic type sites (mg kg1), Si is the amount retained irreversibly by consecutive reaction (mg kg1), Ss is the amount retained irreversibly by concurrent type of reaction (mg kg1), n and m are dimensionless reaction order commonly less than 1, Ke is a dimensionless equilibrium constant, k1 and k2 (h1) are the forward and backward reaction rates associated with kinetic sites, respectively, k3 (h1) is the irreversible rate coefficient associated with the kinetic sites, and ks (h1) is the irreversible rate coefficient associated with solution. For the case n ¼ m ¼ 1, the reaction equations become linear. In the above equations, we assumed n ¼ m since there is no known method for estimating n and/or m independently. According to model formulation of Fig. 14, the total amount of solute retention (S ) by the soil is:

S ¼ Se þ S k þ S i þ S s

ð48Þ

Moreover, the MRM with nonlinear equilibrium and kinetic sorption successfully described the kinetic data of AsO4 adsorption on Olivier loam and Windsor sand. The model was also capable of predicting AsO4 desorption kinetics for both soils (Fig. 15). However, for Sharkey clay, which exhibited strongest affinity for arsenic, an additional irreversible reaction

95

Reaction and Transport of Arsenic in Soils

600 Olivier Adsorption

As(V) sorbed (mg kg−1)

500

Desorption

400 300 200 100 0 750

As(V) sorbed (mg kg−1)

Windsor Adsorption

600

Desorption

450

300

150

0 0

200

600 400 Reaction time (h)

800

1000

Figure 15 Arsenate sorbed versus time during adsorption–desorption for Olivier and Windsor soils. Symbols are for different initial concentrations (Co) of 5, 10, 20, 40, 80, and 100 mg l1 (from bottom to top). Solid curves are two-phase multireaction model (MRM) simulations using parameters obtained from nonlinear optimization with adsorption data. (Reprinted with permission from Zhang and Selim, 2005.)

phase was required to predict AsO4 desorption or release with time (Zhang and Selim, 2005). 6.4.2. Kinetic dissolution Oxidative dissolution of arsenic containing sulfide minerals is a multistep process involving diffusive transport of oxidant to mineral surfaces, adsorption of oxidant, interlattice transfer of oxidant, chemical reaction inside crystalline structure, detachment of reaction product from mineral surface, and diffusive transport of reaction product to the bulk solution. The dissolution kinetics

96

Hua Zhang and H. M. Selim

of arsenic minerals in the environment is determined by one or more ratelimiting processes. Diffusion-controlled dissolution kinetic can be described with the parabolic diffusion rate law in the form of:

r¼

dC ¼ kp t1=2 dt

ð49Þ

where r is the dissolution rate (mol s1), kp is the reaction rate constant (mol s1/2). A zero-order rate law has been used for the simulation of dissolution kinetics under steady state surface condition, that is, concentration of solutes adjacent to the surface is the same as in the bulk solution:

r¼

dC ¼ ks A dt

ð50Þ

where A is the surface area of the reactive mineral phase. Using mixed flow through experiments and for a range of pH and DO values, Lengke and Tempel (2005) calculated steady state oxidation rate of realgar following first-order reaction: r¼

dC q ¼ C dt A

ð51Þ

where q is the flow rate through the system (L s1), A is the total surface area of solid. The calculated dissolution rate is expressed as a function of DO and pH in the form of:

rAs ¼ 109:63ð0:41Þ ½DO0:51ð0:08Þ ½Hþ rAs ¼ 1011:77ð0:36Þ ½DO0:36ð0:09Þ ½Hþ rAs ¼ 1016:77ð0:68Þ ½DO0:42ð0:07Þ ½Hþ

0:28ð0:05Þ

0:47ð0:05Þ

1:26ð0:09Þ

realgar

ð52Þ

orpiment

ð53Þ

amorphous As2 S3 ð54Þ

In addition, the effect of temperature on the rate constant was described using the Arrhenius equation with activation energy Ea ¼ 64.29.8, 59.1 0.44, and 16.8 5.0 kJ/mol for realgar, orpiment, and amorphous As2S3, respectively. Walker et al. (2006) found that dissolution rate r ¼ 10–10.14(0.03) is essentially independent of DO for FeAsS. 6.4.3. Kinetic reduction–oxidation Heterogeneous oxidation of arsenite on the mineral surface is a slow process involving multiple diffusion and reaction steps. Because of the complexity of the reaction, it is implausible to develop mechanistic rate laws describing

97

Reaction and Transport of Arsenic in Soils

element reactions at the molecular level. Instead simple apparent rate laws are commonly employed for the simulation of arsenite depletion in soils. For example, Oscarson et al. (1983a) used a simple first-order rate equation for describing heterogeneous oxidation of arsenite on the surface of manganese oxides, which can be described with dCAsðIII Þ ¼ kox CAsðIII Þ dt

ð55Þ

Integrating this equation gives:

CAsðIII Þ ¼ CAsðIII Þ ekox t

ð56Þ

where kox is the kinetic oxidation rate (h1), CAsðIII Þ is initial AsO3 concentration (mmol L1). Amirbahman et al. (2006) simulated the retention and transformation of arsenite with a set of equilibrium and kinetic reactions. Two types of adsorption sites on the soil surface were characterized. The first represents sites of purely adsorptive type that are not involved in oxidation. The second types of sites are of the oxidative type where AsO3 oxidization takes place. Adsorption on purely adsorptive site was described with fully reversible kinetic equation, whereas AsO3 adsorption on oxidative sites was simulated with equilibrium adsorption followed by first-order irreversible oxidation. Furthermore, concurrent fast and slow reactions were used in combination to simulate reaction kinetics on adsorptive and oxidative sites. The apparent rate constants and density of available oxidative sites were obtained by fitting this model to the experiment data. The transformation between AsO4 and AsO3 as a result of microbial activity is also a kinetic process with great complexity. Similarly, apparent rate laws were adopted by environmental scientists for the simulation of microbial-mediated arsenic transformation ( Jones et al., 2000). Manning and Suarez (2000) treated heterogeneous oxidation and adsorption of arsenite in soils with consecutive kinetic reaction mechanisms, where kox

kad

CAsðIII Þ ! CAsðV Þ ! SAs and give the overall reaction expression CAsðIII Þ SAs ¼ kad 1 ekox t kox 1 ekad t kad kox

ð57Þ

6.5. Transport models The transport of dissolved chemicals through porous media is generally described using the advection–dispersion equation (ADE), sometimes called convection–dispersion equation (CDE). Assuming local equilibrium

98

Hua Zhang and H. M. Selim

condition (e.g., residence time much larger than time required to complete reaction), the equation can be expressed as:

@C @ @C @C ¼ D v R @t @x @x @x

ð58Þ

where C is solute concentration (M L3), x is distance (L), t is time (T), r is soil bulk density (M L3), y is volumetric water content (L3 L3), D is dispersion coefficient (L2 T1), and v is pore water velocity (L T1) where v ¼ q/y, and q is Darcy’s water velocity (L T1). In addition, the solute retardation factor R is determined by the equilibrium distribution of solute between solid and aqueous phases, where

r R ¼ 1 þ Kd y r R ¼ 1 þ Kf NC N 1 y R ¼1þ

r KL Smax y ð1 þ KL C Þ2

for linear sorption; for Freundlich sorption; and for Langmuir sorption:

The dispersion of solute in soils is a combination of hydrodynamic dispersion and molecular diffusion processes. As a result, the dispersion coefficient D is a function of water content, flow velocity, solute property, and other hydraulic parameters, which can be expressed as,

D ¼ dv þ

D0 t

ð59Þ

where d (L) is the longitudinal dispersivity, D0 is the diffusion coefficient for a particular solute diffusing in bulk water, and t is the tortuosity factor for solute diffusing in pore network inside soil (Brusseau, 1993). A common strategy for estimating the value of dispersion coefficient D is to conduct transport experiment using conservative tracer such as bromide or radioactive tracers such as tritium and chloride-36. Chemical nonequilibrium behavior in soils is likely caused by kinetic reactions occurring at the solid–liquid interfaces, which is often described using one-dimensional steady state transport equation of reactive solute in the following form:

r @S @C @ @C @C þ ¼ D v y @t @t @x @x @x

ð60Þ

Reaction and Transport of Arsenic in Soils

99

where the total amount of retention (S ) is accounted for using the MRM kinetic approach described in Fig. 14. Zhang and Selim (2006) evaluated eight formulations of MRM model for the simulation of arsenic transport in soils with different properties. As illustrated in Fig. 16, they found the use of independently derived kinetic retention model parameters underestimated the extent of retention and overpredicted arsenic mobility. However, when utilized in an inverse mode, the MRM model provided good predictions of AsO4 BTCs. Nonlinear reversible along with consecutive or concurrent irreversible reactions were the dominant mechanisms in the MRM model.

6.6. Field application Geochemical models (e.g., MINTEQ, PHREEQC) have been adopted to simulate the reactions among multiple chemical species during their transport in soils and aquifers (e.g., Manning and Goldberg, 1996; Smith and Jaffe, 1998). The common strategy is by coupling transport models with equilibrium thermodynamic models with the main aim to demonstrate the effect various properties such as the adsorbents, pH, redox potential, and competing ions on the chemical species in the soil. For example, Smith and Jaffe (1998) formulated a geochemical model which coupled the transport process with various kinetics reactions and simulated the arsenic transport in Benthic sediment. However, because of the uncertainty associated with numerous geochemical parameters, such models can only be viewed as heuristic tools for exploring possible trends in the fate of contaminants as a result of environmental changes. Sracek et al. (2004) summarized several examples of forward or inverse geochemical modeling of the fate of arsenic in the environments. The application of geochemical models requires detailed description of chemical and mineral composition of solution and porous media, soil matrix properties, as well as numerous reaction constants. However, such information are either unavailable or unreliable under most circumstances (Nitzsche et al., 2000). Heterogeneity of the natural porous media also impedes the application of chemical reaction based models. More importantly, sorption processes are often assumed instantaneous (i.e., equilibrium conditions are assumed) in most geochemical models. An effective way of reducing the uncertainty of model simulation is to estimate input parameters from observed data through inverse modeling. Nonlinear least squares regression method is a commonly employed inverse modeling approach. Other optimization algorithms such as simulated annealing and genetic algorithms have been tested, with various degrees of success, to solve inverse modeling problem in groundwater contamination. To overcome the uncertainty in the input parameters to groundwater flow model, Morse et al. (2003) employed Monte Carlo analysis and the generalized likelihood uncertainty estimator (GLUE) methodology for the

100

Hua Zhang and H. M. Selim

1.0 Olivier column 102 0.8

Measured M1 = k1,k2

C/CO

0.6

M2 = Ke, kirr M3 = Ke,k1,k2

0.4

M4 = k1,k2,k3 M5 = k1,k2,kirr 0.2

0.0 0

20

40 60 Pore volumes (V/VO)

80

100

1.0 Olivier column 102 0.8

C/CO

0.6 Measured M6 = Ke,k1,k2,k3 M7 = Ke,k1,k2,kirr

0.4

M8 = Ke,k1,k2,k3,kirr 0.2

0.0 0

20

40

60

80

100

Pore volumes (V/VO)

Figure 16 Comparison of multireaction model (MRM) model formulations M1–M8 for predicting AsO4 breakthrough curves for Olivier soil column 102. Model parameters were obtained using nonlinear inverse modeling. (Reprinted with permission from Zhang and Selim, 2006.)

Reaction and Transport of Arsenic in Soils

101

stochastic analysis of capture zone for arsenic contaminated wells in the Zimapan Valley of Mexico. Incorporating extra hydrologic and geologic information reduces the uncertainty in parameter estimation. The heterogeneous nature of the soils and geologic porous media often results in highly nonuniform retention and transport processes in contaminated sites, which restricts the application of deterministic approaches which are often based on the assumption of relatively homogeneous material. To overcome these difficulties, several stochastic advection–dispersion approaches have been developed from theoretical study and field experimentation in the past three decades. Stochastic approaches are theoretically attractive because spatial variability of the soil matrix is explicitly defined using parameters with certain statistical distributions. However, the mathematical complexity hinders the practical application of such approaches in the management of contaminant movement in aquifers and field soils. On the contrary, geostatistical models provide a probabilistic framework for predicting the spatiotemporal distribution of contaminants based on statistical analysis of existing observations without reliance on the underlying mechanisms. Goovaerts et al. (2005) employed multi-Gaussian and indicator Kriging for modeling probabilistically the spatial distribution of arsenic concentrations in groundwater of Southeast Michigan. They found that the use of secondary geological information significantly increased the proportion of variance that can be explained. They attribute the inadequacy of the model for prediction purpose with the quality of the sampling data.

7. Remediation of Contaminated Soils Despite the widespread soil contamination of arsenic, there is no nationwide accepted cleanup standard for arsenic due to soil heterogeneity and policy interpretation (Davis et al., 2001). A survey conducted by the Association for the Environmental Health of Soils (AEHS, 1999) revealed that there were large variations among soil arsenic regulations set by different states across the United States. They reported the notification levels of 2–61 mg kg1, soil screening levels of 0.1–250 mg kg1 for residential area and 2.4–200 mg kg1 for industrial sites, cleanup levels of 0.1–250 mg kg1 for residential area, and 0.85–1000 mg kg1 for industrial sites for the 34 states participated in the survey. In addition, the rationale used for setting the regulation levels was widely diversified, including related regulation limits, background levels, human health risk, and migration to the groundwater. From their survey of Records of Decisions (RODs), Davis et al. (2001) divided the studied sites into four risk categories: industrial, residential, background and ecological risk-based decisions with 84% of the sites were risk driven and 16% were background driven. They reported that a

102

Hua Zhang and H. M. Selim

wide range of soil arsenic cleanup standard for residential risk goals (2–305 mg kg1) and a narrower background-based clean up goal of 8–21 mg kg1. In addition, there was no apparent temporal trend but considerable geographic differences were observed for arsenic RODs. Engineering approaches used for the remediation of arseniccontaminated sites include isolation, immobilization, toxicity reduction, physical separation, and extraction (Mulligan et al., 2001). The common action for the control heavy metal pollution at heavily contaminated sites is the removal of the surface soil. This method is not always effective, especially at sites with large affected area and long history of pollution. It was found that after 3 years of remediation (cleanup of the tailings and polluted soils, followed by apply sugar-refinery scum and tilling of the soil) at Aznalco´llar pyrite mine spill in Spain, a large portion of the soils in the area remain highly polluted (Aguilar et al., 2004). Isolation or capping of contaminated soils through the construction of physical barriers is a common technique employed at landfills and superfund sites. Such barriers made of steel, cement, bentonite clay, and grout walls reduce the permeability of the waste and limit the movement of groundwater through the contaminated area. Additional layers of sandy soils are employed to prevent upward movement of groundwater by capillary action. The isolation of contaminated sites is generally less expensive than other techniques. However, a common engineering problem associated with landfill capping is that the aging of the clay liner eventually leads to preferential flow in the fractures. Long-term stability of the solidified material requires frequent monitoring at those contaminated sites. Solidification/stabilization (S/S), also known as chemical fixation or encapsulation, is a set of technologies widely applied to treat soils contaminated with cationic heavy metals. The most common form of S/S uses a cement or pozolanic binder to convert the contaminated soil in order to create a monolithic form that limits the contaminant mobility. Miller et al. (2000) evaluated several combinations of cement binders and reagents in their capability to solidify sandy soils contaminated with arsenic. They reported that a mixture of Type I Portland cement and ferrous sulfate was effective in reducing the leaching of arsenic and improved performance was observed when the soil was pretreated with FeSO47H2O followed by Portland cement. This chemical contaminant strategy was successfully implemented in their field study. In situ S/S processes are most suitable for shallow contamination sites using conventional construction equipment. In addition, liming is a common remediation method for immobilizing trace metals released through acid mine drainage. However, as demonstrated by Jones et al. (1997), liming may also result in enhanced As mobilization due to the pH dependence of As sorption reactions. Soil flushing followed by groundwater extraction is a widely used in situ remediation technique for contaminated sites with relatively high

Reaction and Transport of Arsenic in Soils

103

permeability. Chemical additives are often used to enhance the solubility of the contaminant. Electrokinetic processes, which involve transport of ions and small charged particles through the application of a low intensity electric current between a cathode and an anode imbedded in the soils, are proposed for saturated clay soils with low permeability (Kim et al., 2002). However, several factors limited the suitability of this technique for the treatment of arsenic contaminated soils. First, high sorption capacity and strong binding strength between arsenic and metal oxides substantially retarded the transport of arsenic in soils and limited the extractable amount of arsenic. Second, the rate-limited sorption and transport processes made it practically impossible to achieve the cleanup goal in reasonable time frame with flushing and extraction processes. In addition, arsenic is mixed with other organic or inorganic contaminants in most contaminated sites. But there is no efficient extractant that is capable of simultaneously removing organic and inorganic, anionic, and cationic contaminants. Permeable reactive barrier (PRB) consists of installing a reactive material into the aquifer to induce sequestration and/or transformations of the contaminants and reduce the contaminant concentration in groundwater. Su and Puls (2001, 2003) evaluated several types of zero-valent iron (Fe0) in their capacity to attenuate arsenic concentration using both batch and column experiments. They suggested that Fe0 was a promising material for in situ remediation of contaminated groundwater with relatively low cost. The use of Fe0 in treating arsenic contaminated soils and groundwater have attracted extensive attention over the last 5 years with several new developments appearing in recent literatures. However, to our knowledge, field scale implementations of PRB for arsenic treatment have not been reported. Monitored natural attenuation (MNA) is proposed as an alternative remediation strategy for soils contaminated with inorganic contaminants such as arsenic. In the EPA Directive (USEPA, 1999), MNA is defined as ‘‘physical, chemical, or biological processes that, under favorable conditions, act without human intervention to reduce the mass, toxicity, volume, or concentration of contaminants in soil or groundwater.’’ The establishing of MNA as a potential remediation strategy for a contaminated site requires thorough and adequate site-specific characterization data and analysis. Specifically, there are three tiers of ‘‘lines of evidence’’: (1) historical groundwater and/or soil chemistry data that clearly demonstrate a trend of decreasing contaminant mass and/or concentration; (2) indirect hydrogeologic and geochemical data that demonstrate the type and rate of natural attenuation processes active at the site; and (3) direct field or microbiological data that demonstrate the occurrence of a particular natural attenuation process at the site. Sorption and precipitation are the dominant processes of natural attenuation of arsenic. Biotransformation of arsenic (redox cycling and

104

Hua Zhang and H. M. Selim

methylation) can influence those natural attenuation processes. The key issue of natural attenuation of arsenic is the reversibility of arsenic sequestration (sorption/precipitation) into the solid phase because the intrinsic toxicity of arsenic was not affected by the immobilization process (Reisinger et al., 2005). Therefore, successful implementation of MNA for arsenic contaminated sites requires detailed site-specific characterization including source identification, plume boundary delineation, time-series monitor of arsenic concentration in soil and groundwater. An extensive list of ancillary data (e.g., pH, redox potential, Fe/Al/Mn contents, soil texture, and organic matter) is required for the evaluation of the feasibility of MNA. Reisinger et al. (2005) provided four examples of arsenic contaminated sites where natural attenuation was deemed acceptable by regulators. Phytoremediation is an emerging technology that uses specially selected and engineered metal-accumulating plants for environmental cleanup. Plants, such as Indian mustard (Pickering et al., 2000) and brake fern (Ma et al., 2001), have the capability to accumulate arsenic and is considered as potential method for treating contaminated soils. Since the discovery of arsenic hyperaccumulation capacity of Pteris vittata (brake fern), extensive researches have been conducted to investigate its physiological mechanisms and its effectiveness for remediation of arsenic contaminated soils. In addition to naturally selected plants, scientists used biotechnology to develop engineered plants that have the capability of accumulating arsenic. Dhankher et al. (2002) have developed a genetics-based phytoremediation strategy for arsenic where arsenic is hyperaccumulated in a plant transformed with the ArsC gene [encoding arsenate reductase (ArsC)]. Even though tremendous research effort was devoted to the identification and cultivation of arsenic accumulating plant, there is limited field evidence demonstrating the effectiveness of the phytoremediation for the treatment of arsenic contaminated soils. As a matter of fact, phytoremediation is limited to shallow soils (root zone) and the reliability of the process is heavily dependent on many environmental factors such as climate, terrain, and contamination level.

8. Summary and a Look Ahead Simulating the fate and transport of arsenic requires an in-depth understanding of the physical, chemical, and biological interactions entangled on the liquid–solid interface in soil environment. Predicting the mobility and bioavailability of arsenic is complicated by the multiple organic or inorganic arsenic species in dissolved, adsorbed, or particulate forms coexisting in the soil system. Environmental factors, for example, pH, redox potential, competing ions, soil mineralogy, flow regime, and microbial activity, should be considered in the models because of their substantial

Reaction and Transport of Arsenic in Soils

105

impact on arsenic reaction and transport. Thermodynamic models using equilibrium constants derived from the standard Gibbs energy of the chemical compounds are frequently employed for predicting the environmental behavior of arsenic at contaminated sites. But the capabilities of such models are rather limited because chemical equilibrium rarely exists in soils. Recent studies are focused on developing kinetic models that simulate the dynamic distribution of arsenic species and determining the kinetic reaction rates for various chemical and biological reactions. Research is needed for the application of mechanistically based arsenic transport models in making regulatory decision and designing remediation strategy. Areas of research needs included, 1. Kinetics of reactions/release of arsenic on surfaces of soil constituents (organic matter, metal oxides, clay minerals, sulfides, microbes) and quantitative description of mechanistic processes. 2. Speciation of arsenic in solid and aqueous phases as a function of solution composition, soil composition, flow conditions, and microbial activities. 3. Influence of physical and chemical heterogeneity on arsenic retention and transport and stochastic simulation of the spatial distribution of arsenic in geological porous media. 4. Reactive transport models that incorporates complex biogeochemical processes and capable of predicting the fate of arsenic in the environment. 5. Strategies for the remediation of arsenic contaminated soils using combination of physical, chemical, and biological techniques.

REFERENCES Acharyya, S. K., Chakraborty, P., Lahiri, S., Raymahashay, B. C., Guha, S., and Bhowmik, A. (1999). Arsenic poisoning in the Ganges delta. Nature 401, 545. AEHS. (1999). Study of State Soil Arsenic Regulations. http://www.aehs.com/surveys/ arsenic.Pdf. Aguilar, J., Dorronsoro, C., Fernandez, E., Fernandez, J., Garcia, I., Martin, F., and Simon, M. (2004). Soil pollution by a pyrite mine spill in Spain: Evolution in time. Environ. Pollut. 132, 395–401. Amacher, M. C., Selim, H. M., and Iskandar, I. K. (1988). Kinetics of chromium(VI) and cadmium retention in soils: A nonlinear multireaction model. Soil Sci. Soc. Am. J. 52, 398–408. Amirbahman, A., Kent, D. B., Curtis, G. P., and Davis, J. A. (2006). Kinetics of sorption and abiotic oxidation of arsenic(III) by aquifer materials. Geochim. Cosmochim. Acta 70, 533–547. Allison, J. D., Brown, D. S., and Novo-Gradac, K. J. (1991).MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems: Version 3. 0 user’s manual. EPA/600/3–91/021.

106

Hua Zhang and H. M. Selim

Allinson, G., Turoczy, N. J., Kelsall, Y., Allinson, M., Stagnitti, F., Lloyd-Smith, J., and Nishikawa, M. (2000). Mobility of the constituents of chromated copper arsenate in a shallow sandy soil. New Zeal. J. Agri. Res. 43, 149–156. Anderson, M. A., Ferguson, J. F., and Gavis, J. (1976). Arsenate adsorption on amorphous aluminum hydroxide. J. Colloid Interface Sci. 54, 391–399. Andersen, S., Rasmussen, G., Snilsberg, P, Amundsen, C. E, and Westby, T. (1996). Assessing toxicity and mobilisation of impregnation salts at a contaminated site. Fresenius J. Analytical Chem. 354, 676–680. Appelo, C. A. J., van der Weiden, M. J. J., Tournassat, C., and Charlet, L. (2002). Surface complexation of ferrous iron and carbonate on ferrihydrite and the mobilization of arsenic. Environ. Sci. Technol. 36, 3096–3103. Arai, Y., and Sparks, D. L. (2002). Residence time effects on arsenate surface speciation at the aluminum oxide-water interface. Soil Sci. 167, 303–314. Arai, Y., Lanzirotti, A., Sutton, S., Davis, J. A., and Sparks, D. L. (2003). Arsenic speciation and reactivity in poultry litter. Environ. Sci. Technol. 37, 4083–4090. Arai, Y., Sparks, D. L., and Davis, J. A. (2004). Effects of dissolved carbonate on arsenate adsorption and surface speciation at the hematite-Water interface. Environ. Sci. Technol. 38, 817–824. Association for the Environmental Health of Soils (AEHS). (1999). Study of State Soil Arsenic Regulations. http://www.aehs.com/surveys/arsenic.pdf. Barrow, N. J., Cartes, P., and Mora, M. L. (2005). Modifications to the Freundlich equation to describe anion sorption over a large range and to describe competition between pairs of ions. Euro. J. Soil Sci. 56, 601–606. Barzi, F., Naidu, R., and Mclaughlin, M. (1996). Contaminants in the Australian soil environment. In (Naidu et al., Eds.), Contaminants and the soil environment in the Australasia-Pacific Region. Kluwer Academic Publ. Dordrecht. pp. 451–485. Bednar, A. J., Garbarino, J. R., Ranville, J. F., and Wildeman, T. R. (2002). Presence of organoarsenicals used in cotton production in agricultural water and soil of the southern United States. J. Agric. Food Chem. 50, 7340–7344. Berg, M., Tran, H. C., Nguyen, T. C., Phem, H. V., Scheertenleib, R., and Giger, W. (2001). Arsenic contamination of groundwater and drinking water in Vietnam: A human health threat. Environ. Sci. Technol. 35, 2621–2626. BGS. (2001). Arsenic contamination of groundwater in Bangladesh. Technical Report, WC/00/19. 4 Volumes. British Geological Survey, Keyworth, UK. Bisceglia, K. J., Rader, K. J., Carbonaro, R. F., Farley, K. J., Mahony, J. D., and Di Toro, D. M. (2005). Iron(II)-catalyzed oxidation of arsenic(III) in a sediment column. Environ. Sci. Technol. 39, 9217–9222. Bohn, H. L. (1976). Arsenic Eh-pH diagram and comparison to soil chemistry of phosphorus. Soil Sci. 121, 125–127. Bostick, B. C., and Fendorf, S. (2003). Arsenite sorption on troilite (FeS) and pyrite (FeS2). Geochim. Cosmochim. Acta 67, 909–921. Bothe, J. V., and Brown, P. W. (1999). Arsenic immobilization by calcium arsenate formation. Environ. Sci. Technol. 33, 3806–3811. Bradford, G. R., Chang, A. C., Page, A. L., Bakhtar, D., Frampton, J. A., and Wright, H. (1996).Background Concentrations of Trace and Major Elements in California Soils. Kearney Foundation Special Report. Kearney Foundation of Soil Science, Division of Agriculture and Natural Resources, University Of California http://www.envisci.ucr. edu/faculty/chang/kearney/kearneytext.html. Brouwere, K. D., Smolders, E., and Merckx, R. (2004). Soil properties affecting solid-liquid distribution of As(V) in soils. Euro. J Soil Sci. 55, 165–173. Brusseau, M. L. (1993). The influence of solute size, pore water velocity, and intraparticle porosity on solute dispersion and transport in soils. Water Resour. Res. 29, 1071–1080.

Reaction and Transport of Arsenic in Soils

107

Carbonell-Barrachina, A. A., Carbonell, F. B., and Beneyto, J. M. (1996). Kinetics of arsenite sorption and desorption in Spanish soils. Commun. Soil Sci. Plant Anal. 27, 3101–3117. Chen, M., Ma, L. Q., and Harris, W. G. (2002). Arsenic concentrations in Florida surface soils: Influence of soil type and properties. Soil Sci. Soc. Am. J. 66, 632–640. Chen, S. L., Dzeng, S. R., Yang, M. H., Chiu, K. H., Shieh, G. M., and Wai, C. M. (1994). Arsenic species in groundwaters of the blackfoot disease area, Taiwan. Environ. Sci. Technol. 28, 877–881. Chirenje, T., Ma, L. Q., Chen, M., and Zillioux, E. J. (2003). Comparison between background concentrations of arsenic in urban and non-urban areas of Florida. Adv. Environ. Res. 8, 137–146. Chiu, V. Q., and Hering, J. G. (2000). Arsenic adsorption and oxidation at manganite surfaces. 1. Method for simultaneous determination of adsorbed and dissolved arsenic species. Environ. Sci. Technol. 34, 2029–2034. Corwin, D. L., David, A., and Goldberg, S. (1999). Mobility of arsenic in soil from the Rocky Mountain Arsenal area. J. Contam. Hydrol. 39, 35–58. Cox, C. D., and Ghosh, H. M. (1994). Surface complexation of methylated arsenates by hydrous oxides. Water Res. 28, 1181–1188. Craw, D., Falconer, D., and Youngson, J. H. (2003). Environmental arsenopyrite stability and dissolution: Theory, experiment, and field observations. Chem. Geol. 199, 71–82. Cullen, W. R., and Reimer, K. J. (1989). Arsenic speciation in the environment. Chem. Rev. 89, 713–764. Darland, J. E., and Inskeep, W. P. (1997a). Effects of pore water velocity on the transport of arsenate. Environ. Sci. Technol. 31, 704–709. Darland, J. E., and Inskeep, W. P. (1997b). Effects of pH and phosphate competition on the transport of arsenate. J. Environ. Qual. 26, 1133–1139. Davis, A., Ruby, M. V., Bloom, M., Schoof, R., Freeman, G., and Bergstom, P. D. (1996). Mineralogic constraints on the bioavailability of arsenic in smelter-impacted soils. Environ. Sci. Technol. 30, 392–399. Davis, A., Sherwin, D., Ditmars, R., and Hoenke, K. A. (2001). An analysis of soil arsenic records of decision. Environ. Sci. Technol. 35, 2401–2406. Dhankher, O. P., Li, Y. J., Rosen, B. P., Shi, J., Salt, D., Senecoff, J. F., Sashti, N. A., and Meagher, R. B. (2002). Engineering tolerance and hyperaccumulation of arsenic in plants by combining arsenate reductase and gamma-glutamylcysteine synthetase expression. Nat. Biotechnol. 20, 1140–1145. Dixit, S., and Hering, J. G. (2003). Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: Implication for arsenic mobility. Environ. Sci. Technol. 37, 4182–4189. Dombrowski, P. M., Long, W., Farley, K. J., Mahony, J. D., Capitani, D. F., and Di Toro, D. M. (2005). Thermodynamic analysis of arsenic methylation. Environ. Sci. Technol. 39, 2169–2176. Dzombak, D. A., and Morel, F. M. M. (1990). ‘‘Surface Complexation Modeling.’’ Wiley, New York. Elkhatib, E. A., Bennett, O. L., and Wright, R. J. (1984a). Kinetics of arsenite adsorption in soils. Soil Sci. Soc. Am. J. 48, 758–762. Elkhatib, E. A., Bennett, O. L., and Wright, R. J. (1984b). Arsenite sorption and desorption in soils. Soil Sci. Soc. Am. J. 48, 1025–1030. Farquhar, M. L., Charnock, J. M., Livens, F. R., and Vaughan, D. J. (2002). Mechanisms of arsenic uptake from aqueous solution by interaction with Goethite, Lepidocrocite, Mackinawite, and Pyrite: An X-ray absorption spectroscopy study. Environ. Sci. Technol. 36, 1757–1762. Fendorf, S., Eick, M. J., Grossl, P., and Sparks, D. L. (1997). Arsenate and chromate retention mechanisms on goethite. 1. Surface structure. Environ. Sci. Technol. 31, 315–320.

108

Hua Zhang and H. M. Selim

Folkes, D. J., Helgen, S. O., and Little, R. A. (2001). Impacts of historic arsenical pesticide use on residential soils in Denver, Colorado. In ‘‘Arsenic Exposure and Health Effects’’ (W. R. Chappell, C. O. Abernathy, and R. L. Calderon, Eds.), Elvisier Sciences, Netherland. Foster, A. L. (2003). Spectroscopic investigations of arsenic species in solid phases. In ‘‘Arsenic in Groundwater: Geochemistry and Occurrence’’ (A. H. Welch and K. G. Stollenwerk, Eds.), Kluwer Academic Publishers, Boston, USA. Foster, A. L., Brown, G. E., Tingle, T. N., and Parks, G. A. (1998a). Quantitative arsenic speciation in mine tailings using X-ray absorption spectroscopy. Am. Min. 83, 553–568. Foster, A. L., Brown, G. E., and Parks, G. A. (1998b). X-ray absorption pine-structure spectroscopy study of photocatalyzed, heterogeneous As(III) oxidation on kaolin and anatase. Environ. Sci. Technol. 32, 1444–1452. Frost, R. R., and Griffin, R. A. (1977). Effect of pH on adsorption of arsenic and selenium from landfill leachate by clay minerals. Soil Sci. Soc. Am. J. 41, 53–57. Fuller, C. C., Davis, J. A., and Waychunas, G. A. (1993). Surface chemistry of ferrihydrite: Part 2. Kinetics of arsenate adsorption and coprecipitation. Geochim. Cosmochim. Acta 57, 2271–2282. Ghosh, A., Mukiibi, M., Eduardo Sez, A., and Ela, W. P. (2006). Leaching of arsenic from granular ferric hydroxide residuals under mature landfill conditions. Environ. Sci. Technol. 40, 6070–6075. Goldberg, S. (1986). Chemical modeling of arsenate adsorption on aluminum and iron oxide minerals. Soil Sci. Soc. Am. J. 50, 1154–1157. Goldberg, S. (1992). Use of surface complexation models in soil chemical-systems. Adv. Agron. 47, 233–329. Goldberg, S. (2002). Competitive adsorption of arsenate and arsenite on oxides and clay minerals. Soil Sci. Soc. Am. J. 66, 413–421. Goldberg, S., and Glaubig, R. A. (1988). Anion adsorption on calcareous, montmorillonitic soil—Arsenic. Soil Sci. Soc. Am. J. 52, 1297–1300. Goldberg, S., and Johnston, C. T. (2001). Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectoscopy, and surface complexation modeling. J. Colloid Interface Sci. 234, 204–216. Goovaerts, P., AvRuskin, G., Meliker, J., Slotnick, M., Jacquez, G., and Nriagu, J. (2005). Geostatistical modeling of the spatial variability of arsenic in groundwater of southeast Michigan. Water Resour. Res. 41, W07013, doi:10.1029/2004WR003705. Grafe, M., Eick, M. J., and Grossl, P. R. (2001). Adsorption of arsenate (V) and arsenite (III) on goethite in the presence and absence of dissolved organic carbon. Soil Sci. Soc. Am. J. 65, 1680–1687. Grafe, M., Nachtegaal, M., and Sparks, D. L. (2004). Formation of metal-arsenate precipitates at the goethite-water interface. Environ. Sci. Technol. 38, 6561–6570. Grossl, P. R., Eick, M. J., Sparks, D. L., Goldberg, S., and Ainsworth, C. C. (1997). Arsenate and chromate retention mechanisms on goethite. 2. Kinetic evaluation using a pressure-jump relaxation technique. Environ. Sci. Technol. 31, 321–326. Han, F. X., Kingery, W. L., Selim, H. M., Gerard, P. D., Cox, M. S., and Oldham, J. L. (2004). Arsenic solubility and distribution in poultry waste and long-term amended soil. Sci. Total Environ. 320, 51–61. Herbel, M., and Fendorf, S. (2005). Transformation and transport of arsenic within ferric hydroxide coated sands upon dissimilatory reducing bacterial activity. Adv. Arsenic Res. 915, 77–90. Herbel, M., and Fendorf, S. (2006). Biogeochemical processes controlling the speciation and transport of arsenic within iron coated sands. Chem. Geol. 228, 16–32. Hiltbold, A. E., Hajek, B. F., and Buchanan, G. A. (1974). Distribution of arsenic in soil profile after repeated application of MSMA. Weed Sci. 22, 272–275.

Reaction and Transport of Arsenic in Soils

109

Hingston, F. J., Atkinson, R. J., Posner, A. M., and Quirk, J. P. (1967). Specific adsorption of anions. Nature (London) 215, 1459–1461. Hingston, F. J., Posner, A. M., and Quirk, J. P. (1971). Competitive adsorption of negatively charged ligands on oxide surfaces. Discuss Faraday Soc. 52, 334–342. Hingston, J. A., Collins, C. D., Murphy, R. J., and Lester, J. N. (2001). Leaching of chromated copper arsenate wood preservatives: A review. Environ. Pollut. 111, 53–66. Ho¨hn, R., Isenbeck-Schro¨ter, M., Kent, D. B., Davis, J. A., Jakobsen, R., Jann, S., Niedan, V., Scholz, C., Stadler, S., and Tretner, A. (2006). Tracer test with As(V) under variable redox conditions controlling arsenic transport in the presence of elevated ferrous iron concentrations. J. Contam. Hydrol. 88, 36–54. Hopp, L., Peiffer, S., and Durner, W. (2006). Spatial variability of arsenic and chromium in the soil water at a former wood preserving site. J. Contam. Hydrol. 85, 159–178. Hug, S. J., and Leupin, O. (2003). Iron-catalyzed oxidation of arsenic(III) by oxygen and by hydrogen peroxide: pH-dependent formation of oxidants in the fenton reaction. Environ. Sci. Technol. 37, 2734–2742. Isensee, A. R., Shaw, W. C., Gretner, W. A., Swansen, C. R., Turner, B. C., and Woolsen, E. A. (1973). Revegetation following massive application of selected herbicides. Weed Sci. 21, 409–412. Ishak, C. F., Seaman, J. C., Miller, W. P., and Summer, M. (2002). Contaminant mobility in soils amended with fly ash and flue-gas gypsum: Intact soil cores and repacked columns. Water Air Soil Pollut. 134, 287–305. Jacobs, L. W., Syers, J. K., and Keeney, D. R. (1970). Arsenic sorption by soils. Soil Sci. Soc. Am. Proc. 34, 750–754. Jain, A., and Loeppert, R. H. (2000). Effect of competing anions on the adsorption of arsenate and arsenite by ferrihydrite. J. Environ. Qual. 29, 1422–1430. Jain, A., Raven, K. P., and Loeppert, R. H. (1999). Arsenite and arsenate adsorption on ferrihydrite: Surface charge reduction and net OH- release. Environ. Sci. Technol. 33, 1179–1184. Jia, Y. F., Xu, L. Y., Fang, Z., and Demopoulos, G. P. (2006). Observation of surface precipitation of arsenate on ferrihydrite. Environ. Sci. Technol. 40, 3248–3253. Jones, C. A., Inskeep, W. P., and Neuman, D. R. (1997). Arsenic transport in contaminated mine tailings following liming. J. Environ. Qual. 26, 433–439. Jones, C. A., Langner, H. W., Anderson, K., McDermott, T. R., and Inskeep, W. P. (2000). Rates of microbially mediated arsenate reduction and solubilization. Soil Sci. Soc. Am. J. 64, 600–608. Juillot, F., Ildefonse, P., Morin, G., Calas, G., de Kersabiec, A. M., and Benedetti, M. (1999). Remobilization of arsenic from buried wastes at an industrial site: Mineralogical and geochemical control. Appl. Geochem. 14, 1031–1048. Keon, N. E., Swartz, C. H., Brabander, D. J., Harvey, C., and Hemond, H. F. (2001). Validation of an arsenic sequential extraction method for evaluating mobility in sediments. Environ. Sci. Technol. 35, 2778–2784. Khan, B. I., Solo-Gabriele, H. M., Townsend, T. G., and Cai, Y. (2006a). Release of arsenic to the environment from CCA-treated wood. 1. Leaching and speciation during service. Environ. Sci. Technol. 40, 988–993. Khan, B. I., Jambeck, J., Solo-Gabriele, H. M., Townsend, T. G., and Cai, Y. (2006b). Release of arsenic to the environment from CCA-treated wood. 2. Leaching and speciation during disposal. Environ. Sci. Technol. 40, 994–999. Kim, M. J., Nriagu, J., and Haack, S. (2002). Carbonate ions and arsenic dissolution by groundwater. Environ. Sci. Technol. 34, 3094–3100. Kimber, S. W. L., Sizemore, D. J., and Slavich, P. E. G. (2002). Is there evidence of arsenic movement at cattle tick dip sites? Aust. J. Soil Res. 40, 1103–1114.

110

Hua Zhang and H. M. Selim

Kretzschmar, R., Borkovec, M., Grolimund, D., and Elimelech, M. (1999). Mobile subsurface colloids and their role in contaminant transport. Adv. Agron. 66, 121–193. Kuhlmeier, P. D. (1997a). Partitioning of arsenic species in fine-grained soils. J. Air Waste Manag. Assoc. 47, 481–490. Kuhlmeier, P. D. (1997b). Sorption and desorption of arsenic from sandy soils: Column studies. J. Soil Contam. 6, 21–36. Lafferty, B. J., and Loeppert, R. H. (2005). Methyl arsenic adsorption and desorption behavior on iron oxides. Environ. Sci. Technol. 39, 2120–2127. Langner, H. W., and Inskeep, W. P. (2000). Microbial reduction of arsenate in the presence of ferrihydrite. Environ. Sci. Technol. 34, 3131–3136. Langner, H. W., Jackson, C. R., Mcdermott, T. R., and Inskeep, W. P. (2001). Rapid oxidation of arsenite in a hot spring ecosystem, Yellowstone National Park. Environ. Sci. Technol. 35, 3302–3309. Lin, Z., and Plus, R. W. (2000). Adsorption, desorption and oxidation of arsenic affected by clay minerals and aging process. Environ. Sci. Technol. 34, 3131–3136. Lengke, M. F., and Tempel, R. N. (2005). Geochemical modeling of arsenic sulfide oxidation kinetics in a mining environment. Geochim. Cosmochim. Acta 69, 341–356. Liu, J., Zheng, B., Aposhian, H. V., Zhou, Y., Chen, M., Zhang, A., and Waalkes, M. P. (2002). Chronic arsenic poisoning from burning high-arsenic-containing coal in Guizhou, China. Environ. Health Perspect. 110, 119–122. Livesey, N. T., and Huang, P. M. (1981). Adsorption of arsenate by soils and its relation to selected chemical properties and anions. Soil Sci. 131, 88–94. Lombi, E., Wenzel, W. W., and Sletten, R. S. (1999). Arsenic adsorption by soils and iron-oxide-coated sand: Kinetics and reversibility. J. Plant Nutr. Soil Sci. 162, 451–456. Ma, L. Q., Komar, K. M., Tu, C., Zhang, W. H., Cai, Y., and Kennelley, E. D. (2001). A fern that hyperaccumulates arsenic—A hardy, versatile, fast-growing plant helps to remove arsenic from contaminated soils. Nature 409, 579. Macur, R. E., Jackson, C. R., Botero, L. M., McDermott, T. R., and Inskeep, W. P. (2004). Bacterial populations associated with the oxidation and reduction of arsenic in unsaturated soil. Environ. Sci. Technol. 38, 104–111. Mahimairaja, S., Bolan, N. S., Adriano, D. C., and Robinson, B. (2005). Arsenic contamination and its risk management in complex environmental settings. Adv. Agron. 86, 1–82. Mandal, B. K., and Suzuki, K. T. (2002). Arsenic round the world: A review. Talanta 58, 201–235. Manning, B. (2005). Arsenic speciation in As(III)- and As(V)-treated soil using XANES spectroscopy. Microchimica Acta 151, 181–188. Manning, B. A., and Goldberg, S. (1996a). Modeling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci. Soc. Am. J. 60, 121–131. Manning, B. A., and Goldberg, S. (1996b). Modeling arsenate competitive adsorption on kaolinite, montmorillonite and illite. Clays Clay Miner. 44, 609–623. Manning, B. A., and Goldberg, S. (1997a). Arsenic(III) and arsenic(V) adsorption on three California soils. Soil Sci. 162, 886–895. Manning, B. A., and Goldberg, S. (1997b). Adsorption and stability of arsenic(III) at the clay mineral-water interface. Environ. Sci. Technol. 31, 2005–2011. Manning, B. A., and Suarez, D. L. (2000). Modeling arsenic(III) adsorption and heterogeneous oxidation kinetics in soils. Soil Sci. Soc. Am. J. 64, 128–137. Manning, B. A., Fendorf, S. E., and Goldberg, S. (1998). Surface structure and stability of arsenic(III) on goethite: Spectroscopic evidence for inner-sphere complexes. Environ. Sci. Technol. 32, 2383–2388. Mariner, P. E., Holzmer, F. J., Jackson, R. E., Meinardus, H. W., and Wolf, F. G. (1996). Effects of high pH on arsenic mobility in a shallow sandy aquifer and on aquifer

Reaction and Transport of Arsenic in Soils

111

permeability along the adjacent shoreline, Commencement Bay Superfund Site, Tacoma, Washington. Environ. Sci. Technol. 30, 1645–1651. Masscheleyn, P. H., Delaune, R. D., and Patrick, W. H. (1991). Effect of redox potential and pH on arsenic speciation and solubility in a contaminated soil. Environ. Sci. Techno. 25, 1414–1419. Matera, V., Le Hecho, I., Laboudigue, A., Thomas, P., Tellier, S., and Astruc, M. (2003). A methodological approach for the identification of arsenic bearing phases in polluted soils. Environ. Pullt. 126, 51–64. McGeehan, S. L., and Naylor, D. V. (1994). Sorption and redox transformation of arsenite and arsenate in two flooded soils. Soil Sci. Soc. Am. J. 58, 337–342. Mclaren, R. G., Naidu, R., Smith, J., and Tiller, K. G. (1998). Fractionation and distribution of arsenic in soils contaminated by cattle dip. J. Environ. Qual. 27, 348–354. Meharg, A. A., and Hartley-Whitaker, J. (2002). Arsenic uptake and metabolism in arsenic resistant and nonresistant plant species. New Phytol. 154, 29–43. Melamed, R., Jurinak, J. J., and Dudley, L. M. (1995). Effect of adsorbed phosphate on transport of arsenate through an oxisol. Soil Sci. Soc. Am. J. 59, 1289–1294. Miller, J., Akhter, H., Cartledge, F. K., and McLearn, M. (2000). Treatment of arseniccontaminated soils. II: Treatability study and remediation. J. Environ. Eng. -ASCE 126, 1004–1012. Morse, B. S., Pohll, G., Huntington, J., and Rodriguez Castillo, R. (2003). Stochastic capture zone analysis of an arsenic contaminated well using the generalized likelihood uncertainty estimator (GLUE) methodology. Water Resour. Res. 39(6), 1151, doi: 10.1029/2002WR001470. Mulligan, C. N., Yong, R. N., and Gibbs, B. F. (2001). Remediation technologies for metal-contaminated soils and groundwater: An evaluation. Eng. Geol. 60, 193–207. Myneni, S. C. B., Traina, S. J., Logan, T. J., and Waychunas, G. A. (1997). Oxyanion behavior in alkaline environments: Sorption and desorption of arsenate in ettringite. Environ. Sci. Technol. 31, 1761–1768. Myneni, S. C. B., Traina, S. J., Waychunas, G. A., and Logan, T. J. (1998). Experimental and theoretical vibrational spectroscopic evaluation of arsenate coordination in aqueous solutions, solids, and at mineral-water interfaces—Reevaluation of EXAFS results and topological factors in predicting geometry and evidence for m. Geochim. Cosmochim. Acta 62, 3285–3300. Nesbitt, H. W., Canning, G. W., and Bancroft, G. M. (1998). XPS study of reductive dissolution of 7 angstrom-birnessite by H3AsO3, with constraints on reaction mechanism. Geochim. Cosmochim. Acta 62, 2097–2110. Nesbitt, H. W., Muir, I. J., and Pratt, A. R. (1995). Oxidation of arsenopyrite by air and air-saturated, distilled water, and implications for mechanism of oxidation. Geochim. Cosmochim. Acta 59, 1773–1786. Nickson, R., McArthur, J., Burgess, W., Ahmed, K. M., Ravenscroft, P., and Rahman, M. (1998). Arsenic poisoning of Bangladesh groundwater. Nature 395, 338. Nikolaidis, N. P., Dobbs, G. M., Chen, J., and Lackovic, J. A. (2004). Arsenic mobility in contaminated lake sediments. Environ. Pollut. 129, 479–487. Nitzsche, O., Meinrath, G., and Merkel, B. (2000). Database uncertainty as a limiting factor in reactive transport prognosis. J. Contam. Hydrol. 44, 223–237. Nordstrom, D. K. (2002). Public health—Worldwide occurrences of arsenic in ground water. Science 296, 2143–2145. Nordstrom, D. K., and Archer, D. G. (2003). Arsenic thermodynamic data and environmental geochemistry. In ‘‘Arsenic in Ground Water: Geochemistry and Occurrence’’ (A. H. Welch and K. G. Stollenwerk, Eds.), Kluwer Academic Publishers, Boston, USA. O’Neill, P. (1995). Arsenic. In ‘‘Heavy Metals in Soils’’ (B. J. Alloway, Ed.), 2nd ed., pp. 105–121. Blackie, London.

112

Hua Zhang and H. M. Selim

Onken, B. M., and Hossner, L. R. (1996). Determination of arsenic species in soil solution under flooded conditions. Soil Sci. Soc. Am. J. 60, 1385–1392. Oremland, R. S., and Stolz, J. F. (2003). The ecology of arsenic. Science 300, 939–944. Ori, L. V., Amacher, M. C., and Sedberry, J. E. (1993). Survey of the total arsenic content in soils in Louisiana. Commun. Soil Sci. Plant Anal. 24, 2321–2332. O’Reilly, S. E., Strawn, D. G., and Sparks, D. L. (2001). Residence time effects on arsenate adsorption/desorption mechanisms on goethite. Soil Sci. Soc. Am. J. 65, 67–77. Oscarson, D. W., Huang, P. M., and Liaw, W. K. (1981a). Role of manganese in the oxidation of arsenite by fresh-water lake sediments. Clays Clay Miner. 29, 219–225. Oscarson, D. W., Huang, P. M., Dofosse, C., and Herbillion, A. (1981b). The oxidation power of Mn (IV) and Fe (III) oxides with respect to As (III) in terrestrial and aquatic environments. Nature (London) 291, 50–51. Oscarson, D. W., Huang, P. M., Liaw, W. K., and Hammer, U. T. (1983a). Kinetics of oxidation of arsenite by various manganese dioxides. Soil Sci. Soc. Am. J. 46, 644–648. Oscarson, D. W., Huang, P. M., and Hammer, U. T. (1983b). Oxidation and sorption of arsenite by manganese dioxide as influenced by surface coatings of iron and aluminum oxides and calcium carbonate. Water Air Soil Pollut. 20, 233–244. Parkburst, D. L., and Appelo, C. A. (1999). User’s guide to PHREEQC, version 2. USGS Water Resour. Inv. 9, 9–429. Paktunc, D., Foster, A., and Laflamme, G. (2003). Speciation and characterization of arsenic in Ketza river mine tailings using X-ray adsorption spectroscopy. Environ. Sci. Technol. 37, 2067–2074. Paktunc, D., Foster, A., Heald, S., and Laflamme, G. (2004). Speciation and characterization of arsenic in gold ores and cyanidation tailings using X-ray absorption spectroscopy. Geochim. Cosmochim. Acta 68, 969–983. Pedersen, H. D., Postma, D., and Jakobsen, R. (2006). Release of arsenic associated with the reduction and transformation of iron oxides. Geochim. Cosmochim. Acta 70, 4116–4129. Peryea, F. J., and Creger, T. L. (1994). Vertical-distribution of lead and arsenic in soils contaminated with lead arsenate pesticide-residues. Water Air Soil Pollut. 78, 297–306. Peryea, F. J., and Kammereck, R. (1997). Phosphate-enhanced movement of arsenic out of lead arsenate-contaminated topsoil and through uncontaminated subsoil. Water Air Soil Pollut. 93, 243–254. Pettry, D. E., and Switzer, R. E. (2001). Arsenic concentrations in selected soils and parent materials in Mississippi. MAFES Bulletin 104. Office of Agricultural Communications, Mississippi State University. Pickering, I. J., Prince, R. C., George, M. L., Smith, R. D., George, G. N., and Salt, D. E. (2000). Reduction and coordination of arsenic in Indian mustard. Plant Physcol. 122, 1171–1177. Pierce, M. L., and Moore, C. B. (1980). Adsorption of arsenite on amorphous iron hydroxide from dilute aqueous solution. Environ. Sci. Technol. 14, 214–216. Pierce, M. L., and Moore, C. B. (1982). Adsorption of arsenite and arsenate on amorphous iron hydroxide. Water Res. 16, 1247–1253. Pigna, M., Krishnamurti, G. S. R., and Violante, A. (2006). Kinetics of arsenate sorption– desorption from metal oxides: Effect of residence time. Soil Sci. Soc. Am. J. 70, 2017–2027. Puls, R. W., and Powell, R. M. (1992). Transport of inorganic colloid through natural aquifer material: Implication for contaminant transport. Environ. Sci. Technol. 26, 614–621. Qafoku, N. A., Kukier, U., Sumner, M. E., Miller, W. P., and Radcliffe, D. E. (1999). Arsenate displacement from fly ash in amended soils. Water air Soil Pollut. 114, 185–198.

Reaction and Transport of Arsenic in Soils

113

Radu, T., Subacz, J. L., Phillippi, J. M., and Barnett, M. O. (2005). Effects of dissolved carbonate on arsenic adsorption and mobility. Environ. Sci. Technol. 39, 7875–7882. Rahman, F. A., Allan, D. L., Rosen, C. L., and Sadowsky, M. J. (2004). Arsenic availability from chromated copper arsenate (CCA)-treated wood. J. Environ. Qual. 33, 173–180. Raven, K. P., Jain, A., and Loeppert, R. H. (1998). Arsenite and arsenate adsorption on ferrihydrite: Kinetics, equilibrium, and adsorption envelopes. Environ. Sci. Technol. 32, 344–349. Redman, A. D., Macalady, D. L., and Ahmann, D. (2002). Natural organic matter affects arsenic speciation and sorption onto hematite. Environ. Sci. Technol. 36, 2889–2896. Reigart, J. R., and Roberts, J. R. (1999). Recognition and Management of Pesticide Poisonings 5th Ed. United States Environmental Protection Agency. Reisinger, H. J., Burris, D. R., and Hering, J. G. (2005). Remediating subsurface arsenic contamination with monitored natural attenuation. Environ. Sci. Technol. 39, 458A–464A. Reynolds, J. G., Naylor, D. V., and Fendorf, S. E. (1999). Arsenic sorption in phosphateamended soils during flooding and subsequent aeration. Soil Sci. Am. J. 63, 1149–1156. Robinson, G. R., Larkins, P., Boughton, C. J., Reed, B. W., and Sibrell, P. L. (2007). Assessment of contamination from arsenical pesticide use on orchards in the great valley region, Virginia and West Virginia, USA. J. Environ. Qual. 36, 654–663. Rodriguez, R. R., Basta, N. T., Casteel, S. W., and Pace, L. W. (1999). An in vitro method to estimate bioavailable arsenic in contaminated soils and solid media. Environ. Sci. Technol. 33, 642–649. Rodriguez, R. R., Basta, N. T., Casteel, S. W., Armstrong, F. P., and Ward, D. C. (2003). Chemical extraction methods to assess bioavailability arsenic in soil and solid media. J. Environ. Qual. 32, 876–884. Roy, W. R., Hassett, J. J., and Griffin, R. A. (1986a). Competitive coefficient for the adsorption of arsenate, molybdate, and phosphate mixture by soils. Soil Sci. Soc. Am. J. 50, 1176–1182. Roy, W. R., Hassett, J. J., and Griffin, R. A. (1986b). Competitive interactions of phosphate and molybdate on arsenate adsorption. Soil Sci. 142, 203–210. Saada, A., Breeze, D., Crouzet, C., Cornu, S., and Baranger, P. (2003). Adsorption of arsenic (V) on kaolinite and on kaolinite-humic acid complexes—Role of humic acid nitrogen groups. Chemosphere 51, 757–763. Sadiq, M. (1997). Arsenic chemistry in soils: An overview of thermodynamic predictions and field observations. Water Air Soil Pollut. 93, 117–136. Sadiq, M., Zaidi, T. H., and Mian, A. A. (1982). Environmental behavior of arsenic in soils: Theoretical. Water Air Soil Pollut. 20, 369–377. Sakulpitakphon, T., Hower, J. C., Trimble, A. S., Schram, W. H., and Thomas, G. A. (2003). Arsenic and mercury partitioning in fly ash at a Kentucky power plant. Energy Fuels 17, 1028–1033. Santini, J. M., Sly, L. I., Schnag, R. D., and Macy, J. M. (2000). A new chemolithoautotrophic arsenite-oxidizing bacterium isolated from a gold mine: Phylogenetic, physiological, and preliminary biochemical studies. Appl. Environ. Microbiol. 66, 92–97. Scott, M. J., and Morgan, J. J. (1995). Reaction of oxide surface. I. Oxidation of As(III) by synthetic birnessite. Environ. Sci. Technol. 29, 1898–1905. Senn, D. B., and Hemond, H. F. (2002). Nitrate controls on iron and arsenic in an urban lake. Science 296, 2373–2376. Selim, H. M. (1992). Modeling the transport and retention of inorganics in soils. Adv. Agron. 47, 331–384. Sheindorf, C., Rebhun, M., and Sheintuch, M. (1981). A freundlich-type multicomponent isotherm. J. Colloid Interface Sci. 79, 136–142.

114

Hua Zhang and H. M. Selim

Smedley, P. L., and Kinniburgh, D. G. (2002). A review of the source, behaviour and distribution of arsenic in natural waters. Appl. Geochem. 17, 517–568. Smith, E., Naidu, R., and Alston, A. M. (1998). Arsenic in the soil environment: A review. Adv. Agron. 64, 149–195. Smith, E., Naidu, R., and Alston, A. M. (1999). Chemistry of arsenic in soils: I. Adsorption of arsenate and arsenite by selected soils. J. Environ. Qual. 28, 1719–1726. Smith, E., Naidu, R., and Alston, A. M. (2002). Chemistry of inorganic arsenic in soils: I. Effect of phosphorous, sodium, and calcium on arsenic sorption. J. Environ. Qual. 31, 557–563. Smith, S. L., and Jaffe, P. R. (1998). Modeling the transport and reaction of trace metals in water-saturated soils and sediments. Water Resour. Res. 34, 3135–3147. Sparks, D. L. (1998). ‘‘Soil Physical Chemistry.’’ CRC Press, Boca Raton. Sracek, O., Bhattacharya, P., Jacks, G., Gustafsson, J. P., and von Bromssen, M. (2004). Behavior of arsenic and geochemical modeling of arsenic enrichment in aqueous environments. Appl. Geochem. 19, 169–180. Stadler, S., Jann, S., Ho¨hn, R., Isenbeck-Schro¨ter, M., Niedan, V., Scholz, C., Tretner, A., Davis, J. A., and Kent, D. B. (2001). Tracer tests with As(III) in the oxic and suboxic groundwater zones at the USGS Cape Cod site, Mass., USA. WRI-10, 1013–1016. Su, C. M., and Puls, R. W. (2003). In situ remediation of arsenic in simulated groundwater using zerovalent iron: Laboratory colum tests on combined effects of phosphate and silicate. Environ. Sci. Technol. 37, 2582–2587. Su, C. M., and Puls, R. W. (2001). Arsenate and arsenite removal by zerovalent iron: Effects of phosphate, silicate, carbonate, borate, sulfate, chromate, molybdate, and nitrate, relative to chloride. Environ. Sci. Technol. 35, 4562–4568. Sun, X., and Doner, H. E. (1996). An investigation of arsenate and arsenite bonding structure on goethite by FTIR. Soil Sci. 161, 865–872. Sun, X., and Doner, H. E. (1998). Adsorption and oxidation of arsenite on goethite. Soil Sci. 163, 278–287. Sung, W., and Morgan, J. J. (1981a). Oxidative removal of Mn(II) from solution catalyzed by the gamma-FEOOH (lepidocrocite) surface. Geochim. Cosmochim. Acta 45, 2377–2383. Takahashi, Y., Minamikawa, R., Hattori, K. H., Kurishima, K., Kihou, N., and Yuita, K. (2004). Arsenic behavior in paddy fields during the cycle of flooded and non-flooded periods. Environ. Sci. Technol. 38, 1038–1044. Tamaki, S., and Frankenberger, W. T., Jr. (1992). Environmental biochemistry of arsenic. Rev. Environ. Contam. Toxicol. 124, 79–110. Tessier, A., Campbell, P. G. C., and Bisson, M. (1979). Sequential extraction procedure for the speciation of particulate heavy metals. Anal. Chem. 51, 844–851. Thomas, J. E., and Rhue, R. D. (1997). Volatilization of arsenic in contaminated cattle dipping vat soil. Bull. Environ. Contam. Toxicol. 59, 882–887. Tournassat, C., Charlet, L., Bosbach, D., and Manceau, A. (2002). Arsenic(III) oxidation by birnessite and precipitation of manganese(II) arsenate. Environ. Sci. Technol. 36, 493–500. USEPA. (1999). Use of monitored natural attenuation at superfund, RCRA corrective action, and underground storage tank sites. Report No. 9200. 4–17P. http://www.epa. gov/swerust1/directiv/d9200417.pdf. USEPA. (2001). National primary drinking water regulations; Arsenic and clarifications to compliance and new source contaminants monitoring; Final rule. Federal Register. Vol. 66. No. 14. 6975–7066. Jan 22. 2001. U. S. Gov. Print Office, Washington D. C. USEPA. (2002). Proven alternatives for aboveground treatment of arsenic in groundwater. Report No. EPA-542-S-02–002. USEPA. (2003). A Probabilistic Risk Assessment for Children Who Contact CCA-Treated Playsets and Decks. http://www.epa.gov/tio/tsp/download/arsenic_issue_paper.pdf.

Reaction and Transport of Arsenic in Soils

115

USGS. (2004).The national geochemical survey-Database and documentation. U.S. Geological survey open-file report 2004–1001. Violante, A., and Pigna, M. (2002). Competitive sorption of arsenate and phosphate on different clay minerals and soils. Soil Sci. Soc. Am. J. 66, 1788–1796. Violante, A., Ricciardella, M., Gaudio, S. D., and Pigna, M. (2006). Coprecipitation of arsenate with metal oxides: Nature, mineralogy, and reactivity of aluminum precipitates. Environ. Sci. Technol. 40, 4961–4967. Voigt, D. E., Brantley, S. L., and Hennet, R. J. C. (1996). Chemical fixation of arsenic in contaminated soils. Appl. Geochem. 11, 633–637. Walker, F. P., Schreiber, M. E., and Rimstidt, J. D. (2006). Kinetics of arsenopyrite oxidative dissolution by oxygen. Geochim. Cosmochim. Acta 70, 1668–1676. Waltham, C. A., and Eick, W. J. (2002). Kinetics of arsenic adsorption on goethite in the presence of sorbed silicic acid. Soil Sci. Soc. Am. J. 66, 818–825. Wauchope, R. D. (1975). Fixation of arsenical herbicides, phosphate, and arsenate in alluvial soils. J. Environ. Qual. 4, 355–358. Waychunas, G. A., Rea, B. A., Fuller, C. C., and Davis, J. A. (1993). Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta 57, 2251–2269. Wilkin, R. T., and Ford, R. G. (2002). Use of hydrochloric acid for determining solid-phase arsenic partitioning in sulfidic sediments. Environ. Sci. Technol. 36, 4921–4927. Williams, M. (2001). Arsenic in mine waters: An international study. Environ. Geol. 40, 267–278. Williams, L. E., Barnett, M. O., Kramer, T. A., and Melville, J. G. (2003). Adsorption and transport of arsenic(V) in experimental subsurface systems. J. Environ. Qual. 32, 841–850. Wolthers, M., Charlet, L., Van der Weijden, C. H., Van der Linde, P. R., and Rickard, D. (2005). Arsenic mobility in the ambient sulfidic environment: Sorption of arsenic(V) and arsenic(III) onto disordered mackinawite. Geochim. Cosmochim. Acta 69, 3483–3492. Woolson, E. A., Axley, J. H., and Kearney, P. C. (1971). The chemistry and phytotoxicity of arsenic in soils: I. Contaminated field soils. Soil Sci. Soc. Am. Proc. 35, 938–943. Woolson, E. A., Axley, J. H., and Kearney, P. C. (1973). The chemistry and phytotoxicity of arsenic in soils: Effect of time and phosphorous. Soil Sci. Soc. Am. Proc. 37, 254–259. World Health Organization (WHO). (2004). ‘‘Arsenic in Drinking Water.’’ WHO, Geneva. Xu, H., Allard, B., and Grimvall, A. (1988). Influence of pH and organic substance on the adsorption of As(V) on geologic materials. Water Air Soil Pollut. 40, 293–305. Yu, Y., Zhu, Y., Williams-Jonesb, A. E., Gao, Z., and Li, D. (2004). A kinetic study of the oxidation of arsenopyrite in acidic solutions: Implications for the environment. Appl. Geochem. 19, 435–444. Zhang, H., and Selim, H. M. (2005). Kinetics of arsenate adsorption-desorption in soils. Environ. Sci. Technol. 39, 6101–6108. Zhang, H., and Selim, H. M. (2006). Modeling the transport and retention of arsenic(V) in soils. Soil Sci. Soc. Am. J. 70, 1677–1687. Zhang, H., and Selim, H. M. (2007a). Colloid mobilization and arsenite transport through soil columns. J. Environ. Qual. 36, 1273–1280. Zhang, H., and Selim, H. M. (2007b). Modeling arsenate-phosphate retention and transport in soils: A multi-component approach. Soil Sci. Soc. Am. J. 71, 1267–1277. Zhao, H., and Stanforth, R. (2001). Competitive adsorption of phosphate and arsenate on goethite. Environ. Sci. Technol. 35, 4753–4757.

This page intentionally left blank

C H A P T E R

T H R E E

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia Bijay-Singh,* Y. H. Shan,† S. E. Johnson-Beebout,‡ Yadvinder-Singh,* and R. J. Buresh‡ Contents 1. Introduction 2. Criteria for Evaluating Crop Residue Management Options 2.1. Productivity and profitability 2.2. Environmental impact and sustainability 3. Type and Abundance of Crop Residues 4. Existing and Emerging Residue Management Options 4.1. Rice following rice or a non-flooded crop 4.2. Non-flooded crop following rice 5. Evaluation of Options with Residues Managed During a Rice Crop 5.1. Productivity 5.2. Profitability 5.3. Environmental impact 5.4. Sustainability 6. Evaluation of Options with Residues Managed During a Non-Flooded Crop 6.1. Productivity 6.2. Profitability 6.3. Environmental impact 6.4. Sustainability 7. Crop Residue and Bioenergy Options 8. Summary Acknowledgment References

* { {

118 121 122 122 123 125 129 132 135 135 152 153 158 160 160 174 175 179 181 183 185 186

Department of Soils, Punjab Agricultural University, Ludhiana 141 004, Punjab, India College of Environmental Science and Engineering, Yangzhou University, Yangzhou 225009, China International Rice Research Institute, Los Ban˜os, Philippines

Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00203-4

#

2008 Elsevier Inc. All rights reserved.

117

118

Bijay-Singh et al.

Intensification of rice-based cropping systems in Asia has substantially increased production of food and associated crop residues. The interval between crops in these systems is often brief, making it attractive for farmers to burn residues in the field to hasten and facilitate tillage for the next crop. Open-air burning causes serious air quality problems affecting human health and safety, and it has been banned by many Asian governments. In this chapter, we evaluate for rice-based cropping systems existing and emerging in-field alternatives to burning residues based on criteria of productivity, profitability, environmental impact, and sustainability. In intensive rice monocropping systems, residue is typically managed under conditions of soil flooding and anaerobic decomposition during the rice crop. In systems, where rice is rotated with an upland (non-flooded) crop, there are two major categories: residue of upland crop managed during flooded rice and rice residue managed during the upland crop. One option during the flooded rice crop is incorporation of residues from the previous rice or upland crop into the soil. Many studies have examined incorporation of crop residue during land preparation for flooded rice. In the vast majority of cases there was no significant increase in yield or profit. Residue decomposition in anaerobic flooded soil substantially increases methane (CH4) emission relative to residue removal. Surface retention of residue during rice cropping is challenging to implement because residue must be removed from the field during conventional tillage with soil flooding (i.e., puddling) and then returned. Alternatively, rice must be established without the traditional puddling that has helped sustain its productivity. Mulch is a good option for rice residue management during the upland crop, especially with reduced and no tillage. Mulch can increase yield, water use efficiency, and profitability, while decreasing weed pressure. It can slightly increase nitrous oxide (N2O) emission compared with residue incorporation or removal, but N fertilization and water management are typically more important factors controlling N2O emission than residue management. Long-term studies of residue removal have indicated that removing residue from continuous rice systems with near continuous soil flooding does not adversely affect soil organic matter (SOM). The use of crop residue as a mulch with reduced or no tillage for upland crops should be promoted in rice-based cropping systems. On the contrary, residues from the crop preceding rice on puddled and flooded soil can be considered for removal for off-field uses.

1. Introduction Rice (Oryza sativa L.) is the lifeline of Asia. More than 90% of the world’s total rice crop—or ~570 million tons of the estimated 630 million tons of global rice production in 2006/2007—is produced in Asia (FAO, 2007; USDA, 2007). Modern cultivars of rice with growth duration of

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

119

90–125 days can be cultivated in rotation with one or two additional crops in a year. Intensification and diversification are two main trends of rice-based cropping systems as they have evolved in different agroecological regions in Asia. The most prevalent cropping systems are rice–rice, rice– rice–rice, rice–rice–pulse, rice–wheat (Triticum aestivum L.), rice–oilseed crop, and rice–maize (Zea mays L.). Intensive irrigated rice systems, with two and sometimes three rice crops produced each year in the same field, are a dominant agricultural land use in the lowland tropics and subtropics of Asia (Cassman and Pingali, 1995). Intensive rice-based systems also show great diversity across Asia, where wheat, maize or one of many other secondary crops are grown during the part of the year when rice is not in the field. The rotation of rice and wheat, for example, is a major agricultural production system, which accounts for ~30% of the area of both rice and wheat grown in South Asia (Timsina and Connor, 2001; Ladha et al., 2003). Rice-based cropping systems are the most productive agroecosystems in Asia and produce the most food for the most people. Along with grain yield, these systems generate large amounts of crop residue. Historically, crop residues were often removed from fields for livestock bedding and feed, fuel for cooking, and other off-field purposes. More recently, the off-field uses of crop residues have tended to decrease in parts of Asia even as increasing quantities of crop residues have been produced as crop yields and cropping intensity increase. The intensification of land use results in less time between crops for managing these residues, which can interfere with tillage and seeding operations for the next crop. The lack of alternative uses for crop residues and lack of appropriate mechanization to handle increasing quantities of residue have driven Asian farmers increasingly to burn crop residues as a method of disposal (Flinn and Marciano, 1984; Yadvinder-Singh et al., 2005). Open-field burning of crop residues is recognized as a major contributor to reduced air quality and human respiratory ailments, particularly in China and northwestern India, which represent major irrigated rice ecosystems in Asia. Streets et al. (2003) estimated that 730 Tg of biomass are burned in a typical year from both anthropogenic and natural causes, excluding biofuel. Crop residue burning accounted for 34% of that total. Of the total crop residues burned, China contributed 44%, India 33.6%, Bangladesh 4.4%, Pakistan 4%, Thailand 3.1%, and Philippines 2.8%. The problems of openfield burning straw include atmospheric pollution and nutrient loss. One ton of crop residue on burning releases 1,515 kg CO2, 92 kg CO, 3.83 kg NOx, 0.4 kg SO2, 2.7 kg CH4, and 15.7 kg nonmethane volatile organic compounds (Andreae and Merlet, 2001). These gases and aerosols consisting of carbonaceous matter lead to adverse impacts on human health in addition to contributing to global climate change. Following the IPCC methodology (IPCC, 1996) for estimation of emission from open-field burning of crop residue and assuming 25% of the available residue is burned in the field, the

120

Bijay-Singh et al.

estimated emissions in 2000 from open-field burning of rice and wheat straw in India were 110 Gg CH4, 2306 Gg CO, 2.3 Gg N2O, and 84 Gg NOx (Gupta et al., 2004). For every ton of wheat residue burned, an estimated 2.4 kg of N is lost (Kumar et al., 2001), and up to 60% of the S content is lost (Lefroy et al., 1994). Many governments in Asia have made it illegal to burn crop residues, but these laws have been difficult to enforce. There has been increased realization that crop residues are a resource constituting a readily available source of nutrients and organic material for rice farmers. About 40% of the N, 30–35% of the P, 80–85% of the K, and 40–50% of the S absorbed by rice remain in the vegetative parts at maturity (Dobermann and Fairhurst, 2000). Typical amounts of nutrients in rice straw at harvest are 5–8 kg N, 0.7–1.2 kg P, 12–17 kg K, 0.5–1 kg S, 3–4 kg Ca, 1–3 kg Mg, and 40–70 kg Si per ton of straw on a dry weight basis (Dobermann and Witt, 2000). Residue removal can therefore have a significant effect on soil nutrient depletion. Residue management also influences availability of micronutrients such as zinc and iron, and it is an important factor in maintaining the cumulative Si balance in rice (Dobermann and Fairhurst, 2000, 2002). Residues must be carefully managed for obtaining positive effects on soil and crop production and avoiding negative effects such as interference with the planting of crops, N immobilization, and emission of greenhouse gases. The return of crop residues to flooded soils, which are typical in ricebased cropping systems, influences the chemical, physical, and biological soil environment in different ways than the return of residues to aerobic soils prevalent when crops are grown under non-flooded soil conditions. The desired objectives of adopting a particular crop residue management option can be achieved only if the management option is feasible under a given set of soil, climate, and crop management conditions; is compatible with available machinery; and is socially and economically acceptable. This chapter considers existing and emerging in-field management practices. In addition, there are a variety of potentially attractive and competing offfield uses for crop residues such as animal feed, roof thatch, manufacture of paper or cardboard, and biofuel feedstock. Several reviews on crop residue management of rice systems have appeared in recent years. Kumar and Goh (2000) reviewed crop residues in terms of soil quality, soil N dynamics, crop yields, and N recovery. Their chapter primarily deals with the decomposition and turnover rates of residues in relation to nutrient cycling. Yadvinder-Singh et al. (2005) reviewed work on crop residue management in rice-based cropping systems in the tropics, dealing with short- and long-term effects on cycling of C, N, and other nutrients in order to provide necessary understanding for developing suitable new crop residue management options. They also explained the need for evaluating the relative costs of different residue management options for rice-based cropping systems in terms of environmental impact,

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

121

C sequestration, and long-term soil fertility. In this chapter, we describe important in-field residue management options for rice-based cropping systems including China, which did not receive much attention in the earlier reviews, and we evaluate management options based on criteria of productivity, profitability, environmental impact, and sustainability. We include residue removal in our evaluation of in-field options, but we have not detailed off-field options. We exclude high-N residues, such as from green manures, because they are not currently common in rice-based cropping systems in Asia. This chapter focuses on lowland rice-based cropping systems in Asia, in which fields are typically surrounded by earthen levees or bunds to impound water during the period of rice cropping. ‘‘Lowland’’ indicates a cultivation practice of growing rice under either irrigated or rainfed conditions with impoundment of water to flood the soil—typically during land preparation for rice production and during at least part of the rice growing season. The soil is largely anaerobic during the periods of flooding. ‘‘Upland,’’ in contrast, refers to the cropping period or crop in rice-based cropping systems when the soil is not flooded and aerobic. In this chapter, the terms ‘‘upland crop’’ and ‘‘non-flooded crop’’ synonymously refer to the crop (typically not rice) in the cropping system grown without soil flooding. In this chapter, crop residue is defined as the above-ground part of the plant remaining after the grain is harvested. It includes both the stubble left standing during the harvest process and the leaves and stems left over after threshing. Because harvest practices and nomenclature vary across Asia, ‘‘residue removal’’ and ‘‘straw removal’’ can mean different things in different locations and literature. Sometimes it means removal of all biomass from the soil surface upwards; but often it means removal of biomass except the standing stubble, which can represent an important quantity of biomass depending on the height of crop harvest. Roots are also a source of organic material that crops contribute to soil every season, but they are not included in our definition of crop residue. Roots are almost always retained in the soil, and there are few other management options for them in rice-based cropping systems. Composts, animal wastes, and manures produced from residue removed from the field are outside the scope of this chapter.

2. Criteria for Evaluating Crop Residue Management Options In order to make and implement sound decisions about residue management, it is necessary to scientifically understand the short- and long-term effects of different crop residue management practices and to develop residue management technologies that provide agronomic benefit

122

Bijay-Singh et al.

in a cost-effective and environmentally acceptable fashion. Crop residue management options are evaluated in this chapter using criteria of productivity, profitability, environmental impact, and sustainability for the cropping system. These criteria coincide with those used in the approach of ecological intensification for intensive crop production systems, which aims to satisfy the increasing demand for food, feed, fiber, and fuel while meeting acceptable standards of environmental quality (Cassman, 1999; Witt, 2003). Success in achieving ecological intensification depends greatly on sustaining yield increases in major irrigated and favorable rainfed cereal systems such as the rice-based cropping systems covered in this chapter. This chapter focuses on evaluating potential large-scale effects of residue management options rather than on the effects of residue management on specific processes, which have already been reviewed for soil fertility (Wilhelm et al., 2004; Yadvinder-Singh et al., 2005), C cycling (Martens, 2000), pesticide interactions with soil microorganisms (Moorman, 1989), soil-borne diseases (Chung et al., 1988), and root health (Allmaras et al., 1988).

2.1. Productivity and profitability Productivity and profitability are criteria directly relevant to farmer’s decision making. The quantifiable indicators of short-term productivity (i.e., 1–3 seasons of a given management option) include grain yield, fertilizer use efficiency (grain yield per unit fertilizer applied), water use efficiency (grain yield per unit water applied), and yield loss due to disease, insect, or weed pressure. Profitability indicators include income from yield less inputs (i.e., labor, fertilizer, seed, machinery, irrigation water, and pesticide). Residue management options can differ in their effects on these indicators of productivity and profitability. We have therefore assessed the residue-associated changes for each indicator for different in-field residue management options as compared to either removing or burning residues.

2.2. Environmental impact and sustainability Environmental impact and sustainability are criteria that are not typically important determinants for farmers in their selection of a particular residue management option, but these criteria can be important for policy making such as with the banning of open-field burning of crop residues. The main short-term (i.e., measurable within 1–3 seasons) environmental impacts associated with residue management include changes in air quality and greenhouse gas emission. Sustainability refers to the medium- and longterm (i.e., 5 years or more) ability of a residue management option to maintain or increase the productivity and profitability of the cropping system. Indicators include trends through time in yield, input use efficiency,

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

123

soil parameters (such as N supply capacity, organic matter, K, S, and bulk density), and profitability.

3. Type and Abundance of Crop Residues The geographical distribution of crop residues in Asia is skewed by the large crop production in India and China. These two countries in 2006/2007 accounted for ~51% of global rice production and ~57% of Asian rice production (FAO, 2007; USDA, 2007). Frolking et al. (2002) by combining remote sensing and ground census data to develop maps for the distribution of rice in China showed 25% of the lowland rice cropland was planted as single rice, 15% was a double-crop rotation with two rice plantings per year (rice–rice), 19% was a double-crop rotation with a single rice planting (rice– other crop), and 41% was a triple-crop rotation with two rice plantings per year (rice–rice–other crop). Rice was rotated on an estimated 4.7 million ha with wheat, 4.5 million ha with rapeseed (Brassica napus L.), and 2.2 million ha with oat (Avena sativa L.). More recently, the area of rice–wheat in China was estimated at 3.4 million ha (Dawe et al., 2004). Yadav and Subba Rao (2001) estimated an area in India of 9.2 million ha for rice–wheat, 2.4 million ha for rice–oilseed, 3.5 million ha for rice–pulse. More recently, the area of rice–wheat was estimated at 10 million ha in India and 13.5 million ha for the Indo-Gangetic Plain, which includes Bangladesh, India, Nepal, and Pakistan (Ladha et al., 2003). The cropping patterns in rice-based cropping systems remain dynamic in response to markets, policies, and labor availability. We estimated total production area, grain yield, and production of residues for rice-based cropping systems in Asia from available 2004 data (Table 1). Rice-based cropping systems are defined as those with at least one rice crop per year grown either as a sole crop or in rotation with rice or a non-flooded crop. In these systems the rice would typically be grown on flooded soil, which markedly influences the management options for residues. Available data for area of a given crop in Asia typically does not specify the rotational system in which the crop is grown. Therefore, the areas for nonrice crops in rice-based cropping systems shown in Table 1 are based on estimations on a country basis. Economic yield data (grain or cane yield) were computed from the online databases (FAO, 2007). The residue-toeconomic yield ratios as listed in Table 1 are based on a range of reported data (Barnard and Kristoferson, 1985; Beri and Sidhu, 1996; Koopmans and Koppejan, 1997; Muehlbauer and Tullu, 1997; Yevich and Logan, 2002). According to these estimations for 2004, rice accounted for ~84% of total residue from rice-based cropping systems while wheat and maize accounted for 9% and all other crops accounted for 7%. The estimations highlight the large quantity of residues produced in rice-based cropping systems in Asia.

124

Bijay-Singh et al.

Table 1 Residue production and area for rice and different crops grown in rotation with rice in Asia in 2004

Crop

Rice Wheat Maize Sugarcane Rapeseed Soybean Oats Lentil/Pulses Total

Harvested area (106 ha)

Economic yield (106 t)

Residue-toeconomic yield ratio

134 19 1.4 1.7 7.1 0.3 0.3 2.2

546 57 4.5 107 11 2.7 0.7 1.7

1.4 1.3 2.0 0.17 3.5 2.5 1.5 1.2

Residues (106 t)a

764 74 9 18 38 7 1 2 913

a

Data pertaining to production of residues was computed by multiplying economic yield of different crops with residue-to-economic yield ratios. Source: Data and conversion factors were obtained from FAO (2007), Frolking et al. (2002), Yadav and Subba Rao (2001), Koopmans and Koppejan (1997), Barnard and Kristoferson (1985), Yevich and Logan (2002), Beri and Sidhu (1996), and Muehlbauer and Tullu (1997).

This quantity will increase as yields and intensity of cropping continue to increase. The residues from nonrice crops as a proportion of the total would be expected to increase as rice-based cropping systems diversify. Little data are available on the use of these residues, but there have been several attempts in China and India to estimate the amount returned to soil. Gao et al. (2002) estimated 37% of ~600 million tons of crop residue produced in China are returned to soil, often in the form of rice stubble. In 15 provinces of China with most of the country’s cereal production, crop residues were returned to fields on an estimated 18.2 million ha, comprising of 37% of the total cultivated land in 2000 (Han et al., 2002). In India, an estimated 250 million tons of residues is produced annually in rice–wheat cropping systems in the Indo-Gangetic Plain (Pal et al., 2002). Huge amounts of residues are available either for retaining in fields to enhance productivity and fertility of the soil or for removing from the field for alternative uses, but in many areas of Asia the crop residues produced in rice-based cropping systems have been considered a nuisance by farmers and disposed through burning in fields.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

125

4. Existing and Emerging Residue Management Options Most lowland rice ecosystems in Asia have a rainy season when climatic conditions favor production of rice rather than other crops. Nonflooded crops are often grown in rotation with rice during the drier season. Soil in irrigated and rainfed lowland ecosystems with sufficient water typically remains flooded for most of the rice-cropping season. The choices for managing crop residue can consequently differ between periods with rice cropping, when anaerobic decomposition of residues predominates, and periods with other crops, when the aerobic decomposition of residues predominates. In intensive nonrice cropping systems with reduced or no tillage such as in Europe, America, and Australia, crop residues are often left in the field after combine harvesting, and seed for the next crop is then sown directly into the residue without plowing. In a rice production system in California, where legislation banned traditional open-field burning of residue, many farmers have adopted a system of retaining residues as a habitat for migratory, foraging waterfowl that hasten the decomposition of the residue during a winter-flood fallow (Bird et al., 2000). Tropical rice production systems in Asia, which are characterized by intensive cropping, do not have such long fallow periods with ample water between crops. Rice production systems in Asia also do not lend themselves to reduced or no tillage options because they are characterized by soil puddling— the plowing and harrowing of soil when flooded. Puddling destroys soil structure, restricting downward movement of water to maintain flooding (Sharma and De Datta, 1986), and soil flooding controls weeds and helps sustain the productivity of rice-based cropping systems. Soil puddling, however, restricts options for surface application of crop residues. Mulching in a puddled field typically necessitates removal of the residue before puddling and then returning it afterwards. Management practices for retention of crop residues are listed for different regions of China in Table 2 and for South Asia in Table 3. The recommended management for crop residues during cropping with flooded rice has typically been incorporation into the soil during land preparation. Yet farmers in Asia have not often in recent years followed this recommendation, electing instead the open-field burning of residues. But with increasingly strict legislation against open-field residue burning, a trend of increasing residue return can be anticipated, either through incorporation or as mulch. The retention of residues on the soil surface as a mulch is often an option during the nonrice crop in rice-based cropping systems, which can be established with

126

Bijay-Singh et al.

Table 2 On-field residue management practices in rice cropping systems in different regions of China

Region

Cropping system

Existing residue management practice

Northeast

Rice

Yangtze Valley

Rice–wheat Rice–rapeseed

Southwest

Rice–rice Rice–wheat Rice–rapeseed

South China

Rice–rice Rice–rice– rapeseed

Stubble remaining + shallow plowing Mulching with rice residues in wheat (or rapeseed) season Incorporation in rice season Mulching with rice residues in wheat (or rapeseed) season Incorporation in rice season Mulching with rice residues in rapeseed season Incorporation in rice season

Amount of residue returned per year (t ha–1)

2.3 4.0

4.2

4.9

Sources: Zeng et al. (2001, 2002).

conventional, reduced, or no tillage. A challenge for mulching with reduced and no tillage is to ensure sufficient soil–seed contact after sowing. Methods of harvesting and threshing are critical in determining what happens to crop residue, and are often chosen because of the intended use of the straw. Some of the common harvest methods for rice in Asia include hand-cutting, use of small harvesting machinery, and use of combined harvester-threshers (Gummert and Aldas, 1993; Saunders et al., 1980). When hand-cutting, workers sometimes cut at the soil surface if they want long straw for animal bedding or roof thatch, but they might choose to cut part way up (20–60 cm above soil) to reduce the weight to be carried. The harvested part is carried to a centralized location on- or off-field for threshing to remove the grain from the straw. It can be threshed either manually or mechanically. Manual threshing involves hand flailing, swinging the straw over the head and beating it against a firm object in front that allows easy collection of the grain. If the straw is valuable to the farmer, it can then be stored for future use. With mechanical threshing, the chopped straw accumulates on a pile at the location of threshing. Combined harvesting–threshing machines separate the grain from the straw as they cut the plant and move through the field, retaining the grain and leaving the

Table 3

Existing and emerging in-field residue management practices in rice cropping systems in different regions of South Asia Existing residue management practices and (amount of residue returned per year, t ha–1)

Emerging residue management options and (amount of residue that could be returned per year, t ha–1)

Mulching with rice straw in no-till wheat (5–7) Incorporation of straw and stubble of combine harvested rice in wheat (5–7) Incorporation of combine harvested wheat straw and stubble in rice (1–2) Mulching with rice straw in no-till wheat (~5) Incorporation of manually harvested or combine harvested wheat straw and stubble in rice (~1)

Region

Cropping system

Trans- and Upper IndoGangetic Plain

Rice–wheat

Incorporation of rice and wheat stubble remaining in the field (~1)

Middle- and Lower IndoGangetic Plain

Rice–wheat Rice–oilseed Rice–pulses Rice–jute–rice Rice–vegetable Rice–vegetable– rice Rice–wheat Rice–wheat– pulses

Incorporation of stubble remaining in the field (~1) Rice straw mulch in vegetable production (1–2)

Non-Indo-Gangetic Plain (Terai of Nepal, Bihar, and Uttranchal) South India

Rice–rice Rice–rice– rice Rice–pulses

Source: Based on Gajri et al. (2002) and Pal et al. (2002)

Incorporation of rice and wheat stubble (~1)

Mulching of rice residues in wheat (~4)

Incorporation of rice stubble (~1)

Incorporation of rice straw and stubbles (~4)

128

Bijay-Singh et al.

chopped straw spread on the field. It is usually advantageous to leave high stubble and only move the upper portion of the plant through the machine. Combine-harvested fields consequently tend to have relatively tall standing stubble and short pieces of chopped straw laying on the surface. Methods of harvesting and threshing determine the percentage of total biomass left in the field as standing stubble, and the condition and location of the threshed straw. If standing stubble is not burned in the field, it is typically incorporated during land preparation for the subsequent crop. The primary management options, therefore, usually relate to the threshed straw rather than the stubble, although one management decision could be to adjust the harvesting procedure to increase or decrease the proportion of biomass that is threshed. Interventions for management of residues in rice cropping systems can be categorized based on the type of residue and crop in the system as follows: rice residue for a subsequent rice crop, rice residue for a nonflooded crop, and residue of a non-flooded crop for a subsequent flooded rice crop. These three situations can be further simplified into the following two cropping system categories based on the critical distinction of whether soil is flooded or non-flooded during the crop receiving the residues. 1. Rice following rice (common in South China, Southeast Asia, and southern India) and rice following an upland or non-flooded crop (common in Central and North China and parts of South Asia). In both cases, crop residue is managed during a rice crop, typically established on puddled soil. 2. Upland or non-flooded crop following rice (common in the IndoGangetic Plain in South Asia and many parts of China). In this case, rice residue is managed during a crop grown on non-flooded soil. Intensive rice monocropping systems are often the most challenging for managing crop residues because of the short time interval between rice crops. In rotations with rice and a non-flooded crop, the two crops often differ in the magnitude of the challenges for residue management. Management options are affected by the time of year when residue becomes available and the time before the next crop is planted. In the rice–wheat system in the Indo-Gangetic Plain, there is a relatively short time between harvesting rice and sowing wheat, hence the management of rice straw during wheat cropping (i.e., non-flooded crop following rice) is a critical issue. There is a relatively longer fallow after wheat, and wheat straw is valuable for off-field uses, especially as animal feed (Samra et al., 2003). Hence, the management of wheat straw during rice cropping (i.e., rice following a non-flooded crop) is not as critical. In China, however, there is a very short fallow between wheat and rice, often less than 1 week, because of the typically long growth duration of rice, and wheat straw does not have off-field value as in India. Hence, the management of wheat residue during

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

129

rice (i.e., rice following a non-flooded crop) presents the bigger challenge. In each category of cropping patterns, there are a series of management options differing in relative feasibility and attractiveness.

4.1. Rice following rice or a non-flooded crop Residue management in this category is strongly affected by soil flooding and anaerobic soil conditions during the rice crop. Residues from the previous crop, whether rice or a non-flooded crop, typically must be managed in an environment of historical soil puddling and at least partial anaerobic decomposition. Factors that influence decisions about residue management are harvest method of the previous crop, turnaround time between crops, land preparation practices, availability of implements and labor, and establishment method and water management of the new rice crop. Residue management scenarios vary greatly depending upon whether the previous crop has been harvested manually or by combine harvesters, threshed at locations in field resulting in straw piles, or threshed outside the field. The method of harvesting also determines the extent to which crop residues remain anchored or loose. 4.1.1. Incorporation Residue incorporation into flooded soil has continued to be promoted as an alternative to open-field burning in rice-based cropping systems across Asia. It is a potentially attractive option because rice residues can typically be plowed into the soil as part of the normal tillage operations for preparing the rice field, and residue incorporation therefore does not require an extra step in land preparation. Once incorporated the residue typically decomposes relatively fast thereby potentially providing benefits to the next rice crop. Incorporation also avoids loose residue on the soil surface that could interfere with the preparation of the seedbed or planting of the next crop. Incorporation can reduce the risk of pests and diseases as compared with mulching, and it can potentially increase soil organic C fractions and total organic C. Specific management decisions include tillage method and timing of incorporation relative to rice establishment by transplanting or direct seeding. During land preparation for lowland rice in Asia, the topsoil is typically inverted thereby incorporating crop residues remaining on the soil surface as standing stubble or loose straw. A moldboard plow or disk plow is commonly used for incorporating residues often with a shallow layer of floodwater (Ponnamperuma, 1984). The degree of incorporation varies among tillage systems depending on implement, intensity, and mechanization level (i.e., manual, animal traction, or mechanized) (Sharma and De Datta, 1986). Incorporation of large amounts of fresh residue is labor intensive if suitable machinery is not available (Dobermann and Fairhurst, 2000). When rice is

130

Bijay-Singh et al.

harvested using a combine harvester that leaves straw spread in the field, residue can be incorporated into the soil by disking or plowing. If only stubble is retained, the amount of residue incorporated into the soil is determined by the manner and the height of harvesting. Incorporation of residues might not be feasible when long straw clogs field implements. In rice–rice systems in Hunan and Hubei Province of China with only a few days between early and late season rice, a practice for rapid incorporation of residue before immediate establishment of the next crop is cutting the rice residue into ~20–25 cm lengths followed by shallow mechanical incorporation (Zeng et al., 2001). In rice–rice cropping systems with 2–3 months between rice crops, the shallow incorporation of rice residue into aerobic soil soon after harvest rather than delaying the incorporation of residues until preparation of flooded soil immediately before establishment of the next rice crop has been proposed as a practice to accelerate residue decomposition and release of plant-available nutrients (Dobermann and Witt, 2000). Witt et al. (1998) reported rapid decomposition of rice residue following shallow incorporation into aerobic soil, leading to ~50% loss of residue-C within 30–40 days and increased supply of plant-available N. Thuy (2004) in three cropping seasons in the Philippines consistently found comparable or significantly higher KCl-extractable soil ammonium at rice transplanting and uptake of soil N by rice at panicle initiation when residue from the previous rice crop was incorporated into aerobic soil immediately after harvest rather than incorporated later by the traditional practice during puddling ~3 weeks before transplanting. Shallow incorporation of rice residue into aerobic soil after harvest can also help reduce weed growth, save irrigation water by reducing soil cracking, and allow additional time for phenol degradation under aerobic conditions (Dobermann and Witt, 2000). The time interval between incorporation of crop residue and land preparation, flooding, and transplanting the next rice crop is a crucial factor affecting residue management. This time interval influences the extent of residue decomposition before transplanting, depending upon soil and climatic conditions, thereby affecting the beneficial or adverse effects of residue incorporation on young rice seedlings. In intensive irrigated production systems, two or three short-duration crops are typically grown per year. In the Red River Delta of northern Vietnam, for example, rice–rice–maize is a common cropping system. In the Mekong Delta of southern Vietnam where rice is grown continuously, the intensity of rice cropping can reach 6–7 crops in 2 years (Dobermann et al., 2004). In such intensive triple cropping systems, the fallow between two crops can be only a few days. Whereas a relatively long fallow period of 2–3 months provides opportunities to manage residues for hastened decomposition and nutrient release, a fallow of only a few days does not enable appreciable residue decomposition and release of nutrients before establishment of the next rice crop. In such

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

131

cases the incorporation of large quantities of crop residues poses challenges and potential detrimental effects on the subsequent rice crop. 4.1.2. Mulching Adding mulch to flooded rice is usually not feasible because the traditional practice of puddling will incorporate retained crop residue. Mulching typically requires transfer of biomass off of the field before soil puddling and then return of the biomass after land preparation. In rice–rice systems in South China, some farmers do not drain their fields at harvest of the early rice crop and keep the field flooded without tillage during the brief transition period to transplanting of the late rice crop. Straw from the early rice is placed as mulch in rows along the direction of transplanting for late rice, and late rice is transplanted between the rows of straw covered soil (Li, 1991). The continuous flooding helps ensure the soil is sufficiently moist for easy transplanting of late rice, and the mulch helps control weed growth and prevent ratooning of rice. Other rice farmers and researchers in China have been trying reduced and no tillage flooded rice systems that enable surface retention of crop residue. In Anhui, Guangdong, Hunan, Hubei, Jiangsu, and Zhejiang Provinces, farmers have used either the throwing of rice seedlings or the direct sowing of germinated rice seeds as methods for establishing rice in reduced or no tillage fields (Li, 2005). In such systems crop residue is retained on the soil surface. The soil is saturated or flooded during crop establishment, and herbicides are used to control weeds. In Anhui, Jiangsu, and Zhejiang Provinces, some farmers practice relay cropping whereby rice is sown in wheat fields before combine harvesting. The standing stubble of wheat then slowly decomposes during rice cropping. A ground covering rice production system (GCRPS) to save water and increase N efficiency has been developed in South China. It involves growing a lowland rice variety under non-flooded conditions (70–90% of water holding capacity) with the ground covered by rice straw mulch during growth (Fan et al., 2002; Huang et al., 1997; Lin et al., 2002; Luo, 1997; Zhao et al., 1999). Mulching increased soil organic C and total N (Fan et al., 2002; Liu et al., 2003), but grain yield was sometimes lower than when rice was grown under flooded conditions. 4.1.3. Composting Most composting techniques are off-field residue management options in which the produced compost is not returned to the main production field, and as such are not included in this chapter. Some composting can, however, occur in fields (in situ composting), and a small portion of the compost produced off-field can be returned to the main field. One example of in situ composting is where rice straw is piled in the field at threshing sites (Ponnamperuma, 1984). The straw decomposes slowly, largely aerobically,

132

Bijay-Singh et al.

and it can then be spread and incorporated into the soil at the beginning of the next season. Constraints of this practice include providing a favorable habitat for rodent pests and promoting excessive immobilization of N under the residue piles. Another type of in situ composting, in which the residue of the previous crop is buried in ditches paralleling rows of transplanted rice, has been examined in China for residues of non-flooded crops, such as wheat and barley, grown immediately before rice (Zhong et al., 2003). Crop residue can be removed from the field, composted alone or with other organic materials originating at the farm such as animal wastes, and then returned to soil as manure for the rice crop. The potential of composting to turn on-farm waste materials into a farm resource makes it an attractive proposition. Traditional methods based on a passive composting approach involve simply stacking crop residues in piles or pits to decompose over a long period with little agitation and management (Misra et al., 2003). The time requirement can be reduced through a few turnings, which slightly enhance passive aeration. Chinese rural composting methods also use a passive aeration approach based on turnings and aeration holes, and they provide output in 2–3 months (Ma, 2004). Low turnover and long time span are major bottlenecks. Traditional passive methods require several months from the time of crop harvest until the compost is ready for use. Composting involves labor input, but it is not capital intensive and does not require sophisticated infrastructure and machinery. Small farmers without manual labor constraints are most likely to benefit from composting technology.

4.2. Non-flooded crop following rice Managing rice residue during a non-flooded crop is somewhat easier than managing it during flooded rice because options for reduced or no tillage and mulching are more feasible. In addition, incorporated residue usually decomposes faster in aerobic than in flooded soil. There are, however, challenges. Major factors affecting residue management decisions include method of rice harvest, the time interval between crops, water management, and method of tillage for the subsequent non-flooded crop. The extent to which rice residue remains anchored after harvest also influences the way it can be managed in the following non-flooded crop. The most common options include incorporation with conventional tillage or surface mulching usually with reduced or no tillage. 4.2.1. Incorporation Incorporation of rice residue into the soil before planting a non-flooded crop has been frequently studied, particularly where rice residue does not have off-field economic use such as for animal fodder, fuel, or industrial purposes. Various options are available for farmers to incorporate crop

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

133

residues into the soil depending upon availability of machines, financial resources, and amount of straw (Ball and Robertson, 1990). For example, residues can be directly incorporated using a moldboard plow, or they can be chopped using a straw chopper after harvesting with a combine, and then the chopped residue can be easily incorporated into the soil using a disc plow (Sidhu and Beri, 2005). One reason for the popularity of residue incorporation among scientists is that residue can be incorporated during the preparatory tillage for the non-flooded crop, and hence it does not entail extra cost for implementation. Another option is incorporation of residue with a separate field operation several weeks before land preparation for the following crop. This allows more time for the residue to decompose and helps to control weeds. Residue can be incorporated partially or completely into the soil depending on method of cultivation used. The time interval between residue incorporation and planting of the next crop is determined by the cropping calendars and the time needed for residue decomposition. Yadvinder-Singh et al. (2004b), for example, in a rice–wheat cropping system in the northwestern India, observed that rice residue decomposition of ~25% during the pre-wheat fallow period was sufficient to avoid any detrimental effects on wheat yields. Rice and wheat productivity in a 7-year study was not adversely affected when rice residues were incorporated at least 10 days and preferably 20 days before the establishment of the succeeding crop. 4.2.2. Mulching with reduced or no tillage A reduced or no tillage system makes it relatively easy to retain residue on the surface as mulch simply by leaving it on the field during harvest. It is not necessary to remove the residue before tillage and then return it. However, if residue is threshed off-field, it must be transported to and spread on the field, resulting in no saving in labor for handling residues as compared to mulching with conventional tillage. Direct drilling is a method of sowing the crop after rice harvest without cultivation or incorporation of residue. According to Li (1991), no-till sowing of winter crops including wheat, barley (Hordeum murinum L.), and rapeseed is commonly practiced by farmers in rice-based cropping systems in eastern China. Surface seeding of wheat by making small holes in the soil (dibbling) followed by mulching with rice residue (4–6 t ha–1) is practiced on ~60% of the rice–wheat system in the Sichuan Basin (Humphreys et al., 2004). Because the time between rice harvest and wheat sowing is relatively long (60–85 days) in middle and southern China, the spreading of rice residue as mulch immediately after harvest has been examined as a practice for controlling weeds and reducing evaporation during the fallow before the next non-flooded crop (Zeng et al., 2001, 2002). This technique is attractive for farmers growing rapeseed instead of wheat after rice because they can broadcast the seeds, which are

134

Bijay-Singh et al.

small enough to fall through the rice mulch to the soil surface. A double zero-tillage system of no-till, direct seeded rice and no-till rapeseed with rice residue as mulch for rapeseed has been developed in Hunan Province (Zou et al., 2004b). Reduced and no tillage for wheat has been increasingly adopted by farmers in the Indo-Gangetic Plain in northwestern India since the late 1990s because it leads to large cost savings through reduced use of fuel and labor (Erenstein et al., 2007). In the eastern part of the Indo-Gangetic Plain, it also facilitates early sowing leading to potential yield benefits, especially after late harvested rice. Rice residue mulching in fields seeded with wheat with reduced and no tillage is also practiced by a small number of farmers in the Terai of Nepal and in eastern Uttar Pradesh and Bihar states of India (Humphreys et al., 2004). The area of reduced and no-till wheat in the Indo-Gangetic Plain has expanded at an exponential rate since the late 1990s, increasing to an estimated 20–30% of the rice–wheat area or 2–3 million ha in 2006 (RWC, 2006). No-till sowing of wheat after combine-harvested rice, however, involves some difficulties including residue accumulation in the furrow openers, traction problems with the drive wheel of the seed drill, difficulty with fertilizer metering systems in the loose straw, and nonuniform sowing depth due to frequent lifting of the drill to clear blockages. A number of approaches are currently being tested for direct drilling into rice residue to solve the problem of machinery clogging and ‘‘hair-pinning’’ when the straw bends but is not cut or buried, resulting in seed remaining on the surface. These include double and triple disc systems (Gupta and Rickman, 2002), the straw thrower (Shukla et al., 2002), and the stubble chopper (Garg, 2002); although none of these approaches has been particularly successful to date. A promising new approach is the ‘‘Happy Seeder,’’ which combines the stubble mulching and seed drilling functions into one machine (Blackwell et al., 2004). The stubble is cut and picked up in front of the sowing tines, which therefore engage bare soil, and deposited behind the seed drill as mulch. The evolution of the technology, leading to a machine called the Comboþ Happy Seeder, is described by Humphreys et al. (2006). Results to date from India suggest that wheat can emerge through 8 t ha–1 of evenly spread rice residue mulch with no detrimental effect (Humphreys et al., 2004), although 4–6 t ha–1 is considered optimum in Sichuan, China. 4.2.3. Mulching with conventional tillage Rice residue can be used as mulch for the following non-flooded crop established after conventional tillage. Not many farmers follow this option because it involves temporarily removing the residue from the field and then returning it after the crop has been planted. This option is more feasible for farmers with small land holdings and sufficient labor (Tang

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

135

et al., 2004). There are only sporadic reports of rice residue mulching practiced in wheat planted in conventionally tilled fields in the rice–wheat system in South Asia (Zaman and Choudhuri, 1995). 4.2.4. Transfer of biomass as mulch Rice residue can be removed from the field where it is grown and used as mulch for improved production of vegetable crops (Vos et al., 1995), chickpea (Cicer arietinum L.) and mustard (Brassica rapa L.) (Rathore et al., 1999), and crops like bamboo ( Jiang et al., 2002). In China, it is also being used for mulching horticultural crops and tea. The cost of transportation and labor compared with the profits earned by farmers from increased production and saving in inputs like water are important factors in determining the feasibility of this management option. Rice residue can also be removed from the field for a number of useful purposes such as livestock bedding, composting for mushroom cultivation, bedding for vegetables such as cucumber and melon, and conversion to biofuel and biopower.

5. Evaluation of Options with Residues Managed During a Rice Crop We now evaluate residue management options for the category of cropping described in Section 4.1 (Rice following rice or a non-flooded crop) in which residues are managed during a period of rice cropping, usually on puddled and flooded soil. Our evaluation is based on criteria for productivity, profitability, environmental impact, and sustainability as described in Section 2.

5.1. Productivity As reviewed in Section 4.1.1, the incorporation of residue into soil during rice cropping is one of the most studied and promoted alternatives to residue burning across Asia. Much data are therefore available for evaluating its productivity and profitability. When interpreting these data, it is important to remember that large variations among seasons at a given location and among locations can exist in amount of residue incorporated, the time period for residue decomposition before establishment of the rice crop, and the soil aeration status (i.e., primarily aerobic or anaerobic) during residue decomposition. Soil and plant parameters can be affected by whether residues from only one or all crops in a year are incorporated. Mulching of lowland rice with residue of a preceding rice or upland crop is not a very practical option as explained in Section 4.1.2, and consequently it has not been much studied and only limited data are available.

136

Table 4 Effect of rice residue incorporation on grain yield of rice in rice–rice cropping systems in Asia

Experimental details

Kerala, India: Sandy loam soil with pH 5.5, fertilizer N applied at 100 kg ha–1, continuously flooded Same as above but intermittently flooded Indonesia: Acidic soil Acidic soil, fertilizer N at 120 kg ha–1 Acidic soil, no fertilizer N applied Muara, Indonesia: Latosol

Days residue incorporated before transplanting

Amount of rice residue incorporated (t ha–1)

Rice grain yield (t ha–1) Rice residue removed

Rice residue burned

Rice residue incorporated

Season a

4 4 4

3.04 6.15 2.90

– – –

3.18 6.64* 2.10

WS 1972 DSb 1973 WS 1973

28

4 4 4 20

2.75 6.12 2.45 2.78

– – – –

2.92 6.25 2.45 3.06

WS 1972 DS 1973 WS 1973 –

28 14 7 28 14 7 – –

10 10 10 10 10 10 10 20

4.27 4.22 4.00 3.49 3.22 2.84 2.17 2.17

– – – – – –

3.95 3.97 3.81 3.04 3.15 2.95 2.58 2.44

– – – – – – – –

Reference

Vamadevan et al. (1975)

Ismunadji (1978)

137

West Bengal, India: Sandy clay loam soil with pH 7.8, 60 kg N ha–1 in dry season and 40 kg N ha–1 in wet season Los Ban˜os, Philippines: Maahas clay, averaged for 5 cultivars after 16th crop Tropaqualf, clay soil with pH 6.6, 9-year study Tropaqualf, clay soil with pH 6.6, 9-year study Indonesia: Vertic Tropoquept, pH 6.5, no NPKS applied Same as above with NPKS applied

28 35

10 10

3.88 5.38

– –

4.11 6.04

WS DS

Chatterjee et al. (1979)

–

–

3.2

3.4

4.1*

–

A.B Capati, IRRI, cited by Ponnamperuma (1984)

–

–

8.2c

–

8.7c

WS+DS

–

–

8.3c

8.3c

8.7c

WS+DS

6

2.4

–

2.7

–

6

5.7

–

5.2

–

0

Le Cerff et al. (1985)

(continued)

138

Table 4 (continued)

Experimental details

Uttar Pradesh, India: Soil with pH 8.5 South Korea Shao Shing County, Eastern China: 150 kg N ha–1 Punjab, Pakistan: Soil pH 8.0 Hangzhou, Southeast China: Soil with pH 6.2

Rice grain yield (t ha–1)

Days residue incorporated before transplanting

Amount of rice residue incorporated (t ha–1)

Rice residue removed

Rice residue burned

Rice residue incorporated

30

–

2.21

–

2.92*

– –

4.3 5.97 5.97 5.38

– – – –

4.8* 6.15 6.17 6.04

–

35

7.5 3 6 10

20

5

5.6

–

5.9

WS

Zia et al. (1992)

Equal to 600 kg organic C ha–1 Equal to 600 kg organic C ha–1

6.20

–

6.19

Early rice

Lu et al. (2000)

6.20

–

6.13

0

153

Season

Reference

Pandey et al. (1985) Han et al. (1991) Li (1991)

DS

Los Ban˜os, Philippines: Aquandic Epiqualf, silty clay with pH 6.6 Central Java, Indonesia: Acric Tropoqualf, silt loam soil with pH 4.7, rainfed rice Central Thailand: Vertic Tropaquept with pH 5.8, rice residues (C:N¼67:1) contained 21.7 kg N ha1

–

–

5.4 3.0

– –

3.5 3.0

DS WS

Wassmann et al. (2000a)

– –

– –

4.8 4.7

–

5.3 4.6

WS DS

Setyanto et al. (2000)

7

3.75 (þ70 kg N ha1) 3.75 (þ70 kg N ha1) 3.75 (+no N) 3.75 (+no N) 5 (+70 kg N ha–1) 5 (+70 kg N ha–1) 5 (+no N)

4.7

4.2

4.8

DS 1997

Phongpan and Mosier (2003a)

3.8

3.7

3.9

WS 1997

3.7

3.7

4.0

DS 1997

4.0

3.9

4.1

WS 1997

4.4

–

4.1

DS 1998

4.1

–

4.2

WS 1998

3.0

–

3.4

DS 1998

7

7 7

Central Thailand: Ustic Endoaquerts with pH 6.2, rice residues

7 7 7

Phongpan and Mosier (2003b)

139

(continued)

140

Table 4

(continued)

Experimental details

(C:N=67:1) contained 25.4 kg N ha–1 Central Thailand: Ustic Endoaquerts with pH 6.7, rice residues (C:N=67:1) contained 25 kg N ha–1 Andhra Pradesh, India: Sandy clay loam soil with pH 7.5 Anhui, Guangde, China: Loamy clay soil a

Rice grain yield (t ha–1)

Days residue incorporated before transplanting

Amount of rice residue incorporated (t ha–1)

Rice residue removed

Rice residue burned

Rice residue incorporated

Season

7

5 (+no N)

3.7

–

4.3*

WS 1998

7

5 (+70 kg N ha–1) 5 (+70 kg N ha–1) 5 (+no N) 5 (+no N)

6.0

–

5.6

DS 1999

3.9

–

4.3

WS 2000

4.8 3.9

– –

5.0 4.4

DS 1999 WS 2000

4.4 5.7 4.4 5.7 3

6.2 3.1 6.3 3.2 6.2

6.9 3.4 7.5 3.5 –

7.0 3.5 7.3* 3.7* 6.7

DS 1999 WS 1999 DS 2000 WS 2000

7 7 7

14 3 14 3

Wet season. Dry season. c Yield of dry season+wet season rice crops. * Significantly more than grain yield with residue removed or residue burned treatments at P < 0.05. b

Reference

Phongpan and Mosier (2003c)

Surekha et al. (2003) Li et al. (2003)

Table 5 Effect of incorporation of upland (non-flooded) crop residue on grain yield of rice and residual effects on yield of the following upland crop in rice–upland cropping systems in Asia

Experimental details

Uttar Pradesh, India: Soil with pH 8.5, 2-year study West Bengal, India: Silty clay loam acid laterite soil, 2year study Haryana, India: Clay loam soil, 3-year study Uttar Pradesh, India: Clay loam soil with pH 8.6

Days residue incorporated before transplanting of rice

Kind and amount of upland crop residue incorporated (t ha–1)

30

Grain yield (t ha–1) Upland crop residue removed

Upland crop residue burned

Upland crop residue incorporated

Crop

Reference

Wheat, 0

2.21 4.48

– –

2.82 4.59

Rice Wheat (R)a

Pandey et al. (1985)

10

Wheat, 5

3.74 1.80

– –

4.17* 2.0*

Rice Wheat (R)

Sharma and Mitra (1992)

7–10

Wheat, 5.3

6.97 4.65

7.23 4.84

7.01 4.43

Rice Wheat (R)

30

Wheat, 10

4.10

–

4.45

4.10

–

4.08

2.29

–

2.72

Rice (100% NPK)b Rice (50% NPK)c Wheat (R)

Agrawal et al. (1995) Rajput (1995)

141

(continued)

Table 5

(continued)

142 Experimental details

New Delhi, India: Sandy clay loam with pH 8.1 Guizhou, China: Loamy clay, rice–rapeseed Haryana, India: Sandy loam soil, 25% N applied at the residue incorporation, 3-year study Punjab, India: Typic Ustochrept, sandy loam soil with pH 7.9, 2-year study Punjab, India: Typic Ustochrept loamy sand

Grain yield (t ha–1)

Days residue incorporated before transplanting of rice

Kind and amount of upland crop residue incorporated (t ha–1)

42

Wheat, 5.5

30

Wheat, 6.1

7

Rapeseed, 7.5

60

Wheat, 0

6.83 4.01

51–60

Wheat, 6

5.6 4.8

14

Wheat, 6 (90 kg N+13 kg P+13 kg K ha–1)

6.20c 4.48

Upland crop residue removed

Upland crop residue burned

Upland crop residue incorporated

Crop

Reference

3.3 2.4 3.8 3.6 6.7

3.3 2.3 3.9 3.6

3.6* 2.5 4.0* 3.7 5.9

Rice 1992 Wheat (R) Rice 1993 Wheat (R) Rice

Prasad et al. (1999)

6.85 4.04

Rice Wheat (R)

– –

5.5 4.9

Rice Wheat (R)

Aulakh et al. (2001)

–

5.10 4.33

Rice Wheat (R)

Bhandari et al. (2002)

Zhao and Zhu (2000) Dhiman et al. (2000)

with pH 8.2, 14-year study

Shandong, Lingyi, China: Sandy loam soil Anhui, Guangde, China: Loamy clay soil Shanghai, China: Loamy clay soil Punjab, India: Loamy sand soil with pH 7.6, 12-year study Jiangsu, Wuxi, China: Loamy clay with pH 6.8 a

6.20c 4.48

5.72 3.88

Rice Wheat (R)

7.2

7.8*

Rice

Ma et al. (2003)

Wheat, 1.5

6.2

6.6

Rice

Li et al. (2003)

Wheat, 4.4

8.0

8.8

Rice

52–55

Wheat, 6.4 0.5

5.74 4.41

5.37 4.32

Rice Wheat (R)

10

Wheat, 4.2

7.1

7.3

Rice

Yang et al. (2003) YadvinderSingh et al. (2004a) Zhu et al. (2004)

2

Wheat, 3 (105 kg N +20 kg P +19 kg K ha–1) Wheat, 7.5

– –

– –

(R) denotes that the crop was grown to study the residual effect of crop residues applied to the previous crop listed above. 100% NPK is the blanket recommendation of 120 kg N+13 kg P+25 kg K ha–1 for the region. c 120 kg N+26 kg P+25 kg K ha–1 * Significantly more than grain yield with residue removed or residue burned treatments at P < 0.05. b

143

144

Bijay-Singh et al.

5.1.1. Grain yield for rice A summary of 51 data sets is reported in Table 4 from rice–rice cropping system experiments designed to assess the effect of incorporating from 3 to 20 t ha–1 rice residue at 0–153 days before establishment of the following rice crop. Only 7 (~14% of the experiments) showed statistically significant increases in grain yield associated with residue incorporation. Table 5 lists studies in which 1.5–10 t ha–1 of residue from an upland (non-flooded) crop was incorporated 2–60 days before establishing the rice crop in India and China. Only in 4 out of 17 comparisons was rice yield significantly increased by incorporation of residue. The effect of residue incorporation on yield of lowland rice can depend on incorporation method, amount of residue, soil characteristics, and timing and amount of fertilizer application (Ponnamperuma, 1984). It was difficult to identify a similarity in terms of region, amount of residue incorporated, or time of incorporation among the data sets showing significant yield increases. In most cases, the effect of residue incorporation was assessed with application of fertilizers, suggesting that benefits in nutrient supply from the residue could have been masked by application of sufficient fertilizer to overcome nutrient limitations to rice. On the contrary, in studies with no application of fertilizer N (Le Cerff et al., 1985; Phongpan and Mosier, 2003a,b,c; Thuy, 2004) there is frequently no significant increase in grain yield associated with residue incorporation. Thuy et al. (2008) in a 3-year study at two locations in China found no significant increase in rice yield following incorporation of wheat or rice residue with and without fertilizer N. Thuy et al. (2008) concluded that incorporated residue had no benefit on N supply during the vegetative growth phase of rice, but N supply at later rice growth stages especially with long-duration rice could be slightly increased. A combined analysis of all data sets for effect of incorporating rice and upland crop residues on yield of the following rice (Tables 4 and 5) revealed no significant trend of increasing yield due to residue incorporation (Fig. 1). The slope and intercept of linear regressions were not different than 1 and 0, respectively, at P < 0.001. Residue of wheat incorporated into rice also did not have a residual effect on the wheat crop that followed rice (Table 5, Fig. 1). A multicountry coordinated research project on management of crop residue for sustainable production concluded that residue incorporation did not lead to higher yields (IAEA, 2003). In some cases the incorporation of crop residues, especially without application of fertilizer N, can reduce rice yield (Thuy, 2004). This is often attributed to short-term immobilization of N following the incorporation of residue with high C:N ratio (Bird et al., 2001; Buresh et al., 2008). The anaerobic decomposition of added residue and associated intensively reduced soil conditions can lead to production and accumulation of aliphatic aromatic acids that can inhibit rice root growth (Chung, 2001; Tanaka et al., 1990) particularly under low temperatures (Cho and

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

145

A Rice yield (t ha−1) with rice/upland crop residues incorporated

9 With rice residue With upland cropresidue 1:1 line

6

3

0 0

3

6

9

Rice yield (t ha−1) with no residue incorporation

Wheat yield (t ha−1) with wheat residue incorporated in the preceding rice

B 6 Wheat yield 1:1 line 4

2

0 0

2

4

6

Wheat yield (t ha−1) with no residue incorporation in the preceding rice

Figure 1 (A) Relationship between rice yield with and without incorporation of rice or upland crop residue, and (B) wheat yield with and without incorporation of wheat residue into the preceding rice crop. Data are from the literature listed in Tables 4 and 5. The slope and intercept for the linear regressions from A and B were not different than 1 and 0, respectively, at P < 0.001.

Ponnamperuma, 1971), to production of small-molecular-weight organic acids that can have some toxic properties (Rao and Mikkelsen, 1977), and to induced deficiencies of micronutrients especially zinc (Bijay-Singh et al., 1992; Nagarajah et al., 1989). One option for reducing the potential

Table 6 Effect of returning crop residue as mulch to rice fields on grain yield of rice in rice-based cropping systems in Asia

146 Experimental details

Days residue mulched before planting

Grain yield (t ha1) Amount of residue mulched (t ha–1)

Kind of residue

Crop residue mulching in conventionally tilled rice fields 0 Equal to Hangzhou, Rice 600 kg Southeast organic China: Soil C ha1 with pH 6.2, rice straw mulched before transplanting of rice 0 – Rice Guangzhou, China: Grown as upland rice, lowland rice without mulching yielded 6.96 t ha–1 Bhairahawa, – 1.5 Wheat Nepal

Sichuan, China: Fluvaquent grey flood plain soil with pH 7.8

0

–

Wheat

Crop residue removed

Crop residue incorporated

Crop residue mulched

Crop

Reference

6.31

6.44

6.44

Late rice

Lu et al. (2000)

5.98

–

6.59*

Rice

Fan et al. (2002)

3.75

4.2

4.75*

Rice

5.38 (nonflooded rice) 5.74*

Rice

Duxbury and Lauren (2002) Liu et al. (2003)

6.25 (flooded rice) 4.63

Wheat (R)a with 60 kg N ha–1

147

0 Sichuan, China: Loam soil with pH 6.5, nonflooded rice 0 Jiangxi, Yujiang, China: Fluvisol, pH 5.5, organic matter 25.5 g kg–1, rice–rice rotation, nonflooded rice Crop residue mulching in no-till rice Sichuan, Chengdu, China: Sandy loam soil 0 Sichuan, China: Heavy clay soil, permanent bed planting with double zero tillage for rice and wheat Sichuan, Jianyang, China: Loamy clay soil with pH 6.45, no tillage, rice seedling broadcasting

–

Wheat

6.45b 4.62

– –

6.84 5.29*

Rice Wheat (R)

Liu et al. (2005)

5

Rice

4.72c

–

6.75*

Rice

Qin et al. (2006)

7.5

Wheat

6.6

–

6.8

Dryland rice

Ai et al. (2003)

–

Wheat, rice (in ditches)

5.36

–

5.72

Rice on permanent beds

Tang et al. (2004)

5.3

Wheat

8.9

9.3

Rice

Zheng et al. (2005)

(continued)

Table 6 (continued)

Experimental details

Sichuan, Jianyang, China: Loamy clay soil with pH 6.45, rice– rapeseed, no tillage, rice seedling broadcasting a

Days residue mulched before planting

Grain yield (t ha1) Amount of residue mulched (t ha–1)

Kind of residue

Crop residue removed

5.3

Rapeseed

8.8

Crop residue incorporated

(R) denotes that the crop was grown to study the residual effect of crop residues applied to the previous crop listed above. The no mulch flooded treatment yielded 7.2 t ha–1 and was not significantly different than mulched non-flooded yield. –1 The no mulch flooded treatment yielded 6.8 t ha and was not significantly different than mulched non-flooded yield. * Significantly more than grain yield with residue removed treatments at P < 0.05. b c

Crop residue mulched

Crop

Reference

9.4*

Rice

Zheng et al. (2005)

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

149

detrimental effect of decomposing crop residues on rice seedlings is the type of in situ composting in which the residues are next to but not touching the seedlings in ditches parallel to the rows. This practice, however, did not increase yield of rice in a barley–rice rotation (Zhong et al., 2003). A summary of data from 10 mulching experiments across Asia shows a significant positive effect of rice or upland crop residue applied as mulch on rice grain yield in 4 out of 10 experiments (Table 6 ). In Sichuan Province in China, research has examined the mulching of rice with residue of preceding wheat. The mulch is applied after rice transplanting, and then the mulched rice is not kept flooded to save irrigation water. In two studies (Liu et al., 2003, 2005) the mulch did not increase rice yield, but yield of the wheat crop following rice was significantly increased. The partially decomposed mulch from the rice crop, incorporated during land preparation for the next wheat crop, presumably benefited wheat. The application of mulch after transplanting rice seedlings requires removal of residue from the field and then return as mulch. Following this approach, mulching of rice with wheat residue in Nepal significantly increase rice yield compared to plots without mulch (Duxbury and Lauren, 2002). The application of rice residue before transplanting rice, on the other hand, as examined in Southeast China did not increase rice yield (Lu et al., 2000), probably because the mulch to some extent got incorporated into soil during transplanting of rice. Reduced tillage rather than conventional puddling before rice can result in soil conditions less suitable for transplanting of rice. The throwing of rice seedlings and direct seeding of rice into mulch have consequently been examined as alternatives to transplanting in reduced and no-till rice systems. In a study in Sichuan Province in China, rice yield was significantly increased with establishment by seedling throwing into mulch from residue of a preceding rapeseed crop (Zheng et al., 2005). No benefit of mulching on rice yield was observed when rice and wheat residues were applied as mulch in ditches to rice grown on permanent beds in a double no tillage rice–wheat cropping system (Tang et al., 2004). In a wheat–rice system in Korea, Cho et al. (2001) examined the simultaneous harvesting of wheat and sowing of rice with a sowing device mounted on the combine harvester. Rice seeds broadcast onto the untilled soil surface were covered with wheat straw chopped into mulch by the combine harvester, and comparable rice yields were obtained with no-till, direct sowing of rice and conventional till, transplanting of rice. 5.1.2. Fertilizer use efficiency for rice An important indicator of fertilizer use efficiency is the increase in grain yield per unit of nutrient applied as fertilizer, which is usually referred to as agronomic efficiency. An increase in the agronomic efficiency for a given nutrient occurs when the crop response to the nutrient (i.e., the difference in yield between treatments with and without addition of the nutrient) increases

150

Bijay-Singh et al.

per unit of applied nutrient. As indicated in Section 5.1.1, the incorporation of crop residues seldom significantly increases the yield of rice with and without fertilizer application (Tables 4 and 5). In such cases of no increase in grain yield, the incorporation of crop residues would not increase either the agronomic efficiency of a nutrient or the partial factor productivity of the nutrient (i.e., total grain yield with added nutrient per unit of applied nutrient) unless the incorporation of residue was associated with a reduced rate of applied nutrient. Most studies assessing the effect of residues on rice, however, use similar rates of fertilizer in the treatments with and without the residue. Both increases and decreases in agronomic efficiency of fertilizer N (AEN) have been reported for rice when crop residue is incorporated rather than removed with no change in the rate of fertilizer N. Incorporation of rice residue with 70 kg N ha–1 to dry season rice in Central Thailand reduced AEN by 20–50% (Phongpan and Mosier, 2003a,b,c). This was attributed to a reduced yield response of rice to fertilizer N following incorporation of residue, which slightly increased grain yield without fertilizer N but not with fertilizer N. In a 3-year study at two locations in China, the incorporation of rice or wheat residue typically had no effect on AEN (Thuy et al., 2008). In an experiment in the Philippines, the incorporation of rice residue 20 days before transplanting dry-season rice significantly increased AEN irrespective of whether soil was flooded or aerobic for the 2 months from harvest of the previous rice crop to land preparation for rice (Buresh et al., 2007, unpublished data). This increase in AEN was associated with increased response of rice to fertilizer N with incorporation of residue, which arose because the incorporation of residue significantly reduced yield without fertilizer N. Lower yield without fertilizer N was associated with reduced supply of plant-available N due to immobilization of N following residue decomposition (Thuy, 2004). The higher frequency of yield gains with mulching (Table 6) than incorporation of residue (Tables 4 and 5) suggests mulching might be more likely than incorporation to increase fertilizer use efficiency when fertilizer rates are not adjusted for residue management. An approach for increasing AEN while maintaining or even increasing rice yield is to combine the retention of crop residues with improved management of fertilizer N. Xu et al. (2007) found markedly increased fertilizer N use efficiency for rice when the incorporation of residue from a previous wheat crop was combined with timing and rates of fertilizer N that better matched the needs of the rice crop. This approach of improved matching of fertilizer N to crop needs, referred to as site-specific nutrient management (SSNM) (IRRI, 2007), involves adjusting fertilizer N during early vegetative growth to match crop needs as determined by relatively slow early crop growth and immobilization of N from decomposing residue, and it involves adjusting fertilizer N during tillering and at panicle initiation based on the N status of rice leaves. At locations with excessive use of fertilizer N for rice, such as eastern and southern China, an increase in AEN is largely associated with marked reductions in the use of fertilizer N while

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

151

maintaining or slightly increasing yield (Peng et al., 2006a). At locations in which existing fertilizer N rates are near or below the optimal, an increase in grain yield, often with the same or more fertilizer N, is required to increase AEN. When fertilizer N is optimally managed for rice, the incorporation of crop residue typically has negligible or only small savings in fertilizer N (Linquist et al., 2006; Thuy et al., 2008). The vegetative portion of mature rice contains ~80–85% of the total plant K, and most of this K is not lost during open-field burning of residue (Dobermann and Fairhurst, 2000). The management of crop residue consequently has a strong effect on the requirements of rice for fertilizer K (Witt et al., 2007). Recommended rates for fertilizer K should consequently be adjusted based on the supply of K from crop residues, even when burnt, in order to ensure high rice yields with high efficiency of fertilizer K use. 5.1.3. Water use efficiency for rice There have been several reports in China of savings in irrigation water associated with mulching rice with crop residue and then growing rice under non-flooded conditions. Yields are often comparable for mulched, non-flooded rice and conventional flooded rice, but water use efficiency (i.e., grain yield per unit of water used) can be markedly higher for nonflooded mulched rice (Fan et al., 2002). Qin et al. (2006) found that total water use in non-flooded rice with and without rice straw mulch was 3.3 and 2.4 times less than for rice grown under flooded conditions. Yields were comparable for mulched, non-flooded rice (6.7 t ha–1) and conventional flooded rice grown without mulch (6.8 t ha–1). But growth of non-flooded rice without mulch significantly reduced yield (4.7 t ha–1). Mulching rice to save water can be particularly attractive in regions with limited rainfall or irrigation water. 5.1.4. Pest and disease pressure for rice Weed pressure is typically minimal in flooded soil, decreasing the importance of mulch as a weed suppressant in lowland rice. However, mulch might help minimize weed competition in production systems without soil flooding when rice seedlings are small. In non-flooded rice mulched with wheat residue in the rice–wheat cropping system in southwestern China, Liu (2005) recorded total weed biomass of 1.3 t ha–1 in mulched plots as compared to 4.4 t ha–1 in plots without mulch. The corresponding uptake of N by weeds was 25 kg N ha–1 in mulched plots and 55 kg N ha–1 in plots without mulch. Residue incorporation in rice monocropping systems has been shown to aggravate fungal diseases including stem rot (Sclerotium oryzae) and sheath spot (Rhizoctonia oryzae-sativae), historically leading to the recommendation of infield residue burning as the best means of disease control (Miller and Webster, 2001; Webster et al., 1981). Residue incorporation at the beginning of the flooded winter fallow was identified as an alternative means of control for stem

152

Bijay-Singh et al.

rot, because the pathogen decomposed faster in the flooded soil ecosystem than on the aerobic surface (Cintas and Webster, 2001). Relatively little is known about the effects of mulch on disease in flooded systems because of the impracticality of mulching after puddling the soil. For sheath rot, the sclerotia floating to the surface of the floodwater during the cropping season were most readily able to inoculate and infect the plant (Miller and Webster, 2001), implying that using infected residue as mulch would be much worse than incorporating it. There is still insufficient knowledge about how residue management of rice–upland crop rotations affects rice diseases. Mosquitoes are another important pest affected by residue management in cropping systems with flooded soil. They require standing water to complete their life cycle, and their larvae grow better in flooded fields in which rice residue has been incorporated (Lawler and Dritz, 2006). Therefore, from the perspective of disease and mosquito control, it would be better to remove residue than incorporate or mulch it. When weeds are a significant problem, which usually occurs only when soil is not flooded during crop establishment and early crop growth, mulch would be a good option although it might incur increased disease pressure.

5.2. Profitability Profitability considers the potential economic gains or losses resulting from observed changes in productivity. In situ incorporation of crop residue during normal tillage before establishment of rice results in no extra cost for managing crop residue provided the normal tillage does not involve more time or energy due to the presence of residue. If cost of land preparation is not altered by the incorporation of residue, then any increase in production can result in net profit for the farmer. Because of potential short-term detrimental effects of anaerobic residue decomposition on the young rice crop, such as immobilization of N and release of organic acids, the preferred practice is typically to incorporate residue several weeks before establishing rice. In such case, a change in the timing of tillage or land preparation practices to accommodate the incorporation of crop residue could result in extra expenditure. The profitability of residue incorporation then depends on the extra costs for field operations rather than only the effect of residue incorporation on grain yield. Even though much information is available on the effect of residue on rice yield (Fig. 1), corresponding information on any additional costs associated with residue incorporation is typically not available. Dawe et al. (2003) examined the profitability of incorporating rice and wheat residue using data from two rice–rice experiments (in China and Malaysia) and 12 rice–wheat experiments (in India) conducted for 10–17 years and representing a wide variety of soil types, climatic conditions, and crop management practices. In the two long-term experiments on rice–rice cropping systems, 5–6 t ha–1 of rice residue were incorporated before

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

153

each crop. In the other 12 experiments, wheat residue (6–15.8 t ha–1) was incorporated before the rice crop. Recommended levels of fertilizer N, P, and K were applied to all crops. The break-even cost (BEC), defined as the maximum cost that farmers can incur in managing crop residue without losing money, was computed using the average farm-level prices of US$ 0.15 kg–1 for paddy rice and US$ 0.12 kg–1 for wheat and differences in crop yields between NPK and NPK plus crop residue treatments for the annual rotation. Residue incorporation would be profitable when the sum of all additional costs of incorporating residues above the normal operating cost of the farm are less than the BEC. For rice–wheat experiments, the BEC ranged between –23 and 8 US$ ha–1 and averaged –3 US$ ha–1 per crop, suggesting it was usually not profitable to incorporate wheat to rice. In the two rice–rice experiments, the BEC was 45 and 40 US$ ha–1 per crop, indicating the incorporation of rice residues before transplanting of rice was profitable for rice fertilized with NPK provided there were little or no additional costs associated with incorporation of the residue. These positive BEC arose because of slightly higher yields (mean ¼ 0.2–0.6 t ha–1) when residue was incorporated. Reports of increased yield associated with rice residue incorporation are, however, relatively rare (Table 4). Already in the 1970s, Tanaka (1974) observed the economics of residue incorporation did not encourage farmers to regularly adopt the practice even though the incorporation of residue reportedly improved soil conditions for flooded rice. The tendency for more frequent increases in rice yield when residue of the previous crop is mulched (Table 6) rather than incorporated (Tables 4 and 5) suggests mulching could be profitable when the cost of managing residue as mulch is relatively low. The profitably of mulching could also be influenced by other factors such as potential savings in irrigation water. Savings in irrigation cost without loss in rice yield were reported in China when rice was mulched with wheat residue and grown on non-flooded soil (Fan et al., 2005; Liu et al., 2003). Reports of significantly increased yield for wheat following mulched, non-flooded rice (Liu et al., 2005) suggest additional scope for increased profitability of the production system.

5.3. Environmental impact 5.3.1. Air quality Smoke is one of the most serious environmental problems associated with large-scale, open-field burning of crop residues. It pollutes air with a mixture of gases and fine particles, which can lodge deep in our lungs when we breathe. The peak in asthma admissions to hospitals in India coincides with the annual burning of rice residue in surrounding fields (Bijay-Singh and Yadvinder-Singh, 2003). Smoke particles are less than a micron in diameter, which allows them to remain in the atmosphere up to

154

Bijay-Singh et al.

several weeks (Cooke and Wilson, 1996) and therefore spread hundreds to thousands of kilometers before falling back to earth. Large patches of black C aerosol can be detected by satellites over India and China just after the harvest season during large-scale residue burning. As a result of aerosol–radiation–temperature–CO2 interactions, these patches can lead to reduced biomass and grain yield of field crops. Almost any alternative to open-field burning of residue can substantially reduce harmful environmental and health effects of smoke. Streets (2004) predicted that black C emissions in China could be reduced from 75 Gg in 1995 to 56 Gg by 2020 by enforcing laws against residue burning. 5.3.2. Greenhouse gas emissions for rice Another important environmental impact associated with rice residue management is greenhouse gas emissions, including the three main agricultural contributors to climate change: carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O). Atmospheric concentrations of these gases have been increasing in recent decades due to human activity including agriculture, and they have been shown to contribute to increases in average global temperatures (Houghton et al., 2001). When returned to the field, some of the C in crop residue might be retained in the soil organic matter (SOM) with the rest lost as CO2 or CH4. When residue-C is retained as SOM, there is a net benefit for the greenhouse gas balance because some of the CO2 taken in by the plant is no longer in the atmosphere. When residue-C is lost as CO2, the greenhouse gas balance is neutral because the C came in and out of the plant as CO2. However, when residue-C is lost as CH4, there is a strongly unfavorable consequence for the greenhouse gas balance because each molecule of CH4 has 62 times greater global warming potential than a molecule of CO2 (20-year horizon, Houghton et al., 2001). When soil is anaerobic, the end product of microbial decomposition shifts toward CH4 instead of CO2. Hence, flooded rice systems with retained crop residues are sources of CH4 emission. N2O, which has an even higher global warming potential than CH4 (275 times that of CO2 on a 20-year horizon, Houghton et al., 2001), is formed during aerobic nitrification of ammonium and anaerobic denitrification of nitrate. Under extended periods of soil flooding—as is typical during rice cropping—the predominantly anaerobic soil conditions are not favorable for rapid nitrification– denitrification and emission of N2O (Buresh et al., 2008). The emission of N2O can, however, be important during periods of alternate soil drying and wetting and when soil following a prolonged aerobic period is flooded, such as during land preparation for rice cultivation. The effects of major residue management options on CH4 and N2O emissions from lowland rice are summarized in Table 7. CH4 emission from rice paddies has been measured many times, testing effects of diverse

Table 7 Trends in CH4 and N2O emitted under different residue management options in lowland rice CH4

Management decision

Major options

Best

Crop residue use

Return Removal

Water Continuous management flooding (including preseason fallow or not) Mid-season drainage (one or more drainage periods, intermittent flooding) Mostly aerobic How residue is returned

Incorporation Surface mulch

Timing of residue return

Beginning of preseason fallow Just before rice establishment

N2O

a

Worst

Best

Removal

Return

No clear effect

Aerobic

Continuous flooding

Continuous flooding

No effect Several months before flooding

Close to flooding

Worst

No effectb

Yan et al. (2005) Intermittent flooding

?

References

Bronson et al. (1997b), Yan et al. (2005), Li et al. (2006)

Ishibashi et al. (2005) Yan et al. (2005), Li et al. (2006) (continued)

Table 7

(continued)

Management decision

Type of residue

Overall combination

a

CH4 Major options

Best

High C:N (rice straw) Low C:N (legume residue) Fresh plant material Partially decomposed plant material (compost, manure)

a

N2O Worst

Best

Worst

References

Partially decomposed plant material

Fresh plant material

High C:N

Low C:N

Yan et al. (2005), Zou et al. (2005)

Removal, minimum flooding

Fresh residue applied incorporated just before flooding

Continuous flooding

Mulch with low C:N residue, intermittent flooding

For both gases, ‘‘best’’ means the management practice that results in the lowest gas emission. Lack of timing effect on N2O is based on a model rather than actual data (Li et al., 2006). Note: Methane is in bold because it represents the greater global warming threat in systems that are predominantly flooded b

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

157

variables including residue management, water management, temperature, and soil physical and biochemical properties. In a review of ~1000 seasonal measurements from sites across dominant agroecological zones in Asia, Yan et al. (2005) found the two most important factors controlling CH4 emission were presence or absence of residue followed by water management. Residue return by any method or timing of application or incorporation caused a statistically significant (P < 0.0001) and often very large increase in CH4 emission compared with residue removal. Residue provides a source of easily decomposable organic C, which means (1) anaerobic bacterial populations increase, using up oxygen (O2) followed by other reducible soil components and driving the redox potential down between the transition where CH4 rather than CO2 is end product of decomposition, and (2) methanogenic bacteria have sufficient substrate-C to form CH4 (Yagi and Minami, 1990). The more decomposed the residue before flooding, the less CH4 emitted. The decomposition of residue before soil flooding for rice production can be accomplished by (1) incorporating crop residue soon after harvesting a crop and allowing it to decompose aerobically before soil flooding for the next rice crop (Wassmann et al., 2000a,b,c; Yan et al., 2005), (2) composting the residue off-field (Corton et al., 2000; Yagi and Minami, 1990), or (3) feeding the residue to cattle and returning it as manure (Setyanto et al., 2000; Wang et al., 2000). Although these options were not all directly compared in one study, each of them resulted in lower CH4 emission as compared with the application of fresh straw just before flooding. Because methanogenic bacteria are anaerobes, CH4 formation is minimal under aerobic conditions, meaning that aerobic water management practices mitigate CH4 emission. In many of the experiments cited by Yan et al. (2005), brief mid-season drainage periods as compared to continuous soil flooding significantly reduced total seasonal CH4 emission. The emissions of CH4 from mulched relative to incorporated residue can be strongly influenced by tillage and water management (Hanaki et al., 2002; Harada et al., 2007; Ishibashi et al., 2001, 2005; Xu et al., 2004). When the surface of reduced or no tillage soil was drier than puddled soil due to more rapid percolation, much less CH4 was emitted from the mulched than incorporated residue. But when soil water content was similar between residue management practices, the CH4 emission was also comparable between practices (Ishibashi et al., 2001). We conclude that water management differences have a larger effect than residue placement on CH4 emission, and there is insufficient information to differentiate between incorporated and mulched residue per se (Table 7). In lowland rice cropping systems with continuous soil flooding, N2O emission is not a significant risk regardless of residue management. However, any time the soil is not flooded—during fallow, mid-season drainage, harvest drainage, or water shortage—N2O formation and emission become likely (Table 7). The water management strategies that mitigate CH4 emission simultaneously exacerbate N2O emission. While many have

158

Bijay-Singh et al.

reported significant increases in N2O emission following fertilizer N application in partially flooded systems (Xu et al., 2004), there is not a consistent trend in the effect of residue removal or return. Of the available direct comparisons in field or pot experiments, most are not significant and those that are significant are in both directions with returned residue sometimes showing higher and sometimes lower N2O than with residue removed, occasionally even within the same data set (Bronson et al., 1997b; Lou et al., 2007; Ma et al., 2007; Zheng et al., 2000; Zou et al., 2004a, 2005). In the comparison between residue incorporation and surface mulch, N2O emission is often affected more by the soil water regime than the management of residue per se (Harada et al., 2007). One model predicted, without measured data, no effect of the timing of residue return on N2O emission (Li et al., 2006). After water and fertilizer N management, the most important factor determining N2O emission was the type of residue. Incorporation of a high C:N ratio residue like wheat straw decreased N2O emission, presumably through N immobilization, while a low C:N residue like rapeseed cake increased it (Zou et al., 2005). From the perspective of minimizing greenhouse gas emissions from lowland rice, incorporation of fresh residue would be the worst option for CH4 emission, especially where soil is flooded continuously for the month following incorporation (Table 7). However, regardless of residue management, it is important to keep the soil flooded as much as possible to minimize N2O emission. Therefore, a compromise could be to keep the soil flooded most of the time to minimize N2O emission and to remove residue to minimize CH4 emission. Prolonged aerobic decomposition of residue before rice cropping might not be feasible in intensively cropped rice-based cropping systems, and prolonged fallows with aerobic soil can favor formation of nitrate that is subsequently denitrified with formation of N2O upon soil flooding for rice cultivation (Buresh et al., 2008).

5.4. Sustainability 5.4.1. Yield trends with flooded residue management In 14 long-term experiments (2 on rice–rice and 12 on rice–wheat cropping system) explained in the Section 5.2, Dawe et al. (2003) observed that the value of the F-statistic for testing the null hypothesis of identical yield trends in NPK and NPK plus crop residue treatments was never significant at the 5% level, indicating no statistically distinguishable differences in yield trends between the two treatments. The yield trend in the residue treatment was more positive or less negative than the trend in the NPK treatment for 11 of the 16 cases of rice cropping, which included both rice crops in the two rice–rice systems. These differences in yield trends were not statistically significant at the 5% level. Across all sites, there was no consistent yield

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

159

increase from application of straw. For the rice crops in the rice–rice cropping systems, the average rice yield increase during the 11–12-year duration of the experiments due to crop residue application was 0.43 t ha–1 per crop. For the 12 rice crops in the rice–wheat cropping systems, rice yields during the 11–17-year duration of the experiments were reduced in the crop residue treatments by an average 0.40 t ha–1 per crop. 5.4.2. Soil changes with flooded residue management The long-term incorporation of crop residues in flooded rice soil can increase SOM, total soil N, fractions of soil C, and soil biological activity; but it can decrease the availability of zinc (Yadvinder-Singh et al., 2005). Continuous incorporation of crop residues after each crop can eventually increase the N-supplying capacity of rice soils (Eagle et al., 2000; Verma and Bhagat, 1992). Long-term studies indicate the supply of plant-available soil N can increase after 5–10 years of continuous incorporation of crop residues in tropical (Cassman et al., 1996) and temperate area (Bird et al., 2001). The benefits of incorporated residues on SOM and soil N supply, however, seldom translate into increased yield (Section 5.1) or profit (Section 5.2) for flooded rice. The effect of crop residues on properties of flooded soil has already been extensively reviewed by Yadvinder-Singh et al. (2005), and it is consequently not covered in this chapter. A noteworthy feature of flooded rice soils with continuous and intensive rice cropping is the maintenance and even buildup of SOM (Bronson et al., 1997a; Cassman et al., 1995; Witt et al., 2000). Appreciable inputs of N from biological N2 fixation (BNF) in continuous flooded rice production systems contribute to the maintenance of soil N even in the absence of N fertilization (Ladha et al., 2000). Prolonged soil submergence and anaerobic soil conditions can lead to the buildup of phenolic compounds that can immobilize N abiotically, thereby reducing net N mineralization and supply of plant-available soil N (Olk et al., 1996, 2000). However, long-term experiments with continuous cropping of flooded rice in the Philippines reveal no decline in rice yield during the past 10–20 years in zero–N plots receiving ample supplies of other nutrients (Padilla, 2001). Yield of flooded rice without fertilizer N, which presumably reflects the supply of plant-available soil N, was maintained even when all above-ground crop residues were removed for each crop. The results suggest that the inputs of N during continuous soil flooding via BNF and biological N mineralization matched or exceeded any decline in N-supplying capacity arising from abiotic immobilization of N associated with buildup of phenolic compounds. Long-term experiments in the tropics indicate the incorporation of crop residue is not essential for maintenance of SOM and soil N-supplying capacity in continuous rice cultivation on puddled and flooded soil. Pampolino et al. (2008b) examined trends in total soil C and N during 15 years of

160

Bijay-Singh et al.

continuous cultivation with two or three flooded rice crops per year in three long-term experiments with incorporation of crop residues and one long-term experiment with removal of all above-ground biomass after each crop. Soil C was maintained during the 15 years in each experiment. In the experiment with removal of all above-ground crop residues and four rates of fertilizer N, soil C increased by 1.5–2.3 g kg–1 and soil N increased by 0.09– 0.15 g kg–1 during the 15 years. Soil N-supplying capacity as determined by anaerobic N mineralization was statistically similar at the start and end of the 15-year period regardless of fertilizer N management. The input of N via BNF as estimated from N balances in a treatment without fertilizer N averaged 81 kg N ha–1 year–1 during the 15 years. Whereas SOM can be maintained in flooded rice–rice systems regardless of residue management, SOM significantly decreased when one flooded rice crop was replaced by conventional till maize with retention of rice and maize residues (Witt et al., 2000; Pampolino et al., 2008a). On the basis of above research findings, we conclude residues from the crop preceding rice on puddled and flooded soil can be considered for removal for off-field uses, without loss in productivity or sustainability of the flooded rice provided fertilizer is appropriate increased to compensate for nutrient removal in the residue. The management of K is particularly important when crop residues are removed because crop residues can markedly increase K availability in soil and decrease the crop response to K application (Chatterjee and Mondal, 1996; Ning and Hu, 1990; Patil et al., 1993; Sarkar et al., 1989).

6. Evaluation of Options with Residues Managed During a Non-Flooded Crop In this section, we evaluate residue management options for the category of cropping system described in Section 4.2(non-flooded crop following rice) in which residues are managed during the period of a crop grown on aerobic soil without flooding. Our evaluation is based on criteria for productivity, profitability, environmental impact, and sustainability as described in Section 2.

6.1. Productivity The two main options available for in-field management of crop residue during the non-flooded (upland) crop in rice-based cropping systems are incorporation into the soil and leaving the residue on the soil surface as mulch. Incorporation typically involves conventional tillage of soil, whereas mulching usually involves reduced or no tillage. While these options are comparable to those available for flooded rice (Section 5), their feasibility and

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

161

implementation markedly differ between a non-flooded crop and flooded rice because of differences in land preparation and tillage options between crops. Mulching of crop residue is much more feasible during a non-flooded crop than flooded rice because of greater opportunities for reduced and no tillage. Most of the information on managing rice residue for non-flooded crops in rice-based cropping systems comes from rice–wheat systems. Although mulching rice residue in conventionally tilled wheat has been attempted by some researchers, mulching in no-till wheat is favored and now facilitated by recent developments of appropriate machinery. 6.1.1. Grain yield for non-flooded crop The incorporation of rice residue before wheat or rapeseed significantly increased yield of the nonrice crop in only 1 of 16 data sets examined from China and India in which 3–7.9 t ha–1 of rice residue was incorporated into soil 10–40 days before sowing of wheat or rapeseed (Table 8). A combined analysis of all data sets revealed no trend of greater yield when rice residue was incorporated rather than removed (Fig. 2). The slope and intercept for the linear regression comparing yield with and without residue incorporation were not significantly different from 1 and 0, respectively, at P < 0.001. Combined data for on-farm experiments with rice–wheat cropping in northwestern India similarly reveal no trend of greater yield of wheat when rice residue was incorporated rather than burnt or removed (Fig. 2). Some of the studies reported for rice–wheat systems in Table 8 investigated the residual effect of rice residue incorporated to wheat on the yield of the next rice crop after wheat, but on all seven cases the incorporated residue had not significant residual effect on yield of the following rice. Table 9 summarizes results for rice–wheat systems in which residue from each crop was incorporated before the following crop, resulting in large amounts of incorporated residue. Wheat yield was significantly increased by incorporated residue in only 1 out of 13 cases. When all data for rice and wheat were combined, no effect of residue on grain yield was detected (Fig. 3). The incorporation of crop residue can have adverse effects on the following crop (Cannell and Lynch, 1984), although in some studies the negative effects of residue incorporation in a rice–wheat cropping system diminished after a few initial years (Dhiman et al., 2000). But in other studies the negative effects were not reversed even after 11 years (Beri et al., 1995). The negative effects on wheat yield can result from immobilization of N by the decomposing residue. This is supported by the observation of Beri et al. (1995) of greater decline in wheat yield at a low rate of N application (0.5 t ha–1 decline at 60 kg N ha–1) than at a high rate of N application (0.08 t ha–1 decline at 180 kg N ha–1). The magnitude of N immobilization depends on the extent of straw decomposition before N fertilization (Bhogal et al., 1997). The immobilization of N is temporary, and it can be followed later in the cropping season by release of N through mineralization. In such case the

162 Table 8 in Asia

Effect of rice residue incorporation on grain yield of upland crop and residual effects on the following rice in rice–upland crop systems

Days residue incorporated before planting

Amount of rice residue incorporated (t ha–1)

Himachal Pradesh, India: Acidic clay loam soil, chopped rice straw incorporated Himachal Pradesh, India: Silty clay loam soil with pH 5.9, chopped rice straw incorporated up to 20 cm soil depth

28

30

Punjab, Pakistan Haryana, India: Sandy loam soil, 25% N applied at the time of residue incorporation

Experimental details

Grain yield (t ha1) Rice residue removed

Rice residue burned

Rice residue incorporated

Crop

Reference

–

2.76 2.37

– –

2.79 2.47

Wheat Rice (R)a

Sharma et al. (1985, 1987)

5

2.6

2.6

2.1

Verma and Bhagat (1992)

5

3.7 2.2

3.6 2.2

3.8 2.2

–

–

3.8 2.91

3.7 –

3.9 3.51*

Wheat (first 3 crops) Rice (R) Wheat (next 2 crops) Rice (R) Wheat

30

–

4.01 6.83

– –

3.72 7.11

Wheat Rice (R)

Salim (1995) Dhiman et al. (2000)

Punjab, India: Typic Ustipsamment, loamy sand with pH 7.3 Shanghai, China: Loamy clay soil Anhui, Guangde: China: Loamy clay soil Punjab, India: Sandy loam soil with pH7.2, 7-year study

Punjab, India: Sandy loam—silt loam, on-farm experiments

20 40

6.4

5.06 5.06

7.5

5.1

5.00 4.89

Wheat

4.7

5.1

Wheat

3

1.7

1.9

Rapeseed

40

7.1–7.9

20

7.1–7.9

10

7.1–7.9

21

5.0–7.0

4.94 6.19 4.94 6.19 4.94 6.19 4.3 4.5 4.5 4.6 4.4

5.17 6.34 5.22 6.29 4.95 6.33 4.5 5.1 5.0 4.3 3.7

Wheat Rice (R) Wheat Rice (R) Wheat Rice (R) Wheat Wheat Wheat Wheat Wheat

– – – – – – – – – – –

a (R) denotes that the crop was grown to study the residual effect of crop residues applied to the previous crop listed above. * Significantly more than grain yield with residue removed or residue burned treatments at P < 0.05.

BijaySingh et al. (2001) Yang et al. (2003) Li et al. (2003) YadvinderSingh et al. (2004b) Sidhu et al. (2007)

163

164

Bijay-Singh et al.

Upland crop yield (t ha−1) with rice residue incorporated

6 Upland crop yield 1:1 line

5

4

3

2

1

0 0

1

2

3

4

5

6

Upland crop yield (t ha−1) with rice residue removed or burned

Figure 2 Relationship between yield of a non-flooded crop (wheat in most cases, and oilseed rape) with and without rice residue incorporation. Data are from the literature listed in Table 8 combined with unpublished data from on-farm experiments with recommended NPK levels in northwestern India (P. R. Gajri, Department of Soils, Punjab Agricultural University). The slope and intercept for the linear regression were not different than 1 and 0, respectively, at P <0.001.

incorporation of residue alters the supply of N to the crop, and the optimal distribution of fertilizer N during the growing season to synchronize N supply with N need by the crop can differ when crop residue is incorporated rather than burnt or removed (Thuy et al., 2008). Mulching with crop residues has long been known to improve crop productivity. Farmers in Asia, especially in China, have long practiced the mulching of high-value crops like grapes (Deng and Deng, 2005), tobacco (Liu, 2005), tea (Peng et al., 2005, 2006b; Xiao et al., 2006), and potato (Liu et al., 2006) with rice residue transferred from a nearby field. With increasing availability of rice residue and machinery in recent times, mulching in lower-value field crops is becoming a viable option. Although mulching with rice residue for upland crops established using conventional soil tillage can be labor intensive, mulch can increase crop yields. Mulching of conventionally tilled wheat with rice residue significantly increased wheat yield in all four data sets from China and India reported in Table 10. Reduced and no tillage for sowing upland crops like wheat is gaining increasing acceptance with farmers in Asia because of reduced land preparation costs. An examination of 41 data sets from China and India with reduced or no-till wheat, barley, or rapeseed revealed that

Table 9 Effect of incorporating residues of both rice and wheat in the following crops on the productivity of wheat and rice in the rice–wheat cropping systems in Asia

Experimental details

New Delhi, India: Sandy loam soil, 120 kg N +26 kg P ha–1 applied to residue removed plot, 60 kg N+13 kg P ha–1 applied to residue amended plots Punjab, India: Sandy loam soil, 11-year study Bihar, India: Sandy loam soil with pH 8.5, 100 kg N+26 kg P ha–1 applied Uttar Pradesh, India: Hapludult sandy loamy soil with pH 8.0, 2-year study Punjab, India: Loamy sand soil, 3-year study

Days residue incorporated before planting

Amount of crop residue incorporated (t ha–1)

Grain yield (t ha–1) Residue removed

Residue burned

Residue incorporated

Crop a

Reference

3 14 10 21

5 5 5 5

2.53 4.34 2.53 4.34

– – – –

1.91 3.64 2.37 3.59

Rice Wheata Ricea Wheata

Kavimandan et al. (1987)

15 0

4 4

5.50 4.14

5.65 4.26

4.63 3.97

Rice Wheat

Beri et al. (1995)

30 20

– –

2.95 3.20

– –

3.20 3.43

Riceb Wheatb

Prasad and Sinha (1995)

3.8–3.9 3.7–8.0

4.68 3.50

– –

4.80 3.75

Rice Wheat

Sarkar (1997)

– –

3.19 4.25

3.66 4.41

3.40 4.01

Rice Wheat

Brar et al. (1998)

2 2

30 20

(continued)

Table 9 (continued)

Experimental details

New Delhi, India: Sandy clay loam with pH 8.1

Haryana, India: Sandy loam soil, 25% N applied at the residue incorporation Punjab, India: Typic Ustochrept, sandy loam soil with pH 7.9, 2-year study Bangladesh: Loam with pH 6.1, 3-year study Zhejiang, China: Heavy loam with pH 6.9, 3-year study Jiangsu, Suzhou, China: Loamy clay soil; all straw incorporation, 4-year study a

Days residue incorporated before planting

Amount of crop residue incorporated (t ha–1)

Grain yield (t ha–1) Residue removed

Residue burned

Residue incorporated

10

6

4.1

4.3

3.7

11

6 6

3.1 3.6

3.3 3.7

3.5 4.1*

60 30

6 – –

3.3 6.83 4.01

3.3 6.79 4.06

21–28

6 6

4.8 5.6

–

5 5

–

–

Crop

Reference

Prasad et al. (1999)

3.7 7.36* 3.66

Wheat 1992–93 Rice Wheat 1993–94 Rice Rice Wheat

Dhiman et al. (2000)

– –

5.0 5.6

Wheat Rice

Aulakh et al. (2001)

4.7 2.5

– –

4.9 2.6

Wheat Rice

– –

3.2 5.5

– –

3.5 5.7

Wheat Rice

Rahman et al. (2003) Wang et al. (2003a)

all all

7.4 4.1

7.8* 4.2

Rice Wheat

Wang et al. (2003b)

Rice and wheat residues were inoculated with algae and Azotobacter, respectively. Rice and wheat residues were both chopped, then soaked with 2% urea solution and inoculated with Aspergillus sp. * Significantly more than grain yield with residue removed or residue burned treatments at P < 0.05. b

Yield of rice or wheat (t ha−1) with residues of both rice and wheat incorporated

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

167

9 Rice yield Wheat yield 1:1 line 6

3

0 0 3 6 9 Yield of rice or wheat (t ha−1) with crop residues removed

Figure 3 Effect of incorporating, rather than removing, residues of both rice and wheat in a rice–wheat rotation on grain yield of the subsequent crop of rice or wheat. Data are from the literature listed in Table 9.

mulching rice residue often increased yield (Table 10). Wheat yields increased by up to 1.9 t ha–1 when 3.7–8.6 t ha–1 of rice residue were retained as mulch. In 11 of the 41 data sets, the yield benefits were more than 0.8 t ha–1. A combined analysis of data sets revealed that the slope and intercept for the linear regression comparing yield with and without mulching were significantly different from 1 and 0, respectively (Fig. 4), suggesting that mulching upland crops with rice residue leads to increased productivity. The mulching of rice grown without soil puddling on aerobic, nonflooded soil can sometimes increase yield. For example, the mulching of upland rice with rice residue significantly increased yield in Guangzhou Province of China (Fan et al., 2002), and yield of mulched upland rice was as high as yield of nonmulched flooded rice. However, wheat mulch had no effect on reduced till upland rice grown in Sichuan Province (Ai et al., 2003).The effect of mulch on crop yield depends on the extent of its influence on biophysical conditions that constrain crop growth (Erenstein, 2002). The effects of mulching can be site-, season-, and cropping system-specific. 6.1.2. Fertilizer use efficiency for non-flooded crop Yadvinder-Singh et al. (2004b) studied the in situ decomposition and N release dynamics for incorporated rice residue using the litterbag decomposition technique as described by Beare et al. (2002). Total rice residue input

Table 10

Effect of rice residue mulch in upland (non-flooded) crops on grain yield of the upland crops in rice-based cropping systems in Asia

168 Experimental details

Days residue mulched before planting

Grain yield (t ha–1) Amount of rice residue mulched (t ha–1)

Kind of residue

Rice residue mulching in conventionally tilled wheat fields Himachal Pradesh, 0 5 Rice India: Silty clay loam soil with pH 0 5 Rice 5.9, rice straw mulch applied after sowing of wheat West Bengal, India: – 5 Rice Sandy loam soil with pH 5.7, rice straw mulch applied when wheat crop attained a height of 5 cm Hubei, Jingmen, After 3 Rice China: Loamy clay emergence with pH 5.5 Rice residue mulching in reduced or no-till upland crops Sichuan, Guanghan, Immediately 3.75 Rice China: Clay soil, after wheat rice–wheat seeding rotation

Rice residue removed

Rice residue incorporated

Rice residue mulched

2.6 3.7

2.1 3.8

3.1* 3.6

2.2 3.8

2.2 3.9

1.42

Crop

Reference

Verma and Bhagat (1992)

3.5* 3.8

Wheat Rice (R)a Wheat Rice (R)

–

1.74*

Wheat

Zaman and Choudhuri (1995)

2.9

–

3.6*

Wheat

Hu et al. (2004)

6.2

–

6.6

Wheat

Tang and Huang (2003)

Zhejiang, Jiaxing, China: Loamy clay soil, rice–barley Sichuan, China: Heavy clay soil, permanent-bedplanting with doubling zero tillage for rice and wheat, residues of rice and wheat applied as mulch in ditches Dinajpur, Bangladesh: Noncalcareous acidic (pH 5.3) alluvial loam soil; 2000– 2001 Dinajpur, Bangladesh: Noncalcareous acidic (pH 5.3) alluvial loam soil; 2001–2002

–

all

Rice

3.1

–

3.6*

Barley

Zhong et al. (2003)

0

–

Rice

7.23

–

7.64

Wheat

Tang et al. (2004)

0

4+no N 4+80 kg N ha–1 4+120 kg N ha–1 4+160 kg N ha–1

Rice

0.58 1.93 2.36 3.06

– – – –

0.79 2.92* 3.93* 3.98*

Wheat

Rahman et al. (2005)

0

4+no N 4+80 kg N ha–1 4+120 kg N ha–1 4+160 kg N ha–1

Rice

0.31 1.52 1.96 2.54

– – – –

0.69* 2.21* 3.92* 3.73*

Wheat

(continued) 169

Table 10

(continued)

170 Experimental details

Sichuan, Jiangyang, China: Arid and hilly area, rice– wheat and rice– rapeseed rotations Punjab, India: 2003–2004 Loamy sand Sandy loam Loamy sand Punjab, India: 2004–2005 Loamy sand Sandy loam Loamy sand Punjab, India: Onfarm locations, 2003–2004, Silty loam Punjab, India: Onfarm locations, 2004–2005

Grain yield (t ha–1)

Days residue mulched before planting

Amount of rice residue mulched (t ha–1)

Kind of Rice residue residue removed

Rice residue incorporated

Rice residue mulched

Crop

0 0

7.5 7.5

Rice Rice

4.80 2.74

– –

5.82 3.39

Wheat Zhu et al. Rapeseed (2005)

0

5.3

Rice

4.74 0.44b

–

4.83 0.52

Wheat

6.0 5.7 6.0

4.17 0.42 3.87 0.15 4.73 0.44

– – –

4.31 0.31 3.95 0.26 4.69 0.27

Wheat Wheat Wheat

7.3 5.5 7.8

4.65 0.26 3.90 0.39 4.6

– – –

5.11 0.55 4.22 0.84 5.1

Wheat Wheat Wheat

7.1

3.6

–

4.0

Wheat

Reference

Sidhu et al. (2007)

Silty loam Loam Clay loam Loam Punjab, India: On-farm locations, 2005–2006 Clay loam Clay loam Loam Loam Silty loam Silty loam Silty loam Clay loam Sandy loam Silty loam Sandy loam Loam Punjab, India: Sandy loam—silt loam, on-farm experiments, 2005–2006 a

9.0 7.5 4.4 8.0

5.0 4.9 3.9 5.2

– – – –

5.4 5.2 4.4 5.2

Wheat Wheat Wheat Wheat

7.5 7.4 8.0 8.5 8.1 7.3 7.9 8.6 8.4 6.9 8.2 5.0–7.0

4.7 4.4 4.6 4.7 4.9 4.5 4.9 4.7 4.7 4.4 4.4 4.3 4.5 4.5 4.6 4.4

– – – – – – – – – – – – – – – –

5.1 4.5 5.2 5.8 5.5 5.3 5.2 4.7 4.5 5.4 5.6 5.2 4.9 5.0 5.1 5.3

Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat Wheat

(R) denotes that the crop was grown to study the residual effect of crop residues applied to the previous crop listed above. Means +/– standard deviation. * Significantly more than grain yield with residue removed treatments at P < 0.05. b

171

172

Yield of upland crop (t ha−1) with rice straw mulch

Bijay-Singh et al.

8 y = 0.8853x + 1.0106 R2 = 0.8852

6

4

2

Upland crop yield 1:1 line Linear trendline

0 0

2 4 6 8 Yield of upland crop (t ha−1) with no rice straw mulch

Figure 4 Effect of mulching rice residue rather than removing it on grain yield of the subsequent upland crop of wheat, rapeseed, or barley. Data are from the literature listed in Table 10.

was 7.1 t ha–1, and it contained 40 kg N ha–1 at the time of incorporation. The total amount of N released under different decomposition treatments during the life span of the wheat crop (~150 days) ranged from 6 to 9 kg N ha–1. Yoneyama and Yoshida (1977) found only 8% of the N in the rice leaf sheath containing 8 g N kg–1 was mineralized in 30 days at 30 C under upland conditions. With such small amounts of N released from incorporated residue, a benefit of significant savings in fertilizer N is unlikely. Moreover, crops in the majority of experiments with incorporated residue (Tables 8 and 9) receive recommended or adequate amounts of fertilizer N, P, and K, suggesting the rare increases in yield of crops with incorporated rice residue were most likely not related to the contribution of macronutrients contained in the residue. Although mulch is also unlikely to supply much N to the recipient crop, it might protect against fertilizer loss, especially the volatilization of ammonia from fertilizer N (Bhagat and Verma, 1991). A reduction in fertilizer N loss can lead to recovery by the crop of a larger fraction of the applied fertilizer N. Rahman et al. (2005) observed a significantly larger apparent recovery of fertilizer N under mulch than nonmulch conditions. A reduction in fertilizer N loss and concomitant increase in crop uptake of fertilizer N can lead to increased AEN when mulching is associated with either a reduced rate of fertilizer N or an increase in grain yield, which exceeds any yield gain arising with mulching in the absence of fertilizer N.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

173

6.1.3. Water use efficiency for non-flooded crop Rice residue mulching in non-flooded crops has a significant effect on soil water conservation in reduced and no tillage systems. Experiments conducted by Sidhu et al. (2007) suggest that mulching might reduce the irrigation requirement of wheat during the growing season due to reduced soil evaporation, and some anecdotal reports indicate a saving of one irrigation (OFWM, 2002). Savings in irrigation water in mulched upland crops is attributed to suppression of soil evaporation (Lal, 1989). In a rice– wheat system, Rahman et al. (2005) also showed much less drying of the topsoil (0–15 cm) with rice straw mulch (4 t ha–1) on a loam soil in Bangladesh. Straw mulch enhanced root growth as indicated by higher values of root length density and root weight density under mulching as compared to non-mulched plots. There have been many other reports that crop residue mulch in reduced and no-till soil retained soil water in the surface layer (Dalrymple et al., 1992; Manakul, 1994), resulting in better crop performance. In the Indo-Gangetic Plain in northwestern India, Sidhu et al. (2007) observed a consistent trend for higher gravimetric soil water content in the surface soil under wheat with mulch as compared with no mulch on loamy sand and sandy loam soils. However, a significant texture depth mulch interaction indicated that the effectiveness of mulch on water saving may be influenced by soil texture. 6.1.4. Pest and disease pressure for non-flooded crop Crop residue mulch has the potential to control weed growth (Erenstein, 2002), thereby suppressing the possible negative effect of weed invasion in reduced and no tillage systems (Fisher et al., 2002; Gill and Arshad, 1995; Rahman et al., 2002). Mulch controls weed growth by shading or through allelopathic effects (Erenstein, 2002; Kamara et al., 2000), and it might reduce herbicide requirements and weed competition for nutrients and water. Increasing the amount of rice residue as mulch in wheat can increase the suppression of weeds. At Karnal in northwestern India, weed biomass decreased progressively with increasing mulch: 188 g m–2 for 2 t ha–1, 139 g m–2 for 4 t ha–1, 99 g m–2 for 6 t ha–1, and 82 g m–2 for 8 t ha–1 (R. K. Sharma, Directorate of Wheat Research, Karnal, India unpublished data). Many others have also reported significant and sometimes very large reductions in weed biomass with mulch (Bilalis et al., 2003; Rahman et al., 2005; Sidhu et al., 2007). Mulching in rice–wheat system can also protect the seeds from birds, particularly under no-till sowing (Cho et al., 2001). However, using rice residue to mulch a horticultural crop in the Philippines led to increased insect pest infestation (Aquino and Mabesa, 2002). Retaining residue as mulch in no-till upland cropping systems can exacerbate residue-borne diseases, such that crop rotation becomes the

174

Bijay-Singh et al.

best nonburning option for control (Bockus and Shroyer, 1998). Periodic residue incorporation in a no-till system can inhibit pathogen growth by forcing it into a place with insufficient air and light, but residue incorporation is not considered as effective as crop rotation at controlling disease (Bailey and Duczek, 1996). In addition to residue-borne diseases, mulch can also favor the survival of some soil-borne pathogens because the pathogens are protected from microbial degradation by residence within the crop debris. Their growth is aided by the lower soil temperature, increased soil water, and reduced soil disturbance associated with mulch (Rothrock, 1992; Sturtz et al., 1997). On the contrary, mulch may suppress other soil-borne pathogens, because it increases the population of soil micro- and mesofauna (Kumar and Goh, 2000), which offer potential for biological disease control as many of these species feed on pathogenic fungi (Curl, 1988). In one of the few studies of residue effects on disease in rice-based cropping systems, wheat planted into no-till rice residue had the lowest incidence of Tilletia indica (Karnal bunt) infection (Sharma et al., 2007). Decisions on crop residue management should be made keeping in view the health of previous crops, the potential susceptibility of subsequent crops, cultivar selection, and planting date (Kumar and Goh, 2000). The effects of residue management on pests and diseases in rice-based cropping systems have not received much attention, and systematic work with different management options is needed. It is possible that there would be fewer mulch-related disease problems in a rice–upland crop rotation than in a rotation of two different upland crops, because the flooded season might inhibit survival of the upland crop pathogens. One of the most important advantages of mulch compared with incorporation is weed suppression, but one of the most significant risks of mulching is higher disease incidence. If disease can be adequately controlled by crop rotation, mulching would become an attractive residue management option for mitigation of pest pressure in the upland crop of some rice-based cropping systems.

6.2. Profitability As with rice–rice and upland crop–rice systems (Section 5.2), the incorporation of rice residue before production of upland crops is profitable only if the value of increased production exceeds the increased cost for handling and incorporating residue. Rice residue is typically incorporated 2–6 weeks before sowing of the upland crop (Table 8) to ensure it does not interfere with seedbed preparation and sowing. As a result the residue is not incorporated along with tillage operations for seedbed preparation, and incorporation involves extra cost. No cost savings related to fertilizer use, irrigation water, and weed control are typically anticipated; and no gains in crop yield are typically obtained as shown in Figs. 2 and 3. Hence, the

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

175

incorporation of residues to the non-flooded crop in rice-based cropping systems is unlikely to be profitable. It is not surprising that virtually no farmers use this practice in Asia despite many years of research. In contrast to incorporation, mulching upland crops with rice residue has shown significant yield benefit (Fig. 4). It has been credited for savings in irrigation water and weed control (Section 6.1). An indirect benefit of sowing into rice residue is the possibility of earlier establishment of the upland crop immediately after rice harvest. In the Indo-Gangetic Plain, any reduction in the time between rice harvest and wheat sowing helps to reduce soil evaporation (a real water saving) and might eliminate the need for presowing irrigation. The opportunity for rapid turnaround between crops is particularly beneficial for wheat after late harvested rice crops such as ‘‘basmati,’’ which is transplanted later than other rice varieties. However, there is some risk of increased costs for disease control because loose residue can harbor plant pathogens. As discussed in Section 4.2.2, there have been significant recent developments in machinery allowing reduced and no-till sowing of upland crops in combine-harvested rice fields with rice residue left on the soil surface as mulch. Use of the new machine called the ‘‘Happy Seeder’’ for combined straw mulching and drill seeding in one field operation is compared in Table 11 with incorporation of rice residue for wheat sown in conventionally tilled soil. Considering the cost of custom hiring for cultivation and sowing, retaining residue as mulch using the ‘‘Happy Seeder’’ machine was less than half the cost for incorporation of residue.

6.3. Environmental impact 6.3.1. Air quality There are no anticipated differences between flooded and non-flooded systems in terms of their ability to drastically reduce air pollution associated with residue burning. The information presented in Section 5.3.1 is therefore relevant here. 6.3.2. Greenhouse gas emissions for non-flooded crop Greenhouse gas effects are very different in non-flooded crops than in flooded rice. Methane does not pose a serious environmental risk unless the soil becomes saturated or flooded, which happens rarely during the nonflooded crop in rice–upland crop rotations. Nitrous oxide is the main threat, and it is affected primarily by soil water and N fertilization rather than residue management (Abao et al., 2000; Becker et al., 2007). Several different studies comparing residue return and removal in rice-based or other upland cropping systems have found no significant difference in N2O emission between straw treatments. However, many other studies have

176

Bijay-Singh et al.

Table 11 Comparison of operations required for establishment of wheat after rice and economics under rice residue incorporation and surface mulching in no-till sown wheat using ‘‘Happy Seeder’’a in irrigated rice–wheat cropping system in northwestern Indo-Gangetic Plain

Operation

Rice straw retained on soil surface as mulch on no-till sown wheat

Rice straw incorporated in wheat sown in conventionally tilled soil

Machine used Presowing tillage

Happy Seedera Nil

Presowing irrigations Minimum lag period between rice harvest and sowing Sowing time (h ha–1) Cost of custom hiring for cultivation and sowing (US$ ha–1)

Nil Nil

Conventional seed drill 4 discings 2 harrowings 2 plankings 2 22–25 days

3.75 36.1

2.5 94.4

a

The name ‘‘Happy Seeder’’ is generically applied to all versions of the concept developed by the group from Punjab Agricultural University (Ludhiana, India), Dasmesh Mechanical Works (Amargarh, India) and CSIRO Land and Water (Griffith, Australia). The Combo Happy Seeder combines the straw management and sowing units into a single, light, compact machine. Source: Adapted from Sidhu et al., (2007)

found significant residue management interactions with water management, tillage, and type of residue (Table 12). If the soil is close to saturation after irrigation or a rainfall event, N2O emission is likely via denitrification from the surface (Becker et al., 2007). An increase in N2O emission from fields with mulch compared to those with incorporated residue has often been observed in temperate upland systems, and recently it has also been observed in subtropical Asian rice-based cropping systems (Baggs et al., 2003; Becker et al., 2007). This mulch effect is the result of higher water content under mulch, which leads to more anaerobic conditions, promoting denitrification (Aulakh et al., 1984; Linn and Doran, 1984). Baggs et al. (2000) speculated that timing residue return such that the N becomes available when needed by the upland crop should minimize N2O emission as compared with residue return at the beginning of the preseason fallow. Whether incorporated or mulched, residues with higher N content are likely to increase N2O emission, while those with low N (i.e., high C:N ratio) might decrease it (Baggs et al., 2000; Zou et al., 2004a). CH4 emission is rarely measured in upland experiments because soil is rarely anaerobic enough to produce significant amounts. However, a few

Table 12

Trends in N2O and CH4 emission for rice residue management options in upland (non-flooded) crops N2O

CH4

Management decision

Major options

Best

Crop residue use

Return Removal

No clear overall effect

No effect

Water management

Irrigate to maintain: Saturated soil Field capacity

Field capacity

Saturation

Field capacity

Saturation

How residue is returned

Incorporation Mulch (usually minimum tillage)

Incorporation

Mulch

Incorporationb

Mulch

Baggs et al. (2000, 2003), Becker et al. (2007)

Timing of residue return

Beginning of pre-season fallow Just before crop establishment

Before crop establishment

Fallow

When soil is dry

When soil is very wet

Baggs et al. (2000)

a

Worst

Best

Worst

References

Abao et al. (2000), Jacinthe and Lal (2003, 2004), Jarecki and Lal (2006), Liang et al. (2007), Zou et al. (2004a) Becker et al. (2007), Linn and Doran (1984)

(continued)

Table 12

(continued) N2O

Management decision

Major options

Best

Type of residue

High C:N (rice straw)

CH4 Worst

Best

Worst

References

High C:N, partially decomposed

Fresh, low C:N

Partially decomposed

Fresh

Baggs et al. (2000), Zou et al. (2004a)

Incorporation of rice residue at field capacity

Mulching of low C:N residue at saturation

Any residue management option at field capacity

Incorporation of fresh rice straw at saturation

a

Low C:N (legume residue) Fresh plant material Partially decomposed plant material (compost, manure) Overall combination

a

For both gases, ‘‘best’’ means the action that would minimize emission of the greenhouse gas. Under nearly saturated conditions, mulch would be better than incorporation. Note: Nitrous oxide is in bold because it represents the greater hazard in non-flooded systems b

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

179

researchers have included it for various reasons, and most have shown no significant overall effects of residue management on CH4 emission in upland crops (Table 12). Although there has not been direct testing of CH4 emission from upland fields for most of the residue management options in Table 12, we can infer the basic trends from biochemical principles. For any CH4 to be produced, there must be at least a small number of anaerobic microsites for methanogenic bacteria to grow, so any treatment that makes the soil even slightly more anaerobic is likely to increase the risk of CH4 emission, including a rainfall event or mulch application. If the soil remains very wet for an extended time, the subsurface will become distinctly more anaerobic than the surface, so incorporated residue would produce more CH4 than surface mulch. As in flooded systems, any action that causes residue to decompose before becoming anaerobic will lessen the risk of CH4 emission. From the perspective of mitigating greenhouse gas emissions from upland crops in rice-based cropping systems, residues are not the primary crop management concern. Assuming the soil is not saturated, N2O emission is at much greater risk than CH4 emission, and N2O emission is primarily influenced by the timing and method of N fertilization rather than by any residue-related factor. Rice residue incorporation when the soil is at field capacity water content would result in the lowest risk of N2O emission (Table 12), followed closely by mulching. When soil is at or near field capacity, there would be such little CH4 formation that the effect of residue management would be negligible. Neither mulch nor incorporation of rice residue into an upland crop would be expected to have very significant impact on CH4 emission in the following rice crop, because the incorporated or mulched residue would decompose considerably during the upland crop season (Abao et al., 2000).

6.4. Sustainability The long-term incorporation of crop residues to the non-flooded crop in rice-based cropping systems can increase SOM (Sahoo et al., 1998; Yadvinder-Singh et al., 2004b) and influence a number of soil chemical properties (Yadvinder-Singh et al., 2005). These effects of incorporated residues, however, seldom translate into increased yield (Section 6.1) or profit (Section 6.2) for the non-flooded crop. Incorporation of crop residue with high C:P ratio to non-flooded soil often leads to early immobilization of soil P (Qiu and Ding, 1986), which can be followed by progressively increased content of extractable soil P (Mukherjee et al., 1995; Yadvinder-Singh et al., 1988). In a study by Gupta et al. (2007) with rice–wheat, incorporation of rice residues in wheat and wheat residues in rice increased wheat yield only after 4 years. Significant residueP management interactions observed after the fourth year suggested

180

Bijay-Singh et al.

residues can enhance yield under limited P supply situations. Application of 26 kg P ha1 to wheat increased grain yield by 6–15% compared with no P. Rice did not respond to incorporation of residues or P fertilization. The P balance was negative with removal of rice straw, but when both wheat and rice straw were incorporated, the P balance was positive at the recommended P management of applying 26 kg P ha1 to only wheat. Changes in total soil P suggested that the two crops remove significant P from below 15 cm. Incorporation of residues increased bicarbonate-extractable inorganic and organic P, reduced P sorption, and increased P release. Data showed continuous incorporation of residues substituted for 13 kg inorganic P ha1 year1. Rice residue rapidly releases K during decomposition followed by increased plant-available soil K within 10 days after incorporation (Yadvinder-Singh et al., 2004b). Mishra et al. (2001) found 79% of the total K in rice residue was released within 5 weeks after incorporation into soil and 95% of K in rice residue was released after 23 weeks. The incorporation of crop residues can increase plant-available soil S in rice–wheat cropping system (Singh and Sharma, 2000). About 50–80% of Zn, Cu, and Mn taken up by rice and wheat crops can be recycled through incorporated residue (Prasad and Sinha, 1995). The retention of crop residues can contribute to increased soil water retention and improved soil physical properties during non-flooded crops (Bhagat et al., 2003; Yadvinder-Singh et al., 2005). In a long-term field study in China, the incorporation of rice residue increased soil porosity and formation of soil microaggregates and decreased soil bulk density (Li et al., 1986). The incorporation rather than removal of crop residues during non-flooded crops in rice-based cropping systems can also increase SOM (Yadvinder-Singh et al., 2004b), although the benefit of incorporated crop residues on soil properties seldom translates into increased yield (Section 6.1; Tables 8 and 9) or profit (Section 6.2) for the nonflooded crop (Bhagat et al., 2003; Singh et al., 2005; Yadvinder-Singh et al., 2005). Whereas SOM can be maintained during continuous rice cropping on puddled soil regardless of residue management (Section 5.4.2), the conversion from rice–rice to a rice–upland crop rotation with conventional tillage for the upland crop can result in loss of SOM and a reduction in soil N-supplying capacity even when crop residues are incorporated. In a long-term experiment in the Philippines (Witt et al., 2000), soil organic C significantly decreased by 16 to 17%—on a soil mass basis equivalent to 125 kg m–2—during a 13-year rice–maize rotation with conventional tillage for maize and full fertilization for all crops (Pampolino et al., 2008a). The incorporation of crop residues after each crop failed to prevent the decline in SOM and soil N-supplying capacity, as determined by anaerobic N mineralization and yield trends in no-N plots, in the rice–maize cropping

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

181

system. These results indicate that diversification from rice–rice cropping with traditional soil puddling to rice–upland crops, such as maize, can adversely affect sustainability indicators of SOM and soil nutrient-supplying capacity. Reduced and no tillage combined with surface retention of crop residue is now being examined as a more sustainable alternative for the nonflooded crop than conventional tillage for maize with residue incorporation. The surface retention of crop residues as mulch can increase SOM and improve soil physical properties leading to yield increases (Cho et al., 2001; Verma and Bhagat, 1992). The extent of SOM buildup associated with surface retention of rice residue in rice–wheat systems has not yet been extensively studied because the availability of appropriate machinery for mulching is a relatively recent development. In a recent study in Karnal, India on a soil with initial organic C content of 3.1 g kg–1, the incorporation of rice residue to conventionally tilled wheat increased soil organic C content to 4.0 g kg–1, whereas surface retention in no-till wheat increased soil organic C content to 3.7 g kg–1 (R. K. Sharma, Directorate of Wheat Research, Karnal, India, unpublished data).

7. Crop Residue and Bioenergy Options An emerging issue in residue management is its potential use in the production of bioenergy. There are several different ways in which crop residues in rice-based cropping systems might be used as biofuel (i.e., the conversion of biomass to liquid fuel) or as source of biopower (i.e., the conversion of biomass to electricity) (US DOE, 2007). The main biofuel option for crop residues is cellulosic ethanol production, which involves enzymatically breaking down the cellulose in the straw into its component sugars, which can then be fermented to ethanol. One of the biopower options is combustion—the burning of straw alone (direct-firing) or in combination with coal (co-firing) to produce steam to drive a turbine that turns an electricity generator. Other biopower options are gasification—in which straw is heated in the presence of limited O2 to form a gas mixture of N2, H2, CO, and CO2 known as producer gas or synthesis gas, and pyrolysis—in which straw is heated excluding O2 to form a high-energy liquid. Either the gas produced by gasification or the liquid produced by pyrolysis can be burned to produce electricity. A fourth biopower option is anaerobic digestion, also called biogas, in which straw is mixed with other wastes, such as manure, and allowed to decompose anaerobically—somewhat like when incorporated in a flooded soil—and the CH4 is collected and burned to produce electricity. For any bioenergy process involving crop residue, one of the main challenges is associated with the bulkiness of crop residues, which makes

182

Bijay-Singh et al.

them expensive to transport from the field to the power plant and to store before use (Mani et al., 2006; Shaw and Tabil, 2006). After harvest and threshing, the straw must usually be baled or compacted to facilitate transport to the energy production facility. After arriving at the facility, it usually must be reduced in size, and if destined for combustion, it would then be further compressed into briquettes or other densified forms (Ndiema et al., 2002). In any of the biopower systems, some ash or slag made up of the straw components that do not readily burn or gasify such as Si, K, Cl, and small quantities of other mineral nutrients will remain after the thermal conversion process (Arvelakis and Koukios, 2002). These potential bioenergy applications have three serious implications for residue management if widely adopted: (1) residue would be removed rather than returned to the soil, (2) harvesting and threshing techniques would need to be considerably modified to make it feasible to collect the residue for transportation, and (3) it would become important to grow varieties with desirable straw characteristics. The continuous removal of residue before a puddled and flooded rice crop, as discussed earlier, might be feasible without jeopardizing sustainability provided inputs of K, S, and other nutrients are suitably adjusted to compensate for those removed with residue. A threshing system or combined harvesting–threshing system would need to be developed to chop, compress, and stack straw as it is removed from the grain. Some desirable biochemical straw characteristics that could be bred into popular rice varieties include high lignocellulosic content and low ash content, particularly Si, K, and Cl. Most rice straw is relatively high in K and Si ( Jenkins et al., 1996), leaving much room for improvement. It would be better for the cropping system if these nutrients could be left in the soil rather than removed in the residue. Some studies have considered in-field leaching of the straw with water before using it for bioenergy (Bakker and Jenkins, 2003), but this would be difficult to manage logistically because the residue must be spread in the field, irrigated, dried, and then collected. It might also be possible to develop straw with favorable mechanical properties for easier chopping, grinding, and compressing (Kronbergs, 2000; Yu et al., 2006). A brittle stem mutant rice variety was developed that breaks very easily during threshing (Cabiles et al., 2008), which would make it easier to chop and grind. For cellulosic ethanol production, the step after size reduction is a chemical pretreatment to improve the accessibility of cellulose for enzymatic digestion, so another favorable residue biochemical property might be having forms of hemicellulose and lignin that are more easily digestible. The brittle mutant’s main biochemical difference was much lower cellulose content compared with the nonbrittle variety, although it had similar lignin and total C content ( Johnson et al., 2006), which may make it undesirable for cellulosic ethanol production unless the missing cellulose C is in a more easily digestible form than cellulose, rather than a more recalcitrant form (no data available). While the agronomic

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

183

performance of this particular brittle mutant is not suitable, some of the desirable characteristics might be bred into higher yielding varieties. Rice straw has no particular advantage compared to other residues for bioenergy production, except for its abundance across Asia. There are many cost and energy-efficiency challenges to be overcome before it would be profitable to use rice residue for large-scale energy production (Matsumura et al., 2005). Even if it never becomes profitable for large energy production facilities, it might be a useful source of small-scale energy production for households and small communities. Anaerobic digestion is gaining popularity across China, with 7.64 million households using biogas digestors at the end of 2000 (Zeng et al., 2007). On the basis of our analysis, residues of the crop before production of flooded rice with soil puddling could be removed from the field without detrimental medium-term effects on yield and SOM. Using the residue as bioenergy would have environmental advantages compared to the current practice of open-field burning. It would mitigate the air quality pollution caused by in-field fires while also contributing positively to the shift from fossil fuel to renewable energy sources that minimize greenhouse gas emission.

8. Summary Incorporation of crop residues in both flooded rice and non-flooded crops in rice-based cropping systems usually does not increase yield of the following crop, irrespective of amount of residue, timing of incorporation, amount of fertilizer applied, or incorporation of two or only one season’s residue per year. Productivity of the crop following the one that receives the residue is also usually not benefited. Residue incorporation often involves extra expenditure for tillage operations to turn the residue into the soil, unless the residue can be incorporated along with routine preparatory tillage. Incorporated residue can beneficially influence soil chemical and physical properties, especially in non-flooded soils. However, SOM in flooded rice-based cropping systems without aerobic soil tillage can be maintained even when crop residue is removed. Anaerobic residue decomposition produces much more CH4 than would be emitted from a flooded soil without incorporated residue. It can also have detrimental effects on young rice seedlings because of N immobilization and the production of reputedly toxic organic acids. All of these disadvantages can be overcome by incorporating the residue several months before land preparation and flooding, but this usually involves additional cost—rarely translating into increased productivity or profitability for the farmer—and it is not feasible in intensive cropping systems with only a brief interval between crops.

184

Bijay-Singh et al.

Mulching is an attractive alternative for residue management because it sometimes improves yield, has almost no added expenses (when used in reduced and no tillage systems), and has a number of indirect beneficial effects on crop growth, thereby increasing the likelihood that it will be profitable for farmers. It can be practiced in conventionally tilled (upland or lowland) and reduced or no-tilled (upland) fields, although the former requires extra labor for retaining the residue on the field surface after land preparation. With reduced and no tillage upland crops, extra labor cost for crop establishment in stubble is substantially reduced if appropriate machinery for sowing the crop in the presence of stubble and loose straw is available. Significant breakthroughs have been made in this direction for minimum till wheat into rice residue in the Indo-Gangetic Plain in South Asia. Because puddled soil by definition involves tillage, mulching is much more attractive in non-flooded crops than in flooded rice. Some indirect benefits of mulching include weed suppression, improvement in soil K fertility, and conservation of soil water resulting in savings in irrigation water at some locations (i.e., non-flooded systems). In addition to these two main residue management options (incorporation and mulching), there are a variety of other options practiced by small groups of farmers throughout Asia. In China, rice residue generated in rice–rice cropping systems is sometimes transferred from the field and transported to a location where it can be used as mulch in high-value crops and vegetables. Other alternatives include in situ composting, off-field composting, animal feed, and controlled burning as a household fuel. Local traditions and economic situations influence farmer decisions on use of these alternatives. We made a simplified decision tree (Fig. 5) to illustrate guidelines for managing residues in different parts of rice-based cropping systems, based on our evaluation of the productivity, profitability, and sustainability of the various options. For the systems in which residue from rice or non-flooded crop residues must be managed for rice, the management of residue depends upon whether soil during the recipient rice crop has been puddled. In the less common case of non-puddled rice, a no-till system is recommended in which the residues are left on the surface as mulch. For puddled rice production where crop residue cannot readily be used as mulch, the residue of the preceding crop can typically be safely removed from the field without loss in productivity or sustainability of the system, provided an appropriate increase in fertilizer, particularly K, is added to compensate for nutrient removal in the residue. The removal of crop residue has the potential to reduce the detrimental environmental impacts arising with CH4 emission from incorporating residue in flooded soils. The environmental costs are somewhat less for mulching than incorporation, but the residue must be removed before puddling and then returned, which is often unattractive and unprofitable for the farmer. Therefore, the removal of residue from puddled fields for off-field use rather than incorporating it can be a viable option,

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

185

Cropping system?

Rice residue before rice crop

Non-flooded crop residue before rice

Rice residue before non-flooded crop

Recipient rice crop puddled? Yes

Consider removal of residue for off-field use

No Retain residue as mulch with reduced or no tillage

Figure 5 A decision tree for use by farmers and policy makers to plan strategies for managing crop residues in rice-based cropping systems.

unless the energy costs for removal negate the benefits gained from off-field alternatives. This removal would be easier to accomplish in places where farmers usually thresh off-field. In any non-flooded crop in a rice-based cropping system (non-puddled rice, wheat, maize, legumes, and so on), residue should be retained as mulch in a reduced or no tillage system unless disease pressure becomes high, in which case occasional removal of residue would be appropriate. Consistent residue removal for non-flooded crops with full tillage will result in loss of SOM and soil nutrient supplying capacity because of enhanced oxidation of SOM in tilled, non-flooded fields, even when the non-flooded crop is rotated with a flooded rice crop. Residue incorporation has no significant advantage over mulching in terms of sustainability of SOM, and is less likely to improve productivity. It is important to prioritize continued development of machinery and crop establishment techniques that enable the seeding of non-flooded crops into standing rice stubble and loose mulch.

ACKNOWLEDGMENT This chapter was made possible in part through support by the Federal Ministry for Economic Cooperation and Development, Germany (BMZ) for the project ‘‘Managing Crop Residues for Healthy Soils in Rice Ecosystems’’ with the International Rice Research Institute (IRRI).

186

Bijay-Singh et al.

REFERENCES Abao, E. B., Jr., Bronson, K. F., Wassmann, R., and Singh, U. (2000). Simultaneous records of methane and nitrous oxide emissions in rice-based cropping systems under rain fed conditions. Nutr. Cycl. Agroecosys. 58, 131–139. Agrawal, R. P., Phogat, V. K., Chand, T., and Grewal, M. S. (1995). Improvements of soil physical conditions in Haryana. In ‘‘Research Highlights,’’ pp. 69–75. Department of Soil Science, CCS Haryana Agricultural University, Hisar, India. Ai, Y., Liu, X., Zhang, F., Mao, D., Zeng, X., Lu, S., and Pan, J. (2003). Effect of covering method on nitrogen use efficiency in dryland rice. Plant Nutr. Fert. Sci. 9, 416–419. Allmaras, R. R., Kraft, J. M., and Miller, D. E. (1988). Effects of soil compaction and incorporated crop residue on root health. Annu. Rev. Phytopathol. 26, 219–243. Andreae, M. O., and Merlet, P. (2001). Emissions of trace gases and aerosols from biomass burning. Global Biogeochem. Cycles 15, 955–966. Aquino, A. T., and Mabesa, R. C. (2002). Off-season production of honeydew melon (Cucumis melo L.) using mulch and row cover. Philipp. Agric. Sci. 85, 221–229. Arvelakis, S., and Koukios, E. G. (2002). Physiochemical upgrading of agroresidues as feedstocks for energy production via thermochemical conversion methods. Biomass Bioenergy 22, 331–348. Aulakh, M. S., Rennie, D. A., and Paul, E. A. (1984). The influence of plant residues on denitrification rates in conventional and zero tilled soils. Soil Sci. Soc. Am. J. 48, 790–794. Aulakh, M. S., Khera, T. S., Doran, J. W., and Bronson, K. F. (2001). Managing crop residue with green manure, urea, and tillage in a rice–wheat rotation. Soil Sci. Soc. Am. J. 65, 820–827. Baggs, E. M., Rees, R. M., Smith, K. A., and Vinten, A. J. A. (2000). Nitrous oxide emission from soils after incorporating crop residues. Soil Use Manage. 16, 82–87. Baggs, E. M., Stevenson, M., Pihlatie, M., Regar, A., Cook, H., and Cadisch, G. (2003). Nitrous oxide emissions following application of residues and fertiliser under zero and conventional tillage. Plant Soil 254, 361–370. Bailey, K. L., and Duczek, L. J. (1996). Managing cereal diseases under reduced tillage. Can. J. Plant Pathol. 18, 159–167. Bakker, R. R., and Jenkins, B. M. (2003). Feasibility of collecting naturally leached rice straw for thermal conversion. Biomass Bioenergy 25, 597–614. Ball, A. S., and Robertson, E. A. G. (1990). Straw incorporation and tillage methods: Straw decomposition, denitrification and growth and yield of wheat. J. Agric. Eng. Res. 46, 223–243. Barnard, G., and Kristoferson, L. (1985). ‘‘Agricultural Residues as Fuel in the Third World.’’ Earthscan Publications, London. p. 178 Beare, M. H., Wilson, P. E., Fraser, P. M., and Butler, R. C. (2002). Management effects of barley straw decomposition, nitrogen release, and crop production. Soil Sci. Soc. Am. J. 66, 848–856. Becker, M., Asch, F., Maskey, S. L., Pande, K. R., Shah, S. C., and Shreshtha, S. (2007). Effects of transition season management on soil N dynamics and system N balances in rice–wheat rotations of Nepal. Field Crops Res. 103, 98–108. Beri, V., and Sidhu, B. S. (1996). Management of crop residues for better environment. In ‘‘Agriculture and Environment’’ (B. D. Kansal, G. S. Dhaliwal, and M. S. Bajwa, Eds.), pp. 179–198. National Agriculture Technology Information Centre, Ludhiana, India. Beri, V., Sidhu, B. S., Bahl, G. S., and Bhat, A. K. (1995). Nitrogen and phosphorus transformations as affected by crop residue management practices and their influence on crop yield. Soil Use Manage. 11, 51–54. Bhagat, R. M., and Verma, T. S. (1991). Impact of rice straw management on soil physical properties and wheat yield. Soil Sci. 152, 108–115.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

187

Bhagat, R. M., Bhardwaj, A. K., and Sharma, P. K. (2003). Long-term effect of residue management on soil physical properties, water use and yield of rice in north-western India. J. Indian Soc. Soil Sci. 51, 111–117. Bhandari, A. L., Ladha, J. K., Pathak, H., Padre, A. T., Dawe, D., and Gupta, R. K. (2002). Yield and soil nutrient changes in a long-term rice–wheat rotation in India. Soil Sci. Soc. Am. J. 66, 162–170. Bhogal, A., Young, S. D., and Sylvester-Bradley, R. (1997). Straw incorporation and immobilization of spring-applied nitrogen. Soil Use Manage. 13, 111–116. Bijay-Singh and Yadvinder-Singh (2003). Environmental implications of nutrient use and crop management in rice-based ecosystems. In ‘‘Rice Science: Innovations and Impact for Livelihood’’ (T. W. Mew, D. S. Brar, S. Peng, D. Dawe, and B. Hardy, Eds.), pp. 463–477. International Rice Research Institute, Los Ban˜os, Philippines. Bijay-Singh, Yadvinder-Singh, Sadana, U. S., and Meelu, O. P. (1992). Effect of organic amendments and moisture regimes on the kinetics of micronutrients in a calcareous sandy loam soil. J. Indian Soc. Soil Sci. 40, 114–118. Bijay-Singh, Bronson, K. F., Yadvinder-Singh, Khera, T. S., and Pasuquin, E. (2001). Nitrogen-15 balance as affected by rice residue management in a rice–wheat rotation in northwest India. Nutr. Cycl. Agroecosys. 59, 227–237. Bilalis, D., Sidiras, N., Economou, G., and Vakali, C. (2003). Effect of different levels of wheat straw soil surface coverage on weed flora in Vicia faba crops. J. Agron. Crop Sci. 189, 233–241. Bird, J. A., Pettygrove, G. S., and Eadie, J. M. (2000). The impact of waterfowl foraging on the decomposition of rice straw: Mutual benefits for rice growers and waterfowl. J. Appl. Ecol. 37, 728–741. Bird, J. A., Howarth, W. R., Eagle, A. J., and Van Kessel, C. (2001). Immobilization of fertilizer nitrogen in rice: Effects of straw management practice. Soil Sci. Soc. Am. J. 65, 1143–1152. Blackwell, J., Sidhu, H. S., Dhillon, S. S., and Prashar, A. (2004). The Happy Seeder concept—A solution to the problem of sowing into heavy stubble residues. Rice–Wheat Consort. Newsl. January 2004. Bockus, W. W., and Shroyer, J. P. (1998). The impact of reduced tillage on soil borne plant pathogens. Annu. Rev. Phytopathol. 36, 485–500. Brar, S. S., Kumar, S., Brar, L. S., Walia, S. S., and Kumar, S. (1998). Effect of crop residue management systems on the grain yield and efficacy of herbicides in rice–wheat sequence. Indian J. Weed Sci. 30, 39–43. Bronson, K. F., Cassman, K. G., Wassmann, R., Olk, D. C., van Noordwijk, M., and Garrity, D. P. (1997a). Soil carbon dynamics in different cropping systems in principal ecoregions of Asia. In ‘‘Management of Carbon Sequestration in Soil’’ (R. Lal, J. Kimble, R. F. Follet, and B. A. Stewart, Eds.), pp. 35–57. CRC Press, Boca Raton, FL, USA. Bronson, K. F., Neue, H.-U., Singh, U., and Abao, E. B., Jr. (1997b). Automated chamber measurements of methane and nitrous oxide flux in a flooded soil: I. Residue, nitrogen and water management. Soil Sci. Soc. Am. J. 61, 81–987. Buresh, R. J., Reddy, K. R., and van Kessel, C. (2008). Nitrogen transformations in submerged soils. In ‘‘Nitrogen in Agricultural Systems’’ ( J. S. Schepers and W. R. Raun, Eds.), Agronomy Monograph 49. pp. 401–436. ASA, CSSA, and SSSA, Madison, WI. Cabiles, D. M. S., Angeles, O. R., Johnson-Beebout, S. E., Sanchez, P. B., and Buresh, R. J. (2008). Faster residue decomposition of a brittle stem rice mutant due to finer breakage during threshing. Soil Till. Res. 98, 211–216. Cannell, R. Q., and Lynch, J. M. (1984). Possible adverse effects of decomposing crop residues on plant growth. In ‘‘Organic Matter and Rice’’, pp. 455–475. International Rice Research Institute, Los Ban˜os, Laguna, Philippines.

188

Bijay-Singh et al.

Cassman, K. G. (1999). Ecological intensification of cereal production systems: Yield potential, soil quality, and precision agriculture. In Academy of Sciences colloquium; ‘‘Plants and Population: Is There Time?’’ (Irvine, CA). December 5–6, 1998. Proc. Natl. Acad. Sci. USA 96(11), 5952–5959. Cassman, K. G., and Pingali, P. L. (1995). Intensification of irrigated rice systems learning from the past to meet future challenges. Geo. J. 35, 299–305. Cassman, K. G., De Datta, S. K., Olk, D. C., Alcantara, J., Samson, M., Descalsota, J. P., and Dizon, M. (1995). Yield decline and the nitrogen economy of long-term experiments on continuous irrigated rice systems in the tropics. In ‘‘Sustainable Management of Soils’’ (R. Lal and B. A. Stewart, Eds.), pp. 181–222. Lewis CRC Publishers, Boca Raton, FL, USA. Cassman, K. G., De Datta, S. K., Amarante, S. T., Liboon, S. P., Samson, M. I., and Dizon, M. A. (1996). Long-term comparison of the agronomic efficiency and residual benefits of organic and inorganic nitrogen sources for tropical lowland rice. Exp. Agric. 32, 427–444. Chatterjee, B. N., and Mondal, S. S. (1996). Potassium nutrition under intensive cropping. J. Pot. Res. 12, 358–464. Chatterjee, B. N., Singh, K. I., Pal, I., and Maiti, S. (1979). Organic manures as substitutes for chemical fertilizers for high yielding rice varieties. Indian J. Agric. Sci. 49, 188–192. Cho, D. Y., and Ponnamperuma, F. N. (1971). Influence of soil temperature on the chemical kinetics of flooded soils and the growth of rice. Soil Sci. 112, 184–194. Cho, Y. S., Lee, B. Z., Choe, Z. R., and Ockerby, S. E. (2001). An evaluation of a notillage, unfertilised, direct-sown, wheat–rice cropping system in Korea. Aust. J. Exp. Agric. 41, 53–60. Chung, I. M. (2001). Identification of allelopathic compounds from rice (Oryza sativa L.) straw and their biological activity. Can. J. Plant Sci. 81, 815–819. Chung, Y. R., Hoitink, H. A. H., and Lipps, P. E. (1988). Interactions between organic-matter decomposition level and soilborne disease severity. Agric. Ecosyst. Environ. 24, 183–193. Cintas, N. A., and Webster, R. K. (2001). Effects of rice straw management on Sclerotium oryzae inoculum, stem rot severity, and yield of rice in California. Plant Dis. 85, 1140–1144. Cooke, W. F., and Wilson, J. J. N. (1996). A global black carbon aerosol model. J. Geophys. Res. 101, 19395–19409. Corton, T. M., Bajita, J. B., Grospe, F. S., Pamplona, R. R., Asis, C. A., Jr., Wassmann, R., Lantin, R. S., and Buendia, L. V. (2000). Methane emission from irrigated and intensively managed rice fields in Central Luzon (Philippines). Nutr. Cycl. Agroecosys. 58, 37–53. Curl, E. A. (1988). The role of soil microfauna in plant-disease suppression. CRC Crit. Rev. Plant Sci. 7, 175–196. Dalrymple, A. W., Miller, S. D., and Fornstrone, K. J. (1992). Soil water conservation and winter wheat yield in three fallow systems. J. Soil Water Conserv. 47, 53–57. Dawe, D., Dobermann, A., Ladha, J. K., Yadav, R. L., Bao, L., Gupta, R. K., Lal, P., Panaullah, G., Sariam, O., Singh, Y., Swarup, A., and Zhen, Q. (2003). Do organic amendments improve yield trends and profitability in intensive rice systems? Field Crops Res. 83, 191–213. Dawe, D., Frolking, S., and Li, C. (2004). Trends in rice–wheat area in China. Field Crops Res. 87, 89–95. Deng, Y., and Deng, X. (2005). The ecological effects of rice straw mulching on vineyard. Agric. Sci. Tech. 6(3), 20–24. Dhiman, S. D., Nandal, D. P., and Om, H. (2000). Productivity of rice (Oryza sativa)-wheat (Triticum aestivum) cropping system as affected by its residue management and fertility levels. Indian J. Agron. 45, 1–5.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

189

Dobermann, A., and Fairhurst, T. H. (2000). ‘‘Rice: Nutrient Disorders & Nutrient Management.’’ Potash & Phosphate Institute (PPI), Potash & Phosphate Institute of Canada (PPIC) and International Rice Research Institute (IRRI), East & South Asia Programs of PPI and PPIC, Singapore. Dobermann, A., and Fairhurst, T. H. (2002). Rice straw management. Better Crops Int. 16, 7–9. Dobermann, A., and Witt, C. (2000). The potential impact of crop intensification on carbon and nitrogen cycling in intensive rice systems. In ‘‘Carbon and Nitrogen Dynamics in Flooded Soils’’ (G. J. D. Kirk and D. C. Olk, Eds.), pp. 1–25. International Rice Research Institute, Los Ban˜os, Philippines. Dobermann, A., Simbahan, G. C., Moya, P. F., Adviento, M. A. A., Tiongco, M., Witt, C., and Dawe, D. (2004). Methodology for socioeconomic and agronomic on-farm research in the RTDP project. In ‘‘Increasing Productivity of Intensive Rice Systems Through Site-Specific Nutrient Management’’ (A. Dobermann, C. Witt, and D. Dawe, Eds.), pp. 11–27. Science Publishers, Enfeld, NH (USA) and International Rice Research Institute, Los Ban˜os (Philippines). Duxbury, J. M., and Lauren, J. G. (2002). Sustainability of the rice–wheat cropping system in South Asia. CIFAD Report. http://ciifad.cornell.edu/annualreports/1998-1999/ documents/rice-wheatadj.pdf. Eagle, A. J., Bird, J. A., Horwath, W. R., Linquist, B. A., Brouder, S. M., Hill, J. E., and van Kessel, C. (2000). Rice yield and nitrogen utilization efficiency under alternative straw management practices. Agron. J. 92, 1096–1103. Erenstein, O. (2002). Crop residue mulching in tropical and semitropical countries: An evaluation of residue availability and other technological implications. Soil Till. Res. 67, 115–133. Erenstein, O., Farooq, U., Malik, R. K., and Sharif, M. (2007). Adoption and impacts of zero tillage as a resource conserving technology in the irrigated plains of South Asia. Comprehensive Assessment of Water Management in Agriculture Research Report 19. International Water Management Institute, Colombo, Sri Lanka. Fan, M., Jiang, R., Liu, X., Zhang, F., Lu, S., Zeng, X., and Peter, C. (2005). Interactions between non-flooded mulching cultivation and varying nitrogen inputs in rice–wheat rotations. Field Crops Res. 91, 307–318. Fan, X., Zhang, J., and Wu, P. (2002). Water and nitrogen use efficiency of lowland rice in ground covering rice production system in South China. J. Plant Nutr. 25, 1855–1862. FAO. (2007). FAOSTAT Agricultural Production Database, Food and Agriculture Organization of the United Nations, Rome. http://faostat.fao.org/site/339/default.aspx. Fisher, R. A., Santiveri, F., and Vidal, I. R. (2002). Crop rotation, tillage crop residue management for wheat and maize in the sub-humid and tropical highlands I. Wheat and legume performance. Field Crops Res. 79, 107–122. Flinn, J. C., and Marciano, V. P. (1984). Rice straw and stubble management. In ‘‘Organic Matter and Rice,’’ pp. 593–611. International Rice Research Institute, Los Ban˜os, Philippines. Frolking, S., Qiu, J., Boles, S., Xiao, X., Liu, J., Zhuang, Y., Li, C., and Qin, X. (2002). Combining remote sensing and ground census data to develop new maps of the distribution of rice agriculture in China. Global Biogeochem. Cycles 16, 1091, doi:10.1029/ 2001GB001425. Gajri, P. R., Ghuman, B. S., Samar Singh Mishra, R. D., Yadav, D. S., and Harmanjit Singh, D. S. (2002). Tillage and residue management practices in rice–wheat system in Indo-Gangetic Plains—A diagnostic survey. Technical Report, National Agricultural Technology Project, Indian Council of Agricultural Research, New Delhi and Punjab Agricultural University, Ludhiana, India. p. 12

190

Bijay-Singh et al.

Gao, X., Ma, W., Ma, C., Zhang, F., and Wang, Y. (2002). Analysis on current status of utilization of crop straw in China. J. Huazhong Agric. Univ. 21, 242–247. Garg, I. K. (2002). Design, development and testing of flail type chopper cum spreader for rice straw, quarterly progress report to Indian Council of Agricultural Research. Farm Implement and Machinery Scheme, Punjab Agricultural University, Ludhiana, India. Gill, K. S., and Arshad, M. A. (1995). Weed flora in the early growth period of spring crops under conventional, reduced, and zero tillage systems on a clay soil in northern Alberta Canada. Soil Till. Res. 33, 65–79. Gummert, M., and Aldas, R. E. (1993). A study in the different locally available stationary and transportable rice thresher. Project report GTZ-IRRI III 86.2105.4–03.201. Hohenheim University and IRRI, Los Ban˜os, Philippines. Gupta, R. K., and Rickman, J. (2002). Design improvements in existing zero-till machines for residue conditions. Rice–Wheat Consortium Traveling Seminar Report Series 3, RWC-CIMMYT, New Delhi, India. Gupta, P. K., Sahai, S., Nahar Singh, Dixit, C. K., Singh, D. P., Sharma, C., Tiwari, M. K., Gupta, R. K., and Garg, S. C. (2004). Residue burning in rice–wheat cropping system: Causes and implications. Curr. Sci. 87, 1713–1716. Gupta, R. K., Yadvinder-Singh, Ladha, J. K., Bijay-Singh, Jagmohan-Singh, GurpreetSingh, and Pathak, H. (2007). Phosphorus transformations and yield responses with crop residue and phosphorus management in a rice–wheat system in India. Soil Sci. Soc. Am. J. 71, 1500–1507. Han, H. S., Lee, M. H., and Slum, J. S. (1991). Effects of long-term fertilizer application on growth, yield and grain development of rice. Korean J. Crop Sci. 36, 41–45. Han, L., Yan, Q., Liu, X., and Hu, J. (2002). Straw resources and their utilization in China. J. Chinese Agric. Eng. 18(3), 87–92. Hanaki, M., Ito, T., and Saigusa, M. (2002). Effect of no-tillage rice (Oryza sativa L.) cultivation on methane emission in three paddy fields of different soil types with rice straw application. Japanese J. Soil Sci. Plant Nutr. 73, 135–143 (in Japanese with English summary). Harada, H., Kobayashi, H., and Shindo, H. (2007). Reduction in greenhouse gas emissions by no-tilling rice cultivation in Hachirogata polder, northern Japan: Life-cycle inventory analysis. Soil Sci. Plant Nutr. 53, 668–677. Houghton, J. T., Ding, Y., Griggs, D. J., Noguer, M., van der Linden, P. J., Dai, X., Maskell, K., and Johnson, C. A. (2001). ‘‘Climate Change 2001: The Scientific Basis.’’ Cambridge University Press for the Intergovernmental Panel on Climate Change (IPCC), Cambridge, UK. Hu, H., Wang, Q., Li, S., and Chen, Y. (2004). Production increase effect and mechanism for straw covering of wheat field in Jingmen city. Soils Fert. 5, 30–35. Huang, Y. D., Li, J. C., and Zhang, Z. L. (1997). Primary report of cultivation technique of mulching rice. J. Anhui Agric. Sci. 25, 208–210. Humphreys, E., Meisner, C., Gupta, R. K., Timsina, J., Beecher, H. G., Tang, Y. L., Yadvinder Singh, Gill, M. A., Masih, I., Zheng, J. G., and Thompson, J. A. (2004).Water saving in rice–wheat systems. In ‘‘Proceedings of the 4th International Crop Science Congress,’’ Brisbane, Australia, 26 Sep.–1 Oct., 2004. http://www.cropscience.org.au/ icsc2004/symposia/1/2/237_humphreyse.htm. Humphreys, E., Blackwell, J., Sidhu, H. S., Malkeet Singh, H. S., Sarbjeet Singh, Manpreet Singh, Yadvinder Singh, and Anderson, L. (2006). Direct drilling into stubbles with Happy Seeder. IREC Farmers’ Newsl., No. 172, Autumn 2006: 4–7. IAEA. (2003). Management of Crop Residues for Sustainable Crop Production. IAEATECDOC-1354. Soil and Water Management & Crop Production Section, International Atomic Energy Agency, Wagramer Strasse 5, P.O. Box 100, A-1400 Vienna, Austria. pp. 207–220.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

191

IPCC. (1996). Report of the Twelfth Session of the Intergovernmental Panel on Climate Change. IPCC-XII/RPT. 1. 12-IX-1996. IRRI. (2007). Site-specific nutrient management, www.irri.org/irrc/ssnm. Ishibashi, E., Akai, N., Ohya, M., Ishii, T., and Tsuruta, H. (2001). The influence of notilled direct seeding cultivation on methane emission from three rice paddy fields in Okayama, Western Japan: 2. The relationship between the continuation of no-tilled cultivation and methane emission. Jpn. J. Soil Sci. Plant Nutr. 72, 542–549. (in Japanese with English summary). Ishibashi, E., Yamamoto, S., Akai, N., and Tsuruta, H. (2005). The influence of no-till direct seeding cultivation on methane emissions from rice paddy fields in Okayama, Western Japan. Jpn. J. Soil Sci. Plant Nutr. 76, 629–639 (in Japanese with English summary). Ismunadji, M. (1978). Utilization of cereal crop residues and its agricultural significance in Indonesia. Contributions Central Research Institute for Agriculture, Bogor. (No. 37). Jacinthe, P. A., and Lal, R. (2003). Nitrogen fertilization of wheat residue affecting nitrous oxide and methane emission from a central Ohio Luvisol. Biol. Fert. Soils 37, 338–347. Jacinthe, P. A., and Lal, R. (2004). Effects of soil cover and land-use on the relations fluxconcentration of trace gases. Soil Sci. 169, 243–259. Jarecki, M. K., and Lal, R. (2006). Compost and mulch effects on gaseous flux from an alfisol in Ohio. Soil Sci. 171, 249–260. Jenkins, B. M., Bakker, R. R., and Wei, J. B. (1996). On the properties of washed straw. Biomass Bioenergy 10, 177–200. Jiang, P., Ye, Z., and Xu, O. (2002). Effect of mulching on soil chemical properties and enzyme activities in bamboo plantation of phyllostachy praecox. Comm. Soil Sci. Plant Anal. 33, 3135–3145. Johnson, S. E., Angeles, O. R., Brar, D. S., and Buresh, R. J. (2006). Faster anaerobic decomposition of a brittle straw rice mutant: Implications for residue management. Soil Biol. Biochem. 38, 1880–1892. Kamara, A., Akobundu, I., Chikoye, D., and Jutzi, S. (2000). Selective control of weeds in an arable crop by mulches from some multipurpose trees in Southern Nigeria. Agroforest. Syst. 50, 17–26. Kavimandan, S. K., Gupta, J. P., and Mahapatra, I. C. (1987). Studies on residue management and biofertilizers in rice–wheat sequence. Indian J. Agron. 32, 278–279. Koopmans, A., and Koppejan, J. (1997). Agricultural and forest fires: Generation, utilization and availability. In ‘‘Proceedings of the Regional Consultation on Modern Applications of Biomass Energy,’’ Kuala Lumpur, Malaysia. Kronbergs, E. (2000). Mechanical strength testing of stalk materials and compacting energy evaluation. Ind. Crops Prod. 11, 211–216. Kumar, K., and Goh, K. M. (2000). Crop residues and management practices: Effects on soil quality, soil nitrogen dynamics, crop yields, and nitrogen recovery. Adv. Agron. 68, 197–319. Kumar, K., Goh, K. M., Scott, W. R., and Frampton, C. M. (2001). Effects of 15N-labelled crop residues and management practices on subsequent winter wheat yields, nitrogen benefits and recovery under field conditions. J. Agric. Sci. Camb. 136, 35–53. Ladha, J. K., Dawe, F., Ventura, T. S., Singh, U., Ventura, W., and Watanabe, I. (2000). Long-term effects of urea and green manure on rice yields and nitrogen balance. Soil Sci. Soc. Am. J. 64, 1993–2001. Ladha, J. K., Pathak, H., Tirol-Padre, A., Dawe, D., and Gupta, R. K. (2003). Productivity trends in intensive rice–wheat cropping systems in Asia. In ‘‘Improving the Productivity and Sustainability of Rice–Wheat Systems: Issues and Impacts’’ ( J. K. Ladha, J. E. Hill, J. M. Duxbury, R. K. Gupta, and R. J. Buresh, Eds.), pp. 45–76. ASA Special Publication Number 65. ASA-CSSA-SSSA, Madison, WI, USA.

192

Bijay-Singh et al.

Lal, R. (1989). Conservation tillage for sustainable agriculture: Tropical versus temperate environments. Adv. Agron. 42, 185–197. Lawler, S. P., and Dritz, D. A. (2006). Effects of rice straw and water management on riceland mosquitoes. J. Med. Entomol. 43, 828–832. Le Cerff, R., Le Mufran, R., Buntan, A., and Corpuz, I. T. (1985). Yield response of IR 32 to inorganic and organic fertilizers. Int. Rice Res. Newsl. 10(6), 31–32. Lefroy, R. D. B., Chaitep, W., and Blair, G. J. (1994). Release of sulfur from rice residues under flooded and non-flooded soil conditions. Aust. J. Agric. Res. 45, 657–667. Li, C., Salas, W., DeAngelo, B., and Rose, S. (2006). Assessing alternatives for mitigating net greenhouse gas emissions and increasing yields from rice production in China over the next twenty years. J. Environ. Qual. 35, 1554–1565. Li, H. Z., Han, H. R., Wu, Z. C., Yang, J. C., and Ge, L. M. (1986). A study on the efficacy of organic manures in improvement of paddy soil fertility. J. Soil Sci. China 17, 252–258. Li, S. (1991). Soil management problems in multiply cropped paddy fields in China. Biol. Fert. Soils 12, 213–216. Li, X., Wu, J., Zhu, H., and Wang, Y. (2003). Effect of straw returning on crop yield and soil fertility. J. Anhui Agric. Sci. 31, 870–871. Li, Y. (2005). The status and prospects of rice production mechanization in China. In ‘‘Rice is Life: Scientific Perspectives for the 21st Century’’ (K. Toriyama, K. L. Heong, and B. Hardy, Eds.), pp. 228–229. International Rice Research Institute, Los Ban˜os, Philippines. www.irri.org/publications/wrrc/wrrcPDF/session7-03.pdf. Liang, W., Shi, Y., Zhang, H., Yue, J., and Huang, G. H. (2007). Greenhouse gas emissions from Northeast China rice fields in fallow season. Pedosphere 17, 630–638. Lin, S., Dittert, K., Tao, H., Kreye, C., Xu, Y., Shen, Q., Fan, X., and Sattelmacher, B. (2002). The ground-cover rice production system (GCRPS): A successful new approach to save water and increase nitrogen fertilizer efficiency? In ‘‘Water-Wise Rice Production’’ (B. A. M. Bouman, H. Hengsdijk, B. Hardy, B. Bindraban, T. P. Toung, and J. K. Ladha, Eds.), pp. 187–204. Proceedings of the International Workshop on Waterwise Rice Production, 8–11 April 2002. Los Ban˜os, Philippines. International Rice Research Institute, Los Ban˜os, Philippines. Linn, D. M., and Doran, J. W. (1984). Aerobic and anaerobic microbial populations in notill and plowed soils. Soil Sci. Soc. Am. J. 48, 794–799. Linquist, B. A., Brouder, S. M., and Hill, J. E. (2006). Winter straw and water management effects on soil nitrogen dynamics in California rice systems. Agron. J. 98, 1050–1059. Liu, D., Chen, Y., Luo, L., Yuan, Z., and Li, R. (2006). The effects of sowing date, density and breed on mom-ploughed potato covered with straw. J. Mountain Agric. Biol. 25(1), 1–4. Liu, H. (2005). Effects of straw mulching on growth and development of flue-cured tobacco. Tob. Sci. China 26(1), 31–33. Liu, X., Wang, J., Lu, S., Zhang, F., Zeng, X., Ai, Y. W., Peng, S., and Christie, P. (2003). Effects of non-flooded mulching cultivation on crop yield, nutrient uptake and nutrient balance in rice–wheat cropping systems. Field Crops Res. 83, 297–311. Liu, X., Ai, Y., Zhang, F., Lu, S., Zeng, X., and Fan, M. (2005). Crop production, nitrogen recovery and water use efficiency in rice–wheat rotation as affected by non-flooded mulching cultivation (NFMC). Nutr. Cycl. Agroecosys. 71, 289–299. Lou, Y. S., Ren, L. X., Li, Z. P., Zhang, T. L., and Inubushi, K. (2007). Effect of rice residues on carbon dioxide and nitrous oxide emission from a paddy soil of subtropical China. Water Air Soil Pollut. 178, 157–168. Lu, W. F., Chen, W., Duan, B. W., Guo, W. M., Lantin, R. S., Wassmann, R., and Neue, H. U. (2000). Methane emissions and mitigation options in irrigated rice fields in Southeast China. Nutr. Cycl. Agroecosyst. 58, 65–73.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

193

Luo, S. J. (1997). Study on rice cultivation on dry land or without flooding. J. Anhui Agric. Sci. 25, 333–337. Ma, G. (2004). Ecological organic fertilizers and sustainable development of agriculture. Chinese J. Eco-agric. 12(3), 191–193. Ma, J., Li, X. L., Xu, H., Han, Y., Cai, Z. C., and Yagi, K. (2007). Effects of nitrogen fertiliser and wheat straw application on CH4 and N2O emissions from a paddy rice field. Aust. J. Soil Res. 45, 359–367. Ma, Z., Lu, X., Wan, L., Chen, Z., and Zuo, H. (2003). Effect of wheat straw returning of rice yield and soil fertility. Crops 5, 37–38. Manakul, T. (1994). Response of wheat to straw mulching, M.S. Thesis in Agriculture (Agricultural system) Graduate School of Chiang Mai University, Chiang Mai, Thailand. Mani, S., Tabil, L. G., and Sokhansanj, S. (2006). Specific energy requirement for compacting corn stover. Bioresour. Technol. 97, 1420–1426. Martens, D. A. (2000). Plant residue biochemistry regulates soil carbon cycling and carbon sequestration. Soil Biol. Biochem. 32, 361–369. Matsumura, Y., Minowa, T., and Yamamoto, H. (2005). Amount, availability, and potential use of rice straw (agricultural residue) biomass as an energy resource in Japan. Biomass Bioenergy 29, 347–354. Miller, T. C., and Webster, R. K. (2001). Soil sampling techniques for determining the effect of cultural practices on Rhizoctonia oryzae-sativae inoculum in rice field soils. Plant Dis. 85, 967–972. Mishra, B., Sharma, P. K., and Bronson, K. F. (2001). Decomposition of rice straw and mineralization of carbon, nitrogen, phosphorus and potassium in wheat field soil in western Uttar Pradesh. J. Indian Soc. Soil Sci. 49, 419–424. Misra, R. V., Roy, R. N., and Hiraoka, H. (2003). ‘‘On-farm Composting Methods.’’ Food and Agriculture Organization of the United Nations, Rome. Moorman, T. B. (1989). A review of pesticide effects on microorganisms and microbial processes related to soil fertility. J. Prod. Agric. 2, 14–23. Muehlbauer, F. J., and Tullu, A. (1997). Lens culinaris Medik. Center for new crops and plant products, Purdue University. http://www.hort.purdue.edu/newcrop/cropfactsheets/ lentil.html. Mukherjee, D., Chattopadhyay, M. K., and Chakravarty, A. (1995). Some aspects of chemical changes as influenced by different organic additives in Entisol of Gangetic origin. Adv. Plant Sci. 8, 169–176. Nagarajah, S., Neue, H.-U., and Alberto, M. C. R. (1989). Effect of Sesbania, Azolla and rice straw incorporation on the kinetics of NH4, K, Fe, Mn, Zn and P in some flooded rice soils. Plant Soil 116, 37–48. Ndiema, C. K. W., Manga, P. N., and Ruttoh, C. R. (2002). Densification characteristics of rice straw briquettes. J. Inst. Energ. 75, 11–13. Ning, C. G., and Hu, T. G. (1990). The role of straw-covering in crop production and soil management. Better Crops Int. 6(2), 6–7. OFWM. (2002). ‘‘Impact Assessment of Resource Conservation Technologies (rice–wheat) DFID Project 1999–2002.’’ Directorate General Agriculture Water Management, Lahore, Pakistan. Olk, D. C., Cassman, K. G., Randall, E. W., Kinchesh, P., Sanger, L. J., and Anderson, J. M. (1996). Changes in chemical properties of organic matter with intensified rice cropping in tropical lowland soil. Eur. J. Soil Sci. 47, 293–303. Olk, D. C., Brunetti, G., and Senesi, N. (2000). Decrease in humification of organic matter with intensified lowland rice cropping: A wet chemical and spectroscopic investigation. Soil Sci. Soc. Am. J. 64, 1337–1347.

194

Bijay-Singh et al.

Padilla, J. L. (2001). Analysis of long-term changes in rice productivity under intensive cropping systems in the tropics and improvement of nitrogen use efficiency. Ph.D. Dissertation Graduate School of Agriculture, Kyoto University, Kyoto, Japan. Pal, S. S., Jat, M. L., Sharma, S. K., and Yadav, R. L. (2002). ‘‘Managing Crop Residues in Rice–Wheat System.’’ p. 48. PDCSR Bull. No. 2002–1. Project Directorate for Cropping System Research, Modipuram, Meerut, India. Pampolino, M. F., Larazo, W. M., Buresh, R. J. (2008a). Soil carbon and nitrogen losses following conversion from continuous rice cropping to a rice-maize rotation. Plant Soil (in press). Pampolino, M. F., Laureles, E. V., Gines, H. C., and Buresh, R. J. (2008b). Soil carbon and nitrogen changes in long-term continuous lowland rice cropping. Soil Sci. Soc. Am. J. (in press). Pandey, S. P., Shankar, H., and Sharma, V. K. (1985). Efficacy of some organic and inorganic residues in relation to crop yield and soil characteristics. J. Indian Soc. Soil Sci. 33, 178–181. Patil, M. N., Zade, K. B., Naphade, K. T., and Kharkar, P. T. (1993). Decomposition of organic materials in soil in relation to nutrient mineralization. J. Maharashtra Agric. Univ. 18, 348–351. Peng, W., Song, T., Xiao, R., Yang, Z., Li, S., Xia, Y., and Tang, Y. (2005). Effects of straw mulching and intercropping white clover in tea plantation on soil moisture in subtropical hilly region. J. Soil Water Conserv. 19(6), 97–101. Peng, S., Buresh, R. J., Huang, J., Yang, J., Zou, Y., Zhong, X., Wang, W., and Zhang, F. (2006a). Strategies for overcoming low agronomic nitrogen use efficiency in irrigated rice systems in China. Field Crops Res. 96, 37–47. Peng, W., Song, T., Xiao, R., Yang, Z., Wang, J., Li, S., and Xia, Y. (2006b). Effects of mulching and intercropping on temporal-spatial variation of soil temperature in tea plantation in subtropical hilly region. Chinese J. Appl. Ecol. 17(5), 778–782. Phongpan, S., and Mosier, A. R. (2003a). Effect of crop residue management on nitrogen dynamics and balance in a lowland rice cropping system. Nutr. Cycl. Agroecosys. 66, 133–142. Phongpan, S., and Mosier, A. R. (2003b). Impact of organic residue management on nitrogen use efficiency in an annual rice cropping sequence of lowland Central Thailand. Nutr. Cycl. Agroecosys. 66, 233–240. Phongpan, S., and Mosier, A. R. (2003c). Effect of rice straw management on nitrogen balance and residual effect of urea-N in an annual lowland rice cropping sequence. Biol. Fert. Soils 37, 102–107. Ponnamperuma, F. N. (1984). Straw as a source of nutrients for wetland rice. ‘‘Organic Matter and Rice,’’ pp. 117–135. International Rice Research Institute, Los Ban˜os, Philippines. Prasad, B., and Sinha, S. K. (1995). Nutrient recycling through crop residues management for sustainable rice and wheat production in calcareous soil. Fert. News 40(11), 15–2325. Prasad, R., Gangaiah, B., and Aipe, K. C. (1999). Effect of crop residue management in a rice–wheat cropping system on growth and yield of crops and on soil fertility. Exp. Agric. 35, 427–435. Qin, J., Hu, F., Li, H., Wang, Y., Huang, F., and Huang, H. (2006). Effects of dry cultivation with straw mulching on rice agronomic traits and water use efficiency. Chinese J. Rice Sci. 20(2), 171–176. Qiu, F. Q., and Ding, Q. T. (1986). Role of organic matter in regulating soil fertility. II. Effect of organic matter on availability of phosphorus and trace elements in soil. J. Soil Sci. China 17(Suppl.), 73–76. RWC. (2006). ‘‘A Research Strategy for Improved Livelihoods and Sustainable Rice– Wheat Cropping in the Indo-Gangetic Plains: Vision for 2006–2010 and beyond.’’

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

195

The Rice–Wheat Consortium for the Indo-Gangetic Plains. New Delhi. India. http:// www.rwc.cgiar.org/Pub_Info.asp?ID = 178. Rahman, M. A., Sufian, M. A., Saifuzzaman, M., and Chikushi, J. (2002). Nitrogen management in rice–wheat alternating cropping system and wheat genotype identification preferable to surface seeding condition. J. Fac. Agric. Kyushu Univ. 46, 295–301. Rahman, M. A., Chikushi, J., Saifizzaman, M., and Lauren, J. G. (2005). Rice straw mulching and nitrogen response of no-till wheat following rice in Bangladesh. Field Crops Res. 91, 71–81. Rahman, S. M., Haque, M. E., Ahmed, S., and Wohab Mia, M. A. (2003). Studies of organic matter turnover and nutrient buildup in a Bangladesh soil for sustainable agriculture. In ‘‘Management of Crop Residues for Sustainable Crop Production.’’ (IAEA-TECDOC-1354) pp. 131–148. Soil and Water Management & Crop Production Section, International Atomic Energy Agency. Wagramer Strasse 5, P.O. Box 100, A-1400 Vienna, Austria. Rajput, A. L. (1995). Effect of fertilizer and organic manure on rice (Oryza sativa) and their residual effect on wheat (Triticum aestivum). Indian J. Agron 40, 292–294. Rao, D. N., and Mikkelsen, D. S. (1977). Effects of acetic, propionic, and butyric acids on rice seedling growth and nutrition. Plant Soil 47, 323–334. Rathore, A. L., Pal, A. R., and Sahu, K. K. (1999). Tillage and mulching effects on water use, root growth and yield of rainfed mustard and chickpea grown after lowland rice. J. Sci Food Agric. 78, 149–161. Rothrock, C. S. (1992). Tillage systems and plant disease. Soil Sci. 154, 308–315. Sahoo, D., Rout, K. K., and Mishra, V. (1998). Effect of twenty five years of fertilizer application on productivity of rice–wheat system. In ‘‘Long-term Soil Fertility Management through Integrated Plant Nutrient Supply’’ (A. Swaru, D. D. Reddy, and R. N. Prasad, Eds.), pp. 206–214. Indian Society of Soil Science, New Delhi. Salim, M. (1995). Rice crop residue use for crop production. In ‘‘Organic Recycling in Asia and the Pacific.’’ (RAPA Bulletin Vol. 11, p. 99). Regional Office for Asia and the Pacific, FAO of the United Nations, Bangkok, Thailand. Samra, J. S., Bijay-Singh, J. S., and Kumar, K. (2003). Managing crop residues in the rice– wheat system of the Indo-Gangetic plain. In ‘‘Improving the Productivity and Sustainability of Rice–Wheat Systems: Issues and Impact’’ ( J. K. Ladha, J. E. Hill, J. M. Duxbury, R. K. Gupta, and R. J. Buresh, Eds.), pp. 173–195. ASA, Spec. Publ. 65. American Society of Agronomy, Madison, WI, USA. Sarkar, A. (1997). Energy-use patterns in sub-tropical rice–wheat cropping under short term application of crop residue and fertilizer. Agric. Ecosyst. Environ. 61, 59–67. Sarkar, S., Rathore, T. R., Sachan, R. S., and Ghildyal, B. P. (1989). Effect of wheat straw management on cation status of Tarai soils. J. Indian Soc. Soil Sci. 37, 402–404. Saunders, R. M., Mossman, A. P., Wasserman, T., and Beagle, E. C. (1980). Rice postharvest losses in developing countries. USDA Agricultural Reviews and Manuals ARM-W-12 USDA, Oakland, CA, USA. Setyanto, P., Makarim, A. K., Fagi, A. M., Wassmann, R., and Buendia, L. V. (2000). Crop management affecting methane emissions from irrigated and rainfed rice in Central Java (Indonesia). Nutr. Cycl. Agroecosys. 58, 85–93. Sharma, A. K., Babu, K. S., Sharma, R. K., and Kumar, K. (2007). Effect of tillage practices on Tilletia indica Mitra (Karnal bunt disease of wheat) in a rice-wheat rotation of the Indo-Gangetic Plains. Crop Prot. 26, 818–821. Sharma, A. R., and Mitra, B. N. (1992). Integrated nitrogen management in rice (Oryza sativa)-wheat (Triticum aestivum) cropping system. Indian J. Agric. Sci. 62, 70–72. Sharma, H. L., Modgal, S. C., and Singh, M. P. (1985). Effect of applied organic manure, crop residues and nitrogen in rice–wheat cropping system in north-western Himalayas. Himachal J. Agric. Res. 11, 63–68.

196

Bijay-Singh et al.

Sharma, H. L., Singh, C. M., and Modgal, S. C. (1987). Use of organics in rice–wheat crop sequence. Indian J. Agric. Sci. 57, 163–168. Sharma, P. K., and De Datta, S. K. (1986). Physical properties and processes of puddled rice soils. Adv. Soil Sci. 5, 139–178. Shaw, M., and Tabil, L. (2006). ‘‘Mechanical Properties of Selected Biomass Grinds.’’ American Society of Agricultural and Biological Engineers meeting presentation, Paper No. 066175. American Society of Agricultural and Biological Engineers (ASABE), St. Joseph, MI, USA. Shukla, L. N., Sidhu, H. S., and Bector, V. (2002). Design and development of loose straw thrower attachment for direct drilling machine. Abstract. In ‘‘Proceedings of the XXXVIth Annual Convention of the Indian Society of Agricultural Engineers.’’ 28–30 January 2002. Indian Institute of Technology, Kharagpur, India. Sidhu, B. S., and Beri, V. (2005). Experience with managing rice residues in intensive rice– wheat cropping system in Punjab. In ‘‘Conservation Agriculture—Status and Prospects.” (Abrol, I.P., Gupta, R. K., Malik, R. K. Eds.), pp. 55–63. Centre for Advancement of Sustainable Agriculture (CASA), New Delhi, India. Sidhu, H. S., Manpreet-Singh, H. S., Humphreys, E., Yadvinder-Singh, E., BalwinderSingh, E., Dhillon, S. S., Blackwell, J., Bector, V., Malkeet-Singh, V., and SarbjeetSingh, V. (2007). The Happy Seeder enables direct drilling of wheat into rice stubble. Aust. J. Exp. Agric. 47, 844–854. Singh, G., Jalota, S. K., and Sidhu, B. S. (2005). Soil physical and hydraulic properties in a rice–wheat cropping system in India: Effects of rice–straw management. Soil Use Manage. 21, 17–21. Singh, M., and Sharma, S. N. (2000). Effect of wheat residue management practices and nitrogen rates on productivity and nutrient uptake of rice (Oryza sativa)–wheat (Triticum aestivum) cropping system. Indian J. Agric. Sci. 70, 835–839. Streets, D. G. (2004). Black smoke in China and its climate effects. Paper presented at Special Panel on Alternative Energy Systems and Priority Environmental Issues for Asia, Asian Economic Panel Meeting October 7–8, 2004. Columbia University. www.earth. columbia.edu/events/aep/2004/documents/David_G_Streets.pdf. Streets, D. G., Yarber, K. F., Woo, J.-H., and Carmichael, G. R. (2003). Biomass burning in Asia: Annual and seasonal estimates and atmospheric emissions. Global Biogeochem. Cycles 17(4), 1099. doi:10.1029/2003GB002040. Sturtz, A. V., Carter, M. R., and Johnston, H. W. (1997). A review of plant disease, pathogen interactions and microbial antagonism under conservation tillage in temperate humid agriculture. Soil Till. Res. 41, 169–189. Surekha, K., Padma Kumari, A. P., Narayana Reddy, M., Satyanarayana, K., and Sta Cruz, P. C. (2003). Crop residue management to sustain soil fertility and irrigated rice yields. Nutr. Cycl. Agroecosys. 67, 145–154. Tanaka, A. (1974). Methods of handling of cereal crop residue crop residues in Asian countries and related problems. ASPAC Extension Bull. No. 43. Asian and Pacific Council (ASPAC), Food and Fertilizer Technology Center, Taipei, Taiwan. p. 29. Tanaka, F., Ono, S., and Hayasaka, T. (1990). Identification and evaluation of toxicity of rice root elongation inhibitors in flooded soils with added wheat straw. Soil Sci. Plant Nutr. 31, 97–103. Tang, Y., and Huang, G. (2003). Analysis of biological effect of the wheat sown through surface seeding and mulching straw. Southwest China J. Agric. Sci. 16(2), 37–41. Tang, Y., Zheng, J., Huang, G., and Du, J. (2004). Study on permanent-bed-planting with double zero tillage for rice and wheat in Sichuan Basin. Proceedings of the 4th International Crop Science Congress Brisbane, Australia. 26 Sep.–1 Oct. 2004. http://www. regional.org.au/au/cs/2004/poster/1/2/1320_tangaa.htm.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

197

Thuy, N. H. (2004). Yield trends, soil fertility changes, and indigenous nitrogen supply as affected by crop and soil management in intensive irrigated rice systems. Ph.D Dissertation University of the Philippines at Los Ban˜os, Los Ban˜os, Philippines. Thuy, N. H., Shan, Y., Bijay-Singh Wang, K., Cai, Z., Yadvinder-Singh, Z., and Buresh, R. J. (2008). Nitrogen supply in rice-based cropping systems as affected by crop residue management. Soil Sci. Soc. Am. J. 72, 514–523. Timsina, J., and Connor, D. J. (2001). Productivity and management of rice-wheat cropping systems: Issues and challenges. Field Crops Res. 69, 93–132. United States Department of Energy (US DOE). (2007). Information resources: ABC’s of biofuels. Energy Efficiency and Renewable Energy Biomass Program website. http://www1. eere.energy.gov/biomass/abcs_biofuels.html. USDA. (2007). ‘‘Production, Supply and Distribution Online.’’ United States Department of Agriculture Foreign Agricultural Service. http://www.fas.usda.gov/psdonline/ psdquery.aspx. Vamadevan, V. K., Shinde, J. E., Asthana, D. C., and Chakravarty, S. P. (1975). Management of crop residues. Int. Rice Com. Newsl. 24(1), 53–59. Verma, T. S., and Bhagat, R. M. (1992). Impact of rice straw management practices on yield, nitrogen uptake and soil properties in a wheat–rice rotation in northern India. Fert. Res. 33, 97–106. Vos, J. G. M., Uhan, T. S., and Sutarya, R. (1995). Integrated crop management of hot pepper (Capsicum spp.) under tropical lowland conditions: Effects of rice straw and plastic mulches on crop health. Crop Prot. 14, 445–452. Wang, J. Y., Wang, S. J., Chen, Y., and Zheng, J. Z. (2003a). Management of organic matter and nutrient turnover for increased, sustainable agricultural production and environmental preservation in Chinese rice fields. In ‘‘Management of Crop Residues for Sustainable Crop Production.’’ (IAEA-TECDOC-1354) pp. 207–220. Soil and Water Management & Crop Production Section, International Atomic Energy Agency, Wagramer Strasse 5, P.O. Box 100, A-1400 Vienna, Austria. Wang, Z., Wu, J., Chen, L., and Zhu, P. (2003b). Effects of direct and whole straw manuring method on increasing yield of crop and fertility of soil in rice–wheat double cropping area of Taihu Lake district. Jiangsu J. Agric. Sci. 19(3), 151–156. Wang, Z. Y., Xu, Y. C., Li, Z., Guo, Y. X., Wassmann, R., Neue, H. U., Lantin, R. S., Buendia, L. V., Ding, Y. P., and Wang, Z. Z. (2000). A four-year record of methane emissions from irrigated rice fields in the Beijing region of China. Nutr. Cycl. Agroecosys. 58, 55–63. Wassmann, R., Buendia, L. V., Lantin, R. S., Bueno, C. S., Lubigan, L. A., Umali, A., Nocon, N. N., Javellana, A. M., and Neue, H. U. (2000a). Mechanisms of crop management impact on methane emissions from rice fields in Los Ban˜os, Philippines. Nutr. Cycl. Agroecosys 58, 107–119. Wassmann, R., Lantin, R. S., Neue, H. U., Buendia, L. V., Corton, T. M., and Lu, Y. (2000b). Characterization of methane emissions from rice fields in Asia. III. Mitigation options and future research needs. Nutr. Cycl. Agroecosys. 58, 23–36. Wassmann, R., Neue, H. U., and Lantin, R. S. (2000c). Characterization of methane emissions from rice fields in Asia. 1. Comparison among field sites in five countries. Nutr. Cycl. Agroecosys. 58, 1–12. Webster, R. K., Wick, C. M., Brandon, D. M., Hall, D. H., and Bolstad, J. (1981). Epidemiology of stem rot disease of rice: Effects of burning vs. soil incorporation of rice residue. Hilgardia 49, 1–12. Wilhelm, W. W., Johnson, J. M. F., Hatfield, J. L., Voorhees, W. B., and Linden, D. R. (2004). Crop and soil productivity response to corn residue removal. Agron. J. 96, 1–17.

198

Bijay-Singh et al.

Witt, C. (2003). Fertilizer use efficiencies in irrigated rice in Asia. IFA Regional Conference for Asia and the Pacific, Cheju Island, Republic of Korea, 6–8. October 2003. www. fertilizer.org/ifa/publicat/PDF/2003_regional_cheju_witt.pdf. Witt, C., Cassman, K. G., Otto, J. C. G., and Biker, U. (1998). Soil microbial biomass and nitrogen supply in an irrigated lowland rice soil as affected by crop rotation and residue management. Biol. Fert. Soils 28, 71–80. Witt, C., Cassman, K. G., Olk, D. C., Biker, U., Liboon, S. P., Samson, M. I., and Ottow, J. C. G. (2000). Crop rotation and residue management effects on carbon sequestration, nitrogen cycling and productivity of irrigated rice systems. Plant Soil 225, 263–278. Witt, C., Buresh, R. J., Peng, S., Balasubramanian, V., and Dobermann, A. (2007). Nutrient management. In ‘‘Rice: A Practical Guide to Nutrient Management’’ (T. Fairhurst, et al., Eds.), pp. 1–45. International Plant Nutrition. (International Rice Research Institute), Los Ban˜os, Philippines. Xiao, R., Peng, W., Song, T., Wang, J., Xia, Y., and Tang, Y. (2006). Ecological regulation effects of straw mulching in tea plantation in subtropical hilly red soil region. Chinese J. Ecol. 25(5), 507–511. Xu, G. W., Wu, C. F., Liu, H., Wang, Z. Q., Zhang, M., and Yang, J. C. (2007). Effects of wheat residue incorporation and nitrogen management techniques on formation of the grain yield of rice. Acta Agron. Sinica 33(2), 284–291. Xu, Y. C., Shen, Q. R., Li, M. L., Dittert, K., and Sattelmacher, B. (2004). Effect of soil water status and mulching on N2O and CH4 emission from lowland rice field in China. Biol. Fert. Soils 39, 215–217. Yadav, R. L., and Subba Rao, A. V. M. (2001). ‘‘Atlas of Cropping Systems in India.’’ PDCSR Bulletin No. 2001–02. Project directorate for Cropping System Research, Modipuram, Modipuram, Meerut 250110, India. pp. 96. Yadvinder-Singh, Bijay-Singh, Maskina, M. S., and Meelu, O. P. (1988). Effect of organic manures, crop residues and green manure (Sesbania aculeata) on nitrogen and phosphorus transformations in a sandy loam at field capacity and under waterlogged conditions. Biol. Fert. Soils 6, 183–187. Yadvinder-Singh, Bijay-Singh, Ladha, J. K., Khind, C. S., Gupta, R. K., Meelu, O. P., and Pasuquin, E. (2004a). Long-term effects of organic inputs on yield and soil fertility in the rice–wheat rotation. Soil Sci. Soc. Am. J. 68, 845–853. Yadvinder-Singh, Bijay-Singh, Ladha, J. K., Khind, C. S., Khera, T. S., and Bueno, C. S. (2004b). Effects of residue decomposition on productivity and soil fertility in rice–wheat rotation. Soil Sci. Soc. Am. J. 68, 854–864. Yadvinder-Singh, Bijay-Singh, and Timsina, J. (2005). Crop residue management for nutrient cycling and improving soil productivity in rice-based cropping systems in the tropics. Adv. Agron. 85, 269–407. Yagi, K., and Minami, K. (1990). Effect of organic matter application on methane emission from some Japanese paddy fields. Soil Sci. Plant Nutr. 36, 599–610. Yan, X. Y., Kazuyuki, Y., Akiyama, H., and Akimoto, H. (2005). Statistical analysis of the major variables controlling methane emission from rice fields. Global Change Biol. 11, 1131–1141. Yang, P., Shen, J., and Zhang, W. (2003). Effects of the straw of rice and wheat returned to field wholly and directly on the crop yields and the physical and chemical properties of soil. Acta Agric. Shanghai 19(1), 53–57. Yevich, R., and Logan, J. A. (2002). An assessment of biofuel use and burning of agricultural wastes in the developing world. http://io.harvard.edu/chemistry/trop/publications/ yevich2002.pdf.

Crop Residue Management for Lowland Rice-Based Cropping Systems in Asia

199

Yoneyama, T., and Yoshida, T. (1977). Decomposition of rice residue in tropical soils. 3. Nitrogen mineralization and immobilization of rice residue during its decomposition in soil. Soil Sci. Plant Nutr. 23, 175–183. Yu, M., Womac, A. R., Igathinathane, C., Ayers, P. D., and Buschermohle, M. J. (2006). Switchgrass ultimate stresses at typical biomass conditions available for processing. Biomass Bioenergy 30, 214–219. Zaman, A., and Choudhuri, S. K. (1995). Water use and yield of wheat under unmulched and mulched conditions in laterite soil of the Indian sub-continent. J. Agron. Crop Sci. 175, 349–353. Zeng, M., Zhang, Y., Shan, X., and Liao, C. (2001). Options of returning straw into field of main agricultural areas in China. Soil Fert. 4, 33–36. Zeng, M., Wang, R., Pen, S., Zhang, Y., Cui, Y., Shan, X., Liao, C., and Tian, Y. (2002). Summary of returning straw into field of main agricultural areas in China. Chinese J. Soil Sci. 33, 336–339. Zeng, X., Ma, Y., and Ma, L. (2007). Utilization of straw in biomass energy in China. Renew. Sustain. Energ. Rev. 11, 976–987. Zhao, J., and Zhu, W. (2000). Effect of rapeseed straw returning on rice yield. Tillage Cultivation S1, 68. Zhao, P. R., Li, Z. H., Zheng, M. J., and Gu, J. L. (1999). Cultivation of mulched rice. Tillage Cultivation 3, 32–34. Zheng, J., Xie, H., Jiang, X., and Wu, J. (2005). Effect of straw returning and conservative cultivation on ecology and yield in hilly and double cropping region in southern China. Res. Agric. Modernization 26, 294–297. Zheng, X. H., Wang, M. X., Wang, Y. S., Shen, R. X., Li, J., Heyer, J., Kogge, M., Papen, H., Jin, J. S., and Li, L. T. (2000). Mitigation options for methane, nitrous oxide and nitric oxide emissions from agricultural ecosystems. Adv. Atmos. Sci. 17, 83–92. Zhong, H., Zhang, Y., Lin, C., and Jiang, X. (2003). Means of all straw returning without chopping and tillage and its effect on crop yield and soil fertility. Soil Fert. 3, 34–37. Zhu, P., Chang, Z., Sun, Li., and Xue, X. (2004). Effect of wheat straw returning on concentration of N and P in soil solution and rice yield. Jiangsu Agric. Sci. 6, 151–153. Zhu, Z., Qing, M., Zheng, J., Jiang, X., and Wu, J. (2005). Effect of water utilization efficiency under no tillage and straw mulching to wheat field and rapeseeds field. Southwest China J. Agric. Sci. 18(5), 565–568. Zia, M. S., Munsif, M., Aslam, M., and Gill, M. A. (1992). Integrated use of organic manures and inorganic fertilizers for the cultivation of lowland rice in Pakistan. Soil Sci. Plant Nutr. 38, 331–338. Zou, J., Huang, Y., Zong, L., Zheng, X., and Wang, Y. (2004a). Carbon dioxide, methane, and nitrous oxide emissions from a rice–wheat rotation as affected by crop residue incorporation and temperature. Adv. Atmos. Sci 21, 691–698. Zou, Y., Li, K., and Ren, Z. (2004b). Research development of direct seeding and zerotillage cultivation in multiple cropping systems. Chinese Agric. Sci. Bull. 20(1), 90–95. Zou, J., Huang, Y., and Jiang, J. (2005). A 3-year field measurement of methane and nitrous oxide emissions from rice paddies in China: Effects of water regime, crop residue, and fertilizer application. Global Biogeochem. Cycles 19, GB2021, doi: 10.1029/2004GB002401.

This page intentionally left blank

C H A P T E R

F O U R

Sampling and Measurement of Ammonia at Animal Facilities Ji-Qin Ni and Albert J. Heber Contents 1. Introduction 2. A General View of Ammonia Determination 3. Ammonia Sampling 3.1. Sampling location 3.2. Sampling time 3.3. Sample volume 3.4. Sampling methods and devices 3.5. Selection of sampling method 4. Ammonia Concentration Measurement 4.1. Characteristics of measurement techniques 4.2. Selection of measurement techniques 5. Measurement Methods and Devices 5.1. Wet methods 5.2. Gas detection tubes 5.3. Fourier transform infrared spectroscopy 5.4. Infrared gas analyzer 5.5. Ultraviolet differential optical absorption spectroscopy 5.6. Chemiluminescence analyzer 5.7. Electrochemical sensor 5.8. ChemcassetteÒ detection system 5.9. Solid-state sensor 5.10. Comparison of measurement devices 6. Ammonia Concentration Data 6.1. Advances in data collection 6.2. Data precision and bias 6.3. Error reduction 6.4. Standards for ammonia sampling and measurement

203 205 206 208 208 210 211 218 221 221 224 225 225 228 230 231 234 235 238 239 240 241 243 243 244 245 255

Agricultural and Biological Engineering Department, Purdue University, West Lafayette, Indiana 47907 Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00204-6

#

2008 Elsevier Inc. All rights reserved.

201

202

Ji-Qin Ni and Albert J. Heber

7. Summary and Conclusions Acknowledgments References

255 257 257

Scientific understanding and technical control of ammonia (NH3) at animal facilities, including animal buildings, feedlots, manure storages, and manure treatment plants, depend on reliable sampling and measurement techniques to ensure high quality data that are essential to the study and abatement of NH3 emission. This chapter focuses on the methodology and technology of NH3 sampling and measurement that has been tested or applied under field conditions since the 1960s. It draws a comprehensive and updated picture of the state of the art of NH3 concentration measurement at animal facilities. Ammonia sampling requires selection of location, time, and/or control of sample volume. Three sampling methods, the closed, point, and open-path methods, are summarized. Thirtyone measurement instruments/sensors are identified. They are categorized in nine groups and evaluated according to their technical characteristics. Field studies or applications of these instruments/sensors are reviewed and summarized. Principles, procedures, advantages, and disadvantages of various sampling and measurement techniques are discussed. An overview of data and data quality is provided. Errors resulted from calibration, sampling, measurement, and data processing are discussed. Error reduction methods are presented. Recommendations are made for selection of sampling methods and measurement devices and for future needs including development of methodologies and standards.

Abbreviations AQT, Ammonia Quick Test CL, Chemiluminescence DQIs, Data quality indicators EC, Electrochemical GMIS, Gas Manufacturers Intermediate Standards IMEC, Inter-university Micro Electronic Center, Belgium IR, Infrared MPSS, Multi-point sampling system NAEMS, National Air Emission Monitoring Study NDIR, Non-dispersive infrared NIST, National Institute of Standards and Technology NTRM, NIST Traceable Reference Materials OP-FTIR, open-path Fourier transform infrared PAS, Photoacoustic spectrophotometer PWA, Path-weighted average

Ammonia Sampling and Measurement

203

QAQC, Quality assurance and quality control TWA, Time-weighted average U.S. EPA, United States Environmental Protection Agency UV, Ultraviolet UV-DOAS, Ultraviolet differential optical absorption spectrometer

1. Introduction Ammonia (NH3) is a common substance playing an important role in the nitrogen cycle. Since the 1980s, agricultural NH3 emission has become one of the major worldwide air pollution concerns and has attracted more and more attention from the general public and government regulators. It is believed by some researchers that excessive emissions of NH3 from agriculture to the atmosphere have caused direct and indirect damage to the ecosystem in some regions with intensive animal production (Slanina, 1994; van Breemen et al., 1982). Around 75% of European NH3 emissions come from livestock production (Webb et al., 2005). Most NH3 emissions in Canada are from farm animals, and a 21% increase of NH3 emission from animal husbandry in Canada from 1990 to 1995 was estimated (Kurvits and Marta, 1998). The agricultural sector in New Zealand is also the major contributor to NH3 emissions to the atmosphere (Saggar et al., 2004). The total NH3 emission from agricultural fields in China in 1990 was estimated to be 1.80 Tg N, which accounted for 11% of the applied synthetic fertilizer N (Xing and Zhu, 2000). The European Parliament and the Council on National Emission Ceilings for certain pollutants (NEC Directive) set upper limits for each Member State for the total emissions in 2010 of four pollutants, which included NH3 (Anonymous, 2001). Research in agricultural NH3 in the United States has dramatically increased in recent years symbolized by large-scale and multi-institution projects, including the six-state research project funded by the U.S. Department of Agriculture (Heber et al., 2006; Hoff et al., 2006) and the ongoing world’s largest study: the 14-million-dollar National Air Emission Monitoring Study (NAEMS) supervised by the U.S. Environmental Protection Agency (EPA). High concentrations of NH3 inside animal houses represent potential health hazards to humans and animals (Crook et al., 1991; Kirkhorn and Garry, 2000; Portejoie et al., 2002; Reece et al., 1980). Chronic respiratory diseases of swine production facility workers have been attributed to dust and NH3 (Donham et al., 1995). The U.S. Occupational Safety and Health Administration (OSHA) has set a 15-min exposure limit for gaseous ammonia of 35 ppm by volume in the environmental air and an 8-h exposure limit

204

Ji-Qin Ni and Albert J. Heber

of 25 ppm by volume (ATSDR, 2004). There is a great need to evaluate health effects of exposures to air pollutants including toxic gases emitted into the general environment by confined animal feeding operations (Heederik et al., 2007), and this depends on sufficient and high quality measurement data. However, reliable field data of NH3 at animal facilities that include animal houses, feedlots, manure storages, and manure treatment plants are still a major need. The understanding and control of NH3 at animal facilities depend on sampling/measurement techniques, including devices, instruments, and procedures. Accurate and reliable techniques provide high quality data that are essential to research, emission abatement, and policy-making. There are some publications that summarize techniques for measuring NH3 at agricultural operations. Some of these articles provide detailed descriptions of specific measurement setups (Berckmans and Ni, 1993; Gates et al., 2005; Heber et al., 2001; van’t Klooster and Heitlager, 1992) while others presented general overviews of the topic. The first comprehensive introduction of NH3 measurement and monitoring techniques was published in 1979 (Kamin et al., 1979). There have been new and advanced techniques developed and employed in agriculture since then. Van Ouwerkerk (1993) reviewed various techniques of NH3 emission measurement at animal houses. McGinn and Janzen (1998) presented techniques to determine the loss of NH3 from manure-amended soils. Monteny and Erisman (1998) discussed NH3 measurement in dairy cow buildings including naturally ventilated barns. Phillips et al. (2001) provided a comprehensive review of techniques for NH3 emission measurement. Bolan et al. (2004) reviewed the sampling and measurement techniques applied at grazed pastures. Shah et al. (2006) presented the techniques for agricultural land and liquid surface measurement. Mosquera et al. (2005) and Losada (2007) studied NH3 measurement methods in the Netherlands. Part of the overview by Bunton et al. (2007) focused on NH3 measurement. However, much of the useful information about the methodologies and technologies associated with field sampling and measurement is still scattered in the literature and needs more careful synthesis. A thorough and updated evaluation of NH3 measurement techniques related to animal facilities will benefit users and researchers in selecting, employing, and developing such methods. The objectives of this chapter were to (1) draw an accurate picture of the state of the art of NH3 sampling/measurement methods/devices based on their tests and applications at animal facilities, (2) address technical and practical issues in their implementation, and (3) delineate future research needs in NH3 measurement methodology.

205

Ammonia Sampling and Measurement

2. A General View of Ammonia Determination Ammonia at animal facilities generally comes from open feedlots, confined animal buildings, manure storages (e.g., lagoons and storage tanks), and manure treatment sites (e.g., composts and anaerobic digesters). Confined animal buildings are typically either mechanically ventilated using fans or naturally ventilated without fans. Manure storages can be uncovered or have a physical cover to limit NH3 emission into the atmosphere. Manure treatment sites can be indoors or outdoors. The design of animal housing and methods of manure storage and manure handling reflect large differences in climate and production objectives and affect the nitrogen flow and NH3 release and emission (Sommer et al., 2006). Although agricultural NH3 emission is also related to manure application in the fields, it does not occur at animal facilities and is not within the scope of this chapter. To obtain reliable information of NH3 at animal facilities, suitable techniques must be adopted and one or more measurement variables must be chosen depending on measurement objectives. These variables include NH3 concentration, air temperature, air pressure, and air exchange rate or air speed (Fig. 1). To obtain atmospheric NH3 concentrations at animal facilities for determining human and animal exposure, measurement of concentrations at required locations is indispensable while all the other variables are optional, because they are relatively less important. Comparison of animal facilities, management, and abatement techniques and their effects on NH3 usually involves determining not only NH3 concentrations but also its emissions (Heber et al., 2000b). To obtain NH3 emissions, whether for comparison or for baseline determination, from animal buildings or manure storages, the measurement of NH3 concentration difference between the outgoing and

Measurement variables Ammonia concentration

Human/animal exposure

Air temperature

Comparison study

Air pressure

Emission baseline

Air exchange rate or air speed Indispensable

Figure 1

Measurement objectives

Dispersion modeling Optional

Determination of ammonia concentration and emission at animal facilities.

206

Ji-Qin Ni and Albert J. Heber

incoming air is essential along with the air temperature, air pressure, and air exchange rate (or ventilation rate). Techniques for measuring building air exchange rates were reviewed by Phillips et al. (2001). Measurements of NH3 concentration, air temperature, air pressure, and air or wind speed are necessary for modeling NH3 dispersion because it requires quantitative information of NH3 mass at the source and at the sink. Most NH3 concentration measuring devices provide volumetric concentrations directly. However, mass concentrations are required to calculate NH3 emissions. The volume of gas depends on both temperature and pressure and is therefore generally not constant. When converting to mass concentration, the volumetric concentration is multiplied by the molecular weight and the pressure, and divided by the gas constant and the temperature. Temperature and pressure therefore needs to be known. However, although measurement of air temperature was often reported in published works, air pressure measurements are seldom found. Atmosphere pressure varies between 980 and 1040 mbar, a 6% variation, or 3% from standard atmosphere, which was often assumed. The measurements of temperature and air pressure are relatively easy with few technical challenges. Reported sampling and measurement methodologies applied at animal facilities are summarized in Fig. 2. Details of these techniques are discussed in the following sections.

3. Ammonia Sampling Because of the large volume of air flowing through animal facilities, it is impossible to capture all the air for gas concentration determination. Therefore, air must be sampled in order to perform gas concentration measurement. The technical term ‘‘sampling’’ has different definitions in different research fields. In the case of gases at animal facilities, we define ‘‘sampling’’ as: The technique and procedure that specifies the locations where air samples are taken, controls the time, interval, frequency, and duration of sample taking, and regulates the volume or mass of the sample air to be measured.

This definition includes three elements: location, time, and volume/ mass. However, inclusion of all three factors is not always necessary depending on sampling and measurement techniques that are used. For example, in open-path sampling (Section 3.4.3), volume/mass does not need to be known because there is no need to draw air from the sampling space. Another example, in which volume/mass may not be needed, is when using passive diffusion sampling devices, such as pH test paper (Section 5.1.2) and diffusion gas tubes (Sections 5.2.2) for NH3 sampling and concentration determination.

207

Ammonia Sampling and Measurement

Sampling device Converter Collection medium Measuring device VECHTA Denuder

Aqueous acid trap Acid coating Distilled water

Sampling chamber

Open-path

A m m o n i a

Sampling chamber Stream controller

Sampling chamber Open-path Sampling chamber NH3→NO Stream converter controller Stream NH3→NO converter controller Sampling chamber

Sampling chamber

Wet chemistry Titrimetry Photometry and colorimetry Conductimetry pH paper pHydrion test strips Ammonia quick test Gas detection tubes Dräger tube Kitagawa tube Gastec tube Sensidyne tube MSA tube FTIR spectroscopy K300 M 2401 Bomen-100 Infrared analyzer PAS type 1302, 1314, 1412 MSA analyzer Miran 203 Rosemount Beckman UV-DOAS system Opsis monitor WSU system CL NOx analyzer Monitor-labs analyzer THIS, TEI, TE NOx analyzer API analyzer Electrochemical sensor Dräger sensor Quadscan gas monitor Twistik transmitter iTX multi-gas monitor Chemcassette monitor Single point monitor Solid-state sensor Solidox sensor IMEC sensor

Figure 2 An overview of techniques applied to sample and measure ammonia at animal facilities.

Specific devices are often required during NH3 sampling. Integrated systems that regulate the time and volume of air intake inside measurement instruments are not within the scope of this chapter.

208

Ji-Qin Ni and Albert J. Heber

3.1. Sampling location Spatial variations of NH3 concentration that exist at animal facilities make the selection of sampling location very important. An animal building is a ventilated but imperfectly mixed air space with nonuniform sources, resulting in temperature and concentration gradients. Although ventilation in the building creates air mixing, it can also increase the spatial concentration variations in situations when it dilutes NH3 at some locations more than at others. At uncovered manure storages and lagoons, in addition to the less than uniform source across the surface, localized NH3 concentrations depend to a great extent on weather conditions especially the wind speed, wind direction, and temperature. Large concentration gradients exist at these areas when there is a poor dispersion due to slow air movement. Consequently, different sampling locations may result in wide variations in measurement data because of spatial NH3 differences. Sampling at particular locations is needed if the NH3 concentrations at these locations are to be obtained. The number of sampling locations determines the spatial resolution of the NH3 concentration profile. The more the sampling locations are, the better the spatial resolution of the data. Unfortunately, in practical situations, the ideal number of sampling locations is often compromised due to the limitation of budget, equipment, time, and manpower. Many recent long-term emission research projects at large animal buildings with continuous measurement equipment can only afford about a dozen shared sampling locations at a farm. Therefore, choosing the most representative sampling locations has great statistical importance. Selection of sampling locations depends on the measurement objectives. If the study is to determine human or animal exposure to NH3, the sampling locations should be at the human or animal breathing zones. If the primary objective of sampling is to determine emissions, the sampling locations should be located to represent not only the ventilation exhaust air but also the incoming air to account for background concentrations. The sampling locations were at the building air inlets and ventilation fan exhausts for mechanically ventilated livestock buildings (Hartung et al., 1997; Heber et al., 2001), at the ridge exhaust of a naturally ventilated dairy buildings (Monteny et al., 2002; Snell et al., 2003), and upwind and downwind of an uncovered manure storage (Sommer et al., 2004a). Sampling at both inlet/ upwind and exhaust/downwind is to subtract background NH3 concentrations and obtain the net emission or generation from these facilities. For NH3 dispersion modeling, sufficient samples should be taken to represent a large area downwind as well as the NH3 generation sources.

3.2. Sampling time Temporal variation of NH3 concentration at animal facilities is the reason that time of sampling is critical for obtaining dependable data. Results of NH3 concentrations acquired at specific times usually are only valid for that

209

Ammonia Sampling and Measurement

20

250

16

200

12

150

8

100

4

50

0

Ventilation, 1000 m3/h

Concentration, ppm

time. Extended reference of these concentrations could introduce errors. For example, NH3 concentrations in animal buildings measured in winter are not suitable to represent those in summer because of seasonal differences. Seasonal temperature change is the most important factor influencing seasonal NH3 concentration variations. The difference between summer and winter daily mean temperatures can be significant in agricultural regions. Temperature plays an important role in NH3 release from liquid manure because it determines the convective mass transfer coefficient (Ni, 1999). In open animal feedlots and manure storages/lagoons, higher temperature means faster NH3 release from liquid manure and hence higher NH3 concentrations above the manure surface. In confined animal buildings, higher ambient temperature in summer requires more ventilation to reduce animal heat stress. Daily mean ventilation rates in summer can be more than 20 times than those in winter in mechanically ventilated animal buildings. Seasonal variations of temperature and ventilation at animal houses change not only the NH3 concentrations but also the NH3 emission rates (Harper et al., 2004). Seasonal wind velocity and direction changes have greater effects on NH3 concentration in open facilities than in buildings. In addition, precipitation and solar radiation can quickly change the solid content in manure in open fields, consequently changing the NH3 release rate from manure. Diurnal variations of NH3 concentration at animal facilities are also caused by weather conditions. In fair weather, the recorded diurnal temperature varied as much as 15 C on a swine farm in Midwest, USA, and the temperature-based ventilation in the afternoon was about 5 times as high as in early morning (Fig. 3). The diurnal NH3 concentrations in the same

0 0

3

6

9 12 15 18 Time, hour of day

WF

PF

PH

21

24

V

Figure 3 A 24-h record of temporal and spatial variations of ammonia concentrations at wall fans (WF), pit fans (PF), pit headspace (PH), and ventilation rate (V) in a 66 m (L) 13 m (W) mechanically ventilated swine finish building. Source: Ni et al. (2000b).

210

Ji-Qin Ni and Albert J. Heber

study closely followed the diurnal variation of ventilation rates (Ni et al., 2000b). Moreover, certain daily animal activities affected the NH3 concentrations, for example, pig urination was followed by NH3 concentration peaks (Aarnink et al., 1993a). It is clear that low frequency sampling results in poor representation of the true NH3 fluctuation patterns and therefore unreliable mean NH3 concentrations. Measurements of varying concentrations that cover excessively short periods may produce data with serious temporal limitations. To depict an accurate picture of NH3 at animal facilities, its temporal variations should be resolved by selecting sufficiently high sampling frequency. The higher the sampling frequency is, the better the data resolution and the more accurate the mean value. According to the Nyquist Theorem, the sampling frequency should be at least twice the maximum frequency of the signal that is being sampled (Finkelstein and Grattan, 1994). However, high frequency sampling is subject to some technical restrictions especially the response time of the measuring device. The temporal variations of NH3 concentrations shown in Fig. 3 demonstrate that it is important to select proper sampling time. Sampling time should be arranged to cover peak and valley concentrations during the day, especially when there are significant temperature and airflow rate fluctuations, to obtain daily mean concentration, whether short-duration sampling (e.g., active gas detection tube) or long-duration sampling (e.g., passive gas tube or wet chemistry) techniques are used. Based on the same principle, sampling should be designed to cover the low concentration season (usually summer) and high concentration season (usually winter) if an annual mean concentration or emission rate is to be obtained.

3.3. Sample volume Accurate control of air sample volume is indispensable for some measurement techniques such as active gas tubes and wet chemistry with acid traps, in which the chemical reaction depends on the total mass of NH3 in the sample. In order to get volume-specific concentration, the volume of the air sample in the chemical reaction must be accurately known. A special sampling/control device is often needed to control sample volume. For active gas tubes, a handheld pump provided by the gas tube manufacturer is employed (Section 5.2.1). For wet chemistry, several different air volume control devices with different sample volumes and sampling durations were reported in the literature. Pfeiffer et al. (1993) and Krieger et al. (1993) adopted an air collection system called VECHTA to control sampling volume. The system continuously provided 1.0 l/min of sample air to an acid collection medium. In Guiziou and Beline (2005), 5 l/min sampling air was flown through acid trap for 2–3 days.

211

Ammonia Sampling and Measurement

3.4. Sampling methods and devices Three main sampling methods are currently applied in field tests. They are closed, point, and open-path methods (Fig. 4). The differences among the three methods are the spatial coverage by the sampling devices. The closed method collects samples from an enclosed surface area. The point method and the open-path method target air at certain points and in a narrow optical path within a three-dimensional zone, respectively. Depending on the sampling devices, the point sampling method can also be divided as two sub-methods: the passive exposure method and the active extraction method, which can be localized or centralized. 3.4.1. Closed sampling The closed sampling method involves a physical enclosure or chamber to create a limited headspace over a selected piece of NH3 release surface. The ‘‘static’’ chamber (Sommer et al., 2004b) does not have air exchange between the outside and inside of the chamber and has thus far only been used in investigations of NH3 release from soil. The sampling chambers discussed in this chapter are all ‘‘dynamic.’’ They have an open bottom and are equipped with one or more air inlets and one or more outlets. The chamber is placed on the floors of animal buildings or on the surfaces of liquid or solid manure that releases NH3 thus isolating the release surface from its surroundings. The samples are drawn from the inlet and outlet of the chamber (Fig. 5). Sampling method

Open-path

Point

Closed Exposure

Extraction Localized

Figure 4

Centralized

Ammonia sampling methods at animal facilities.

Flow controller

Stirring fan

Sampling chamber

Zero-air Ambient air Filtered air

In situ analysis Outlet air Lab analysis

Inlet air Ammonia release surface

Figure 5

Schematic of a closed method for ammonia sampling.

212

Ji-Qin Ni and Albert J. Heber

Sampling chambers have been given various names, for example, Lindvall box (Scholtens, 1990), dynamic chamber (Elwinger and Svensson, 1996; Loyon et al., 2007), wind tunnel (Misselbrook et al., 1998), flux chamber (Heber et al., 2000a), measuring chamber (Andersson, 1996), floating chamber (Willers et al., 1996), and emission hood (Chadwick, 2005). The earliest sampling chamber was the Lindvall Box (1.5 1.0 0.4 m3) developed in Sweden for odor analysis, during which cleaned air was blown through the box and brought back to a mobile laboratory for measurement (Lindvall et al., 1974). The Lindvall Box was later used for NH3 sampling in animal building by Scholtens (1990), Kant et al. (1992), Elzing and Swierstra (1993), and Kroodsma et al. (1993). The volumes of the sampling chambers were between 0.02 and 108 m3, and most of them had box-like shapes. The largest chamber covering area was 160 m2 in a pig house converted from manure culverts (Andersson, 1995). A special design, a movable sampling chamber attached to a steel trolley that could be rolled back and forth along a steel runner system was reported by Chadwick (2005) to sample NH3 from beef cattle farmyard manure. Airflow through the chamber was supplied by a fan (installed in the inlet or outlet) or a compressed air cylinder, hence sometimes emphasized as ‘‘dynamic.’’ Most tests involved blowing ambient air, filtered (Heber et al., 2000a) or unfiltered (Svensson et al., 1997), through the chamber. A few tests blew compressed zero-air (Aneja et al., 2000). In one study, the fan was in the chamber air exhaust ( Jeppsson, 1999). Some projects used stirring fans inside the chamber to mix air (Aneja et al., 2000; Ferguson et al., 1997; Jeppsson, 1999; Svensson et al., 1997). A comparison of 14 reported sampling chambers is given in Table 1. Many of these sampling chambers were used in animal houses. Elwinger and Svensson (1996) described a battery-powered chamber, developed by Svensson and Ferm (1993), in an experimental broiler house. Ferguson et al. (1997) developed an isolated 19-liter container laid over a portion of the litter in a broiler room. Brewer and Costello (1999) used a Plexiglas chamber to cover the litter in a commercial broiler house. Misselbrook et al. (1998) conducted sampling on concrete floors in a commercial dairy yard. In a study by Blanes-Vidal et al. (2007), a 30-cm diameter stainless steel chamber was constructed with circulation air delivered via nine inlet holes into the chamber for dairy manure emission in an experimental test room. The NH3 concentration was measured with an Innova Gas Monitor. Sampling chambers were also used in manure composting sites. Amon et al. (1997) described two large-scale chambers developed in Austria to collect emission data from aerobically composted and anaerobically stored solid manure. Osada and Fukumoto (2001) developed a 13-m3 chamber for monitoring a composting mixture.

Table 1

Summary of sampling chambers

Chamber name

Emission source

Stir fan

Cleaned

Animal houses

No

0.40 0.30 0.18 m3

Ambient

No

Box Not specified

Length 5.4 m Volume 19 liter

Ambient None

Broiler house Pig house Broiler litter

Chamber

Box

1.22 0.76 0.41 m3

Dynamic chamber Open dynamic chamber Wind tunnel

Box

Bottom 0.25 m2

Ambient (0.2 m/s) Ambient

Broiler litter Pig house

Box

6 3 2 m3 and 9 6 2 m3

Ambient (1–11 m3/h)

Not specified

2 0.5 m2

Ambient (1.0 m/s)

Shape

Dimension

Lindvall box

Box

1.5 1.0 0.4 m

Dynamic chamber Chamber Container

Box

Incoming air 3

Note

Yes

Elzing and Swierstra (1993), Kant et al. (1992), Kroodsma et al. (1993), Scholtens (1990) Elwinger and Svensson (1996) Andersson (1995) Ferguson et al. (1997). Shape of the chamber was not described Brewer and Costello (1999) Svensson et al. (1997)

Treated manure

No

Amon et al. (1997)

Concrete floor

No

Misselbrook et al. (1998). Shape and height of the wind tunnel was not given by the authors

No Yes

No

(continued)

Table 1

(continued)

Chamber name

Shape

Dimension

Incoming air

Emission source

Stir fan

Hood

Box

Open face: 0.9 m2

Dairy farm

No

1.22 0.61 0.25 m3

Filtered (1.0 m/s) Filtered (1.1 m/s)

Lagoon

No

Misselbrook et al. (1998) Heber et al. (2000a)

Buoyant convective flux chamber Dynamic chamber Emission hood Flux chamber

Box

Cylinder

Zero-air (2.4– 4.7 l/min) Not specified (130 l/s) Circulation air

Lagoon

Yes

Aneja et al. (2000)

Not specified Cylinder

0.27 m (d) 0.42 m (h) 1.15 m (h), volume 7 m3 30 cm

Farmyard manure Dairy farm

No

Chadwick (2005)

No

Box

0.4 0.4 0.6 m3

Clean air (1.5 m3/h)

Manure treatment

Blanes-Vidal et al. (2007). Only partial cylinder dimension was provided Loyon et al. (2007)

Dynamic chamber

No

Note

Ammonia Sampling and Measurement

215

The buoyant sampling chamber was a tool for NH3 monitoring in lagoons. Heber et al. (2000a) and Lim et al. (2003) developed and employed a buoyant chamber using 1.0 m/s air speed in the chamber. It floated ~0.17 m above the water and covered 0.74 m2 of swine lagoon surface. Aneja et al. (2000) and Bajwa et al. (2006) obtained NH3 data from lagoons using a cylinder (27-cm diameter, 42-cm height) with a Teflon-coated wall and a stirring fan. 3.4.2. Point sampling Point sampling is the method in which samples are taken at a selected single point or at multiple points at animal facilities. Unlike closed sampling, the sampling location of this method can be at different heights from the NH3 release surface or at the air inlet/exhaust of a building. It can also be used to sample upwind or downwind ambient air around the facility. There are two different ways of point sampling: exposure and extraction. Exposure sampling uses passive sampling devices or sensors and therefore does not require sampling pumps. It can be a simple procedure when using measuring devices, such as detection tubes, where sample air is diffused to passive NH3 samplers/sensors for obtaining a small number of time weighted average (TWA) concentration data. The micrometeorological technique is a point-sampling method and has been used to determine NH3 concentrations at lagoons (Zahn et al., 2001), on animal farms (Cassel et al., 2005a,b), and at a manure storage (Sommer et al., 2004a). In micrometeorological sampling, multiple passive samplers are set up at different height in towers or poles located upwind and downwind of the source. The samplers provide TWA NH3 concentrations. This method is usually applied to large area sources. Extraction sampling employs a pump or pumps to pull air from sample locations. Localized extraction uses a pump and sampler at the sampling location, for example, sucking air with a handheld pump to a gas tube (Section 5.2.1), or with an electrical pump to an acid trap (Section 5.1.1). It does not engage transporting the sample air long distances to a measurement location. Centralized extraction is a more complex method for multi-point sampling. It extracts and transports sample air from different points to a single set of measurement devices, which can be as far as several hundred meters away from the sampling points. Sample air is transported continuously via tubing, usually TeflonÒ tubing (Ogink and Kroodsma, 1996). The tubing may require insulation or heating to avoid condensation inside the tubing when it passes through zone that is cooler than the dew point of the sample location (Heber et al., 2006; Ogink and Kroodsma, 1996). The instruments used in the system measure the NH3 concentration in the sample air from one sampling point for a certain duration, for example, 10 min, before switching to air from another point. A multi-point sampling system

216

Ji-Qin Ni and Albert J. Heber

(MPSS), usually computer controlled, is needed to regulate the time and location of sampling. Figure 6 presents schematics of three commonly used MPSSs. The system on top of the figure is a negative pressure system. It has a vacuum sampling pump that applies vacuum pressure to a stream controller and tubing network. Sample air is withdrawn into the pump and then delivered to the gas measurement instrument. This system requires that the stream controller and the tubing network must be airtight. It was first reported in Europe in the early 1990s. An air stream controller placed between several NH3 ! NO converters and an NOx analyzer was described by van Ouwerkerk (1993). Berckmans and Ni (1993) employed a similar system that regulated air sample streams, 2 min for each of six converters. This system requires multiple converters, one at each sampling point, but has fast response because it avoids the transportation of the ‘‘sticky’’ NH3 in the long tubing between the converter and the analyzer. In the study by Ogink and Kroodsma (1996), air samples of each sampling location were analyzed over 1 min at intervals of 5 min. The system shown in the middle of Fig. 6 is also a negative pressure system. It was used in the United States (Heber et al., 2001). Heber et al. (2006) described a gas sampling system, which included a set of solenoids, NH3 → NO converter

Point 1 n

.. .

1

n

.. .

n

.. .

1 n

.. .

To exhaust Analyzer

.. .

Pump Manifold

Stream controller

Heated tubing

NOx Analyzer

Pump

.. . Pump

To exhaust

Stream controller

Heated tubing

Point 1

Stream controller

.. .

n

Point 1

Heated tubing

1 n

To exhaust Analyzer 1

.. .

Manifold

n

Figure 6 Schematics of three MPSS. Top, negative pressure system with multiple NH3 ! NO converters and single NOx analyzer. Middle, negative pressure system with multiple analyzers. Bottom, positive pressure system with multiple analyzers.

217

Ammonia Sampling and Measurement

manifolds, pumps, mass flow meter, and vacuum pressure transducer in a semi-enclosure. It was used to sample and regulate air streams from multiple sampling locations, each lasting for 10 min, to a set of analytical instruments, including a single NH3 ! NO converter that was followed by an NH3 analyzer and other gas analyzers for hydrogen sulfide (H2S) and carbon dioxide (CO2) measurement. The system at the bottom of Fig. 6 is a positive pressure system that uses one air compressor at each sampling location to pump air to the measurement instrument (Hoff, 2005; Zhang et al., 2005). This system avoids most of the air leakage problem but requires multiple air pumps. 3.4.3. Open-path sampling Open-path sampling uses optical detection devices, which consist of an emitter telescope and a receiver/detector. The source light from the emitter, ultraviolet (UV) or infrared (IR), is beamed in one direction over a certain distance (hence an open path), which contains gaseous NH3, to the receiver/ detector (Fig. 7). The open-path technique provides path-weighted average (PWA) gas concentrations of samples taken in a one-dimensional path within a three-dimensional zone. Depending on the technology, the length of the open path between the light source and receiver can be 100–150 m (Secrest, 2001a), 500 m (Amon et al., 1997), and up to 750 m (Mount et al., 2001). The source and the receiver can be at two ends of the open path. The detection path is between the source and the receiver. They can also be at the same end with a reflection mirror at the other end of the open path. In this case, the light source beamed to the distant mirror reflects back to the receiver. The path is between the mirror and the source/receiver. Open-path sampling for NH3 concentration was conducted downwind of an animal facility for assessment of neighbor exposure to NH3 concentrations. Secrest (2001a) oriented the monitoring paths east–west when the prevailing winds were from the south at 0.8 km north of a cluster of eight swine barns in Missouri using an ultraviolet differential optical absorption Light source Open-path Detector Ammonia source

Figure 7 Schematic of open-path sampling. According to Amon et al. (1997).

218

Ji-Qin Ni and Albert J. Heber

spectrometer (UV-DOAS) and a Fourier transform interferometer. Harris et al. (2001) employed a single-beam open-path Fourier transform infrared (OP-FTIR) system along several paths to measure all the exhaust plumes from nine finishing swine barns in North Carolina. The OP-FTIR beam passed through the fan plumes 1 m from the fan outlets. This technique was also used to measure anaerobic lagoon NH3 emissions by Shores et al. (2005).

3.5. Selection of sampling method The selection of a sampling method depends on several factors, including cost, sampling objective, and technical pros and cons of the method. Financial resource is always one of the most important factors in choosing a sampling method. Sampling methods differ in initial cost, which includes equipment cost, setup cost, and operation cost. Cost of sampling equipment and setup can vary greatly depending on the scale of the study. To reduce the equipment and setup cost, many research institutions tend to take advantage of existing equipment, laboratory facilities, and technical expertise. The cost per sample depends on the measurement method. Wet chemistry (Section 5.1) and gas tubes (Section 5.2) can have high cost per sample although their initial sampling cost is relatively low compared with some other measurement devices, for example, chemiluminescence (CL) gas analyzers (Section 5.6), which offer continuous measurement hence large numbers of sample concentration data. Sampling objectives may vary from defining NH3 exposure to human and animals, determining NH3 emission rate, comparing facilities or abatement techniques, to modeling NH3 dispersion. Some studies may need only a few samples to get a quick and preliminary result. Others may require large number of samples taken at different locations and cover a long period of time to fulfill the objective. Technical characteristics of the methods are reflected in their intrusiveness, controllability, source isolation, sampling area, measurement instrument sharing, and automation. Intrusion of the sampling equipment may alter airflow patterns and gas concentration gradients that would naturally occur at the sampling location. Controllable sampling allows simulation of the animal facility to sample NH3 concentrations at different physical conditions, for example, air speed (Lim et al., 2003) and manure temperature (Andersson, 1996). Animal facilities differ in size and complexity. Many modern animal facilities include multiple large animal buildings, manure storages, and processing plants. Some research institution- or universityowned experimental facilities and traditional family farms are relatively small. Large size complexes have multiple NH3 sources that are difficult to isolate if only some components are of research interest. Sampling area refers to the size of area at the animal facility for which the method is

Ammonia Sampling and Measurement

219

suitable to use. Measurement instrument sharing allows a single set of instruments to measure NH3 concentrations sampled at multiple locations. It reduces the cost of instruments as well as the errors among the instruments if multiple instruments are used. Automation is necessary for comprehensive projects to obtain large number of samples over long periods. Table 2 compares various aspects of the three sampling methods. The cost of each method in the table is relative to each other and is split into equipment and setup. The per-sample cost is not listed because it does not only depend on the sampling method. The equipment and setup costs of the closed method are low (
Table 2

Comparison of different ammonia sampling methods Closed sampling

Point sampling

Open-path sampling

Cost Equipment

Low

Medium to very high

Setup

Medium

Very low for exposure method. Medium for localized extraction method. High for centralized extraction method. Very low to high for exposure and localized extraction methods. Very high for centralized extraction method.

Treatment comparison and surface release Small release surface

Animal and human exposure, baseline emission, treatment comparison, and dispersion modeling Flexible

Human ambient exposure, baseline emission, treatment comparison, and dispersion modeling Large

Intrusive Controllable for airflow at release surface Best Yes

Little intrusive Controllable for sampling flow

Nonintrusive Not controllable

Better Yes in centralized extraction method

Poor Yes for scanning system

Technical aspect Study objective

Size of sampling area Intrusiveness Controllability

Source isolation Instrument sharing

Low to medium

Ammonia Sampling and Measurement

221

Open-path sampling has a relatively shorter history than point sampling for agricultural NH3 study. Its sampling equipment and setup cost can be from low to medium depending on the complexity of the research objectives. An obvious advantage of this method is that it is not intrusive to the system being measured. There is also no adsorption of NH3 on sample transporting system (e.g., tubing and fitting). Large areas can be investigated and the detection limit is very low. Disadvantages of the method lie in the determination of emission rates by the inverse dispersion models. Weather conditions must be known during the measurement period. Different emission sources lying close to each other cannot be distinguished from each other (Amon et al., 1997). It is also not easy to use for animal exposure study inside the barns.

4. Ammonia Concentration Measurement 4.1. Characteristics of measurement techniques The first reported measurements of NH3 concentrations at animal facilities were in the 1960s using wet chemistry (Moum et al., 1969; Valentine, 1964). With the advancement of sensor and analytical technology, more and more techniques were developed. Thus far, nine groups of technologies have been introduced and used at animal facilities. They are summarized in Fig. 2, which also illustrates the sampling methods that these measurement devices were reportedly used with. The devices and setups were chosen to meet various objectives, different technical requirements, and budget constraints based on their characteristics, which are categorized in Table 3 and discussed below. 4.1.1. Wet or dry According to Kamin et al. (1979), analytical methods of NH3 can be categorized as ‘‘wet methods,’’ which use aqueous media, and ‘‘dry methods,’’ which is a direct analysis of NH3 in the gas phase. 4.1.2. Active or passive An active measurement device needs a pump, whether hand or electric powered, to provide controlled sample air flowing to the device. A passive measurement device does not require a pump. It allows air to diffuse into the sensor. Passive measurement devices need to be placed right at the sampling location during measurement. Passive techniques that depend on diffusion take longer to finish a measurement. Because of this, it can provide TWA concentration with a single point measurement. Some active methods (e.g., wet chemistry) only provide TWA concentrations.

Table 3 Characteristics of some reviewed ammonia measurement devices and methods

a

Method

Wet or dry

Active or passivea

Sensitivityb

Readout

Sensor lifec

Response timed

Coste

Standardized wet chemistry pH paper/test strip Denuder and wet chemistry Active gas tubes Passive gas tubes Chemcassette EC sensors NOx analyzers FTIR spectroscopy Rosemount NDIRg PASh Chillgard MSA analyzer UV-DOAS Solid-state sensor

Wet Wet Wet Dry Dry Dry Dry Dry Dry Dry Dry Dry Dry Dry

A P P A P A A; P A P A A A P P

0.01–1 mg/l ppm ppb 0.5 ppm ppm ppm ppm 1 ppb ppb ppm 0.01–1 ppm 1 ppm ppb ppm

Indirect Direct Indirect Direct Direct Direct Direct Direct Direct Direct Direct Direct Direct Direct

S S S S S S M M M M M M M M

h s h m h s m m m s s– mi m s s

Lf VL L L L H H VH VH H–VH VH H H M

A, active, P, passive. Expressed in concentration level. c S, single use, M, multiple use. d Response time in the ranges of: h = hours, m = minutes, s = seconds. e Costs are approximate and may vary depending on time and location of purchase, providers, and accessories. Equipment cost only, not including maintenance and operation. VL, very low (US $15,000/device). f Analytical instrument used to analyze the collection medium is not counted. g RosemountÒ nondispersive infrared analyzer. h Bru¨el & Kjær PAS and Innova photoacoustic gas monitor. i Depending on the number of gases to measure and the instrument configuration. EC, electrochemical; FTIR, Fourier transform infrared; MSA, Mine Safety Appliances; NDIR, nondispersive infrared; PAS, photoacoustic spectrophotometer; UVDOAS, ultraviolet differential optical absorption spectrometer. b

Ammonia Sampling and Measurement

223

4.1.3. Sensitivity Sensitivity is the capability of a measuring device to discriminate between measurement responses representing different levels of a variable of interest. Sensitivity is determined from the value of the standard deviation at the concentration level of interest. It represents the minimum difference in concentration that can be distinguished between two samples with a high degree of confidence (USEPA, 1998). Some techniques were reported to be highly sensitive, for example, the method of converting NH3 to NO followed by NOx analysis has lower detectable limit of 1 ppb NH3. Others provide sensitivity at ppm reading, for example, NH3 detection tubes. Sometimes high sensitivity techniques are called analytical techniques, which provide quantitative data, and low sensitivity ones are called detection techniques, which provide qualitative or semi-quantitative data. 4.1.4. Readout Direct readout is an important feature, especially for field measurements. Techniques with direct readouts provide an immediate visual display of NH3 concentration right after the measurement is completed. Most of the techniques reviewed in this study provide direct readout. Some of them are followed by automatic data retrieving and calculation. Techniques with indirect readouts require a chain of procedures and devices, for example, trapping NH3 in acid collection medium followed by laboratory analysis of the medium with wet chemistry methods. Compared with the direct readout, the indirect readout method takes more time to obtain results and is less suitable for large numbers of measurements. 4.1.5. Sensor life The sensor in an NH3 measuring device is the material or part that undergoes a physical or chemical change when exposed to sample air. Single measurement techniques adopt disposable sensor materials that cannot be reused. The gas detection tube is a typical single use sensor. In wet chemistry methods, the NH3 collection medium is used only once. Continuous measurement techniques can provide many concentration readings over time, usually in the form of electronic signals that can be easily recorded and processed. Sensor materials for these devices usually do not need to be replaced, with the exception of EC sensors (e.g., Dra¨ger unit that has a sensor life of 18 months). The cassette in the Chemcassette gas monitor is a single-use sensor material, although one cassette can provide multiple readings. 4.1.6. Response Response is a measure to evaluate how quickly or slowly a measuring system can react to NH3 and present correct concentration readings. Manufacturers use response time, time needed for an instrument to reach from

224

Ji-Qin Ni and Albert J. Heber

0 to 90% (t90) at zero to span difference in gas concentrations, to describe the instrument specifications. The response of passive diffusion sampling can be as long as 8 –24 h. 4.1.7. Cost Costs of the reviewed techniques vary widely and also differ with time and geographical locations. The lowest cost of single measurement is
4.2. Selection of measurement techniques Selection of measurement techniques should be based on research objectives, coupled with the capabilities of the research institution and the research budget constraints. Cost of the techniques is one of the most important factors to be considered in almost all research projects. Thus, capital and operating costs may need to be assessed with the performance of the technique. Standard wet chemistry requires analytical instruments that may already exist at many institutions. When the capital investment of analytical instruments does not need to be considered, wet chemistry methods are inexpensive and affordable. They are especially useful for small numbers of samples. The pH-paper-based test kits are appropriate for obtaining mere indications of in-building NH3 concentrations when accuracy is not a high priority. However, small numbers of samples and short-term measurements cannot satisfy the quality requirements in many field investigations. Techniques that produce large amounts of data should therefore be considered. The cost per sample using high-priced instruments with multiuse sensors may be less expensive than using low-price single-use sensors. Nevertheless, some expensive instruments, like IR analyzers and NH3 analyzers, are usually only used at institutions conducting intensive research on agricultural NH3. Applications of high and low sensitivity measuring devices are generally related to indoor and outdoor NH3 measurements, respectively. Indoor NH3 concentrations at animal facilities are usually at least several ppm, and almost all the techniques reviewed in this study are compatible. Because outdoor NH3 may be at ppb level, some sensors, for example, gas detection tubes, are not appropriate due to their low sensitivity and high minimum detection limit. Measuring devices with quick responses, for example, having a t90 less than 2 min, are required to properly study the dynamic behavior and diurnal

Ammonia Sampling and Measurement

225

variations of NH3 concentrations. Sensors with long response times, for example, passive gas tubes, are good only when TWA data are needed. There appears to be an increasing use of dry methods with direct readouts in recent years. Some of these methods can handle large numbers of measurements with immediate results, provide high frequency and high precision measurements, and enable computer-aided data logging. The successful application of these techniques advanced the characterization, quantification, and modeling of agricultural NH3 emission significantly, for example, by establishing NH3 emission factors for different sources and by revealing dynamic behaviors of NH3 release from various components of animal production systems.

5. Measurement Methods and Devices 5.1. Wet methods 5.1.1. Standardized wet methods Most wet methods are standardized methods (Table 4) that rely on collecting gaseous NH3 into a suitable acid solution (acid trap or scrubber) and then performing concentration determination in the laboratory (Fig. 8). Wet methods can be active or passive. In the active method, air is pumped through the acid solution during sampling and the volume of air passing through the solution is recorded (Nicholson et al., 2004). The NH3 concentration in the air is calculated based on the volume. This method has a long history in NH3 measurement at animal facilities started in the 1960s (Valentine, 1964) and is still in use (Guiziou and Beline, 2005; Loyon et al., 2007; Shah et al., 2006). In the passive method, NH3 in air is diffused into the acid and no pump is required. The most commonly used acid traps for measuring NH3 at animal facilities include boric acid (Curtis et al., 1975), orthophosphoric acid (Kay and Lee, 1997; Misselbrook et al., 1998; Nicholson et al., 2004), nitric acid (Willers et al., 1996), and sulfuric acid (Guingand, 1997, #1681; Guiziou, 2005, #2272; Jiang and Sands, 2000; Krieger et al., 1993; Loyon et al., 2007; Pfeiffer et al., 1993; Valli et al., 1991). Three general methods of colorimetric techniques have accounted for the majority of the determinations of NH3 concentration in the acid trap solutions. They are Nessler, indophenol, and pyridine–pyrazolone techniques; each of them has modifications and adaptations. However, only the first two, the Nessler and indophenol techniques, have been applied to the NH3 measurement at animal facilities. Photometry constitutes one of the most important methods in air analysis. The air samples must be converted into colored compounds that are then determined. There are specific and very sensitive color reactions

226

Ji-Qin Ni and Albert J. Heber

Table 4 Standardized wet methods used for measuring ammonia in acid traps Methods

Sensitivity

Comments

References

ColorimetryNessler

0.02 mg/l

Hashimoto (1972), Kamin et al. (1979), Kay et al. (1992), Valentine (1964), Valli et al. (1991)

Colorimetryindophenol

0.01 mg/l

Photometry

NA

Traditional method, widely used in past; numerous interferences by other compounds, including aldehydes, sulfur dioxide, amines, and metals; prepurification by distillation often recommended Widely used; adapted for automated analysis; less sensitive to interference than Nessler method; pH dependent Relatively simple but cannot handle large amounts of samples

PhotometryNA Nessler Conductimetry 0.1 mg/l

Titrimetry

1 mg/l

NA Potential interference from other redox species All acids and bases interfere

Kamin et al. (1979), Hesse (1994), Hoy et al. (1994)

Leithe (1971), Schmidt-Van Riel (1991), Wang et al. (1991) Jiang and Sands (2000) Kamin et al. (1979), Mannebeck and Oldenburg (1991) Curtis et al. (1975), Kamin et al. (1979), Verstegen et al. (1976)

Notes: Some references cited in this table provided technical information of the methods but did not report NH3 measurement in animal houses. NA: not available from the cited references.

227

Ammonia Sampling and Measurement

Sample air

Flow meter

Pump

Exhaust Flow rate and time control

Acid solution

To concentration determination

Figure 8 Schematic of sampling and analysis with wet chemistry.

available and the time required to perform a comparatively accurate measurement is short. Jiang and Sands (2000) analyzed NH3 concentration in broiler buildings using the Nessler method and a spectrophotometer. Noncolorimetric wet-chemistry techniques used at animal houses include acid conductimetry (Mannebeck and Oldenburg, 1991) and titrimetry (Curtis et al., 1975; Verstegen et al., 1976). Generally these techniques tend to be less sensitive than the colorimetric method and are subject to a host of interferences (Kamin et al., 1979). 5.1.2. pH test paper The method of pH test paper is based on the fact that NH3 is readily dissolvable in water and converted into the ammonium cation (NH4+) thus changes the pH of the water. It is so far the simplest and least expensive way to obtain discrete ammonia measurements. Moum et al. (1969) developed a very simple method by employing pH test paper and neutral distilled water as an NH3 trap. The measuring range was 0–100 ppm and the resolution was 5 ppm. One follow-up use was tested by Seltzer et al. (1969). The method was inexpensive and provided direct in situ readout. However, it had low sensitivity and precision. Using a commercialized version of this method, Dewey et al. (2000) tested pHydrionTM NH3 test strips (Micro Essential Laboratory, Brooklyn, NY) in a swine unit in 1994. The pHydrion cost only US$0.06 per test. It involved placing a drop of distilled water on a paper test strip, waving the strip in the air for one minute, and estimating NH3 concentrations by matching the color change with a calibrated color chart. A similar product, Ammonia Quick Test (AQT), distributed by Vineland Laboratories (Vineland, NJ), has a measurement range of 0–100 ppm NH3. An inch of the strip is tore off a roll of the test paper, wetted with 1 or 2 drops of distilled water, shaken off of excess water, and exposed for 15 s in sampling air. The paper strip changes color and is compared with the color chart to get the concentration reading. Skewes and Harmon (1995) tested the AQT in eight commercial broiler houses and concluded that it estimated NH3 levels accurately in the 20–25 ppm range.

228

Ji-Qin Ni and Albert J. Heber

5.1.3. Denuder The denuder is a device used to separate gases and aerosols over a given diameter based upon the difference in diffusion velocity between gases and aerosol particles. It is usually a tube containing a selective internal wall coating that removes the gaseous compounds at the wall. Two different denuder systems are available for sampling and determination of gaseous NH3. The first procedure uses simple cylindrical tubes, sometimes called Ferm tube, as introduced by Ferm (1979). The second procedure uses so-called annular denuders, where the air is passed through the annular space between two concentric cylinders. This arrangement allows the airflow rate to be increased, and makes the subsequent chemical analyses somewhat less demanding (EMEP, 2001). Preparation of denuders involves a careful laboratory procedure of coating an acid solution (usually citric acid, oxalic acid, or phosphorous acid) on the denuder internal wall and evaporating the liquid film. The NH3 collected in the denuder is extracted with a chemical solution and analyzed in the laboratory, for example, using colorimetry (McCulloch et al., 1998; Sommer et al., 2004a). Sampling can be passive using a diffusion denuder as described by Fitz et al. (2003) for evaluating NH3 emissions from a dairy lagoon. A revised passive ‘‘Ferm tube’’ was reported by Phillips et al. (2000). The essential feature of this device was that a precision orifice in a disc of very thin material was installed in the mid-section of the tube so that the air flowing through the tube was proportional to the wind speed. A similar device, passive flux sampler, was described by Scholtens et al. (2003, 2004) and Mosquera et al. (2005) for NH3 sampling and measurement at mechanically and naturally ventilated animal houses. It contains an orifice and collects NH3 at a rate proportional to the air velocity of the air stream passing it without a pump. The passive flux samplers were also reported by Rodhe and Karlsson (2002) in a micrometeorological mass balance setup for NH3 emission at a broiler manure storage. Ferm et al. (2005) presented a passive sampler consisting of two glass tubes internally coated with an acid. One end has a probe with a thin (0.05 mm) stainless steel disc with a 1-mm diameter hole. The average airflow in two opposite directed samplers is proportional to the wind speed vector component along the tubes. Sampling with a denuder can be active with a sampling pump as outlined by EMEP (2001). Denuders can also be combined with NOx monitors for continuous sampling/measurement (Mennen et al., 1996).

5.2. Gas detection tubes Gas detection tubes are based on adsorption of target air pollutants on solid surfaces accompanied by a color reaction. There are two types of disposable tubes: active and passive. Tubes with different measurement ranges are

229

Ammonia Sampling and Measurement

available. Some have suitable measurement ranges for NH3 in animal buildings. Usually, the sensitivities of the tubes are too low for measuring outdoor NH3 concentrations. The most obvious advantage of the gas detection tube is its operational and functional simplicity. Therefore, it was widely used in agricultural NH3 measurements. Meyer and Bundy (1991) measured NH3 concentrations in 200 farrowing pig houses with gas detection tubes. Gas tubes are relatively low cost, usually US$5–10 per tube and US $300–500 for a hand pump. Dewey et al. (2000) reported US$306 for the hand pump (Dra¨ger Accuro Pump, Dra¨ger Safety Inc., Pittsburgh, PA) and US$4.12 for each tube. Leithe (1971) stated that the standard deviation of the results of Dra¨ger gas detection tubes is ~6–8% in favorable and 10–20% in less favorable cases. Skewes and Harmon (1995) found that the passive tubes (Gastec Passive Dosimeter Tube No. 3D) estimated average NH3 levels accurately at low levels of NH3 as compared with the Gastec Low Range Ammonia Detector Tube No. 3La. Liu et al. (1993) concluded that the accuracy of NH3 detector tubes from the Mine Safety Appliance Company in Pennsylvania was 1 ppm. According to Scholtens (1993), Dra¨ger tubes Type 2A and Kitagawa tubes Type 105SD for NH3 measurement both had precision (coefficient of variation) less than 5% and an inaccuracy (systematic error) less than 10%. The precision of Dra¨ger gas detector tubes (Type 5A) was ~10% and the inaccuracy was ~15%. Both precision and inaccuracy of Kitagawa tube (Type 105SC) were ~2%. The precision and inaccuracy of the gas tubes became worse at lower NH3 concentration levels. 5.2.1. Active tubes Active gas tubes require a hand-pump that sucks a predefined volume of air per stroke (Fig. 9). Both ends of the test tube are sealed when manufactured and are cut open just before measurement. The open-end tube is inserted tightly into the pump connector. By pumping the hand-pump, the air sample flows through the tube. The color that arises is evaluated to assess the NH3 concentration. Active gas tubes from five different manufacturers have been used for NH3 measurement at animal facilities. Gas tube

Concentration scale

Direction of sample airflow and tube color change

Figure 9

Tube connector

Hand pump

Chain

Dr€ager gas test tube and hand pump. Source: Ni (1998)

230

Ji-Qin Ni and Albert J. Heber

The Dra¨ger tube is probably the most widely applied gas tube product in dairy cattle barns (Patni and Clarke, 1991), swine barns (Amon et al., 1995; Hayes et al., 2004; Stowell and Foster, 2000), and caged-layer barns (Patni and Clarke, 1991). Other detection tubes used in animal houses include the KitagawaÒ gas detector (De Praetere and van Der Biest, 1990; Jeppsson, 1999; Johnston et al., 1981; Reece et al., 1979; Scholtens, 1993; Svensson et al., 1997), Gastec tubes ( Jacobson et al., 1992), Sensidyne detector tubes (Xin et al., 1996), and Mine Safety Appliances (MSA) detector tubes with a Kwik-DrawÒ Pump (Mine Safety Appliances Company, Part number#487500) (Wheeler et al., 1999). 5.2.2. Passive tubes Like active tubes, passive sampling tubes are also sealed before using. However, only one sealed end of the tube is broken open to commence measurement. The opened tube is exposed at the selected sampling location for a specific time, usually several hours. The gas concentration indicated in the tube should be interpreted with the exposure time. Passive NH3 sampling tubes from different manufacturers have also been used at animal facilities. Patni and Clarke (1991) used Dra¨ger tubes for TWA NH3 concentrations in dairy, swine, and poultry barns. Busse (1993) used a passive Dra¨ger system with a diffusion tube and a holder, in which the tube was inserted for 24-h measurements. Nicks et al. (1993) reported using 8-h diffusion tubes. Nicks et al. (1997), Choinie`re et al. (1997), and Wheeler et al. (1999) employed Gastec diffusion tubes. Pratt et al. (2000) used diffusion tubes in horse stalls in Kentucky, but did not indicate the tube manufacturer.

5.3. Fourier transform infrared spectroscopy Fourier transform infrared (FTIR) spectroscopy is a technique involving the interaction of IR electromagnetic radiation with the test sample. The technology has been called interferential spectroscopy, multiplex spectroscopy, Fourier spectroscopy, interferometric spectrophotometry, or Fourier transform spectroscopy through its development by physicists and manufacturers over the years. The acronym FTIR is almost universally used by chemists to refer to the technique ( Johnston, 1991). The Fourier transformation is a mathematical manipulation that relates a signal, curve, or algebraic function to its frequency content. In Fourier spectroscopy, the output signal is known as an interferogram and is produced by an interferometer (Fig. 10). Interferometers used in FTIR instruments manufactured in recent years are similar in design to the one built at the end of nineteenth century (White, 1990). As the movable mirror is gradually displaced, a cycle of maximum and minimum intensity recurs. It yields specific information about the chemical structure of organic and inorganic compounds based on the unique vibrational modes of

231

Ammonia Sampling and Measurement

Fixed mirror Light source

Moving mirror

Beam splitter

Sample Detector

Figure 10 Schematic diagram of interferometer. Drawn according to White (1990) and Johnston (1991).

different chemical bonds. The FTIR spectrum is rich with information because each vibrational mode absorbs a specific wavelength of IR radiation. Each bond within a molecule may have several vibrational modes. The FTIR absorption spectrum is a ‘‘fingerprint’’ for a particular molecule that can be compared with reference spectra of known compounds, thereby aiding in the identification of unknowns and providing unambiguous confirmation of the identity of ‘‘known’’ materials. A few of the many models of FTIR spectroscopy have been used for NH3 measurement at animal facilities. An FTIR spectroscope K300 with a White-Cell was used in Germany (Neser et al., 1997). Air samples were pumped into or through a special optic cell (White-cell), which used a series of mirrors to create a lengthened light path of 8 m (Amon et al., 1997). A Midac Model M2401 FTIR spectroscope was used in an outdoor open-path measurement of NH3 emission from swine buildings in North Carolina (Harris, 2001; Harris et al., 2001). An FTIR ETG w/ Bomen-100 interferometer and sterling cycle cooled detector was used in Missouri (Secrest, 2001a,b). Other uses of FTIR spectroscopy included an NH3 gas analyzer in the IR/VIS and UV spectral regions (Keck et al., 1994), an IR spectrometer, and a data logging system measuring NH3 emission from pig and dairy barns in Germany (Hartung et al., 1997; Jungbluth and Bu¨scher, 1996; Jungbluth et al., 1997), and two other reports by Hauser and Fo¨lsch (1993) and Gallmann and Hartung (2000) without technical details.

5.4. Infrared gas analyzer An infrared gas analyzer is an infrared spectroscope, a subset of spectroscopy that deals with the infrared region of the electromagnetic spectrum. Nondispersive infrared (NDIR) analyzers measure the spectral absorption of a gas at one spectral band of the IR spectrum. The spectral dispersion of the absorption spectrum of the gas is not used (Phillips et al., 2001).

232

Ji-Qin Ni and Albert J. Heber

¨el and Kjær and Innova photoacoustic spectrophotometer 5.4.1. Bru In a photoacoustic spectrophotometer (PAS), the gas to be measured is irradiated by intermittent light of preselected wavelengths. The gas molecules absorb some of the light energy and convert it into an acoustic signal, which is detected by a single or double microphone(s). The general principle of the PAS system (Innova AirTech Instruments A/S, Ballerup, Denmark) is illustrated in Fig. 11. The PAS monitor can automatically measure multiple gases with a single instrument, hence the name of multi-gas monitor. When gas samples are drawn from ambient air around the analyzer, the measurement time is ~30 s for one gas or water vapor, and ~120 s if five gases plus water vapor are measured. The measurement time is configurable in the monitor. Increasing the length of the sampling tube increases the time required to pump in a new air sample and therefore increases the measurement time. The PAS requires less frequent calibration as compared with NOx analyzers. However, its investment is relatively high. A Type 1302 PAS (Bru¨el and Kjær, Nærum, Denmark) was used intensively in the Netherlands in different studies (den Brok and Verdoes, 1997; Groenestein et al., 2006; Hendriks and Vrielink, 1997; van der PeetSchwering et al., 1999; van’t Klooster and Heitlager, 1992; Verdoes and Ogink, 1997). Its measurement range was 0.2–20,000 ppm (Groenestein et al., 2006). Use of PAS in Germany included Hoy (1995), who measured five gases quasi-continuously; Snell and Van den Weghe (1999), who used it in a pig building; Brunsch (1997), who measured gas concentrations in a poultry farm by combining a Type 1302 monitor with a multi-point sampler Type 1309; and Snell et al. (2003), who measured NH3 at naturally ventilated dairy buildings.

Infrared source

Chopper wheel

Optical Analysis filters cell

Microphone 1

Air outlet

Outlet valve

Flush valve

Inlet valve Mirror

Optical filter Optical carousel window

Microphone 2

Air inlet

Figure 11 Schematic diagram of the general principle of Photo Acoustic Multi-gas Monitor (Type 1302). Courtesy of Innova AirTech Instruments A/S, Denmark.

Ammonia Sampling and Measurement

233

Osada et al. (1998) and Zhang et al. (2005) reported an Innova Type 1312 gas monitor (Ballerup, Denmark) and a multiplexer 1303 in a pig house and a dairy building, respectively, in Denmark. The 1312 is a newer version of the Type 1302. It costs between US$28,000 and 35,000 depending on options (Einfeld and Billets, 1998). The PAS has been used in the United States by Ferguson et al. (1997) in a broiler house and by Pratt et al. (2000) in horse stalls in Kentucky. The ongoing NAEMS project (PAAQL, 2007) has obtained 18 units of Model 1412 photoacoustic field gas monitors (Ballerup, Denmark) for multiple gas measurement, mainly NH3, in animal barns and lagoons. 5.4.2. Chillgard Refrigerant Leak Detection System The Chillgard Refrigerant Leak Detection System is a product of the Mine Safety Appliances Company (Pittsburgh, PA) and operates on the photoacoustic principle, allowing continuous measurement of NH3 concentration. The Chillgard has two models. The IR model has an NH3 detection range of 0–1000 ppm and the new RT model has selectable measurement ranges of 0–100 ppm and 0–1000 ppm. The instrument has a low detection limit of 1 ppm. The applications of this instrument in agricultural NH3 emission study were reported in Canada (Cortus, 2006; Godbout et al., 2000) and in the United States (Sun et al., 2007). Although this instrument has low resolution, our field comparison study (unpublished data) demonstrates some of its advantages over other NH3 analyzers, including stability, low cost, and low maintenance, when used in high NH3 concentration situations. 5.4.3. Rosemount gas analyzer According to RosemountÒ Analytical (Orrville, OH), the Model 880A analyzer produces IR radiation from two separate energy sources. A chopper modulates this radiation into 5 Hz pulses. Depending on the application, the radiation may then pass through optical or gas filters to reduce background interference from other IR-absorbing components. Each IR beam passes through a separate cell. One cell contains a continuous flowing sample while the other cell is either sealed or contains a continuous flowing reference gas. A portion of the IR radiation is absorbed by the component of interest in the sample, with the quantity of IR radiation absorbed being proportional to the component concentration. The detector is a ‘‘gas microphone’’ based on the Luft principle. It converts the difference in energy between sample and reference cells to a change in capacitance (Fig. 12). Brewer and Costello (1999) used a Rosemount Analytical Model 880A gas analyzer with a sampling chamber to measure NH3 concentration in a broiler house in Arkansas. Hartung et al. (2001) described an IR gas analyzer Model BINOSTM-IR-2 (Rosemount GmbH & Co.), but did not provide details.

234

Ji-Qin Ni and Albert J. Heber

Component of interest Other molecules

Infrared sources Chopper

Air sample in

Sample cell

Reference cell

Air sample out

Detector Signal

Diaphragm distended

Figure 12 Schematic of operation principle of nondispersive infrared analyzer Model 880A. Source: Product Data Sheet PDS 103–880A.A01, Rosemount Analytical. Courtesy of Rosemount Analytical Inc.

5.4.4. Beckman Industrial models Maghirang et al. (1991) and Maghirang and Manbeck (1993) used Beckman Industrial models 770 and 780 (Beckman Industrial Corporation, La Habra, CA) NDIR analyzers in a commercial egg laying house in Pennsylvania. The Beckman models have been discontinued. 5.4.5. Miran 203 infrared analyzer The Miran 203 infrared analyzer (Foxboro Company, East Bridgewater, MA) is a portable analyzer designed primarily to promote safety in some areas of the hospital that can have high levels of toxic gases. It went to the gas analyzer market at the end of 1991. This analyzer was only reported in Sweden for measurement of NH3 in pig houses (Andersson, 1996, 1998; Jeppsson, 2002) and layer houses (Gustafsson and von Wachenfelt, 2005; Nimmermark and Gustafsson, 2005), but no details were provided. Miran infrared analyzers are currently products by Thermo Electron Corporation (Franklin, MA).

5.5. Ultraviolet differential optical absorption spectroscopy In the ultraviolet differential optical absorption spectroscopy (UV-DOAS) method, an emitter–receiver set creates a light path in a detection zone. Light is generated by a xenon lamp in the emitter and projected to the receiver. Each gas in the detection path absorbs different parts of the light spectrum in a unique way. The absorption is recorded using a spectroscope.

235

Ammonia Sampling and Measurement

5.5.1. Opsis AR-500 UV open-path monitor The Opsis AR-500 open-path monitoring system consists of a light source emitter, a target gas cell, a receiver, a fiber optic cable, and an analyzer. The fiber optic cable connects the receiver and the analyzer (Fig. 13). According to an environmental technology verification test (Myers et al., 2000), the AR-500 had an NH3 detection limit between 2.8 and 5.8 ppb with good linearity. Its relative accuracy was 3.3–11% over the range of 24–200 ppb. Secrest (2001a,b) used a UV-DOAS Model AR-500 (Opsis AB, Furulund, Sweden) with a Czerny-Turner spectrometer and ‘‘ER-110’’ series (110 mm optics) telescopes to measure ambient NH3 concentration on swine farms in Missouri and Maryland. 5.5.2. Washington State University system An open-path measurement system was developed by Washington State University and has been used for over 10 years in challenging field conditions. Mount et al. (2001) described its use on a dairy farm, where NH3 was measured in the UV bands near a wavelength of 210 nm at an integration time of several seconds to an accuracy of approximately 20%. The system was conceptually simple and consisted of (1) a UV light source, (2) a telescope to beam the UV light into the atmosphere, (3) a mirror system to reflect the light back toward the light source, (4) a receiver telescope to focus the light spectrally absorbed by the atmosphere onto a dispersing spectrograph, (5) a multielement multiplexing digital detector, and (6) a data analysis system. The path length ranged from a few meters to 750 m, and the sensitivity limit was 1 ppbv (parts per billion based on volume).

5.6. Chemiluminescence analyzer Chemiluminescence (CL) NH3 analyzers involve an indirect measurement of NH3 based on converting NH3 to nitric oxide (NO) and then performing NO analysis by the CL method. The NH3 content is obtained by either chemical or mathematical subtraction of the background NO Target gas cell Light source

Receiver Open detection path Fiber optic cable Analyzer

Figure 13 (2000).

Schematic of Opsis monitoring system. Drawn according to Myers et al.

236

Ji-Qin Ni and Albert J. Heber

signal (Pranitis and Meyerhoff, 1987). This technique requires two instrument modules: an NH3 converter and an NOx analyzer. The converter is kept at a temperature of 795 C, and stainless steel is usually used as the catalytic active metal. At this temperature, NH3 is converted into NO by the following reaction (Aneja et al., 1978):

4NH3 + 5O2

Δ →

4NO + 6H2O

(1)

Fe2O3 CrO3

The gas phase reaction of NO and ozone (O3) produces NO2 and a characteristic luminescence in the analyzer (Fig. 14). When electronically excited, NO2 molecules decay to lower energy states and light emission occurs:

NO þ O3 ! NO2 þ O2 þ hv

ð2Þ

Pressure in the NO2 detection chamber is kept at least 31 kPa below atmospheric pressure and temperature is kept at 50 C. Under these conditions, the NO concentration is directly proportional to the photon emission intensity. During transport of gas in tubes from the converter to the NOx analyzer, some NO may oxidize to NO2. A molybdenum converter at 325 C converts NO2 into NO prior to entering the reaction chamber of the NOx analyzer (van’t Klooster and Heitlager, 1992). The advantages of this technique include high sensitivity (1 ppb), high precision (0.5 ppb), linearity (1% full scale), and automation. The disadvantages include highly priced parts (e.g., internal gas scrubbers) by NO detection pathway NO2* detection chamber

NO signal [NO]

PMT

O3

Sample air

Ozone generator

NO2 + NO

[NO2] = [NOx] − [NO]

O3

NO2 → NO Converter

Exhaust

“NOx” as NO

NO2* detection chamber

PMT

NOx signal

[NO]

NOx detection pathway

Figure 14 Schematic diagram of the general principle of chemiluminescent detector for nitrogen dioxide and nitric oxide. PMT, photo-multiplier tube. Source: Stern et al. (1984).

Ammonia Sampling and Measurement

237

some manufacturers, and relatively large initial investment and complicated maintenance. The CL method has been used for NH3 measurement at animal facilities in the Netherlands, UK, Belgium, and the United States. 5.6.1. Matthe¨us-IMAG converter and Monitor-Labs analyzer van’t Klooster and Heitlager (1992). (1992), Verdoes and Ogink (1997), and den Brok and Verdoes (1997) used NH3 ! NO converters type Matthe¨usIMAG and a dual-channel NOx analyzer (Monitor-Labs Model 8840) in the Research Institute for Pig Husbandry (RIPH), the Netherlands. Scholtens (1990), Aarnink et al. (1993b, 1995, 1996, 1997), Smits et al. (1995), Braam et al. (1997), Groenestein and Faassen (1996), Groenestein et al. (1997, 2003, 2006, 2007), and Swierstra et al. (1995) used an instrument of the same model in the Institute of Agricultural and Environmental Engineering, the Netherlands. Once a week, the monitor was calibrated with a gas of 40 ppm NO in N2 and the flow of the different channels was checked. Dust filters were changed when necessary. The converter heated the sample air to 775 C (Ogink and Kroodsma, 1996). A similar system was reported by Demmers et al. (1999) in the UK. 5.6.2. Matthe¨us-IMAG converter and THIS NOx analyzer Berckmans and Ni (1993) and Ni et al. (1999, 2000d) described another system in Belgium that had five NH3 ! NO converters (Type Matthe¨usIMAG) installed at five different sampling locations. Through some heated tubing and a stream selector, the converted air from all the converters was conducted to a single NOx analyzer [Model 42-I, THIS (Thermal Instrument System, currently Thermo Electron Corporation), Franklin, MA]. The converters had conversion efficiencies between 95 and 99% at NH3 concentrations below 30 ppm according to the manufacturer. 5.6.3. TEI converter and analyzer Heber et al. (2001) reported that four NH3 analyzers [Model 17C, TEI (Thermal Environmental Instruments, currently Thermo Electron Corporation), Inc., Franklin, MA]. were used in a comprehensive field study of gas emission measurement in Indiana and Illinois. Each NH3 analyzer consisted of two separate modules, a converter module and an analyzer module. The difference between these systems and the systems in Europe is that only one converter was used with each analyzer. An air stream controller was installed before the converter to facilitate multi-point sampling (Fig. 6). Several papers on NH3 emissions using these analyzers have been published (e.g., Heber et al., 1997; Lim et al., 2004; Ni et al., 2000a,b). Harris et al. (2001) and Walker (2001) reported using two TEI analyzers (Model 17C) in tunnel-ventilated swine finishing houses in North Carolina. At each house, one analyzer was dedicated to the primary exhaust fan and the other was periodically moved from fan to fan in order to ascertain

238

Ji-Qin Ni and Albert J. Heber

variability in exhaust NH3 concentrations. McCulloch et al. (2000) tested three TEI Model 17C analyzers and found that two of them exhibited excellent system linearity across the 0–1000 ppb range. In a third analyzer, the converter efficiency varied from 54 to 77% at 12 and 46 ppb, respectively. 5.6.4. MT converter and API analyzer Aneja et al. (2000) transferred sampling air to a Measurement Technologies 1000N stainless steel NH3 converter when measuring lagoon NH3 emission in North Carolina. The sample flow from the converter was routed to a CL NOx analyzer (Model 200, Advanced Pollution Instrumentation, San Diego, CA).

5.7. Electrochemical sensor Electrochemical (EC) NH3 sensors consist of two electrodes and detect NH3 with the following EC reactions: on the measuring electrode,

2NH3 ! N2 þ 6Hþ þ 6e

ð3Þ

and on the counter electrode,

3=2O2 þ 6Hþ þ 6e ! 3H2 O

ð4Þ

EC sensors provide direct readout and continuous measurements. Several EC sensors have been tested or used at animal facilities. ¨ger sensor 5.7.1. Dra The most frequently reported EC sensor was the Dra¨ger sensor, including a Dra¨ger apparatus for measuring NH3 in a fattening pig house (Heinrichs and Oldenburg, 1993), a Dra¨ger Polytron 2 used in broiler houses (Wheeler et al., 2000a), and a Dra¨ger Pac III for testing in dairy cow and pig housing systems (Kavolelis, 2006). The Dra¨ger Polytron 2 (Dra¨ger Safety, Inc., Pittsburgh, PA) was battery powered. Its scale was 0–300 ppm and its precision was 3% or 9 ppm. The unit consisted of a multi-gas body (Polytron 2) and a sensing unit. The multi-gas body can combine with specific sensor units to measure over 60 toxic gases including NH3. The expected life of an NH3 sensor is equal to or longer than 18 months. Xin et al. (2002) developed a portable and relatively low-cost unit for continuous measurement of NH3 at animal facilities. The unit utilizes sampling and purging cycles to overcome the shortcoming of sensor

Ammonia Sampling and Measurement

239

saturation inherent with electrochemical NH3 sensors (Li et al., 2005). Twenty-eight portable monitoring units were fabricated and successfully used in a multistate and multidisciplinary poultry house research project (Gates et al., 2005). This device was also used in monitoring NH3 emission from turkey houses (Gay et al., 2006). 5.7.2. Other EC sensors Three other EC sensors were found in the literature. Hoy et al. (1992) and Hoy and Willig (1994) described a Series 6004 QuadScan Gas monitoring system, connected with a printer, for continuous measurement of NH3. This system consisted of two main components, the Series 6004 gas receiver and three Series 4485 NH3 gas transmitters. The Series 4485 was a two-wire transmitter designed for monitoring ambient NH3 gas concentration. The transmitter itself consisted of two components, an EC sensor, and an electronic transmitter. The standard 4485 had a measurement range of 0–100 ppm. Jiang and Sands (2000) measured NH3 concentrations in broiler buildings with an EC sensor (ETI series 4700 Twistik Transmitter), but did not give details about the sensor and its performance. An iTX Multi-gas monitor with a biased sensor (iTX Multi-gas Monitor ISC, PA) was used by Hayes et al. (2006) to determine NH3 concentrations at four integrated pig units over a 2-year period.

5.8. ChemcassetteÒ detection system The colorimetric principle is employed by the ChemcassetteÒ detection system (Zellweger Analytics Inc., Lincolnshire, IL). A carefully prepared reel of porous paper tape is impregnated with a chemical. The paper acts as both a trapping and analysis medium, detecting and measuring nanogram amounts of target gas. Upon exposure to the target gas, the paper tape changes color in direct proportion to the sample gas concentration. A photo-optical system measures its color intensity change and determines the sampled gas concentration (Fig. 15). The system can measure different gases with different Chemcassettes, which are individually formulated for specific gases or a family of gases. The low-level NH3 detection Chemcassette has a range of 0.5–30 ppm. The system measures gas concentrations continuously and the response time is several seconds. A single-point Chemcassette system cost US$5200 and an NH3 cassette tape cost US$46 in 2000. According to the manufacturer, this system is the only gas detection method providing physical evidence of gas presence and the technique virtually eliminates in-field calibration. Bicudo et al. (2000) described a Zellweger MDA single-point continuous air monitor at animal facilities in Minnesota. The accuracy of the instrument was 20% of the actual reading.

240

Ji-Qin Ni and Albert J. Heber

Exhaust air

Chemcassette Cable

Photo-optical detector Tape

Sample air

Figure 15 Schematic of ammonia measurement with Chemcassette Monitor. Source: Product description, Zellweger Analytics Inc.

5.9. Solid-state sensor The solid-state or electronic NH3 sensor is a relatively new measurement method. It benefits from the boom of the electronic sensor technology in the latter part of the twentieth century. There are several types of these sensors that are sensitive to NH3 (Go¨pel and Schierbaum, 1991; Timmer et al., 2005). There exist some advantages of solid-state NH3 sensors, including simplicity, low price, quick response, and automatic measurement, compared with other available NH3 measurement technologies. Their limitations include low accuracy, drifting, and interference by humidity and other gases. Several types of NH3 sensors have been tested in animal houses. However, they were still in the development stage. Krause and Janssen (1990, 1991) used a chemical NH3 sensor to measure NH3 distribution in animal houses. The sensor had a detection range of 1–1000 ppm, a response time of <1 s and an accuracy of 10% between 4 and 500 ppm. Krause (1993) described a test of a semiconductor NH3 sensor, but did not indicate whether it was the same sensor reported in 1990 and 1991. Berckmans et al. (1994) conducted a test of a solid-state NH3 sensor, developed by the Inter-university Micro Electronic Center (IMEC), Belgium, in livestock buildings. The sensor had a detection range of 0–100 ppm NH3 and a response time of 10–15 s. It was a thick film semiconducting metal oxide sensor consisting of a heater element, a dielectric layer, a contact layer, and a gas-sensitive semiconductive metal oxide layer (Fig. 16). The conductivity of semiconducting metal oxide films at a certain temperature was influenced by the presence of NH3 gas in the surrounding atmosphere. The sensor’s optimum operating temperature was around 350–400 C. Hess and Hu¨gle (1994) tested an NH3 measuring system called SOLIDOX-NH3, manufactured in Germany, in animal houses a few times but with difficulties.

241

Ammonia Sampling and Measurement

Contact pad heater

Sensor material

Dielectric layer

Sensor contact

Sensor contact

Dielectric layer

Contact pad heater 0 1 2 3 mm

Figure 16

Ammonia sensor developed by IMEC, Belgium. Source: Ni (1998).

5.10. Comparison of measurement devices When different devices are employed for NH3 concentration measurement at animal facilities, their comparability is an interesting question. So far, there have been several studies conducted to answer this question (Table 5). Skewes and Harmon (1995) evaluated the AQT and Gastec passive tubes against Gastec active tubes in eight broiler houses. The authors concluded that the AQT estimated NH3 levels accurately at 20–25 ppm and the passive tubes estimated average NH3 levels accurately at low levels of NH3. Their study indicated that the passive and the active tubes could not agree well, though both were products from the same manufacturer. Mennen et al. (1996) compared six devices, including a DOAS system (OPSIS, Sweden), a photoacoustic monitor (KEMA, the Netherlands), and a CL NOx monitor with NH3 converter, in a test chamber and in the field conditions for the Netherlands National Air Quality Monitoring Network. The converter method was found to have interference by NH4+. The noise level of the DOAS system appeared high. The photoacoustic monitor suffered from many problems, leaving only 1% useful data. All three were rejected as network monitors. Wheeler et al. (2000b) compared two Dra¨ger EC sensors, the KwikDraw gas detector tube, and the Sensidyne passive gas tube in a lab, three environmental chambers, and three high-rise poultry houses. Although all devices agreed well in the lab test, there was a significant difference between the Dra¨ger sensor and the active tube in environmental chambers. There were poor correlations among the three types of sensors in poultry houses. Ni (unpublished data) analyzed NH3 concentration data measured in the exhaust chimney in a pig house with a Dra¨ger EC sensor and with a CL analyzer. The correlation coefficient of NH3 concentrations was 0.55 from 3900 paired data recorded by the two devices over a month. Other studies showed satisfactory comparison results. Dewey et al. (2000) compared pHydrionTM NH3 test strips against Dra¨ger tubes, and concluded that the test strips provided a precise and cost-effective means of detecting NH3 concentrations in swine confinement buildings.

242

Ji-Qin Ni and Albert J. Heber

Table 5 Summary of comparison studies of ammonia measurement devices Compared device

Test condition and conclusion

Reference

1. AQT 2. Gastec passive tubes 3. Gastec active tubes 1. DOAS 2. NH3 converter +CL NOx monitor 3. Photoacoustic monitor 1. Dra¨ger EC sensors 2. Kwik-Draw active tubes

Eight broiler houses: AQT and passive tubes agreed well with active tubes at 20–25 ppm and lower levels, respectively Test chamber and field: All rejected as network monitors

Skewes and Harmon (1995)

Laboratory: All three devices agreed well Environmental chambers: EC sensors and active tubes differed significantly Layer houses: Poor correlations existed among three devices Pig house: Correlation coefficient r = 0.55 for 3900 paired data points Pig house: Test strips were precise and cost-effective as compared with Dra¨ger tubes Pig house: Gas tubes could accurately measure NH3 using three sample averages Pig farm: Agreement was reasonably good

Wheeler et al. (2000b)

3. Sensidyne passive tubes 1. Dra¨ger EC sensor 2. CL analyzer 1. pHydrionTM NH3 test strips 2. Dra¨ger tubes 1. Gas tubes 2. CL analyzer 1. Open-path FTIR 2. Open-path UV-DOAS

Mennen et al. (1996)

Ni (unpublished data) Dewey et al. (2000) Parbst et al. (2000)

Secrest (2001a)

AQT, Ammonia Quick Test; CL, chemiluminescence; FTIR, Fourier transform infrared; NH3, ammonia; UV-DOAS, ultraviolet differential optical absorption spectrometer.

Parbst et al. (2000) evaluated 0.25–3 ppm gas detection tubes (Dra¨ger Model 6733231) in finishing swine buildings in summer and winter against a CL analyzer (TEI Model 17C) being used in a long-term study (Heber et al., 2001). The authors concluded that NH3 concentrations could be accurately evaluated using the mean of three gas detection tube samples.

Ammonia Sampling and Measurement

243

Secrest (2001a) collocated the optical paths of an open-path FTIR system and an open-path UV-DOAS system. The FTIR path was 159 m and the UV-DOAS path was 150 m. The author concluded that the agreement was reasonably good (R2 = 0.94), although there was a 15-ppb difference between their responses at lower concentrations. To date, however, it is still unknown which technique provides results that are the closest to the ‘‘true’’ NH3 concentrations under agricultural field conditions. The tests summarized in this section only compared selected techniques with unknown characteristics against each other, not against a standard technique, which does not yet exist. Therefore, standard techniques and relevant methodologies need to be developed, and the existing NH3 concentration measuring devices need to be tested, compared, and evaluated.

6. Ammonia Concentration Data 6.1. Advances in data collection Research into NH3 at animal facilities has undergone dramatic changes since the 1960s. This is clearly demonstrated in change observed in reported NH3 concentration data. First, the level of research into NH3 at animal facilities has expanded from small tests to multi-institutional and international projects. The data quantity has increased exponentially. The study conducted in broiler pens by Valentine (1964) contained ~300 wet-chemistry samples of NH3 concentrations. The six-university project from 2004 to 2005 with continuous measurement systems (Heber et al., 2006) produced 2 million NH3 concentration data points, each with a 1-min average. The ongoing NAEMS project is expected to generate 13 million NH3 concentration data points from field measurement at animal barns and lagoons (PAAQL, 2007). Second, sampling and measurement covers longer periods of time and more animal facilities, especially when the study objective is to determine baseline emissions. A research project in Northern Europe included 14 housing types for cattle, pigs, and poultry in England, the Netherlands, Denmark, and Germany (Groot Koerkamp et al., 1998). The NAEMS is monitoring two to four mechanically or naturally ventilated buildings continuously for two years at each of five dairies, five pork production sites, three egg layer operations, and one broiler ranch. It is also monitoring 10 lagoons and a dairy corral for 20 days in each season at 11 farms. Third, high-frequency concentration data is providing richer information. While TWA data are still useful, more and more research is employing advanced sampling and measurement techniques to acquire data that contain dynamics of NH3 concentration and allows obtaining insight into

244

Ji-Qin Ni and Albert J. Heber

the complicated NH3 generation and emission mechanism at animal facilities (Mosquera et al., 2005). Lastly, most of the research on NH3 is coupled with measurements of other pollutants, including other gases (e.g., greenhouse gases and H2S), particulate matter, and odor, at the animal facilities. Multi-pollutant studies reduce the overall costs and increase understanding of the interaction of pollutant generation and emission. An example of this is the effect of pH changes on NH3 and H2S in liquid manure. Higher pH promotes NH3 release while reducing H2S release and vice versa. Another example is the effect of CO2 release that accelerates NH3 release at certain dynamic conditions (Ni et al., 2000c,d). Furthermore, multi-pollutant studies are favorable for developing and testing abatement technologies that could produce multiple benefits. The increase in the quantity and complexity of NH3 data has created new challenges to ensure the quality of data collection and processing.

6.2. Data precision and bias Collection of high quality data is critical to any research program. Erroneous data are worse than no data because bad data misleads scientific conclusions, regulatory decisions, abatement technique evaluations, and health risk assessments. Much of the information accompanying NH3 data reported in the literature was insufficient to provide a reasonable assessment of data quality. This was due to two problems. First, data quality did not receive enough attention in many studies. In most published articles, there was little information about quality assurance and quality control (QAQC), for example, calibration of measurement systems, assessment of precision and bias. Second, proper and consistent data quality terminology is lacking, adding to the difficulty in comparing inter-project results. Different data quality indicators (DQIs) were used to describe the measurement quality in various reports. These DQIs include ‘‘accuracy,’’ ‘‘inaccuracy,’’ ‘‘precision,’’ ‘‘error,’’ ‘‘sensitivity,’’ and ‘‘standard deviation.’’ This increases the difficulty of comparing results. The U.S. EPA recommends using the terms ‘‘precision’’ and ‘‘bias,’’ rather than ‘‘accuracy,’’ to convey information usually associated with accuracy. According to U.S. EPA (1998), accuracy is a measure of the closeness of an individual measurement or the average of a number of measurements to the true value. Accuracy includes a combination of random error (precision) and systematic error (bias) components that result from sampling and analytical operations (Fig. 17). Precision is a measure of agreement among replicate measurements of the same property, under prescribed similar conditions. Bias is the systematic or persistent distortion of a measurement process that causes errors in one direction.

245

Ammonia Sampling and Measurement

* * * * * * * *

*

*

* *

*

*

*

* High bias + low precision = low accuracy

* **** * ** High bias + high precision = low accuracy

Low bias + low precision = low accuracy

** * **** *

Low bias + high precision = high accuracy

Figure 17 Random measurement uncertainties and measurement bias: Shots at a target. Source: USEPA (1998).

There are several potential error sources that can lower precision and increase bias when conducting NH3 studies at animal facilities. Understanding these errors and the related preventive or corrective measures will help improve data quality.

6.3. Error reduction Errors are common in physical measurements and methods of error analysis are well established (Taylor, 1997). Errors in NH3 concentration data can be introduced from instrument calibration, air sampling, sample measurement, and data processing. While most of these error sources are universal in scientific research, a few of them deserve particular attention. Table 6 summarizes these sources that most frequently occur, and provides suggestions to reduce them. 6.3.1. Errors from calibration Calibration is the setting or correcting of a measuring device or base level, usually by adjusting it to match or conform to a dependably known and unvarying measure. Calibration of measurement systems assures data quality and provides information about characteristics of these systems, such as response, drift, linearity, stability, and precision. Errors can be easily introduced into NH3 data by improper calibration. Calibration of an NH3 measuring system is usually realized by challenging the system with known concentrations of NH3 mixed in dry nitrogen or in dry air. The difference between the known NH3 concentrations and the

246

Ji-Qin Ni and Albert J. Heber

Table 6 Potential error sources and reduction methods in ammonia concentration measurement Potential error sources

Calibration Inaccurate gas

Inadequate procedure

Incorrect operation Sampling Spatial variation

Temporal variation

Gas adsorption/desorption at contacting surfaces of sampling system Condensation in sampling system

Dust in sampling system Insufficient sample air flow Inadequate sampling duration

Incorrect sample volume

Disturbance at gas release

Comments and reduction methods

It is a potentially serious source-level error. Select high quality gas from specialty gas vendor and use unexpired gas. It may be design error. Select adequate gas introduction location, calibration frequency, duration, and flow rate. It is human error. Practice QAQC and use automated calibration setup and software. It is inevitable and might be at great extent. Carefully select sufficient number of representative sampling locations. It always exists. Use TWA sampling device or high frequency sampling to cover day and night to reduce diurnal errors. Use long-term sampling to cover winter and summer to reduce seasonal errors. It affects system response time, especially at low NH3 concentrations. Reduce sampling tubing and increase sampling system temperature. It absorbs NH3 and affects sensor operation. It could be harmful for some instruments. Insulate or heat sampling system to keep the temperature above that at the sampling locations. It causes leaks in the sampling system. Use dust filters at the sampling inlets. It will introduce errors to the analyzer. Increase and monitor sample air flow. It causes insufficient sample volume for acid trap samplers and introduces temporal errors. It will also be too short for analyzers to reach equilibrium. Select proper sampling duration. It introduces errors to acid traps. Calibrate and monitor the device for volume measurement. Too much disturbance at gas release surface causes unrealistic data when using sampling chambers. Simulate natural

x

247

Ammonia Sampling and Measurement

Table 6 (continued) Potential error sources

Leaks in sampling system

Incorrect sampling operation

Measurement Low instrument precision and sensitivity

Slow instrument dynamic response Poor instrument stability

Interferences of water vapor and other gases Incorrect operation

Data processing Incorrect temperature data

Incorrect atmospheric pressure data

Improper validation and treatment of data

Comments and reduction methods

conditions if the objective is to determine baseline NH3 emission. It mixes air from nonsampling locations with the air from sampling locations in negative pressure sampling system. Improve the design of the sampling system and check it periodically or monitor it continuously. It is human error. Practice QAQC and use automatic sampling monitoring equipment and software. It is expected random error. Select instrument of high precision and suitable sensitivity may reduce error. Avoid using expired disposable sensors. It may introduce serious errors in MPSS. Use instrument with small time constant. It may be random or systematic errors. Avoid using unstable instrument and perform periodic calibration. It is random or systematic error. Develop and use new calibration methodology and select better devices. It is human error. Practice QAQC and use automated measurement monitoring software. Change from 20 C to 10 C increases 3.5% in converted mass concentrations. Use correct temperature for concentration conversion. A variation of 0.06 atm may introduce 6% of error when converting volumetric to mass concentrations. Use correct pressure for concentration conversion. It leads to erroneous data selection and calculation. Have a good understanding about the field test and the data that it generates. Only use valid data in result calculation. (continued)

248

Ji-Qin Ni and Albert J. Heber

Table 6 (continued) Potential error sources

Comments and reduction methods

Improper interpolation and extrapolation of data

It creates data that are unreliable. Understand the limitation of data and use interpolation and extrapolation carefully. Incomplete data They have limitations to represent real conditions. Avoid using the data that do not satisfy data quality objectives. Incorrect selection of gas It occurs in centralized multi-point sampling equilibrium time data. Select the equilibrium time based on test results. Insufficient or missing test It brings difficulty for data validation, notes correction, and interpretation. Make good test notes about the system status and human activity that has effected or will affect data. Misinterpretation of test notes It will introduce various errors. Direct experience and good understanding about the test setup and operation is needed. MPSS, multi-point sampling system; QAQC, Quality Assurance and Quality Control; TWA, time weighted average.

system outputs guides the correction of the measurement by adjusting system hardware or software, or by correcting concentration data during data processing. Calibration of NH3 sampling and measurement devices at animal facilities has not been emphasized enough. Only a small portion of the publications described the procedure. They include calibrating NH3 analyzers (Heber et al., 2001; Mosquera et al., 2005; Ogink and Kroodsma, 1996; van’t Klooster and Heitlager, 1992), PAS gas analyzers (Rom, 1993), EC sensors (Wheeler et al., 1999), and UV-DOAS and FTIR systems (Secrest, 2001a). Commercial measuring devices are typically calibrated by the manufacturer or supplier before shipping to the consumer following purchase, repair, or maintenance. Although there are devices with claimed selfcalibration capacity (Mount et al., 2001), calibration is required for most analytical instruments before initial use. Re-calibration after using the device for a certain time is necessary because instrument drift occurs. Manufacturers usually recommend calibration intervals, specify calibration procedure, and provide calibration accessories. An initial check of a new instrument with reference NH3 is advisable. Disposable measuring devices, such as active and passive gas detection tubes, cannot be calibrated by users. However, they can be verified for their

249

Ammonia Sampling and Measurement

precision and bias by randomly selecting some tubes, for example, 5% from a batch of tubes and using them to measure calibration gas with known NH3 concentrations. There are three potential error sources that can lead to inadequate calibration. The first one is from the source level, the calibration gas, or the ‘‘standard.’’ We evaluated three gas cylinders of 53.1, 33.2, and 9.33 ppm certified NH3 in air and found significant errors in them. The gases had been ordered from the same gas supplier at the same time requesting that they be analyzed twice with at least one week apart between analyses and both analyses should agree to within 1%. The gas company used FTIR as the analytical method. Upon receiving the gases, we compared the gas cylinders with each other. An NH3 analyzer (TEI Model 17C) was first confirmed for its linearity by using the NH3 directly from the cylinders and after 50% dilution with certified zero air. The analyzer was then calibrated against the 53.1 ppm cylinder. However, when the 53.1 ppm NH3 cylinder was used as a ‘‘standard,’’ the 40.5 and 5.77 ppm cylinders surprisingly resulted in 33.2 and 9.33 ppm readings, respectively, with the same analyzer. The 53.1 ppm gas was confirmed to be the most reliable based on measurements with two other devices: Dra¨ger NH3 detection tubes and a Chillgard Refrigerant Leak Detection System. The cylinders with 40.5 and 5.77 ppm certified concentrations were therefore returned to the gas provider, who admitted their inaccurate concentrations, for recertification. However, the recertified gases still showed significant differences compared with the 53.1 ppm NH3 cylinder (Table 7). Unfortunately, this type of source-level error was not an isolated incident. According to the U.S. EPA audit results for EPA Protocol Gas from April 2003 to February 2004, 57% of specialty gas vendors failed the audit and the overall failure rate was 11% or 14 out of 126 cylinders containing blends of sulfur dioxide (SO2), nitric oxide (NO), and CO2 (Scott Specialty Gases, 2004). Moreover, unlike some other gases with agricultural origin, for example, H2S and methane (CH4), NH3 is not yet available as a compressed gas with Standard Reference Materials from the National Institute of Standards and Table 7 Verification of three certified ammonia gas cylinders Concentrations measured at NH3 analyzer (ppm)

Cylinder

Certified concentration (ppm)

No dilution

50% dilution

A B C B (recertified) C (recertified)

53.1 33.2 9.33 29.6 9.6

53.1 40.5 5.77 16.9 6.0

26.6 20.0 2.80

250

Ji-Qin Ni and Albert J. Heber

Technology (NIST) (USEPA, 1997). The NIST provides U.S. industry, government, and the public with measurements, standards, and information services. The U.S. EPA traceability protocol for assay and certification of gaseous calibration standards (USEPA, 1997) allows for Gas Manufacturers Intermediate Standards (GMIS) and NIST Traceable Reference Materials (NTRM) Gas Calibration Standards to be prepared by gas manufacturers and certified by NIST. If a measuring system is calibrated against an incorrect NH3 concentration, the systematic error passes on to the collected and processed data. The cause of this error goes back to NH3 gas standards and is related to the gas suppliers. The accuracy of calibration gas is out of control by the users conducting NH3 measurements at animal facilities. However, selecting products from the most reputable gas companies and using unexpired gas will reduce the risk of introducing this error to the NH3 measurement data. Additionally, checking the new calibration gases with a high quality stable instrument such as the FTIR helps to assure that there were no gross errors by the supplier. The second main error comes from improper procedure and operation of calibration. Ammonia measuring devices used at animal facilities are usually field-calibrated under steady-state conditions by applying a zero input using zero air and a step input using span gas. Multi-point calibration, which consists of a zero and several different levels of spans, can be used to determine the linearity of the device. Calibration gases can be introduced at the inlets of NH3 sensors/analyzers or at the sampling location, from which the entire sampling and measurement system can be checked. Calibration procedure defines the calibration frequency, duration, gas flow rate, and so on. Calibration frequency should be based on characteristics of different instruments. Excessively frequent calibration is not only expensive, but also causes loss of normal measurement time. However, belated calibration will increase uncertainty in the measured data. Carefully determining the procedure according to the characteristics of the sampling and measurement devices will ensure better calibration data quality. Errors resulting from various improper operations, for example, incorrect concentration recording, are human errors. They can be reduced by practicing QAQC and using automated calibration devices and software. The third error is related to interferences from other components in air. Calibration gas is usually a mixture of pure and dry gases, different from the actual air at animal facilities that is a mixture of water vapor, other gases, and particulate matter. This difference can introduce errors in measurement data when the instrument is adjusted to the calibration gas but measures the actual air. Theoretically, these errors can be corrected if the instrument manufacturer provides reliable data about the interferences and the concentrations of interfering components are known. However, in practice, it is not always easy to realize.

Ammonia Sampling and Measurement

251

6.3.2. Errors from sampling Inadequate sampling design, setup, and operation can introduce spatial and temporal errors. Lefcourt (2002) suggested that the most accurate method for estimating NH3 emission rates would be to collect and mix all of the exhaust streams prior to continuously measuring NH3 concentration and airflow. However, this method is unfeasible at animal facilities. Because of the nature of dynamic situations and the limited air samples at animal facilities, errors are inevitable in practical situations. Spatial and temporal NH3 concentration variations could occur to a great extent in animal barns as shown in Fig. 3. More sample locations will reduce these errors but the number of samples is usually limited by the availability of equipment. In a centralized MPSS using a single set of analytical instruments, there is a tradeoff between temporal and spatial information. More sampling points mean less frequent sampling. Therefore, sampling locations should be carefully selected to optimize representativeness. To reduce errors due to diurnal concentration variations, sampling should cover day and night using TWA sampling/measurement devices or high frequency sampling. Similarly, long-term measurement to cover winter and summer should be used to reduce seasonal errors. Condensation and dust in the sampling system not only introduces errors in NH3 concentration data, but is also harmful to the instrument and sampling system. The centralized point-sampling system, which includes air transportation tubing, and the measurement instrument should be kept at a temperature at least the same as the air temperature at sampling locations. Dust filters can be used to prevent dust from coming into the sampling system and causing leakage. Inadequate sample airflow and sampling duration can introduce errors and should be avoided. For acid trap sampling, low sample airflow and short sampling duration result in low sample air volume, which may produce undetectable NH3 concentrations in the trap. On the contrary, too much sample air will cause saturation of the acid solution. Sampling airflow rate should satisfy the requirement of the analytical instrument. Most of these instruments need 0.3–1.0 l/min of sample air. Required sampling duration depends on the response time of the measurement instrument and should allow the analyzer to reach equilibrium in multi-point sampling. In the reported studies, the sampling duration varied from 1 min (Ogink and Kroodsma, 1996), 2 min (Berckmans et al., 1998), 15 min (Neser et al., 1997), to 30 min (Heber et al., 2005). Errors from sampling can also be introduced by improper interference of the air flow in sampling chambers, leakage of the negative pressure sampling system, and incorrect operation of the sampling system. These errors can usually be reduced by careful design, maintenance, and execution of QAQC.

252

Ji-Qin Ni and Albert J. Heber

6.3.3. Errors from measurement Measurement errors are related to the measurement devices and their operations. Some of the error sources are inherent to instruments themselves. Ammonia concentrations at animal facilities are almost always under transient conditions with high frequency and large magnitude of variations (Ni et al., 2000b). These variations do not introduce serious temporal concentration errors to sensors that provide TWA concentrations (e.g., passive gas tubes). However, they do introduce errors to instruments that provide ‘‘real-time’’ NH3 concentrations (e.g., FTIR, PAS, and NOx analyzer) during measurement but have inadequate frequency responses compared with the dynamic changes in concentrations. Accurate dynamic measurement requires a small time constant of the instrument (Doebelin, 1983). The magnitude and correction of this type of error and its relationship with specific measuring devices, sampling method, animal facility size, and so on need more investigation. Furthermore, the true frequencies of NH3 concentration changes under various conditions (e.g., in open air, inside buildings, with different sizes of buildings, at different seasons, under different weather) have not yet been fully characterized and deserve further study. Nevertheless, measurement instruments with poor response time should be avoided, if possible. Figure 18 presents an example of a problematic gas analyzer during our study at a multiroom animal building. The NH3 concentrations obtained by analyzer B could have caused serious misinterpretation if it had not been compared with analyzer A, which reflected good dynamics following the concentration change when the system switched sampling locations every 10 min. Location #3, measured from 0:40 to 0:50 and from 2:10 to 2:20, 10

9

9 8

A

7

7

6

6 5

5 4

4

3

3

2 0 0:00

2 1

B

1 0:30

1:00

Sample location #

Concentration, ppm

8

Location #

1:30

2:00

2:30

0 3:00

Time, h:mm

Figure 18 Comparison of a fast response analyzer (A) with a slow response analyzer (B) when sampling ammonia at nine different locations.

Ammonia Sampling and Measurement

253

had the highest NH3 concentration according to analyzer A. Instead, analyzer B, which had slow response, showed that this location had the lowest concentration. Except for the pH paper method, almost all of the NH3 measurement techniques were initially developed for nonagriculture use. They were introduced to the animal industry after successful use in other fields. For instance, the Chemcassette monitor that had existed for over 30 years was first reported in agriculture use in 2000 (Bicudo et al., 2000). The conditions under which the instruments were developed and factory calibrated might be very different from the conditions at animal facilities. Air produced from animal facilities is a mixture of a large number of gases plus relatively high moisture content compared with the commercial calibration gases. Some of the NH3 measuring devices are sensitive to water vapor and gases other than NH3. Interferences are therefore possible during field measurements. Four of the seven comparison studies summarized in Table 5 demonstrated at least some degree of disagreement between the tested measuring devices. The increasing inconsistency of the three devices tested in the lab, environmental chamber, and commercial layer house reported by Wheeler et al. (2000b) could probably be explained by increasing interferences among the three test locations. The gases that interfere with detection tubes may be more prevalent in commercial poultry houses than one might expect. Discussion about development or improvement of NH3 measuring devices specifically targeting the agricultural environment has not been found in the literature. The potential errors caused by agricultural air interference are not fully understood and are not compensated in field data. Therefore, there is a need to develop test methodology and conduct studies to determine field performance of available measuring devices. Recommendations on the use of these devices and the field data they produced will help to improve agricultural NH3 measurement significantly. 6.3.4. Errors from data processing Data processing is mainly a statistical procedure that involves data validation and calculation with predefined criteria and algorithms. It is different from data management and data analysis. Data management is an administrative procedure that engages data organization, distribution, retrieval, storage, and backup. Data analysis is a scientific procedure that requires insightful knowledge for interpreting, exploiting, and synthesizing the information carried by or hidden behind the data. Data processing can be complex in some comprehensive research projects, of which the data sets can consist of large amount of information that the NH3 concentrations are dependant on. To satisfy these conditions, customized computer program has been developed to process the data (Eisentraut et al., 2004).

254

Ji-Qin Ni and Albert J. Heber

When calculating NH3 release or emission, sample air temperature and atmospheric pressure are used to convert the volumetric NH3 concentration to mass concentration. The air temperature and pressure should be those at the location where the samples are taken, not where the samples are analyzed, which may be different from the sampling locations. Sampling location temperatures at animal facilities vary diurnally and seasonally, especially outdoors. Changing the temperature from 20 C to 10 C increases the converted NH3 mass concentration by 3.5%. A variation of 0.03 atmospheric pressure introduces 3% of difference in NH3 concentration conversion. Therefore, using accurate temperature and pressure data can reduce errors in NH3 concentration processing. Not all the collected NH3 concentration data are usable due to various reasons. Raw data are invalid until they are validated. Only valid data should be used for compilation of final results because data quality is more important than data quantity. Validation is a process that may introduce errors. Suspicious data should be linked to possible reasons before validating or invalidating them. Missing data can sometimes be filled by interpolation or extrapolation. However, these methods have their limitations. Their improper use can create data that are unreliable. Data completeness is a measure of the amount of valid data obtained from a measurement system, expressed as a percentage of the number of valid measurements that should have been collected, that is, measurements that were planned to be collected (USEPA, 1998). Data completeness is often set at 75% or more (PAAQL, 2007). Incomplete data have limitations for representing real situations at animal facilities. Errors will be introduced if using NH3 concentrations covering only a few hours in the morning to represent the mean concentration of the day. Similarly, spatially deficient data, relative to the original sampling design, are inadequate to represent the entire sampling space. Selection of gas equilibration time affects the results of the NH3 processed data for the MPSS using single measurement instrument. The equilibration time should be determined based on the response time of the sampling and measurement system. Correct data processing requires understanding of the actual circumstances under which the data were obtained, for example, the sampling and measurement system and their operations. The test circumstances and procedures are recorded in test notes or test logs. Insufficient or missing test notes and misinterpretation of these notes can all introduce errors. Test notes should aim at assisting data processing and interpretation. They should record all test operations and measurement system changes that have affected or will affect the data. They should be complete, concise, and easy to understand.

Ammonia Sampling and Measurement

255

6.4. Standards for ammonia sampling and measurement Because methodologies that are proven to be scientifically sound were not always available or not always used, many of the field studies did not produce reliable data. Some European researchers believe that ~80% of the publications of agricultural NH3 emissions are not useable to establish annual emission factors (Gallmann and Hartung, 2000). This has created a lack of confidence in reported data and has made comparison of research results difficult or unreliable. The present situation leads to the conclusion that there is an urgent need to develop standards for agricultural NH3 measurement and relevant methodologies and technologies in this specific field. Although some attempts to establish standards in air pollution monitoring have been made (Heber et al., 2004), much effort is still needed to make the following standards: Technical terms: Standardize technical terms for this specific research field including those for describing processes (e.g., NH3 production, generation, volatilization, release, emission), presenting data (e.g., concentration, emission flux, emission rate, emission per animal unit, emission per animal place, annual emission factor), and indicating data qualities (e.g., precision, bias, accuracy). Calibration gases: Define approved analytical instruments and procedures used by calibration gas providers for certifying gases. This standard will guide researchers on selecting gas providers and will assure quality of certifications. Sampling devices and procedures: Standardize closed and open sampling devices (e.g., dimensions and structures of sampling chambers, configuration of micrometeorological sampling, materials used in sampling systems), and sampling procedures (e.g., sampling location, interval, frequency, duration, season). Measuring devices: Standardize performance requirements (e.g., precision, sensitivity, response time) of measuring devices (including those for NH3 concentration and airflow rate measurements) for various measurement objectives (e.g., compliance, emission factor determination, abatement evaluation, and health risk assessment).

7. Summary and Conclusions Tremendous advances in sampling and measurement of NH3 at animal facilities have been achieved during the past four decades. These advances have taken place under the increasing awareness of the environmental problems related to agricultural NH3 and with the efforts to address these problems at different levels, from the general public, animal producers,

256

Ji-Qin Ni and Albert J. Heber

scientists and engineers, to the local and national governments. However, these advances could not have happened without the development of the sampling and measurement methodology, the computer technology, and the analytical instruments. The state of the art of NH3 sampling and measurement at animal facilities displays a variety of technical possibilities to satisfy different research and monitoring requirements. Nevertheless, all the sampling and measurement technologies reviewed in this chapter have their advantages and limitations. Selection of the methodology and technology should be based on study objectives, budget limit, and available equipment and expertise. The three sampling methods (closed, point, and open path) can satisfy most of the research objectives to assess human and animal exposure, determine baseline emissions, compare building structure and mitigation technologies, and model pollutant dispersions. However, spatial variation of NH3 concentrations at large animal facilities is still a major technical difficulty for NH3 sampling. The current sampling technology can only cover limited sampling points or sampling paths, leaving significant uncertainties for the NH3 concentrations at uncovered spaces. The 31 measurement instruments and sensors tested or applied at animal facilities are featured with different costs, sensitivities, and response times, providing more choices for selection than the sampling devices. Measurement devices exhibit different levels of performance and should be selected according to research objectives. While some low-cost sensors offer TWA NH3 concentrations, the high-end analytical instruments enable continuous measurement with fast response, thus supplying data containing dynamic information of NH3 concentrations. Most of these high-end and expensive instruments were found in universities and research institutions where intensive research on agricultural NH3 was carried out. However, temporal variations of NH3 concentrations at animal facilities are still not sufficiently characterized and require more investigation. Future study is also needed to assess the performances of measurement instruments under agricultural field conditions. Particular errors associated with NH3 concentration data can be introduced by system calibration, air sampling, NH3 measurement, and data processing. While some of these errors are inevitable, others can be reduced by selecting gas vendors of high reputation, better research design and performance, and execution of QAQC. Nonetheless, some sources of errors associated with sampling and measurement deserve further investigation. Because of the incomplete technical standards in research conducted worldwide, confidence in the quality of reported NH3 data is not yet fully established. Comparison of research results is difficult or impossible in some cases. Development of standards for technical terms, calibration gases, sampling devices and procedures, and measuring devices is needed. Standard

Ammonia Sampling and Measurement

257

testing methods for assessing sampling and measuring devices and procedures for agricultural NH3 studies wait to be developed. Cooperation among instrument developers/manufacturers, calibration gas providers, regulators, standard agencies, and agricultural research scientists is necessary to improve agricultural NH3 sampling and measurement. Similar to what has advanced the NH3 research at animal facilities in the past decades, the future breakthrough in this field will largely depend on emerging new methodology and technology.

ACKNOWLEDGMENTS This study was partially supported by (1) Catholic University of Leuven, Belgium; (2) Hangzhou Rural Energy Office, China; (3) Purdue Center for Animal Waste Management Technologies, Purdue University, USA; (4). The U.S. EPA; and (5). Purdue University Agricultural Research Program, USA. A preliminary version of this study was presented at the 2001 ASAE Annual International Meeting, Sacramento, California, USA.

REFERENCES Aarnink, A. J. A., Koetsier, A. C., and van den Berg, A. J. (1993a). Dunging and lying behaviour of fattening pigs in relation to pen design and ammonia emission, In ‘‘Fourth International Symposium of Livestock Environment’’ (E. Collins and C. Boon, Eds.), pp. 1176–1183. ASAE St. Joseph, Michigan, and University of Warwick, Coventry, England. Aarnink, A. J. A., Wagemans, M. J. M., and Keen, A. (1993b). Factors affecting ammonia emission from housing for weaned piglets, In ‘‘Proceedings of the First International Symposium on Nitrogen Flow in Pig Production and Environmental Consequences’’ (M. W. A. Verstegen, L. A. den Hartog, G. J. M. van Kempen, and J. H. M. Metz, Eds.) pp. 286–294. Pudoc Scientific Publishers Wageningen (Doorwerth), The Netherlands. Aarnink, A. J. A., Keen, A., Metz, J. H. M., Speelman, L., and Verstegen, M. W. A. (1995). Ammonia emission patterns during the growing periods of pigs housed on partially slatted floors. J. Agric. Eng. Res. 62, 105–116. Aarnink, A. J. A., van den Berg, A. J., Keen, A., Hoeksma, P., and Verstegen, M. W. A. (1996). Effect of slatted floor area on ammonia emission and on the excretory and lying behaviour of growing pigs. J. Agric. Eng. Res. 64, 299–310. Aarnink, A. J. A., Swierstra, D., vandenBerg, A. J., and Speelman, L. (1997). Effect of type of slatted floor and degree of fouling of solid floor on ammonia emission rates from fattening piggeries. J. Agric. Eng. Res. 66, 93–102. Amon, B., Boxberger, J., Amon, T., Gronauer, A., Depta, G., Neser, S., and Scha¨fer, K. (1997). Methods for measuring emissions from agrarian sources: FTIR measurement techniques with White-Cell, large chamber or open-path, In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), pp. 161–167. NVTL Rosmalen, The Netherlands, Vol. I. Amon, M., Dobeic, M., Misselbrook, T. H., Pain, B. F., Phillips, V. R., and Sneath, R. W. (1995). A farm scale study on the use of De-Odorase(R) for reducing odor and ammonia emissions from intensive fattening piggeries. Bioresour. Technol. 51, 163–169.

258

Ji-Qin Ni and Albert J. Heber

Andersson, M. (1995). ‘‘The Effect of Different Manuring Systems on Ammonia Emissisns in Pig Buildings,’’ Rep. No. 100. Institutionen fo¨r Jordbrukets Biosystem och Teknologi Sveriges Lantbruksuniversitet, Lund, Sweden. Andersson, M. (1996). Performance of bedding materials in affecting ammonia emissions from pig manure. J. Agric. Eng. Res. 65, 213–222. Andersson, M. (1998). Reducing ammonia emissions by cooling of manure in manure culverts. Nutr. Cycl. Agroecosyst. 51, 73–79. Aneja, V. P., Stahel, E. P., Rogers, H. H., Witherspoon, A. M., and Heck, W. W. (1978). Calibration and performance of a thermal converter in continuous atmospheric monitoring of ammonia. Anal. Chem. 50, 1705–1707. Aneja, V. P., Chauhan, J. P., and Walker, J. T. (2000). Characterization of atmospheric ammonia emissions from swine waste storage and treatment lagoons. J. Geophys. Res. 105, 11535–11545. Anonymous. (2001). Directive 2001/81/EC of the European parliament and of the council. Official J. Eur. Communities L309, 22–30. ATSDR. (2004). ‘‘Toxic FAQ Sheet for Ammonia.’’ Agency for Toxic Substances and Disease Registry (ATSDR), U.S. Department of Health and Human Services, Public Health Service, Washington, D.C. Bajwa, K. S., Aneja, V. P., and Arya, S. P. (2006). Measurement and estimation of ammonia emissions from lagoon-atmosphere interface using a coupled mass transfer and chemical reactions model, and an equilibrium model. Atmos. Environ. 40, S275–S286. Berckmans, D., and Ni, J.-Q. (1993). Field test installation for on-line continuous measurement of total ammonia emission from animal houses. In ‘‘Proceedings of the International Conference for Agricultural Machinery & Process Engineering,’’ Vol. 2 pp. 393–402. Seoul, South Korea. Berckmans, D., Ni, J.-Q., Goedseels, V., Roggen, J., Huyberechts, G., and Van Overstraeten, R. (1994). ‘‘Feasibility Study of the Use of an Existing Sensor for the Measurement and Control of Ammonia Emission from Animal Production Unit,’’ Rep. No. VLIM/H/9032. K.U. Leuven and IMEC Leuven, Belgium. Berckmans, D., Vinckier, C., Hendriks, J., Ni, J. Q., Gustin, P., Urbain, B., and Ansay, M. (1998). ‘‘Ammonia Emission and Impact at Pig Farms,’’ Ministry of Middle Classes and Agriculture Brussels, Belgium. Bicudo, J. R., Tengman, C. L., Jacobson, L. D., and Sullivan, J. E. (2000). Odor, hydrogen sulfide and ammonia emissions from swine farms in Minnesota. In ‘‘Conference Proceedings Odors and VOC Emissions 2000,’’ Session 8, p. 20. Water Environment Federation, Cincinnati, Ohio. Blanes-Vidal, V., Topper, P. A., and Wheeler, E. F. (2007). Validation of ammonia emission from dairy cow manure estimated with a non-steady-state, recirculation flux chamber with whole-building emissions. Trans. ASABE 50, 633–640. Bolan, N. S., Saggar, S., Luo, J. F., Bhandral, R., and Singh, J. (2004). Gaseous emissions of nitrogen from grazed pastures: Processes, measurements and modelling, environmental implications, and mitigation. In ‘‘Advances in Agronomy’’ (D. L. Sparks, Ed.), 84, pp. 37–120. Elsevier Academic Press Inc. Amsterdam. Braam, C. R., Smits, M. C. J., Gunnink, H., and Swierstra, D. (1997). Ammonia emission from a double-sloped solid floor in a cubicle house for dairy cows. J. Agric. Eng. Res. 68, 375–386. Brewer, S. K., and Costello, T. A. (1999). In situ measurement of ammonia volatilization from broiler litter using an enclosed air chamber. Trans. ASAE 42, 1415–1422. Brunsch, R. (1997). Methodical aspects relating the results of multigas monitoring. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. I, pp. 185–191. NVTL, Rosmalen, The Netherlands.

Ammonia Sampling and Measurement

259

Bunton, B., O’Shaughnessy, P., Fitzsimmons, S., Gering, J., Hoff, S., Lyngbye, M., Thorne, P. S., Wasson, J., and Werner, M. (2007). Monitoring and modeling of emissions from concentrated animal feeding operations: Overview of methods. Environ. Health Perspect. 115, 303–307. Busse, F.-W. (1993). Comparison measurements of the house climate in swine stables with and without respiratory diseases or cannibalism. In ‘‘Livestock Environment. Fourth International Symposium’’ (E. Collins and C. Boon, Eds.), pp. 904–908. ASAE St. Joseph, Michigan, and University of Warwick, Coventry, England. Cassel, T., Ashbaugh, L., Flocchini, R., and Meyer, D. (2005a). Ammonia emission factors for open-lot dairies: Direct measurements and estimation by nitrogen intake. J. Air Waste Manage. Assoc. 55, 826–833. Cassel, T., Ashbaugh, L., Flocchini, R., and Meyer, D. (2005b). Ammonia flux from openlot dairies: Development of measurement methodology and emission factors. J. Air Waste Manage. Assoc. 55, 816–25. Chadwick, D. R. (2005). Emissions of ammonia, nitrous oxide and methane from cattle manure heaps: Effect of compaction and covering. Atmos. Environ. 39, 787–799. Choinie`re, Y., Marquis, B., and Gingas, G. (1997). Ammonia and contaminant concentrations with conventional versus pit ventilation in finishing pig units. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ (J. A. M. Voermans and G. J. Monteny, Eds.), Vol. I, pp. 365–372. NVTL Rosmalen, The Netherlands. Cortus, E. L. (2006). A Dynamic Model of Ammonia Production within Grow-Finish Swine Barns. Ph.D. Thesis, University of Saskatchewan Saskatoon, SK, Canada. Crook, B., Robertson, J. F., Glass, S., Botheroyd, E. M., Lacey, J., and Topping, M. D. (1991). Airborne dust, ammonia, microorganisms, and antigens in pig confinement houses and the respiratory health of exposed farm-workers. Am. Ind. Hyg. Assoc. J. 52, 271–279. Curtis, S. E., Anderson, C. R., Simon, J., Jensen, A. H., Day, D. L., and Kelley, K. W. (1975). Effects of aerial ammonia, hydrogen sulfide and swine-house dust on the rate of gain and respiratory-tract structure in swine. J. Anim. Sci. 41, 735–739. De Praetere, K., and van Der Biest, W. (1990). Airflow patterns in piggeries with fully slatted floors and their effect on ammonia distribution. J. Agric. Eng. Res. 46, 31–44. Demmers, T. G. M., Burgess, L. R., Short, J. L., Philips, V. R., Clark, J. A., and Wathes, C. M. (1999). Ammonia emissions from two mechanically ventilated UK livestock buildings. Atmos. Environ. 33, 217–227. den Brok, G. M., and Verdoes, N. (1997). Slurry cooling to reduce ammonia emission from pig houses. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. II, pp. 441–447. NVTL Rosmalen, The Netherlands. Dewey, C. E., Cox, B., and Leyenaar, J. (2000). Measuring ammonia concentrations in the barn using the Draeger (TM) and pHydrion (TM) tests. Swine Health Prod. 8, 127–131. Doebelin, E. O. (1983). ‘‘Measurement Systems: Application and Design,’’ 3rd ed. McGrawHill, New York. Donham, K. J., Reynolds, S. J., Whitten, P., Merchant, J. A., Burmeister, L., and Popendorf, W. J. (1995). Respiratory dysfunction in swine production facility workers dose-response relationships of environmental exposures and pulmonary-function. Am. J. Ind. Med. 27, 405–418. Einfeld, W., and Billets, S. (1998). ‘‘Environmental Technology Verification Report: Photoacoustic Spectrophotometer Innova AirTech Instruments Type 1312 Multi-gas Monitor,’’ Rep. No. EPA/600/R-98/143. Sandia National Laboratories Albuquerque, New Mexico. Eisentraut, M. A., Heber, A. J., and Ni, J.-Q. (2004). CAPECAB Part I: Processing barn emission data. In ‘‘A&WMA’s 97th Annual Conference & Exhibition,’’ p. 18. A&WMA Indianapolis, Indiana.

260

Ji-Qin Ni and Albert J. Heber

Elwinger, K., and Svensson, L. (1996). Effect of dietary protein content, litter and drinker type on ammonia emission from broiler houses. J. Agric. Eng. Res. 64, 197–208. Elzing, A., and Swierstra, D. (1993). Ammonia emission measurements in a model system of a pig house. In ‘‘Proceedings of the First International Symposium on Nitrogen Flow in Pig Production and Environmental Consequences’’ (M. W. A. Verstegen, L. A. den Hartog, G. J. M. van Kempen, and J. H. M. Metz, Eds.) pp. 280–285. Pudoc Scientific Publishers Wageningen, The Netherlands. EMEP. (2001). ‘‘EMEP Manual for Sampling and Chemical Analysis,’’ Rep. No. EMEP/ CCC-1/95. Norwegian Institute for Air Research Kjeller, Norwa. Ferguson, N. S., Gates, R. S., Cantor, A. H., Taraba, J. L., Pescatore, A. J., Straw, M. L., Ford, M. J., Turner, L. W., and Burnham, D. J. (1997). Effects of dietary crude protein on growth, ammonia concentration and litter composition in broilers. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 974069. ASAE St. Joseph, Michigan. Ferm, M. (1979). Method for determination of atmospheric ammonia. Atmos. Environ. 13, 1385–1393. Ferm, M., Marcinkowski, T., Kieronczyk, M., and Pietrzak, S. (2005). Measurements of ammonia emissions from manure storing and spreading stages in Polish commercial farms. Atmos. Environ. 39, 7106–7113. Finkelstein, L., and Grattan, K. T. V. (Eds.). (1994). ‘‘Concise Encyclopedia of Measurement & Instrumentation,’’ pp. 1–434. Pergamon Press, Oxford. Fitz, D. R., Pisano, J. T., Malkina, I. L., Goorahoo, D., and Krauter, C. F. (2003). A passive flux denuder for evaluating emissions of ammonia at a dairy farm. J. Air Waste Manage. Assoc. 53, 937–945. Gallmann, E., and Hartung, E. (2000). Evaluation of the emission rates of ammonia and greenhouse gases from swine housings. In ‘‘Proceedings of 2nd International Conference on Air Pollution from Agricultural Operations, pp. 92–99. ASAE St. Joseph, Michigan, Des Moines, Iowa. Gates, R. S., Xin, H., Casey, K. D., Liang, Y., and Wheeler, E. F. (2005). Method for measuring ammonia emissions from poultry houses. J. Appl. Poult. Res. 14, 622–634. Gay, S. W., Wheeler, E. F., Zajaczkowski, J. L., and Topper, P. A. (2006). Ammonia emissions from US tom turkey growout and brooder houses under cold weather minimum ventilation. Appl. Eng. Agric. 22, 127–134. Godbout, S., Lemay, S. P., Joncas, R., Larouche, J. P., Martin, D. Y., Leblanc, M., Marquis, A., Bernier, J. F., Zijlstra, R. T., Barber, E. M., and Masse, D. (2000). Reduction of odour and gas emissions from swine buildings using canola oil sprinkling and alternate diets. In ‘‘Air Pollution from Agricultural Operations. Proceedings of the Second International Conference,’’ pp. 211–219. ASAE St. Joseph, Michigan, Des Moines, Iowa. Go¨pel, W., and Schierbaum, K.-D. (1991). Chemical and biochemical sensors, Part 1. In ‘‘Sensors, A Comprehensive Survey’’ (W. Go¨pel, T. A. Jones, M. Kleitz, J. Lundstro¨m, and T. E. Seiyama, Eds.) pp. 119–158. VCH Weinheim. Groenestein, C. M., and Faassen, H. G. V. (1996). Volatilization of ammonia, nitrous oxide and nitric oxide in deep-litter systems for fattening pigs. J. Agric. Eng. Res. 65, 269–274. Groenestein, C. M., Vermeer, H. M., and Hol, J. M. G. (1997). Ammonia emission and feeding-induced activity from houses with sows kept individually and in groups. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ (J. A. M. Voermans and G. J. Monteny, Eds.), Vol. II. pp. 553–560. NVTL, Rosmalen, The Netherlands. Groenestein, C. M., Hendriks, M. M. W. B., and den Hartog, L. A. (2003). Effect of feeding schedule on ammonia emission from individual and group-housing systems for sows. Biosyst. Eng. 85, 79–85.

Ammonia Sampling and Measurement

261

Groenestein, C. M., den Hartog, L. A., and Metz, J. H. M. (2006). Potential ammonia emissions from straw bedding, slurry pit and concrete floors in a group-housing system for sows. Biosyst. Eng. 95, 235–243. Groenestein, C. M., Monteny, G. J., Aarnink, A. J. A., and Metz, J. H. M. (2007). Effect of urinations on the ammonia emission from group-housing systems for sows with straw bedding: Model assessment. Biosyst. Eng. 97, 89–98. Groot Koerkamp, P. W. G., Metz, J. H. M., Uenk, G. H., Phillips, V. R., Holden, M. R., Sneath, R. W., Short, J. L., White, R. P., Hartung, J., Seedorf, J., Schroder, M., Linkert, K. H., et al. (1998). Concentrations and emissions of ammonia in livestock buildings in Northern Europe. J. Agric. Eng. Res. 70, 79–95. Guiziou, F., and Beline, F. (2005). In situ measurement of ammonia and greenhouse gas emissions from broiler houses in France. Bioresour. Technol. 96, 203–207. Gustafsson, G., and von Wachenfelt, E. (2005). Measures against ammonia release in a floor housing system for laying hens. Agric. Eng. Int.: CIGR Ej, VII, Manuscript BC0 5003. Harper, L. A., Sharpe, R. R., and Simmons, J. D. (2004). Ammonia emissions from swine houses in the southeastern United States. J. Environ. Qual. 33, 449–457. Harris, D. B. (2001). ‘‘Personal communication.’’ Harris, D. B., Shores, R. C., and Jones, L. G. (2001). Ammonia emission factors from swine finishing operations. In ‘‘10th Annual International Emission Inventory Conference, One Atmosphere, One Inventory, Many Challenges.’’ U.S. EPA, Denver, Colorado. Hartung, E., Jungbluth, T., and Bu¨scher, W. (1997). Reduction of ammonia and odor emissions from a piggery with biofilters. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 974126. ASAE St. Joseph, Michigan. Hartung, E., Jungbluth, T., and Bu¨scher, W. (2001). Reduction of ammonia and odor emissions from a piggery with biofilters. Trans. ASAE 44, 113–118. Hashimoto, A. G. (1972). Aeration under caged lying hens. Trans. ASAE 15, 1119–1123. Hauser, R. H., and Fo¨lsch, D. W. (1993). The quality of poultry-house in alternative systems for farming laying hens. In ‘‘Fourth International Symposium of Livestock Environment’’ (E. Collins and C. Boon, Eds.), pp. 671–677. ASAE, St. Joseph, Michigan, and University of Warwick, Coventry, England. Hayes, E. T., Leek, A. B. G., Curran, T. P., Dodd, V. A., Carton, O. T., Beattie, V. E., and O’Doherty, J. V. (2004). The influence of diet crude protein level on odour and ammonia emissions from finishing pig houses. Bioresour. Technol. 91, 309–315. Hayes, E. T., Curran, T. P., and Dodd, V. A. (2006). Odour and ammonia emissions from intensive pig units in Ireland. Bioresour. Technol. 97, 940–948. Heber, A. J., Duggirala, R. K., Ni, J.-Q., Spence, M. L., Haymore, B. L., Adamchuk, V. I., Bundy, D. S., Sutton, A. L., Kelly, D. T., and Keener, K. M. (1997). Manure treatment to reduce gas emissions from large swine houses. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. II pp. 449–458. NVTL, Rosmalen, The Netherlands. Heber, A. J., Lim, T. T., Ni, J.-Q., and Sutton, A. L. (2000a). ‘‘Odor Emission from Anaerobic Treatment of Swine Manure. Final Report to Indiana’s Value-Added Agricultural Grant Program, Office of the Commissioner of Agriculture.’’ Purdue University West Lafayette, Indiana. Heber, A. J., Ni, J.-Q., Lim, T. T., Diehl, C. A., Sutton, A. L., Duggirala, R. K., Haymore, B. L., Kelly, D. T., and Adamchuk, V. I. (2000b). Effect of a manure additive on ammonia emission from swine finishing buildings. Trans. ASAE 43, 1895–1902. Heber, A. J., Ni, J. Q., Haymore, B. L., Duggirala, R. K., and Keener, K. M. (2001). Air quality and emission measurement methodology at swine finishing buildings. Trans. ASAE 44, 1765–1778. Heber, A. J., Ni, J.-Q., Lim, T.-T., Beasley, D. B., Hoff, S. J., Jacobson, L. D., Koziel, J., and Zhang, Y. (2004). Proposed standard for measuring air emissions from livestock

262

Ji-Qin Ni and Albert J. Heber

housing. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 044181. ASAE St. Joseph, Michigan. Heber, A. J., Tao, P.-C., Ni, J.-Q., Lim, T. T., and Schmidt, A. M. (2005). Air emissions from two swine finishing building with flushing: Ammonia characteristics. In ‘‘Seventh International Livestock Environment Symposium,’’ pp. 436–443. ASAE Beijing, China. Heber, A. J., Ni, J.-Q., Lim, T.-T., Schmidt, A. M., Koziel, J. A., Tao, P. C., Beasley, D. B., Hoff, S. J., Nicolai, R. E., Jacobson, L. D., and Zhang, Y. (2006). Quality assured measurements of animal building emissions: Part 1. Gas concentrations. J. Air Waste Manage. Assoc. 56, 1472–1483. Heederik, D., Sigsgaard, T., Thorne, P. S., Kline, J. N., Avery, R., Bonlokke, J. H., Chrischilles, E. A., Dosman, J. A., Duchaine, C., Kirkhorn, S. R., Kulhankova, K., and Merchant, J. A. (2007). Health effects of airborne exposures from concentrated animal feeding operations. Environ. Health Perspect. 115, 298–302. Heinrichs, P., and Oldenburg, J. (1993). Effect of protein feeding on gaseous ammonia emissions and slurry loading with nitrogen in fattening pigs. In ‘‘Proceedings of the First International Symposium on Nitrogen Flow in Pig Production and Environmental Consequences’’ (M. W. A. Verstegen, L. A. den Hartog, G. J. M. van Kempen, and J. H. M. Metz, Eds.) pp. 336–339. Pudoc Scientific Publishers Wageningen (Doorwerth), The Netherlands. Hendriks, J. G. L., and Vrielink, M. G. M. (1997). Reducing ammonia emission from pig houses by adding or producing organic acids in pig slurry. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. II, pp. 493–451. NVTL, Rosmalen, The Netherlands. Hess, H. J., and Hu¨gle, T. (1994). Comparative measuring on ammonia emissions. Landtechnik 49, 362–363. Hesse, D. (1994). Comparison of different old and new fattening pig husbandrys with focus on environment and animal welfare. In ‘‘Proceedings of the XII World Congress on Agricultural Engineering,’’ pp. 559–566. Milano. Hoff, S. J. (2005). ‘‘Personal Communication’’. Hoff, S. J., Bundy, D. S., Nelson, M. A., Zelle, B. C., Jacobson, L. D., Heber, A. J., Ni, J.-Q., Zhang, Y. H., Koziel, J. A., and Beasley, D. B. (2006). Emissions of ammonia, hydrogen sulfide, and odor before, during and after slurry removal from a deep-pit swine finisher. J. Air Waste Manage. Assoc. 56, 581–590. Hoy, S. (1995). Studies on the use of multi-gas monitoring in animal houses. Tierarztliche Umschau 50, 115–123. Hoy, S., and Willig, R. (1994). Results of continuous measurement of ammonia in the air of pig houses, using electrochemical sensors [Ergebnisse von Verlaufsmessungen der Ammoniak-Konzentration in Schweinestallen mit Hilfe der Sensortechnik]. Monatshefte fu¨r Veterina¨rmedizin 49, 37–41. Hoy, S., Willig, R., and Buchholz, I. (1992). Results from continuous measurements of ammonia in keeping fattening pigs on deep litter with additives in comparison with housing on slatted floors, In ‘‘Workshops: Deep Litter Systems for Pigs’’ ( J. A. M. Voermans, Ed.), pp. 37–50. Rosmalen The Netherlands. Hoy, S., Mu¨ller, K., and Willig, R. (1994). Ammonia concentrations in pig houses with different types of floors—Deep-litter husbandry with bioactivator compared to fully slatted floor husbandry (Zur Ammoniak-Konzentration bei zwei Systemen der Mastschweinehaltung auf Tiefstreu mit Bioaktivator im Vergleich zur Vollspaltenbodenhaltung). Tiera¨rztliche Umschau 49, 414–416, 418–420.

Ammonia Sampling and Measurement

263

Jacobson, L. D., Janni, K. A., Arellano, P. E., and Pijoan, C. J. (1992). Winter swine ventilation evaluation using air quality criteria. In ‘‘ASAE International Summer Meeting.’’ ASAE Paper No. 924039. ASAE St. Joseph, Michigan. Jeppsson, K. H. (1999). Volatilization of ammonia in deep-litter systems with different bedding materials for young cattle. J. Agric. Eng. Res. 73, 49–57. Jeppsson, K. H. (2002). Diurnal variation in ammonia, carbon dioxide and water vapour emission from an uninsulated, deep litter building for growing/finishing pigs. Biosyst. Eng. 81, 213–223. Jiang, J. K., and Sands, J. R. (2000). ‘‘Odour and Ammonia Emission from Broiler Farms,’’ Rep. No. Publication No. 0012. Rural Industries Research and Development Corporation Kingston, Australia. Johnston, S. (1991). ‘‘Fourier Transform Infrared: A Constantly Evolving Technology.’’ E. Horwood New York. Johnston, N. L., Quarles, C. L., Fagerberg, D. J., and Caveny, D. D. (1981). Evaluation of yucca saponin on broiler performance and ammonia suppression. Poult. Sci. 60, 2289–2292. Jungbluth, T., and Bu¨scher, W. (1996). Reduction of ammonia emissions from piggeries. In ‘‘1996 ASAE Annual International Meeting.’’ ASAE Paper No. 964091. ASAE St. Joseph, Michigan. Jungbluth, T., Brose, G., and Hartung, E. (1997). Ammonia and greenhouse gas emissions from dairy barns. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 974127. ASAE St. Joseph, Michigan. Kamin, H., Barber, J. C., Brown, S. I., Delwiche, C. C., Grosjean, D., Hales, J. M., Knapp, J. L. W., Lemon, E. R., Martens, C. S., Niden, A. H., and Wilson, R. P. (1979). ‘‘Ammonia,’’, 384 p. University Park Press, Baltimore, Maryland. Kant, P. P. H., Verboon, M. C., and Huis in ‘t Veld, J. W. H. (1992). ‘‘Ammonia Emission Measurements with the Lindvall Box,’’ A report on the measurements carried out at Waiboerhoeve Experimental Farm in 1989–1991. (Ammoniak-emissiemetingen met de Lindvalldoos - Inventarisatie van de metingen op de Waiboerhoeve in 1989–1991). Lelystad en IMAG Wageningen, The Netherlands. Kavolelis, B. (2006). Impact of animal housing systems on ammonia emission rates. Polish J. Environ. Stud. 15, 739–745. Kay, R. M., and Lee, P. A. (1997). Ammonia emission from pig buildings and characteristics of slurry produced by pigs offered low crude protein diets. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), pp. 253–260. NVTL Rosmalen, The Netherlands. Kay, R. M., Pain, B. F., Clarkson, C. R., and Misselbrook, T. H. (1992). The effect of a slurry additive on odour and ammonia emissions from pig buildings. Anim. Prod. 54, 2p–22. Keck, M., Bu¨scher, W., and Jungbluth, T. (1994). Influence on the ammonia emission by under- and overfloor extraction in a pig house. In ‘‘AgEng ’94. International Conference on Agricultural Engineering,’’ pp. 230–231, Milano. Kirkhorn, S. R., and Garry, V. F. (2000). Agricultural lung diseases. Environ. Health Perspect. 108, 705–712. Krause, K.-H. (1993). On dispersion of ammonia emissions (Zur Ausbreitung van Ammoniakemissionen). Landtechnik 48, 562–569. Krause, K.-H., and Janssen, J. (1990). Measuring and simulation of the distribution of ammonia in animal houses. ‘‘Room Vent ’90, Session C-7,’’ pp. 1–12, Oslo, Norway. Krause, K.-H., and Janssen, J. (1991). Modelling the dispersion of ammonia within animal houses. In ‘‘Cost 681 Expert Odours Group Workshop: Odour and Ammonia Emissions from Livestock Farming’’ (V. C. Nielsen, J. H. Voorburg, and P. L’Hermite, Eds.) pp. 71–80. Elsevier Applied Science London/New York, Silsoe, UK.

264

Ji-Qin Ni and Albert J. Heber

Krieger, R., Hartung, J., and Pfeiffer, A. (1993). Experiments with a feed additive to reduce ammonia emissions from pig fattening housing - preliminary results. In ‘‘Proceedings of the First International Symposium on Nitrogen Flow in Pig Production and Environmental Consequences’’ (M. W. A. Verstegen, L. A. den Hartog, G. J. M. van Kempen, and J. H. M. Metz, Eds.) pp. 295–300. Pudoc Scientific Publishers, Wageningen (Doorwerth), The Netherlands. Kroodsma, W., HuisintVeld, J. W. H., and Scholtens, R. (1993). Ammonia emission and its reduction from cubicle houses by flushing. Livestock Prod. Sci. 35, 293–302. Kurvits, T., and Marta, T. (1998). Agricultural NH3 and NOx emissions in Canada. Environ. Pollut. 102, 187–194. Lefcourt, A. M. (2002). Some potential problems for measuring ammonia emissions from farm structures. Trans. ASAE 45, 1585–1591. Leithe, W. (1971). ‘‘The Analysis of Air Pollutants.’’ Ann Arbor Science Publishers Ann Arbor, Michigan. Li, H., Xin, H., Liang, Y., Gates, R. S., Wheeler, E. F., and Heber, A. J. (2005). Comparison of direct vs. indirect ventilation rate determinations in layer barns using manure belts. Trans. ASAE 48, 367–372. Lim, T. T., Heber, A. J., Ni, J.-Q., Sutton, A. L., and Shao, P. (2003). Odor and gas release from anaerobic treatment lagoons for swine manure. J. Environ. Qual. 32, 406–416. Lim, T. T., Heber, A. J., Ni, J.-Q., Kendall, D., and Richert, B. T. (2004). Effects of manure removal strategies on odor and gas emission from swine finishing. Trans. ASAE 47, 2041–2050. Lindvall, T., Nore´n, O., and Thyselius, L. (1974). Odor reduction for liquid manure systems. Trans. ASAE 17, 510–512. Liu, Q., Bundy, D. S., and Hoff, S. J. (1993). Utilizing ammonia concentrations as an odor threshold indicator for swine facilities. In ‘‘Livestock Environment. Fourth International Symposium’’ (E. Collins and C. Boon, Eds.), pp. 678–685. ASAE, St. Joseph, Michigan, and University of Warwick, Coventry, England. Losada, J. M. (2007). Measurement methods and strategies. In ‘‘Ammonia, the Case of the Netherlands’’ (D. A. J. Starmans and K. W. V. D. Hoek, Eds.), pp. 103–123. Wageningen Academic Publishers, Wageningen, The Netherlands. Loyon, L., Guiziou, F., Beline, E., and Peu, P. (2007). Gaseous Emissions (NH3, N2O, CH4 and CO2) from the aerobic treatment of piggery slurry - Comparison with a conventional storage system. Biosyst. Eng. 97, 472–480. Maghirang, R. G., and Manbeck, H. B. (1993). Dust, ammonia, and carbon dioxide emissions from a poultry house. In ‘‘ASAE International Summer Meeting.’’ ASAE Paper No. 934056. ASAE St. Joseph, Michigan. Maghirang, R. G., Manbeck, H. B., Roush, W. B., and Muir, F. V. (1991). Air contaminant distributions in a commercial laying house. Trans. ASAE 34, 2171–2180. Mannebeck, H., and Oldenburg, J. (1991). Comparison of the effects of different systems on ammonia emissions. In ‘‘Cost 681 Expert Odours Group Workshop: Odour and Ammonia Emissions from Livestock Farming’’ (V. C. Nielsen, J. H. Voorburg, and P. L’Hermite, Eds.) pp. 42–49. Elsevier Applied Science London/New York, Silsoe, UK. McCulloch, R. B., Few, G. S., Murray, G. C., and Aneja, V. P. (1998). Analysis of ammonia, ammonium aerosols and acid gases in the atmosphere at a commercial hog farm in eastern North Carolina, USA. Environ. Pollut. 102, 263–268. McCulloch, R. B., Walker, J. T., Chauhan, J. S., and Miller, K. (2000). Performance characteristics of a chemiluminescence ammonia analyzer. In ‘‘Proceedings of the International Symposium on Measurement of Toxic and Related Air Pollutants.’’ Air and Waste Management Association Research Triangle Park, North Carolina. McGinn, S. M., and Janzen, H. H. (1998). Ammonia sources in agriculture and their measurement. Can. J. Soil Sci. 78, 139–148.

Ammonia Sampling and Measurement

265

Mennen, M. G., VanElzakker, B. G., VanPutten, E. M., Uiterwijk, J. W., Regts, T. A., VanHellemond, J., Wyers, G. P., Otjes, R. P., Verhage, A. J. L., Wouters, L. W., Heffels, C. J. G., Romer, F. G., et al. (1996). Evaluation of automatic ammonia monitors for application in an air quality monitoring network. Atmos. Environ. 30, 3239–3256. Meyer, V., and Bundy, D. (1991). Farrowing building air quality survey. In ‘‘International Summer Meeting ASAE.’’ ASAE Paper No. 914012. ASAE St. Joseph, Michigan. Misselbrook, T. H., Pain, B. F., and Headon, D. M. (1998). Estimates of ammonia emission from dairy cow collecting yards. J. Agric. Eng. Res. 71, 127–135. Monteny, G. J., and Erisman, J. W. (1998). Ammonia emission from dairy cow buildings: A review of measurement techniques, influencing factors and possibilities for reduction. Neth. J. Agric. Sci. 46, 225–247. Monteny, G. J., Smits, M. C. J., van Duinkerken, G., Mollenhorst, H., and de Boer, I. J. M. (2002). Prediction of ammonia emission from dairy barns using feed characteristics part II: Relation between urinary urea concentration and ammonia emission. J. Dairy Sci. 85, 3389–3394. Mosquera, J., Monteny, G. J., and Erisman, J. W. (2005). Overview and assessment of techniques to measure ammonia emissions from animal houses: The case of the Netherlands. Environ. Pollut. 135, 381–388. Moum, S. G., Seltzer, W., and Goldhaft, T. M. (1969). A simple method of determining concentrations of ammonia in animal quarters. Poult. Sci. 48, 347–348. Mount, G. H., Rumburg, B. P., Lamb, B. L., Havig, J. R., Westberg, H., Johnson, K. A., and Kincaid, R. L. (2001). DOAS measurement of atmospheric ammonia emissions at a dairy. In ‘‘10th Annual International Emission Inventory Conference, One Atmosphere, One Inventory, Many Challenges.’’, U.S. EPA Denver, Colorado. Myers, J., Kelly, T., Lawrie, C., and Riggs, K. (2000). ‘‘Environmental Technology Verification Report: Opsis Inc. AR-500 Ultraviolet Open-Path Monitor.’’ Battelle Columbus, Ohio. Neser, S., Depta, G., Stegbauer, B., Gronauer, A., and Scho¨n, H. (1997). Mass balance of the compounds nitrogen and carbon in housing systems for laying hens. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ (J. A. M. Voermans and G. J. Monteny, Eds.), Vol. I pp. 129–136. NVTL, Rosmalen, The Netherlands. Ni, J.-Q. (1998). Emission of Carbon Dioxide and Ammonia from a Mechanically Ventilated Pig House. Ph.D. Thesis, Catholic University of Leuven Leuven, Belgium. Ni, J.-Q. (1999). Mechanistic models of ammonia release from liquid manure, a review. J. Agric. Eng. Res. 72, 1–17. Ni J.-Q. (Unpublished). Comparison of a Drager sensor and an ammonia analyzer for NH3 measurement at a fattening pig house. Ni, J. Q., Vinckier, C., Coenegrachts, J., and Hendriks, J. (1999). Effect of manure on ammonia emission from a fattening pig house with partly slatted floor. Livestock Prod. Sci. 59, 25–31. Ni, J.-Q., Heber, A. J., Diehl, C. A., and Lim, T. T. (2000a). Ammonia, hydrogen sulphide and carbon dioxide from pig manure in under-floor deep pits. J. Agric. Eng. Res. 77, 53–66. Ni, J.-Q., Heber, A. J., Lim, T. T., Diehl, C. A., Duggirala, R. K., Haymore, B. L., and Sutton, A. L. (2000b). Ammonia emission from a large mechanically-ventilated swine building during warm weather. J. Environ. Qual. 29, 751–758. Ni, J.-Q., Hendriks, J., Vinckier, C., and Coenegrachts, J. (2000c). Development and validation of a dynamic mathematical model of ammonia release in pig house. Environ. Int. 26, 105–115. Ni, J.-Q., Hendriks, J., Vinckier, C., and Coenegrachts, J. (2000d). A new concept of carbon dioxide accelerated ammonia release from liquid manure in pig house. Environ. Int. 26, 97–104.

266

Ji-Qin Ni and Albert J. Heber

Nicholson, F. A., Chambers, J., and Walker, A. W. (2004). Ammonia emissions from broiler litter and laying hen manure management systems. Biosyst. Eng. 89, 175–185. Nicks, B., Marlier, D., and Canart, B. (1993). Air pollution levels in pig houses. In ‘‘Fourth International Symposium of Livestock Environment’’ (E. Collins and C. Boon, Eds.), pp. 635–642. ASAE St. Joseph, Michigan, and University of Warwick, Coventry, England. Nicks, B., De´siron, A., and Canart, B. (1997). Deep litter materials and the ammonia emissions in fatterning pig houses. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. I, pp. 335–342. NVTL, Rosmalen, The Netherlands. Nimmermark, S., and Gustafsson, G. (2005). Influence of temperature, humidity and ventilation rate on the release of odor and ammonia in a floor housing system for laying hens. Agric. Eng. Int.: CIGR Ej VII, Manuscript BC 04008. Ogink, N. W. M., and Kroodsma, W. (1996). Reduction of ammonia emission from a cow cubicle house by flushing with water or a formalin solution. J. Agric. Eng. Res. 63, 197–204. Osada, T., and Fukumoto, Y. (2001). Development of a new dynamic chamber system for measuring harmful gas emissions from composting livestock waste. Water Sci. Technol. 44, 79–86. Osada, T., Rom, H. B., and Dahl, P. (1998). Continuous measurement of nitrous oxide and methane emission in pig units by infrared photoacoustic detection. Trans. ASAE 41, 1109–1114. PAAQL. (2007). ‘‘Quality Assurance Project Plan for the National Air Emissions Monitoring Study (Barns Component),’’ Rep. No. Revision 1.0. Agriculture Air Research Council, Purdue Agricultural Air Quality Laboratory, Department of Agricultural and Biological Engineering, Purdue University West Lafayette, Indiana. Parbst, K. E., Keener, K. M., Heber, A. J., and Ni, J.-Q. (2000). Comparison of a low cost and high cost odor monitoring equipment in a commercial swine finishing house. Appl. Eng. Agric. 16, 693–699. Patni, N. K., and Clarke, S. P. (1991). Transient hazardous conditions in animal buildings due to manure gas released during slurry mixing. Appl. Eng. Agric. 7, 478–484. Pfeiffer, A., Arends, F., Langholz, H. J., and Steffens, G. (1993). The influence of various pig housing systems and dietary protein levels on the amount of ammonia emissions in the case of fattening pigs. In ‘‘Proceedings of the First International Symposium on Nitrogen Flow in Pig Production and Environmental Consequences’’ (M. W. A. Verstegen, L. A. den Hartog, G. J. M. van Kempen, and J. H. M. Metz, Eds.), pp. 313–317. Pudoc Scientific Publishers Wageningen (Doorwerth), The Netherlands. Phillips, V. R., Lane, S. J., and Burgess, L. R. (2000). A technique for measuring ammonia emissions from the individual parts of a livestock building. In ‘‘Air Pollution from Agricultural Operations, Proceedings of the Second International Conference,’’ pp. 84–91. ASAE Des Moines, Iowa. Phillips, V. R., Lee, D. S., Scholtens, R., Garland, J. A., and Sneath, R. W. (2001). A review of methods for measuring emission rates of ammonia from livestock buildings and slurry or manure stores, part 2: Monitoring flux rates, concentrations and airflow rates. J. Agric. Eng. Res. 78, 1–14. Portejoie, S., Martinez, J., and Landmann, G. (2002). Ammonia of farm origin: Impact on human and animal health and on the natural habitat. Prod. Anim. 15, 151–160. Pranitis, D., and Meyerhoff, M. (1987). Continuous monitoring of ambient ammonia with a membrane-electrode-based detector. Anal. Chem. 59, 2345–2350. Pratt, S. E., Lawrence, L. M., Barnes, T., Powell, D., and Warren, L. K. (2000). Measurement of ammonia concentrations in horse stalls. J. Equine Vet. Sci. 20, 197–200.

Ammonia Sampling and Measurement

267

Reece, F. N., Bates, B. J., and Lott, B. D. (1979). Ammonia control in broiler houses. Poult. Sci. 58, 754–755. Reece, F. N., Lott, B. D., and Deaton, W. (1980). Ammonia in the atmosphere during brooding affects performance of broiler chickens. Poult. Sci. 59, 486–488. Rodhe, L., and Karlsson, S. (2002). Ammonia emissions from Broiler manure—Influence of storage and spreading method. Biosyst. Eng. 82, 455–462. Rom, H. B. (1993). Ammonia emission from livestock buildings in Denmark. In ‘‘Fourth International Symposium of Livestock Environment’’ (E. Collins and C. Boon, Eds.), pp. 1161–1168. ASAE, St. Joseph, Michigan, and University of Warwick, Coventry, England. Saggar, S., Bolan, N. S., Bhandral, R., Hedley, C. B., and Luo, J. (2004). A review of emissions of methane, ammonia, and nitrous oxide from animal excreta deposition and farm effluent application in grazed pastures. New Zealand J. Agric. Res. 47, 513–544. Schmidt-Van Riel, C. J. M. (1991). Experiences with automatic ammonia analyses. In ‘‘Cost 681 Expert Odours Group Workshop: Odour and Ammonia Emissions from Livestock Farming’’ (V. C. Nielsen, J. H. Voorburg, and P. L’Hermite, Eds.) pp. 67–70. Elsevier Applied Science, London/New York, Silsoe, UK. Scholtens, R. (1990). Ammonia emission measurements in animal housing with forced ventilation. In ‘‘Ammoniak in der Umwelt. Kreisla¨ufe, Minderung. Proceedings of a Conference’’ (H. Do¨hler and H. van den Weghe, Eds.), p. 9. Braunschweig Germany. Scholtens, R. (1993). ‘‘Gas Detector Tubes for Ammonia Dra¨ger and Kitagawa (Gasdetectiebuisjes voor ammoniak: Dra¨ger en Kitagawa),’’ Rep. No. Nota V93-54. Instituut voor Mechanisatie Arbeid en Gebouwen, Dienst Landbouwkundig Onderzoek. Scholtens, R., Hol, J. M. G., Wagemans, M. M., and Phillips, V. R. (2003). Improved passive flux samplers for measuring ammonia emissions from animal houses, part 1: Basic principles. Biosyst. Eng. 85, 95–100. Scholtens, R., Dore, C. J., Jones, B. M. R., Lee, D. S., and Phillips, V. R. (2004). Measuring ammonia emission rates from livestock buildings and manure stores—part 1: Development and validation of external tracer ratio, internal tracer ratio and passive flux sampling methods. Atmos. Environ. 38, 3003–3015. Scott Specialty Gases. (2004). 57% of specialty gas vendors fail audit of EPA protocol gases. Stott Tech Newsline 1, 1. Secrest, C. D. (2001a). Field measurement of air pollutants near swine confined-animal feeding operations using UV DOAS and FTIR. In ‘‘Proceedings of SPIE: Water, Ground, and Air Pollution Monitoring and Remediation’’ (T. Vo-Dinh and R. L. Spellicy, Eds.), Vol. 4199, pp. 98–104. Society of Photo Optical, Boston, MA, USA. Secrest, C. D. (2001b). ‘‘Personal communication.’’. Seltzer, W., Moum, S. G., and Goldhaft, T. M. (1969). A method of treatment of animal waste to control ammonia and other odours. Poult. Sci. 48, 1912–1918. Shah, S. B., Westerman, P. W., and Arogo, J. (2006). Measuring ammonia concentrations and emissions from agricultural land and liquid surfaces: A review. J. Air Waste Manage. Assoc. 56, 945–960. Shores, R. C., Harris, D. B., Thompson, E. L., Vogel, C. A., Natschke, D., Hashmonay, R. A., Wagoner, K. R., and Modrak, M. (2005). Plane-integrated openpath Fourier transform infrared spectrometry methodology for anaerobic swine lagoon emission measurements. Appl. Eng. Agric. 21, 487–492. Skewes, P. A., and Harmon, J. D. (1995). Ammonia quick test and ammonia dosimeter tubes for determining ammonia levels in broiler facilities. J. Appl . Poult. Res. 4, 148–153. Slanina, S. (1994). Forest dieback and ammonia—A typical Dutch problem. Chem. Int. 16, 2–3. Smits, M. C. J., Valk, H., Elzing, A., and Keen, A. (1995). Effect of protein nutrition on ammonia emission from a cubicle house for dairy-cattle. Livestock Prod. Sci. 44, 147–156.

268

Ji-Qin Ni and Albert J. Heber

Snell, H. G. J., and Van den Weghe, H. (1999). Reduction of ammonia emissions and optimization of the stable air temperature by using earth-tube heat exchangers. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 994133. ASAE St. Joseph, Michigan. Snell, H. G. J., Seipelt, F., and van den Weghe, H. F. A. (2003). Ventilation rates and gaseous emissions from naturally ventilated dairy houses. Biosyst. Eng. 86, 67–73. Sommer, S. G., McGinn, S. M., Hao, X., and Larney, F. J. (2004a). Techniques for measuring gas emissions from a composting stockpile of cattle manure. Atmos. Environ. 38, 4643–4652. Sommer, S. G., Schjoerring, J. K., and Denmead, O. T. (2004b). Ammonia emission from mineral fertilizers and fertilized crops. In ‘‘Advances in Agronomy’’ (D. L. Sparks, Ed.), 82, pp. 557–622. Elsevier Academic Press Inc. Amsterdam. Sommer, S. G., Zhang, G. Q., Bannink, A., Chadwick, D., Misselbrook, T., Harrison, R., Hutchings, N. J., Menzi, H., Monteny, G. J., Ni, J. Q., Oenema, O., and Webb, J. (2006). Algorithms determining ammonia emission from buildings housing cattle and pigs and from manure stores. In ‘‘Advances in Agronomy’’ (D. L. Sparks, Ed.), 89, pp. 264–335. Elsevier Academic Press Inc. Amsterdam. Stern, A. C., Boubel, R. W., Turner, C. B., and Fox, D. L. (1984). ‘‘Fundamentals of Air Pollution,’’ 2nd ed. Academic Press Inc. San Diego, California. Stowell, R. R., and Foster, S. (2000). Ammonia emissions from a High-Rise swine finishing facility. In ‘‘ASAE Annual International Meeting.’’ ASAE Paper No. 004080. ASAE St. Joseph, Michigan. Sun, H., Lim, T.T., Heber, A.J., Zhao, L., Ni, J.-Q., Diehl, C. A., Tao, P. C., and Hanni, S. M. (unpublished data). Ammonia Emissions from a Belt-battery Layer House in Ohio. Svensson, L., and Ferm, M. (1993). Mass transfer coefficient and equilibrium concentration as key factors in a new approach to estimate ammonia emission from livestock manure. J. Agric. Eng. Res. 56, 1–11. Svensson, L., Jeppsson, K. -H., and Gustafsson, G. (1997). Evaluation of different methods of measuring NH3 emission in naturally ventilated animal houses with deep litter. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ ( J. A. M. Voermans and G. J. Monteny, Eds.), Vol. I, pp. 209–217. NVTL, Rosmalen, The Netherlands. Swierstra, D., Smits, M. C. J., and Kroodsma, W. (1995). Ammonia emission from cubicle houses for cattle with slatted and solid floors. J. Agric. Eng. Res. 62, 127–132. Taylor, J. R. (1997). ‘‘An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements,’’ University Science Books Sausalito, California. Timmer, B., Olthuis, W., and van den Berg, A. (2005). Ammonia sensors and their applications—A review. Sens. Actuators B-Chem. 107, 666–677. USEPA. (1997). ‘‘EPA Traceability Protocol for Assay and Certification of Gaseous Calibration Standards,’’ National Exposure Research Laboratory, Human Exposure and Atmospheric Science Division, U.S. EPA Research Triangle Park, North Carolina. USEPA. (1998). ‘‘EPA Guidance for Quality Assurance Project Plans,’’ Office of Research and Development, U.S. EPA Washington, DC. Valentine, H. (1964). A study of the effect of different ventilation rates on the ammonia concentrations in the atmosphere of broiler houses. Br. Poult. Sci. 5, 149–159. Valli, L., Piccinini, S., and Bonazzi, G. (1991). Ammonia emission from two poultry manure drying systems. In ‘‘Cost 681 Expert Odours Group Workshop: Odour and Ammonia Emissions from Livestock Farming’’ (V. C. Nielsen, J. H. Voorburg, and P. L’Hermite, Eds.) pp. 50–58. Elsevier Applied Science London/New York, Silsoe, UK. van Breemen, N., Burrough, P. A., Velthorst, E. J., van Dobben, H. F., de Wit, T., Ridder, T. B., and Reijnders, H. F. R. (1982). Soil acidification from atmospheric ammonium sulphate in forest canopy throughfall. Nature 299, 548–550.

Ammonia Sampling and Measurement

269

van der Peet-Schwering, C. M. C., Aarnink, A. J. A., Rom, H. B., and Dourmad, J. Y. (1999). Ammonia emissions from pig houses in The Netherlands, Denmark and France. Livestock Prod. Sci. 58, 265–269. van Ouwerkerk, E. N. J. (1993). ‘‘Methods of Measuring NH3 Emission from Animal Housing (Meetmethoden Ammoniak Emissie uit Stallen),’’ IMAG Wageningen, The Netherlands. van’t Klooster, C. E., and Heitlager, B. P. (1992). ‘‘Measurement Systems for Emissions of Ammonia and Other Gasses at the Research Institute for Pig Husbandry,’’ Research Institute for Pig Husbandry The Netherlands. Verdoes, N., and Ogink, N. W. M. (1997). Odour emission from pig houses with low ammonia emission. In ‘‘International Symposium on Ammonia and Odour Control from Animal Production Facilities’’ (J. A. M. Voermans and G. J. Monteny, Eds.), pp. 317–325. NVTL Rosmalen, The Netherlands. Verstegen, W. A., Van der Hel, W., Jongebreur, A. A., and Enneman, G. (1976). The influence of ammonia and humidity on activity and energy balance data in groups of pigs. Z.Tierphysilo., TierernShrg.u.Futtermittelkde 37, 255–263. Walker, J. T. (2001). ‘‘Personal communication.’’. Wang, Z., Van Cleemput, O., Demeyer, P., and Baert, L. (1991). Effect of urease inhibitors on urea hydrolysis and ammonia volatilization. Biol. Fertil. Soils 11, 43–47. Webb, J., Menzi, H., Pain, B. F., Misselbrook, T. H., Dammgen, U., Hendriks, H., and Dohler, H. (2005). Managing ammonia emissions from livestock production in Europe. Environ. Pollut. 135, 399–406. Wheeler, E. F., Weiss, R. W. J., and Weidenboerner, E. (1999). Evaluation of instrumentation for the measurement of aerial ammonia in poultry houses. In ‘‘1999 ASAE Annual International Meeting.’’ ASAE Paper No. 993188. ASAE St. Joseph, Michigan. Wheeler, E. F., Smith, J. L., and Hulet, R. M. (2000a). Ammonia volatilization from litter during nine broiler flocks. ‘‘Air Pollution from Agricultural Operations, Proceedings of the Second International Conference pp. 25–32. ASAE Des Moines, Iowa. Wheeler, E. F., Weiss, R. W. J., and Weidenboerner, E. (2000b). Evaluation of instrumentation for measuring aerial ammonia in poultry houses. J. Appl. Poult. Res. 9, 443–452. White, R. (1990). ‘‘Chromatography/Fourier Transform Infrared Spectroscopy and Its Applications.’’ Marcel Dekker New York. Willers, H. C., Derikx, P. J. L., Have, P. J. W. T., and Vijn, T. K. (1996). Emission of ammonia and nitrous oxide from aerobic treatment of veal calf slurry. J. Agric. Eng. Res. 63, 345–352. Xin, H., Berry, I. L., and Tabler, G. T. (1996). Minimum ventilation requirement and associated energy cost for aerial ammonia control in broiler houses. Trans. ASAE 39, 645–648. Xin, H., Tanaka, A., Wang, T., Gates, R. S., Wheeler, E. F., Casey, K. D., Heber, A. J., Ni, J.-Q., and Lim, T. (2002). A portable system for continuous ammonia measurement in the field. In ‘‘ASAE Annual International Meeting.’’, ASAE Paper No. 024168. ASAE St. Joseph, Michigan. Xing, G. X., and Zhu, Z. L. (2000). An assessment of N loss from agricultural fields to the environment in China. Nutr. Cycl. Agroecosyst. 57, 67–73. Zahn, J. A., Tung, A. E., Roberts, B. A., and Hatfield, J. L. (2001). Abatement of ammonia and hydrogen sulfide emissions from a swine lagoon using a polymer biocover. J. Air Waste Manage. Assoc. 51, 562–573. Zhang, G., Strom, J. S., Li, B., Rom, H. B., Morsing, S., Dahl, P., and Wang, C. (2005). Emission of ammonia and other contaminant gases from naturally ventilated dairy cattle buildings. Biosyst. Eng. 92, 355–364.

This page intentionally left blank

C H A P T E R

F I V E

Will Stem Rust Destroy the World’s Wheat Crop? Ravi P. Singh,* David P. Hodson,* Julio Huerta-Espino,† Yue Jin,‡ Peter Njau,§ Ruth Wanyera,§ Sybil A. Herrera-Foessel,* and Richard W. Ward*

Contents 1. Introduction 2. Stem Rust Disease, Pathogen, and Epidemiology 3. Breeding for Resistance 3.1. Historical account 3.2. International cooperation 3.3. Spread of semidwarf wheats with resistance to stem rust 3.4. Current knowledge of resistance to stem rust 4. Race Ug99 and Why it is a Potential Threat to Wheat Production 4.1. Avirulence/virulence genes in Ug99 4.2. Current distribution of race Ug99 4.3. Predicting Ug99 migration to other wheat areas 4.4. Resistance/susceptibility of current wheat germplasm 4.5. Can a Ug99 pandemic be predicted? 5. Breeding Strategies to Mitigate the Threat from Ug99 and Achieve a Long-Term Control of Stem Rust 5.1. Prevalence of Sr24 and its breakdown 5.2. Race-specific resistance genes effective to Ug99 5.3. Strategy to use race-specific resistance genes in wheat improvement 5.4. Adult plant resistance to Ug99 in old and new wheat 5.5. High-yielding wheat lines with adult plant resistance to stem rust 5.6. Mexico-Kenya shuttle to breed high-yielding spring wheat with near-immune level of adult plant resistance

* { { }

272 274 277 277 277 278 279 281 281 282 284 285 287 288 289 290 292 295 296 298

International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico INIFAP-CEVAMEX, 56230 Chapingo, Mexico USDA-ARS, Cereal Disease Laboratory, St. Paul, Minnesota 55108 Kenya Agricultural Research Institute, Njoro Plant Breeding Research Center (KARI-NPBRC), Njoro, Kenya

Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00205-8

#

2008 Elsevier Inc. All rights reserved.

271

272

Ravi P. Singh et al.

5.7. Reducing the world’s wheat area under susceptible cultivars 5.8. Efforts to identify and develop resistant wheat varieties in secondary risk areas 6. Conclusion and Future Outlook Acknowledgments References

300 304 305 306 306

Race Ug99, or TTKSK, of fungus Puccinia graminis tritici, causing stem or black rust disease on wheat (Triticum aestivum), first identified in Uganda in 1998 has been recognized as a major threat to wheat production. Its spread in 2006 to Yemen and Sudan and further spread towards North Africa, Middle East and West-South Asia is predicted -aided by predominant wind currents and large areas of wheat varieties that are susceptible and grown under environments favorable for survival and multiplication of the pathogen. This has raised serious concerns of major epidemics that could destroy the wheat crop in these primary risk areas. Detection in Kenya of a new variant TTKST in 2006 with virulence to gene Sr24, which caused severe epidemics in 2007 in some regions of Kenya and rendered about half of the previously known Ug99-resistant global wheat materials susceptible, has further increased the vulnerability globally. Rigorous screening since 2005 in Kenya and Ethiopia of wheat materials from 22 countries and International Centers has identified low frequency of resistant materials that have potential to replace susceptible cultivars. Diverse sources of resistance, both race-specific and adult-plant type, are now available in high-yielding wheat backgrounds and are being used in breeding. The proposed strategy is to deploy spring wheat varieties possessing durable, adult plant resistance in East Africa and other primary risk areas to reduce inoculum and selection of new virulences capable of overcoming undefeated race-specific resistance genes. Race-specific resistance genes can then be deployed in secondary risk areas preferably in combinations. We believe that Ug99 threat in most countries can be reduced to low levels by urgently identifying, releasing and providing seed of new high yielding, resistant varieties.

1. Introduction Wheat, one of the most important staple food crops, is grown on about 225 million ha worldwide from the equator to latitudes of 60 N and 44 S and at altitudes ranging from sea level to more than 3000 m. Approximately 600 million tons of wheat is produced annually, roughly half of which is in developing countries (Aquino et al., 2002). The only limitation to production is humid and high-temperature areas in the tropics and highlatitude environments where fewer than 90 frost-free days are available for crop growth. The world’s largest producers of wheat are China, India, and the USA, producing annually 100, 70, and 64 million tons with productivities

Stem Rust Threat to Wheat

273

of 3.8, 2.6, and 2.9 t/ha, respectively (Aquino et al., 2002). Only 10% of total wheat produced is sold on the export market, the primary exporting countries are USA, Canada, Australia, and France, and developing countries consume most of the wheat sold on the export market (Aquino et al., 2002). In some countries, such as those in North Africa, per capita consumption of wheat is as high as 240 kg (FAO, 2001). Globally important fungal diseases of wheat, caused by biotrophs (obligate parasites), include the three rusts, powdery mildew, and the bunts and smuts; whereas, those caused by hemibiotrophs (facultative parasites) include Septoria tritici leaf blotch, Septoria nodorum blotch, spot blotch, tan spot, and Fusarium head blight (scab). The obligate parasites are highly specialized and significant variation exists in the pathogen population for virulence to specific resistance genes. Evolution of new virulence through migration, mutation, recombination of existing virulence genes and their selection has been more frequent in rust and powdery mildew fungi. Therefore, enhancement of the knowledge on the genetic basis of resistance and breedingresistant cultivars to these diseases has received larger attention. Stem or black rust of wheat, caused by fungus Puccinia graminis Pers. f. sp. tritici Eriks. & E. Henn., historically is known to cause severe devastation periodically and was most feared disease in various countries in all continents where wheat is grown. The fear from stem rust is understandable because an apparently healthy looking crop of a susceptible cultivar about 3 weeks prior to harvest could reduce to a black tangle of broken stems and shriveled grains by harvest. There are several major wheat production areas worldwide in which stem rust can cause severe losses due to environments that are conducive to disease development. According to Saari and Prescott (1985), stem rust was historically a major problem in all of Africa, the Middle East, all of Asia except Central Asia, Australia and New Zealand, Europe, and the Americas (both North and South). Although the last major stem rust epidemics occurred in Ethiopia during 1993 and 1994 (Shank, 1994) when a popular wheat variety ‘‘Enkoy’’ suffered major losses, the rest of the world has practically remained unhurt from stem rust for over three decades. Because worldwide epidemics of other two rust diseases, leaf (or brown) rust caused by P. triticina and stripe (or yellow) rust caused by P. striiformis, were more frequent in recent years, there has been a major shift in priority and resources away from stem rust research and breeding to such an extent that in some countries testing and breeding for stem rust resistance have been suspended. Consequently, it is not surprising to find new wheat pathologists and breeders who have not even seen stem rust infection in the field. In 1998, severe stem rust infections were observed on wheat in Uganda, and a race, designated as Ug99 with virulence on Sr31, was detected (Pretorius et al., 2000). Race Ug99 was subsequently detected in Kenya and Ethiopia in 2005 (Wanyera et al., 2006), and in Sudan and Yemen in 2006 ( Jin et al., unpublished data). A new variant of this race with virulence

274

Ravi P. Singh et al.

to Sr24 was detected in Kenya in 2006 ( Jin et al., 2007a). It is predicted that these races will migrate to North Africa, Middle East, Asia, and beyond and challenge wheat scientists and policy makers to identify, develop, and replace most of the wheat varieties currently grown in these areas that are susceptible. Although major losses to stem rust have already occurred in Kenya during the 2007 crop season in the ‘‘Narok’’ area and fungicides had to be used heavily to protect the wheat crop in other later-sown areas, the authors strongly believe that the major threat of stem rust once again destroying the vast span of wheat crops in its predicted migration path can be mitigated. This optimism arises from the fact that wheat scientists and policy makers worldwide have begun to respond positively to the alarm raised by Noble Laureate Dr. Norman E. Borlaug in 2005 (CIMMYT, 2005), which resulted in the formation of the ‘‘Global Rust Initiative’’ (www.globalrust.org).

2. Stem Rust Disease, Pathogen, and Epidemiology Stem rust is caused by fungus P. graminis Pers. f. sp. tritici Eriks & E. Henn and belongs to one of several formae speciales in P. graminis. Stem rust appears as elongated blister-like pustules, or uredinia, most frequently on the leaf sheaths of a wheat plant, but also on true stem tissues, leaves, glumes, and awns. Stem rust pustules on leaves develop mostly on the lower side, but may penetrate and produce limited sporulation on the upper side. On the leaf sheath and glumes, pustules rupture the epidermis and give a ragged appearance. Masses of urediniospores produced on the pustules are brownish red in color, and easily shaken off the plants. As infected plants mature, uredinia convert into telia, changing color from red into dark brown to black, thus the disease is also called black rust. Teliospores are firmly attached to plant tissue. The fungus is heteroecious, alternating between a telial host in Poaceae and an aecial host in Berberidaceae, and macrocyclic, with five spore states that are distinct in morphology and function. Crop species as primary hosts include bread wheat, durum wheat, barley, and triticale. There are a large number of species in Berberis and Mahonia that are susceptible to P. graminis (Roelfs, 1985), but the common barberry, B. vulgaris, is considered to be the most important alternate host. Aeciospores arising from an alternate host can be a source of inoculum. Historically, the source of inoculum from B. vulgaris was important in North America and northern and eastern Europe. This source of inoculum has generally been eliminated or greatly reduced by removal of common barberry from the proximity of wheat fields.

Stem Rust Threat to Wheat

275

Urediniospores disseminate to newly emerged tissues of the same plant or adjacent plants to cause new infections, or can be transported through wind in long distances. Long-distance transport through prevailing winds is known to occur across the North American Great Plains (Roelfs, 1985), from Australia to New Zealand, and rarely to a distance of about 8000 km from southern Africa to Australia (Luig, 1985). In the case of long-distance dispersal, spore depositions on crops in a new area are often associated with rain showers. Stem rust urediniospores are rather resistant to atmospheric conditions if their moisture content is moderate (20–30%). The minimum, optimum, and maximum temperatures for urediniospore germination are 2, 15–24, and 30 C; and for sporulation 5, 30, and 40 C (Roelfs et al., 1992), thus providing a vast range of favorable environmental conditions. Urediniospores initiate germination within 1–3 h of contact with free moisture over a range of temperatures. In field conditions, 6–8 h of dew period or free moisture from rains is required for the completion of infection process. After two devastating stem rust epidemics in North America in 1904 and 1916, an important finding came from the pioneering work of E. C. Stakman (Stakman and Piemeisel, 1917) who showed that the stem rust pathogen had various forms or races. These races varied in their ability to infect different wheat varieties which later were found to carry distinct resistance genes or combinations thereof. At present wheat scientists use wheat lines that usually carry a single race-specific resistance gene to determine avirulence/virulence characteristics of a race. Mutation toward virulence in existing populations followed by selection on susceptible hosts is at present considered to be the most important evolution mechanism for stem rust pathogen to acquire new virulence to overcome resistance conferred by race-specific resistance genes. Where an alternate host is present, it is possible to have new combinations of virulences through sexual recombination; however, it is limited at present to few areas of the world. Rare asexual recombination is also known to occur through exchange of nuclei between conjugating hypha of two races that have by chance infected same tissues. Wheat rust pathogens are biotrophs and therefore need living wheat plants or other secondary hosts for survival in the absence of alternate hosts. They produce large numbers of urediniospores during the crop season and wind dispersion transmits these urediniospores onto the same or new host plants in the vicinity or distantly. Typically, most spores will be deposited close to the source (Roelfs and Martell, 1984); however, longdistance dispersal is well documented with three principal modes of dispersal known to occur. The first mode of dispersal is single event, extremely longdistance (typically cross-continent) dispersal that results in pathogen colonization of new regions. Dispersion of this type is rare under natural conditions and by nature inherently unpredictable. It is also difficult to specifically attribute long-distance dispersal. However, rusts are one pathogenic group with reasonably strong evidence for unassisted, long-distance dispersal under

276

Ravi P. Singh et al.

natural airborne conditions. Several examples of long-distance dispersal have been described by Brown and Hovmller (2002), including the introduction of sugarcane rust into the Americas from Cameroon in 1978 and a wheat stem rust introduction into Australia from southern Africa in 1969. Both these examples provide strong evidence for being unassisted natural longdistance wind-borne dispersals. The enabling factor in this mode of dispersal for rusts is the robust nature of spores ensuring protection against environmental damage (Rotem et al., 1985). Deposition in new areas is primarily through rain-scrubbing of airborne spores onto susceptible hosts (Rowell and Romig, 1966). Assisted long-distance dispersal, typically on travelers clothing or infected plant material, is another increasingly important element in the colonization of new areas by pathogens. Despite strict phytosanitary regulations, increasing globalization and air travel both increase the risk of pathogen spread. Evidences strongly support an accidental introduction of wheat stripe rust into Australia in 1979, probably on travelers clothing, from Europe (Steele et al., 2001). More recently, concerns over nonaccidental release of plant pathogens as a form of ‘‘agricultural bio-terrorism’’ have arisen, with wheat stem rust considered one pathogen of concern (Hugh-Jones, 2002) primarily due to its known ability to cause devastating production losses to a major food staple (Leonard, 2001). The second major mode of dispersal for pathogens like rusts is stepwise range expansion. This typically occurs over shorter distances, within a country or a region, and has a much higher probability than the first described dispersal mode. This probably represents the most common or normal mode of dispersal for rust pathogens. A good example of this type of dispersal mechanism would include the spread of a Yr9-virulent race of P. striiformis that evolved in eastern Africa and migrated to South Asia through the Middle East and West Asia in a stepwise manner over about 10 years, and caused severe epidemics along its path (Singh et al., 2004b). The third mode of dispersal, extinction, and recolonization, could perhaps be considered a sub-mechanism of stepwise range expansion. This mechanism occurs in areas that have unsuitable conditions for year-round survival. Typically these are temperate areas where hosts are absent during winter or summer. A good example of this mechanism is the ‘‘Puccinia pathways’’ of North America—a concept that arose from another pioneering work of Stakman (1957) in which rust pathogens over winter in southern USA or Mexico and recolonize wheat areas in the Great Plains and further north following the prevailing south-north winds as the wheat crop season progresses. The second well-documented extinction–recolonization example is that of wheat stripe rust survival and spread from mountains in the Gansu province of China (Brown and Hovmller, 2002) and wheat rusts in the Himalayas and Nilgiri Hills in northern and southern India, respectively (Nagarajan and Joshi, 1985) where susceptible hosts can be found year

Stem Rust Threat to Wheat

277

round and environmental conditions are favorable for the pathogen to survive. Urediniospores from these areas are then blown to wheat fields in other areas and initiate disease.

3. Breeding for Resistance 3.1. Historical account It was not until the beginning of twentieth century and soon after the rediscovery of Mendel’s laws, that Biffen (1905) demonstrated that inheritance of resistance to wheat stripe rust followed Mendel’s laws. Strong emphases to identify resistance to stem rust and to breed resistant wheat varieties were initially given in the USA, Canada, Australia, and Europe. Although the major epidemic of 1916 in the USA and Canada had already triggered extensive research on stem rust, efforts in the USA, Canada, and Australia were intensified further with subsequent epidemics in the following decades. Although resistance present in some hexaploid wheat sources were used in breeding during early years, the most successful control of stem rust came when H. K. Hayes in the University of Minnesota and E. S. McFadden in South Dakota State University transferred the stem rust resistance from tetraploid sources ‘‘Iumillo’’ durum and ‘‘Yaroslav’’ emmer, respectively, into bread wheat that gave rise to hexaploid wheat varieties ‘‘Thatcher’’ and ‘‘Hope’’ (Kolmer, 2001). Although several race-specific genes are present in Hope and Thatcher, the most effective component of the resistance in these two varieties is due to adult plant resistance. Thatcher and Hope, Hope sib ‘‘H44–24a,’’ and other varieties derived from these parents such as ‘‘Selkirk’’ and ‘‘Chris’’ that combined resistance to stem rust from other sources including gene Sr6 found to be present in a plant selection by J. McMurachy in 1930. ‘‘Kenya 58’’ and other Kenyan varieties carrying the same gene Sr6 were also used extensively in Australia by I. A. Watson and in Mexico by N. E. Borlaug. Efforts to find a solution to the stem rust problems facilitated global collaboration amongst wheat scientists who shared, grew, and evaluated wheat germplasm in the quest of finding different sources of resistance to stem rust. Resistant wheat materials developed at Njoro, Kenya through the support from Canadian scientists in 1960s and 1970s contributed substantially to international breeding efforts. Resistance from Hope and Chris formed the foundation of the high-yielding, semidwarf wheat varieties that led to ‘‘Green Revolution’’ in the 1970s.

3.2. International cooperation Although germplasm exchange was common among wheat scientists, the International Spring Wheat Rust Nursery Program, initiated in 1950 by B. B. Bayles and R. A. Rodenhiser of USDA-ARS (United States

278

Ravi P. Singh et al.

Department of Agriculture—Agricultural Research Services), Beltsville, formed the basis of a true international collaboration and operated continuously until the mid 1980s. The objectives of the program were (1) to find new genes or combinations of genes in wheat which condition field resistance to rusts throughout the world, and (2) to test new varieties and promising selections of wheat developed by plant breeders and pathologists for resistance to rusts. The germplasm and information generated were made available to the global wheat community. This nursery was the foundation of numerous other international nurseries and led to global cooperation to achieve resistance to diseases and pests of several crops. CIMMYT (International Maize and Wheat Improvement Center) and several other international research centers continue to use this methodology to not only distribute improved germplasm they develop but also to evaluate their performance for agronomic and disease resistance attributes.

3.3. Spread of semidwarf wheats with resistance to stem rust The semidwarf wheat varieties developed by Dr. N. E. Borlaug in Mexico during early 1960s under the program sponsored by the Mexican Government and the Rockefeller Foundation were also resistant to stem rust and earlier in maturity compared to tall varieties grown previously. The two semidwarf ‘‘Green Revolution’’ mega-varieties, ‘‘Sonalika’’ and ‘‘Siete Cerros,’’ continue to have moderate levels of resistance to race Ug99 even today; however, they were mostly replaced as they succumbed to leaf and yellow rusts and better varieties became available. These semidwarf varieties significantly reduced stem rust incidence in many areas, which is often attributed to a combination of resistance and early maturity that avoided stem rust inoculum buildup (Saari and Prescott, 1985). The tall variety ‘‘Yaqui 50,’’ released in Mexico during the 1950s, and other Sr2-carrying semidwarf varieties released since then had stabilized the stem rust situation in Mexico and possibly in many other countries where modern semidwarf wheats were adopted. Changes in stem rust races have not been observed in Mexico for almost 40 years and natural infections are nonexistent. Successful transfers and utilization of alien resistance genes Sr24 and Sr26 from Agropyron elongatum (Thinopyrum ponticum), Sr31 located in the 1BL.1RS translocation from ‘‘Pektus’’ rye and an undesignated gene on 1AL.1RS translocation from ‘‘Insave’’ rye, Sr36 from T. timopheevi and more recently Sr38 from T. ventricosum further reduced stem rust incidence in various countries around the world in 1970s and 1980s. The alien resistance gene Sr31 has been used in agriculture on the largest scale since 1980s in spring, facultative and winter wheat breeding programs worldwide except Australia. Its use in CIMMYT wheat improvement resulted in the release of several popular cultivars worldwide. The use of 1BL.1RS translocation was initially associated with increased grain yields and resistance to all three rusts

Stem Rust Threat to Wheat

279

and powdery mildew as it carried resistance genes for all these diseases on the same translocation. Large-scale deployment of Sr31 surprisingly did not result in its breakdown until the detection of race Ug99 in Uganda. In fact this gene probably further reduced the already low stem rust survival to almost nonexistent levels in most wheat growing regions to the extent that stem rust started to become a forgotten curse. The decrease in incidence of stem rust to almost nonsignificant levels by the mid-1990s throughout most of the wheat producing areas worldwide were coincident with a decline in research and breeding emphasis to such a level that in many countries breeding was done in the absence of this disease. CIMMYT scientists continued to select for stem rust resistance in Mexico using artificial inoculation with six P. graminis tritici races of historical importance. New stem rust races have rarely occurred since the ‘‘Green Revolution’’ in Mexico (Singh, 1991). Moreover, a majority of wheat lines selected in Mexico remained resistant at international sites either due to absence of disease, inadequate disease pressure, or presence of races that lacked necessary virulence for the resistance genes contained in CIMMYT wheat germplasm. Frequency of 1BL.1RS translocation went up to 70% at one stage in CIMMYT’s spring wheat germplasm but has declined to about 30% in more recent advanced lines. Such alien chromosome segments on the one hand are very useful for controlling multiple diseases, but on the other hand could lead to ‘‘vertifolia’’ or a masking effect (Vanderplank, 1963) resulting in decrease in frequency or even loss of other useful genes, especially minor types, in breeding materials. All wheat lines of CIMMYT origin evaluated in Kenya since 2005, irrespective of the presence or absence of 1BL.1RS translocation, were highly resistant to stem rust in Mexico and remain highly resistant in other parts of the world, indicating that the high frequency of this translocation in 1980s and 1990s cultivars explains only a portion of the current susceptibility of wheat germplasm to race Ug99 in Kenya. Jin and Singh (2006) compared seedling reactions of US wheat cultivars and germplasm with highly virulent races present in the USA and race Ug99. Several wheat lines, especially spring wheat that were highly resistant to US races and did not carry the1BL.1RS translocation, were also found to be susceptible to Ug99. This further supports the hypothesis that race Ug99 carries a unique combination of virulence to known and unknown resistance genes present in wheat germplasm. The major susceptibility is due to the specific nature of avirulence/virulence combination that Ug99 possesses, which had led to the susceptibility of many wheat materials irrespective of where they were developed.

3.4. Current knowledge of resistance to stem rust At present 46 different stem resistance genes are catalogued and multiple alleles are known for three gene loci (Table 1). There are a few additional resistance genes that need further research before they can receive designation

280

Ravi P. Singh et al.

Table 1 Origin and usefulness of designated Sr-genes in conferring seedling and/or adult plant resistance to Ug99 race of stem rust pathogen Puccinia graminis f. sp. tritici Origin of Sr genes

Triticum aestivum Triticum turgidum Triticum monococcum Triticum timopheevi Triticum speltoides Triticum tauschii Triticum comosum Triticum ventricosum Triticum araraticum Thinopyrum elongatum Thinopyrum intermedium Secale cereale a b

Stem rust resistance (Sr) genes Ineffective

Effective

5, 6, 7a, 7b, 8a, 8b, 9a, 9b, 9f, 10, 15, 16, 18, 19, 20, 23, 30, 41, 42, Wld-1 9d, 9e, 9g, 11, 12, 17 21

28a, 29b, Tmpa 2b, 13a,b, 14a 22, 35 36a, 37 32, 39 33b, 45

34 38

31

40 24a, 25, 26, 43 44 27a, 1A.1Ra, R

Virulence for the gene is known to occur in other races. Level of resistance conferred in the field usually not enough.

(McIntosh et al., 1995). Several of these genes were incorporated into wheat from alien wheat relatives (Table 1). All designated genes, except Sr2, are race specific and are expressed in both seedling and adult plants. Race specificity derives from the gene-for-gene relationship between the host plant resistance gene and corresponding avirulence genes in the pathogen. With avirulent races a majority of stem resistance genes allows formation of tiny- to mediumsized uredinia, with limited sporulation, which are surrounded by a necrosis or chlorosis (McIntosh et al., 1995). Genes that allow development of only microscopic or macroscopic hypersensitive reactions include Sr5, Sr17, Sr27, Sr36, and Sr6 at cooler temperatures. The adult plant resistance gene Sr2 confers slow rusting (Sunderwirth and Roelfs, 1980). Combination of Sr2 with other unknown slow rusting resistance genes possibly originating from Thatcher and Chris, commonly known as the ‘‘Sr2-Complex,’’ provided the foundation for durable resistance to stem rust in germplasm from the University of Minnesota in the United States, Sydney University in Australia, and the spring wheat germplasm developed by Dr. N. E. Borlaug (McIntosh, 1988; Rajaram et al., 1988). Unfortunately, not much is known about the other genes in the Sr2 complex and their interactions. Knott (1988) has shown that adequate levels of multigenic resistance to stem rust can be achieved by accumulating approximately five minor genes.

Stem Rust Threat to Wheat

281

US wheat cultivar Chris, which is not known to carry Sr2 but possesses several seedling resistance genes including Sr7a (Singh and McIntosh, 1987) displayed adequate level of resistance to Ug99 in the field in Kenya. Preliminary studies of inheritance of seedling resistance to Ug99 in Chris indicated that Ug99 resistance in Chris is controlled by two complementary recessive genes ( Jin, 2007), and the same seedling resistance is present in AC Barrie (a Canadian spring wheat cultivar), Thatcher, and Bonza 65 (a CIMMYT-derived cultivar). Singh and McIntosh (1987) indicated the possibility that the adult plant resistance to Sr7a-avirulent Australian races may involve interaction of the moderately effective gene Sr7a and other unknown adult plant resistance genes. Seedling tests indicated that Ug99 is virulent on the Sr7a-tester line ( Jin et al., 2007b) although Chris did show seedling resistance. Singh and McIntosh (1987) indicated that resistance conferred by Sr7a is difficult to evaluate both in seedlings and adult plants when the gene is present alone. Therefore, at this stage we cannot determine the role Sr7a may have played in resistance of ‘‘Chris’’ observed in Kenya. Even though seedling tests indicate that Sr23, another gene whose expression is difficult to evaluate in seedlings and adult plants when present alone, may be ineffective against Ug99, adequate resistance in ‘‘Selkirk’’ may involve interactions of moderately effective genes Sr2 and Sr23 (linked to leaf rust resistance gene Lr16) and perhaps additional unknown adult plant resistance genes. These observations, although they still require validation through genetic analyses, indicate that complex resistance to stem rust present in some tall cultivars developed in the 1960s and 1970s continue to remain effective.

4. Race UG99 and Why it is a Potential Threat to Wheat Production 4.1. Avirulence/virulence genes in Ug99 Race Ug99, that emerged in Uganda in 1998 and was identified in 1999 (Pretorius et al., 2000), is the only known race of P. graminis tritici that has virulence for gene Sr31 known to be located in the translocation 1BL.1RS from rye (Secale cereale). It was designated as TTKS by Wanyera et al. (2006) using the North American nomenclature system (Roelfs and Martens, 1988) and more recently as TTKSK after a fifth set of differentials was added to further expand the characterization ( Jin et al., 2008). The most striking feature of race Ug99 is that it not only carries virulence to gene Sr31 but also this unique virulence is present together with virulence to most of the genes of wheat origin, and virulence to gene Sr38 introduced into wheat from Triticum ventricosum that is present in several European and Australian cultivars and a small portion of new CIMMYT germplasm (Table 1, Jin et al.,

282

Ravi P. Singh et al.

2007b). This virulence combination might have accounted for the widespread Ug99 susceptibility in wheat varieties worldwide. A variant of Ug99 with added virulence to Sr24 was detected in 2006 in Kenya. It is anticipated that mutation toward more complex virulence will likely occur as the fungal population size increases and selection pressure is placed on the population by resistant varieties.

4.2. Current distribution of race Ug99 As described by Singh et al. (2006) Ug99 was first identified in Uganda in 1998, although there is some evidence indicating that the race may have been present in Kenya since 1993, and had spread to most of the wheat growing areas of Kenya and Ethiopia by 2003. In 2005, Ethiopian reports confirmed its presence in at least six dispersed locations (Fig. 1). The East African highlands are a known ‘‘hot-spot’’ for the evolution and survival of new rust races (Saari and Prescott, 1985). The favorable environmental conditions and the presence of host plants year-round favor the survival and buildup of pathogen populations. Available evidence emerging from

SUDAN 2006

YEMEN

2006 2006

2006 Legend +

2006 sites

Lake Tana

wheat zone 2003 2003 2003

2003 2003

2003 2003

Ethiopia

N E

W

Lake Turkana

S

2001

Uganda Lake Albert

1993

1996

1998/99

1998

Lake Victoria

1996

Kenya 1993

2001

1998 1999

2001

0 Lake Tanganyika

235

470

940 Kilometers CIMMYT

Figure 1 Regional wheat production areas and known distribution of stem rust pathogen race Ug99 as of September 2007.

Stem Rust Threat to Wheat

283

the East African countries indicates that Ug99 has exhibited a gradual stepwise range expansion, following the predominant west-east airflows. The confirmed range of Ug99 continues to expand, with new sites being recorded beyond the previously confirmed three East African countries Uganda, Kenya, and Ethiopia. In early 2006 (February/March), stem rust—tentatively caused by the Ug99 race—was reported from a site near New Halfa in eastern Sudan. Later the same year (October/November), reports were obtained from at least two sites in western Yemen (Fig. 1). Subsequent race analysis of samples from these sites, undertaken by the USDA-ARS Cereals Disease Laboratory, St. Paul, MN, USA confirmed the presence of Ug99 in these countries. The observed expansion into new areas is in-line with previous predictions on the likely movement of Ug99 (Hodson et al., 2005; Singh et al., 2006) and fits the stepwise dispersal model following prevailing winds as outlined by Singh et al. (2006). The exact route taken by Ug99 to reach Yemen is unknown, but neither the possibility of transfer from south-eastern/eastern Ethiopia on the fringes of the southwestern monsoon system nor the transfer from eastern Sudan/Eritrea/ northern Ethiopia can be excluded (Fig. 2).

Figure 2 Updated potential migration routes of Ug99 based on historical precedence and recent studies of actual wind movements.

284

Ravi P. Singh et al.

4.3. Predicting Ug99 migration to other wheat areas Crossing of the Red Sea into Yemen by Ug99 is regarded as being particularly significant, as the pattern of regional airflows, combined with historical recorded migration of Yr9-virulent stripe rust race (Singh et al., 2004b), both support the potential for onward movement from Yemen into significant wheat production areas of the Middle East and West-South Asia. On the basis of airflow patterns, Fig. 2 updates the potential migration routes A and B described in Singh et al. (2006). Nothing in the current observed spread of Ug99 indicates any basis for changing this hypothesis. Given this situation, the buildup of significant levels of Ug99 urediniospores in Yemen would be a cause for concern. More detailed analysis of further potential onward movements undertaken using the HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) airborne particle trajectory model developed by NOAA (Draxler and Rolph, 2003) supports the hypothesis that Yemen could be a staging post for onward movement into the Middle East and Asia. Figure 3 illustrates 72-h airborne particle trajectories, derived from HYSPLIT, using SYRIA

JORDAN

AFGHANISTAN

IRAQ

IRAN

KUWAIT

Legend PAKISTAN

2006 sites Wind trajectory BAHRAIN

wheat zone

QATAR

SAUDI ARABIA

UNITED ARAB EMIRATES

OMAN N W

SUDAN

E S

2006 ERITREA

ETHIOPIA Lake Tana

YEMEN 2006

DJIBOUTI

0 SOMALIA

235

470

940 Kilometers CIMMYT

Figure 3 Air-borne particle trajectories, derived from the HYSPLIT model, originating from the confirmed Ug99 site of Al Kedan, Yemen (trajectories represent weekly 72-h movements for the period 1st December 2006 to 28th February 2007).

Stem Rust Threat to Wheat

285

the confirmed Ug99 site Al Kedan in Yemen as a source. The trajectories shown are for weekly intervals during the period 1st December 2006 to 28th February 2007—a period in which wheat would be present and at a potentially susceptible growth stage in areas north of Yemen. During this period there was a clear tendency for airborne trajectories, originating at Al Kedan, to follow a north-easterly routing heading toward the wheat producing areas of Saudi Arabia, Iraq, and Iran. Similar results were obtained from an identical analysis covering the period 1st December 2005 to 28th February 2006, supporting the notion that the possibility of onward movements from sites in Yemen in the direction of key wheat areas occurs on a regular basis. Immediate onward movements from eastern Sudan are potentially less problematic as airflow models indicate that direct movements in a northerly direction into the important wheat areas of the Nile valley are unlikely. However, given the uncertainty and complexity of airflows in this region the possibility of spores reaching these areas can never be totally excluded. In addition, there is a very real risk that spores could move northwards up the Arabian Peninsula from Yemen, enter the Nile Delta and then cycle back south down the Nile Valley. The Yr9-virulent stripe rust race did reach Egypt soon after its detection in Yemen (Singh et al., 2004b). Sudan had escaped stripe rust because wheat is grown under relatively warm conditions, which is unfavorable for stripe rust survival. At present, no known long-distance, single event ‘‘random jump’’ type movement (assisted or natural) has been recorded for Ug99. But with an expanding known range for the pathogen and the high mobility of people both regionally and internationally, there is a clear need for continued monitoring and surveillance in wheat areas beyond the immediate at risk region. Presence of the Sr24-virulent variant of Ug99 first identified in Kenya in 2006 has not yet been detected beyond Kenya, even though it was widespread in epidemic form in Kenyan highlands on the Sr24 carrying variety ‘‘Kenya Mwamba.’’

4.4. Resistance/susceptibility of current wheat germplasm Reynolds and Borlaug (2006) estimated that the potential area under the risk from Ug99 along the natural migration path in North Africa, Middle East and Asia (excluding China) might amount to 50 million ha of wheat, that is, about 25% of the world’s wheat area and accounting for an estimated 19% of global production amounting to about 117 million tons. An estimated 1 billion people live within these wheat production areas. Extensive screening of global wheat varieties for resistance to Ug99 has been undertaken at key sites in Kenya and Ethiopia (principally Njoro, Kenya and Kulumsa, Ethiopia) and results summarized by Singh et al. (2006). Available screening data has been linked via known pedigrees to databases

286

Ravi P. Singh et al.

on areas planted to known varieties (CIMMYT, unpublished data). Prior to 2006, data were available for 10 countries in the Africa/Asia region with limited susceptibility ratings for Ug99 recorded for wheat varieties covering an estimated 44 million ha. By the end of 2006, the screening dataset had been extended to include germplasm from 18 countries in the region, including China, with more detailed resistance/susceptibility ratings obtained on varieties covering an estimated 75 million ha (Fig. 4). A summary of the area by susceptibility rating data is given in Table 2. Varieties exhibiting any observed resistance to Ug99 only account for 5% of the total estimated area in the 18 countries. The huge areas observed in India and Pakistan result from the predominance of ‘‘mega-cultivars’’ ‘‘PBW343’’ and ‘‘Inqualab 91’’ in the two countries, both of which are susceptible to Ug99. Further screening of additional varieties from 22 countries undertaken in Kenya during 2007 indicated a similar low frequency of resistant materials; however, the database mentioned above has not been updated yet.

5.2 M Ha

5.4 M Ha 12.5 M Ha

Legend Variety Susceptibility Susceptible Unknown Mod. Susceptible Mod. Resistant Resistant

750 1,500

E S

Wheat Production Wheat zone

0

N W

Kilometers 3,000

Figure 4 Summary of areas planted to known wheat varieties from 18 countries by their resistance/susceptibility groupings to Ug99 from screening results in Kenya and Ethiopia during 2005 and 2006 (symbols are scaled according to recorded areas by country).

287

Stem Rust Threat to Wheat

Table 2 Estimated areas planted to wheat varieties in 18 African and Asian countries by resistance/susceptibility grouping to Ug99, based on screening data 2005, 2006 in Njoro, Kenya Estimated area Resistance/susceptibility group

million ha

%

Unknown Susceptible Moderately susceptible Resistant Moderately resistant

30.71 38.54 1.73 1.04 2.72

41.1 51.6 2.3 1.4 3.6

The seemingly favorable environmental conditions, coupled with the extensive coverage of susceptible wheat varieties is a grave cause of concern if Ug99 does spread unchecked. Conditions favorable for outbreaks of epidemics currently exist in the migration path of highest probability. If an epidemic from Ug99 does occur, extremely large numbers of wheat farming families may be seriously affected—especially those who have few alternative livelihoods. In these circumstances, landless laborers dependent on agricultural jobs would also be seriously affected, and one could anticipate an increase in the rural–urban migration of landless laborers and small farmers. If large production losses occur there would be significant implications for rural and national economic growth rates in seriously affected countries, and could even affect global wheat markets. Given the serious implications, there is a clear need for improved monitoring, surveillance, and early warning to directly support the efforts of wheat researchers and national policy makers working to prevent the spread of the disease and to alleviate its impacts.

4.5. Can a Ug99 pandemic be predicted? A frequently asked question these days is when will Ug99 cause a major epidemic? However, the answer is very difficult because combination of factors occurring together is necessary for any epidemic to build up, and these could be different in different areas. Obviously susceptible host, virulent pathogen races, and a conducive environment are primary factors, but other more important factors in many areas are where and how much inoculum survives during the off-season, and when disease establishes in commercial fields. Let us take a few examples to illustrate this further. In Kenya, where stem rust is endemic, that is, it is present throughout the year in the vicinity of wheat crops sown at different times and at different elevations, it took only one year from the detection in 2006 of the TTKST variant of Ug99 to cause

288

Ravi P. Singh et al.

epidemics in 2007 on a previously resistant variety ‘‘Kenya Mwamba’’ and other varieties that were already susceptible to the original Ug99. Weather conditions during the 2006/07 period, October-April, were unexpectedly favorable for stem rust due to higher than normal rainfall, which allowed multiplication of race TTKST not only on wheat crop but also on offseason voluntary wheat plants. The specific combination of both changing virulence patterns and favorable weather conditions may have been the drivers of the observed epidemic. Conversely, a Ug99 epidemic has not occurred in Ethiopia even though it was first detected in 2003 and sporadic infections were found on the commercial wheat crop. Huerta-Espino (1992) reported the presence of a Sr36-virulent race from samples collected in 1987 in Ethiopia; however, not until 1993 did an epidemic occur on the variety Enkoy that was previously protected by gene Sr36. The 1993 epidemic probably left enough inoculum to survive during the off-season on susceptible voluntary plants to cause a repeat second epidemic during 1994. Occurrence of epidemics had stopped by 1995, as the area under Enkoy was significantly reduced. Stakman (1957) summarized the difficulties he faced in preventing stem rust epidemics in North America. It took 11 years for race 15B, first identified in collections from barberries in Iowa in 1939, to cause a pandemic in 1950 in Mexico, and USA, and Canada in 1953, 1954. However, it may not be necessary for a rust pathogen to take such a long time to cause epidemics. In northwestern Mexico, an exotic race of leaf rust detected in 2001 on durum wheat was able to cause an epidemic during the same year and the two subsequent years until the susceptible cultivar was removed from cultivation (Singh et al., 2004a). It is therefore difficult to predict when, or even if, an epidemic will occur once Ug99 is detected in a country. Since the ‘‘Green Revolution,’’ agronomic practices have changed in many parts of the world; creating a more favorable environment for stem rust buildup due to the higher use of nitrogen fertilizers and irrigation. Expansion of wheat cultivation into new areas and the expansion of conservation agriculture, allowing survival of more volunteer plants in the off-season, are all likely to change stem rust epidemiology that we know at present.

5. Breeding Strategies to Mitigate the Threat from UG99 and Achieve a Long-Term Control of Stem Rust Reducing the area planted to susceptible cultivars in ‘‘Primary Risk Areas’’ of East Africa, Arabian Peninsula, North Africa, Middle East, and West-South Asia is the best strategy if major losses are to be avoided.

Stem Rust Threat to Wheat

289

The ‘‘Global Rust Initiative’’ (www.globalrust.org), launched in 2005, is using the following strategies to reduce the possibilities of major epidemics: (1) monitoring the spread of race Ug99 beyond eastern Africa for early warning and potential chemical interventions, (2) screening of released varieties and germplasm for resistance, (3) distributing sources of resistance worldwide for either direct use as varieties or for breeding, and (4) breeding to incorporate diverse resistance genes and adult plant resistance into high-yielding adapted varieties and new germplasm. The best long-term strategy to mitigate the threat from Ug99 is to identify resistant sources among existing materials, or develop resistant wheat varieties that can adapt to the prevalent environments in countries under high risk, and release them after proper testing while simultaneously multiplying the seed. An aggressive strategy to promote these resistant varieties in farmers’ fields is the only viable option as resource-poor as well as commercial farmers in most of Africa, Middle East, and Asia cannot afford chemical control or may not be able to apply chemicals in the event of large-scale epidemics due to their unavailability for timely application. A reduction in disease pressure in East Africa and Yemen will likely reduce chances of migration beyond these areas to other primary risk areas; however, it is unlikely that further range expansion of Ug99 can be stopped at this stage. Reduction of susceptible varieties throughout the primary risk area should reduce wind dispersal of spores from these areas to ‘‘Secondary Risk Areas.’’ For a long-term control, we like to discuss strategies that are already implemented or can be applied to identify, develop, and deploy varieties with race-specific resistance genes or with adult plant resistance.

5.1. Prevalence of Sr24 and its breakdown A high frequency of the highly resistant wheat materials from South America, Australia, USA, and CIMMYT identified from 2005 to 2006 screening with Ug99 in Kenya possess Sr24 indicating it as an important resistance gene especially due to its presence in adapted genetic backgrounds. Sr24 is located on the Thinopyrum elongatum translocation on chromosome 3DL together with leaf rust resistance gene Lr24. There are three distinct Sr24 carrying translocations: the original one linked to a gene for red grain color, the shorter segment with white grain, and a third segment where a very small segment has been retranslocated onto chromosome 1BS. In all three segments both Sr24 and Lr24 are present together. Therefore, selection for Lr24 with avirulent leaf rust isolates can be used as an indirect selection strategy. Virulence for Sr24 is known in South Africa (Le Roux and Rijkenberg, 1987) and India (Bhardwaj et al., 1990) in local races and it arose from the deployment of this gene. Detection of race TTKST with Sr24 virulence in

290

Ravi P. Singh et al.

Ug99 lineage during 2006 in low frequency ( Jin et al., 2007a) resulted in rapid buildup to cause an epidemic on Sr24 carrying variety Kenya Mwamba in 2007, which occupied about 30% of the Kenyan wheat area. Three Sr24based resistant varieties, ‘‘ETBW19,’’ ‘‘ETBW21,’’ and ‘‘ETBW22’’ were multiplied under the emergency program in Ethiopia during the 2006 and 2007 to obtain several tons of seed. However, these varieties are now susceptible to the Sr24-virulent race TTKST present in Kenya indicating their genetic vulnerability if they occupy significant areas of rust-prone areas in Ethiopia in coming years. The situations described above have once again questioned and reminded us of the consequence of dependence on single race-specific genes in the control of stem rust in areas where rust is endemic. Adoption of new varieties is a very slow process in many of the target countries and often the change of varieties is triggered by rust epidemics as happened in Ethiopia after the 1993 and 1994 stem rust epidemics on wheat variety Enkoy, which once was a leading variety but was soon replaced by other available varieties such as ‘‘Pavon 76,’’ ‘‘Kubsa,’’ and others. A similar effect was seen in Pakistan where stripe rust epidemics in northwestern Pakistan in 1995 and 1996 on varieties ‘‘Pak 81’’ and ‘‘Pisabak 85’’ led to the almost complete replacement by ‘‘Inqualab 92,’’ which was resistant to stripe rust at that time.

5.2. Race-specific resistance genes effective to Ug99 Resistance gene Sr25 is located on a Th. elongatum translocation together with leaf rust resistance gene Lr19 on chromosome 7DL. Despite the fact that this translocation is known to enhance yield potential (Singh et al., 1998), it was not used widely because it is linked to a gene associated with the accumulation of undesirable levels of yellow pigment. A white floured mutant of the translocation, developed by D.R. Knott (1980), was recently transferred into some Australian and CIMMYT wheat backgrounds. Sr25 conferred high level of resistance only in some genetic backgrounds, especially when the adult plant resistance gene Sr2 was also present, for example, lines ‘‘Super Seri#1’’ (yellow flour), ‘‘Wheatear’’ (white flour), and several lines derived by crossing Wheatear. Virulence to Sr25 was detected in the Nilgiri Hills of India during 2007 (M. Prashar, personal communication). Gene Sr26, also of Th. elongatum origin, translocated to chromosome 6AL, has been used successfully in Australia and remains effective despite its large-scale deployment in the 1970s and 1980s (McIntosh, 1988). It is not known to be present in cultivars from other countries and the translocation used initially may confer a yield penalty (The et al., 1988). The size of this translocation has been reduced by I. Dundas, University of Adelaide, and work is currently underway in Australia to determine if the negative effects were also removed.

Stem Rust Threat to Wheat

291

Gene Sr27 of rye origin has not been used in wheat improvement. Its deployment in triticale in Australia resulted in a rapid evolution of virulence (McIntosh et al., 1983). This gene has also become ineffective in South Africa. Strategically, this gene should be left for triticale improvement in areas where virulence is not known. Gene Sr36, derived from T. timopheevi, exhibits almost an immunity (no symptoms) to race Ug99 at both seedling and adult plant stages ( Jin et al., 2007b). This gene occurs in a high frequency in the US soft winter wheat ( Jin and Singh, 2006) and in some Australian wheat varieties. Although races with virulence to Sr36 are common worldwide including East Africa, this gene is effective to Ug99. Susceptible pustules were seen during 2007 in Kenya on wheat lines known to carry this gene indicating that Ug99 has evolved further, which was later confirmed. Genes Sr22 and Sr35, derived from T. monococcum and located on chromosomes 7AL and 3AL, respectively are also highly effective and can be backcrossed to modern wheats. Virulence to Sr35 was identified in a laboratory culture in Australia (McIntosh et al., 1995). Although race Ug99 is avirulent on gene Sr28, numerous races virulent to this gene are known to occur worldwide. Genes Sr29, Sr32, Sr33, Sr37, Sr39, Sr40, and Sr44 have not been tested widely for their effectiveness to other races and also not used in breeding. Attempts at CIMMYT to transfer gene Lr35 linked to Sr39 in four spring wheat backgrounds resulted in a 15–20% reduction in grain yield potential (Singh, unpublished data). Sizes of alien chromosome segments must be reduced before Sr32, Sr37, Sr39, Sr40, and Sr44 can be used successfully. The undesignated resistance gene SrTmp from ‘‘Triumph 64’’ is present in some US wheat cultivars ( Jin and Singh, 2006) and can be used in breeding. However, virulence to it is known in North America ( Jin and Singh, 2006). An additional undesignated resistance gene, Sr1A.1R, located in rye chromosome translocation 1AL.1RS, is present in some US winter wheats such as ‘‘Amigo,’’ ‘‘TAM107,’’ ‘‘TAM200,’’ ‘‘Nekota,’’ ‘‘Prairie Red,’’ and other hard red winter wheat cultivars ( Jin and Singh, 2006), and can also be used as it confers moderate resistance to Ug99 ( Jin and Singh, 2006). Virulence to Sr1A.1R was detected in a sample collected in Yemen but not in the Ug99 lineage ( Jin, unpublished data). Translocation carrying resistance gene SrR, introduced to wheat from ‘‘Imperial’’ rye in chromosomes 1BL.1RS and 1DL.1RS (Mago et al., 2004) is also likely to be effective against race Ug99 as the original chromosome addition line ‘‘TAF 2’’ was found to be resistant. The allelic relationship between the stem rust resistance genes Sr1A.1R, SrR, and Sr31 is unknown. The Sr1A.1R translocation is present in a few CIMMYT spring wheat lines derived from the crosses with TAM200. Certain hexaploid synthetic (Triticum turgidum x Aegilops tauschii) wheatderived advanced lines, some lines where certain Chinese cultivars such as

292

Ravi P. Singh et al.

‘‘Shanghai#7’’ and ‘‘Chuanmai 18’’ are parents, and a few more lines where resistance genes can be tracked to the US line ‘‘ND643’’ and a CIMMYT cross ‘‘HUW234þLr34/Prinia’’ also have shown adequate level of resistance. Resistance in synthetic wheat-derived lines can be due to the presence of Sr13 and Sr14 in chromosome 6AL and 1BL, respectively, originating from durum wheat parents of synthetic wheats and Sr33 and a newly designated resistance gene Sr45 on chromosomes 1DL and 2DS, respectively deriving from T. tauschii parents. Genes Sr13, Sr14, and Sr33 confer only moderate levels of resistance ( Jin et al., 2007b) and they will be useful in areas where stem rust pressure remains at moderate levels. Virulence to both Sr13 and Sr14 are known among races different from Ug99 (McIntosh et al., 1995). Response of Sr45 with Ug99 in seedling and field conditions is yet to be determined.

5.3. Strategy to use race-specific resistance genes in wheat improvement The fastest way to reduce the susceptibility of important wheat cultivars and the best new germplasm is to systematically incorporate diverse sources of resistance through limited or repeated backcrossing. Because most of these Ug99-effective genes are of alien origin, co-segregating molecular markers for some of them are already available (Mago et al., 2005; Prins et al., 2001) and can aid selection. Where the alien stem rust resistance genes are linked to leaf rust resistance genes, screening for leaf rust in seedlings or adult plants can also be practiced in countries where Ug99 is absent. To avoid fast breakdown, the best strategy is to use race-specific resistance genes in combinations. Molecular markers provide a powerful tool to identify plants that carry combinations of resistance genes. Table 3 lists available molecular markers that can be used in marker-assisted breeding. Markers for other genes need to be developed to facilitate their utilization. To transfer two or more effective resistance genes into an adapted cultivar the better crossing strategy would be to first cross the resistance sources and then cross the F1 plants with the adapted cultivar. Molecular markers can then be used to select top-cross plants that have desirable agronomic features and carry the targeted resistance genes. Because such plants are expected to be in a low frequency, it is desirable to maintain large family size of 400, which can be obtained by emasculating and pollinating 20 spikes. Further backcross on selected plants will help to restore the characteristics of the recurrent parent. One major issue remains that various currently effective resistance genes are already present in some advanced spring breeding materials that are being tested in various countries to mitigate the immediate threat from Ug99. Should they not be deployed until their combinations are developed, is a difficult issue to resolve with wide range of opinions.

Table 3 PCR (polymerase chain reaction)-based markers associated to stem rust resistance genes effective to Puccinia graminis f. sp. tritici Race Ug99 Sr gene

Chromosome

Marker

Size (bp)

Marker sequence

Reference

Sr2

3BS

gwm533

120

Hayden et al. (2004)

stm598tcac

61

stm559tgag

85

cfa2123

245

cfa2019

234

Sr24#12

500

Sr24#50

200

barc71

85, 103

STSLr19– 130 wmc221

130 190

F 50 GTTGCTTTAGGGGAAAAGCC 30 R 50 AAGGCGAATCAAACGGAATA 30 F 50 GTTGCTTTAGGGGAAAAGCC 30 R 50 TCTCTCTCTCTCTCACACACAC 30 F 50 AAGGCGAATCAAACGGAATA 30 R 50 TGTGTGTGTGTGTGAGAGAGAG 30 F 50 CGGTCTTTGTTTGCTCTAAACC 30 R 50 ACCGGCCATCTATGATGAAG 30 F 50 GACGAGCTAACTGCAGACCC 30 R 50 CTCAATCCTGATGCGGAGAT 30 F 50 -CACCCGTGACATGCTCGTA -30 R 50 - AACAGGAAATGAGCAACGATGT -30 F 50 - CCCAGCATCGGTGAAAGAA -30 R 50 - ATGCGGAGCCTTCACATTTT -30 F 50 - GCGCTTGTTCCTCACCTGCTCATA -30 R 50 - GCGTATATTCTCTCGTCTTCTTGTTGGTT -30 F 50 - CATCCTTGGGGACCTC -30 R 50 - CCAGCTCGCATACATCCA -30 F 50 - ACGATAATGCAGCGGGGAAT -30

Sr22

Sr24/ Lr24

Sr25/ Lr19

7AL

3DL/1BS

7DL

Sr26

6AL

Sr26#43

207

Sr36

2BS

gwm271

171

stm773

195

R 50 - GCTGGGATCAAGGGATCAAT -30 F 50 - AATCGTCCACATTGGCTTCT -30 R 50 - CGCAACAAAATCATGCACTA -30 F 50 CAAGATCGTGGAGCCAGC 30 R 50 AGCTGCTAGCTTTTGGGACA 30 F 50 ATGGTTTGTTGTGTTGTGTGTAGG 30 R 50 AAACGCCCCAACCACCTCTCTC 30

Hayden et al. (2004) Hayden et al. (2004) Khan et al. (2005) Khan et al. (2005) Mago et al. (2005) Mago et al., 2005 Mago et al. (2005) Prins et al. (2001) H. Bariana (personal communication) Mago et al. (2005) Bariana et al. (2001) Bariana et al. (2001) (continued)

Table 3

(continued)

Sr gene

Chromosome

Marker

Size (bp)

Marker sequence

Reference

Sr39/ Lr35

2BS

900

F 50 - AGA GAG AGT AGA AGA GCT GC -30 R 50 - AGA GAG AGA GCA TCC ACC -30

Gold et al. (1999)

Sr1A1R

1AL.1RS

230, 310

F 50 AACGAGGGGTTCGAGGCC 30 R 50 GAGTGTCAAACCCAACGA 30

Mater et al. (2004)

SrR

1BL.1RS

Sr39/Lr35 (Sr39F2/ R3) R173.R (Paw S5/ Paw S6) IB-267

200–300

Mago et al. (2002)

IB-262

200–300

F 50 GCAAGTAAGCAGCTTGATTTAGC 30 R 50 AATGGATGTCCCGGTGAGTGG 30 F 50 GTAGGTAATGTATCAGAGTTGTAC 30 R 50 GTCTTTGTGCTCGGTAGCTCC 30

Mago et al. (2004, 2005)

Stem Rust Threat to Wheat

295

Devoting resources during the next 4–6 years to develop gene combinations knowing that these genes are already in hands of many wheat breeding groups, with no legislation to stop the release of cultivars carrying single resistance genes has provoked the CIMMYT wheat improvement group to focus their breeding effort toward breeding minor genes-based adult plant resistance, especially for areas considered to be under high risk and where survival of the pathogen for several years is expected due to the presence of susceptible hosts and favorable environmental conditions. It is thought that this strategy will allow other areas of the world, especially facultative and winter wheat growing regions to use race-specific resistance genes more successfully in their breeding program.

5.4. Adult plant resistance to Ug99 in old and new wheat Durable stem rust resistance of some older US, Australian, and CIMMYT spring wheats is believed to be due to the deployment of Sr2 in conjunction with other unknown minor, additive genes that could have originated from Thatcher and Thatcher-derived line Chris. Sr2 can be detected through its complete linkage with pseudo-black chaff phenotype, which can be prominently expressed under certain environments leading to its elimination in some breeding programs. However, under the same environmental conditions negligible to high expression of pseudo-black chaff is observed in advanced breeding materials indicating that it is possible to select lines with Sr2 with negligible pseudo-black chaff. On wheat lines that displayed pseudo-black chaff, we observed varying degrees of disease severity in Kenya ranging from traces to about 60–70% compared to 100% severity for highly susceptible materials. Reaction types varying from MR to S (moderately resistant to susceptible) on the same internodes of Sr2 bearing plants clearly indicated that Sr2 did confer at least some resistance. Sr2 was detected in several highly resistant old, tall Kenyan cultivars, including ‘‘Kenya Plume’’ (Singh and McIntosh, 1986), and CIMMYTderived semidwarf wheats ‘‘Pavon 76,’’ ‘‘Parula,’’ ‘‘Kritati,’’ and ‘‘Kingbird.’’ Pavon 76 and Kiritati were resistant since the initiation of rigorous screening in 2005 at Njoro, Kenya with maximum disease scores of 20MRMS. Kingbird, a new advanced line, is at present the best known source of adult plant resistance in semidwarf wheat with maximum score recorded to be 5 MR-MS during the same period. Because these wheats are susceptible as seedlings with race Ug99, their resistance is speculated to be based on multiple additive genes where Sr2 is an important component. With the exception of Sr2, little is known on the genes involved in durable adult plant resistance; however, earlier work done by Knott (1982), knowledge on durable resistance to leaf and yellow rusts (Singh et al., 2004b), and observations made on breeding materials and a F6 mapping population involving Pavon 76 all indicate that the rate of rust progress is a

296

Ravi P. Singh et al.

100

Susceptible

80 % rust severity

1 to 2 minor genes 60

40 2 to 3 minor genes 20 4 to 5 minor genes 0 0

10

20

30

40

50

Days data recorded

Figure 5 Graphical representation of the additive effects from estimated number of minor genes in retarding rust progress in the field.

function of the cumulative effect of the number of minor genes present in a genotype and individual effects of each gene (Fig. 5). Accumulation of between 4 and 5 genes is therefore expected to retard disease progress to rates that result in negligible disease levels at maturity under high disease pressure, described as ‘‘near-immunity’’ by Singh et al. (2000). Accumulating such complex resistance will be cumbersome but not impossible in the absence of disease pressure caused by race Ug99 at most breeding sites and lack of molecular markers associated with genes contributing to resistance. Molecular markers linked to the slow rusting resistance gene Sr2 are known and can be used in selection; however, this gene can also be identified in the field under most environments from its linkage with pseudo-black chaff phenotype. Sr2 is present in about 60% of the current CIMMYT spring wheat germplasm including some of the most recent highyielding wheats that have high level of resistance to leaf and stripe rusts and desirable end-use quality characteristics.

5.5. High-yielding wheat lines with adult plant resistance to stem rust Stem rust screening of CIMMYT advanced breeding materials in Kenya since 2005 has resulted in the identification of 15–20% lines that carried adequate level of resistance under heavy disease pressure. Table 4 summarizes the information on resistance to stem rust in resistant spring wheat germplasm distributed worldwide by CIMMYT during 2006 and 2007, through the newly initiated 1st and 2nd Stem Rust Resistance Screening Nurseries (1st and 2nd SRRSN), and those currently under multiplication

297

Stem Rust Threat to Wheat

Table 4 Stem rust resistance of entries included in 1st and 2nd SRRSN (Stem Rust Resistance Screening Nursery), and candidate entries under multiplication for inclusion in 3rd SRRSN

1st SRRSN

Resistance Category

Number

3rd SRRSN candidates

2nd SRRSN %

Number

%

Number

%

a

Adult plant R (5–10% 4 severity) R-MR (15–20% 19 severity) MR (30% 6 severity) MR-MS (40% 2 severity) MS (50–60% 0 severity) S (70–100% 0 severity) Race specific Sr24 39 Sr25 17 Sr36 (þSr24) 0 Sr1A.1R (þSr24) 2 SrTmp 0 SrSynt 4 SrSha7 9 SrND643 0 SrUnknown 1 Total 103 a

4

0

0

6

4

18

26

20

42

28

6

22

17

36

24

2

15

12

0

0

0

17

13

0

0

0

4

3

0

0

38 17 0 2 0 4 9 0 1

0 0 0 0 25 8 8 0 3 128

0 0 0 0 20 6 6 0 2

0 18 5 0 11 6 8 12 6 150

0 12 3.3 0 7.3 4 5.3 8 4

Adult plant resistance categories include lines that are susceptible in seedling greenhouse tests and with highest rating recorded during multiple years/seasons testing when the susceptible entries annihilated following 100% stem rust severity based on the modified Cobb Scale (Peterson et al., 1948).

to form the 3rd SRRSN for distribution in 2008. A total of 29 (28%), 48 (37%), and 84 (56%) lines in these three nurseries, respectively have shown from high to moderate levels (up to 30% stem rust severity when the susceptible materials show annihilation following 100% severity) of resistance in at least two seasons of evaluation under high disease pressure in Kenya. Entries included in the 2nd and 3rd SRRSN have high yield potential in combination with various other desirable traits. These improved wheat materials have the potential to be released directly or be used by breeding programs worldwide.

298

Ravi P. Singh et al.

Frequency of high-yielding spring wheat materials with stem rust resistance for distribution through other international yield and screening nurseries, such as Elite Spring Wheat Yield Trial (ESWYT) and International Bread Wheat Screening Nursery (IBWSN), is also rising (Table 5). Of the 190 entries included in 29th ESWYT and 41st IBWSN being prepared for growing in 2008–2009 crop seasons, 50 (26%) lines have adequate adult plant resistance. A total of 111 lines that have shown adequate stem rust resistance are being evaluated further for yield performance in Mexico and seed is being multiplied simultaneously during the 2007–2008 crop season to form the 30th ESWYT and 42nd IBWSN. Frequency of entries with race-specific resistance genes is much lower than those with adult plant resistance (Table 5).

5.6. Mexico-Kenya shuttle to breed high-yielding spring wheat with near-immune level of adult plant resistance Because a large portion of CIMMYT high-yielding spring wheat germplasm does not carry effective race-specific stem rust resistance genes to Ug99 and several lines were identified to carry at least moderate levels of resistance, this Table 5 Stem rust resistance of entries included in 29th ESWYT and 41st IBWSN, and entries being multiplied for possible inclusion in 30th ESWYT and 42nd IBWSN

Resistance Category

29th ESWYT and 41st IBWSN

Multiplied 30th ESWYT and 42nd IBWSN

Number

%

Number

%

0 13 37 14 65 44

0.0 6.8 19.5 7.4 34.2 23.2

9 38 64 72 83 59

2.4 10.3 17.4 19.6 22.6 16.0

1 10 3 1 0 2 190

0.5 5.3 1.6 0.5 0.0 1.1

26 6 2 0 3 6 368

7.1 1.6 0.5 0.0 0.8 1.6

a

Adult plant R (5–10% severity) R-MR (15–20% severity) MR (30% severity) MR-MS (40% severity) MS (50–60% severity) S (100% severity) Race specific Sr25 SrTmp SrSynt SrSha7 SrND643 SrUnknown Total a

Adult plant resistance categories include lines that are susceptible in seedling greenhouse tests and with highest rating recorded during multiple years/seasons testing when the susceptible entries annihilated following 100% stem rust severity based on the modified Cobb Scale (Peterson et al., 1948).

Stem Rust Threat to Wheat

299

was viewed as a perfect opportunity to reconstitute high levels of adult plant resistance in newer wheat materials. In the absence of molecular markers for adult plant resistance genes and the absence of Ug99 race in Mexico, a shuttle breeding scheme between two Mexican sites and Njoro, Kenya was initiated in 2006 to transfer adult plant resistance identified in semidwarf CIMMYT wheats to a range of important wheat germplasm. Two crop seasons per year in Mexico and Kenya will accelerate the breeding. The ‘‘single-backcross, selected-bulk’’ breeding approach (Singh et al., 2004b) is being applied for transferring multiple minor genes to adapted backgrounds. Simple and three-way crosses, where one or more parents carry adult plant resistance, are being used to breed new high-yielding, near-immune wheat materials. In the single-backcross approach, we crossed resistance sources with the adapted high-yielding wheats and then a single backcross was made with the recurrent parent to obtain about 400 BC1 seeds. BC1 plants were then selected for desired agronomic features and resistance to leaf and stripe rusts, and harvested as bulk. F2 plants derived from BC1, simple and three-way crosses with desired agronomic features and resistance to leaf and stripe rusts were selected for agronomic traits and resistance to other diseases at CIMMYT research stations in Ciudad Obregon in northwestern Mexico or Toluca in the highlands near Mexico City and harvested as bulk. If F2 populations were grown in Ciudad Obregon, where quarantine disease ‘‘Karnal bunt’’ is known to occur, the F3 populations are grown at Toluca for another round of selection. About 1000 seeds of each of the F3 and F4 populations obtained from harvesting materials at Toluca were grown densely in Njoro, Kenya for selection under high stem rust pressure during the off-season. After removing tall plants, the remaining populations were bulk harvested and about thousand plump grains selected to grow F4 and F5 populations during the main season in Kenya under high disease pressure. Because stem rust affects grain filling, we expect that plants with insufficient resistance will have shriveled grains (Fig. 6). About 400 plump seeds harvested from the selected plants were sent back to Mexico for final selection as individual plants in the F5 and F6 generations at Ciudad Obregon. Individual plant selections will also be made in Kenya. This is the current status as of the 2007–2008 crop season. Selected plants in Ciudad Obregon with good characteristics will be grown as small plots in Toluca and El Batan field sites in Mexico and selected lines will be grown in Kenya for stem rust screening. Selected plants in Kenya with good grain characteristics will be grown in F6 as hill plots or short rows in Kenya as well as small plots in Mexico for final selection. Finally, the resistant F6 plots will be harvested for conducting yield trials in the following crop season in Ciudad Obregon and simultaneously evaluated for stem rust resistance in Kenya. The single-backcross, selected-bulk scheme is also being applied to transfer resistance from old, tall Kenyan cultivars into adapted semidwarf wheats. A shuttle breeding scheme

300

Ravi P. Singh et al.

Figure 6 Effect of stem rust pathogen race Ug99 on wheat spikes and grains of a susceptible (left) and a resistant (right) variety in Njoro, Kenya during 2007.

was also implemented in 2006 between Aleppo in Syria and Kulumsa, or Melkasa, in Ethiopia by the bread wheat improvement program of ICARDA (International Center for Agricultural Research in the Dry Areas). We expect that the frequency of advanced lines which carry high yield potential, maintain wide adaptation, end-use quality characteristics, and high level of resistance to all three rusts will increase over time through the use of the Mexico–Kenya shuttle. Moreover, the proposed approach is expected to rebuild the durable resistance in modern wheat germplasm. Genetic analyses are underway to understand the number and type of resistance genes involved in sources contributing the slow rusting, adult plant resistance. Genomic locations of minor, additive resistance genes, determined through molecular mapping, is expected to not only result in molecular markers for some of the slow-rusting genes but also will be useful to establish and enhance genetic diversity for such genes in the global spring wheat germplasm and will allow their incorporation in facultative and winter wheat materials.

5.7. Reducing the world’s wheat area under susceptible cultivars Potential epidemics following the spread of Ug99 or its variants can be avoided if current susceptible cultivars occupying most of the wheat areas in the primary risk areas in the predicted migration path are reduced. Screening in Kenya during 2005, 2006, and 2007 has identified a few resistant released varieties or advanced breeding materials at various stages of testing in most of the countries that submitted their materials for screening.

301

Stem Rust Threat to Wheat

One strategy is to find ways to ensure that the best, high-yielding resistant materials occupy at least 5% of total wheat area distributed throughout the wheat region and are readily available. This might be via seed supply through procurement in the case that Ug99 establishment is evident in a particular country. However, it will be very difficult to promote resistant varieties on a large scale if they are inferior to the current popular varieties or because farmers have not seen stem rust. Moreover, growing inferior, resistant varieties is not an option as it will affect wheat production at a time when global wheat supply is at its lowest level causing sharp increases in wheat prices. Wheat production must continue to increase at the rate of 2% annually to meet predicted demand of about 800 million tons by 2020 from a current production of about 600 million tons (Fig. 7). Identification and promotion of new stem rust resistant varieties that have significantly enhanced yield potential than current varieties, in conjunction with other desirable traits is probably the best strategy to ensure their fast adoption and thus to succeed in replacing the existing popular but susceptible varieties. This is an achievable objective as most of the current popular varieties were developed during early-to-mid-1990s and yield potential of current CIMMYT spring wheat germplasm has progressed significantly since then (Singh et al., 2007). Yield performances of 14 new high yielding, Ug99 resistant wheat lines together with local check and 15 additional high-yielding entries were determined in replicated yield trials, 2nd Elite Bread Wheat Yield Trial (2nd EBWYT), planted during the 2006–2007 crop season at 27 sites in India, Pakistan, Nepal, Afghanistan, Iran, Egypt, Sudan, Syria, and Mexico. Results from seven sites in India for 10 lines derived from different crosses with high yield potential and adequate resistance for their successful deployment in the northwestern Gangetic Plains, the wheat basket of India, are given in Table 6. The best performing entry Waxwing*2/Kiritati,

Production (million tons)

900

Projected demand until 2020: 2% annual

800 700 600 500

Growth since 1982: 1.3% annual

400 82

86

90

94

98 2 Year

6

10

14

18

Figure 7 Historical and projected future wheat production requirements to meet the demand by year 2020.

302 Table 6 Mean grain yield performance for entries included in the 2nd EBWYT grown at seven sites in northwestern India during 2006–2007 crop season and stem rust response evaluated on three dates at Njoro, Kenya during 2007 Stem rust responsea

Mean grain yield Entry no.

Cross

kg ha

501 510 514

Local check Weebill_1*2/Kiritati Oasis/SKauz// 4*Bcn/3/ 2*Pastor Pfau/Seri.1B// Amadina/3/ Waxwing Weebill_1*2/ Brambling Munia/Chto/3/ Pfau/Bow// Vee#9/4/Chen/ Ae. Sq.//Bcn/5/ Babax/Lr42// Babax

516

517 518

1

Resistance

Rank

% Check

25/09/2007

1/10/2007

11/10/2007

Gene/type

4472.2 4688.6 5055.9

27 19 4

100.0 104.8 113.1

– 5MR-MS 5RMR

– 20MS-S 10RMR

30MS-S 10MR

– APR Sr25

5074.4

3

113.5

20MR-MS

40MR-MS

40MR-MS

APR

4590.8

23

102.7

5MR-MS

15MR-MS

40MR-MS

APR

4953.3

8

110.8

10MS-S

20MR-MS

20MR-MS

APR þ (Sr24)

519 521 527

528 530

a

Babax/LR42// Babax*2/3/Vivitsi Waxwing*2/Kiritati Hpo/Tan//Vee/3/ 2*Pgo/4/Milan/ 5/SSeri1 Pfau/Weaver*2// Kiritati SKauz/Bav92 LSD, P ¼ 0.05 CV (%)

5208.2

2

116.5

15R-MR

30MR-MS

40MR-MS

SrTmp

5222.4 4737

1 18

116.8 105.9

5MR-MS 5MS

20MS-S 10MS

30MS-S 20MS-S

APR APR

4596.6

22

102.8

1MS-S

5MS-S

15MR-MS

APR

4605.9 410.1 11.1

21

103.0

5RMR

10RMR

15RMR

Sr25

Stem rust response has two components: severity based on modified Cobb Scale (Peterson et al., 1948) and host reaction as described in Roelfs et al. (1992). Host reactions are R, resistant; MR, moderately resistant; MS, moderately susceptible; and S, susceptible.

303

304

Ravi P. Singh et al.

numbered 521 in the trial, had16.8% mean grain yield advantage over the local check cultivars PBW502 or mega-cultivar PBW343. PBW502 is a stripe rust resistant descendent of PBW343 and has begun replacing PBW343. Because the adult plant resistance of this entry involves the ‘‘Sr2-complex’’ derived from ‘‘Kiritati’’ parent, and it also has excellent end-use quality characteristics and adult plant resistance to leaf and stripe rusts (data not presented), it could be an excellent option as a potential replacement cultivar. The 2nd and 4th best performing entries numbered 519 and 514 with grain yield advantages of 16.5 and 13.1% over the check have resistance based on race-specific resistance genes SrTmp and Sr25, respectively. Similar results were also seen in other countries where trials were grown indicating that replacement of current cultivars with new Ug99 resistant materials should also result in enhancing productivity even in the absence of stem rust. Yield performances of 27 additional high yielding, stem rust resistant materials are being evaluated as 3rd EBWYT at 50 sites in many more countries in the primary risk area during the 2007–2008 crop season.

5.8. Efforts to identify and develop resistant wheat varieties in secondary risk areas Identifying existing varieties or advanced breeding materials through screening in Kenya or Ethiopia is the top priority at present for developing and developed countries where spring, facultative and winter wheats are grown without the use of fungicides. Developed countries that routinely test their materials for stem rust in Kenya or Ethiopia are the USA, Canada, and Australia. Characterization of a small number of spring wheat materials from Switzerland and Sweden in Kenya showed that they were moderately to highly susceptible. Resistance in materials from other European countries is yet to be characterized; however, intense use of modern broad spectrum fungicides in most of these countries to control other important diseases will also likely control stem rust. Screening of US materials in Kenya are coordinated by the USDA-ARS. Resistance is also identified in important materials through seedling tests with Ug99 (TTKSK) and its variant race TTKST with Sr24 virulence at the USDA-ARS Cereal Disease Laboratory, St. Paul, MN during winter months in quarantine greenhouses. A similar effort is also underway at the Cereal Research Center, Winnipeg, Canada. The resistance status of US materials was described by Jin and Singh (2006). Most breeding programs in the USA and Canada are using this information and diverse sources of resistance in their crossing program. Since the creation of the ‘‘Australian Cereal Rust Control Program’’ in 1973 following the major stem rust epidemic, it is a routine exercise to incorporate new resistance genes in Australian varieties and breeding materials and determine their agronomic suitability (The et al., 1988). This effort

Stem Rust Threat to Wheat

305

has enhanced genetic diversity for resistance in Australian wheat germplasm and therefore new varieties are likely to be identified to replace those susceptible to Ug99. Resistance to Ug99 in spring, facultative and winter wheats from China, Russia, Turkey, Iran, and some countries from Central Asia is also limited. Breeding efforts have been initiated to incorporate diverse race-specific resistance genes in important cultivars and promising advanced wheat breeding materials through marker-assisted backcrossing. A simultaneous effort to introduce adult plant resistance is also planned to be undertaken in China through close collaboration between the ‘‘CIMMYT-CAAS Facultative and Winter Wheat Breeding Program’’ and breeding programs of various other academies, and the ‘‘Turkey-CIMMYT-ICARDA Winter and Facultative Wheat Breeding Program’’ based in Turkey to target winter and facultative wheats grown in Turkey, Iran, Afghanistan, and other countries of Central Asia. Spring wheat germplasm developed by CIMMYT in Mexico are well adapted in most of Central and South America and several South American breeding programs heavily use CIMMYT materials in their breeding program. This will allow identification of adapted stem rust resistant materials for deployment in the near future.

6. Conclusion and Future Outlook Considering the progress made in identifying stem rust resistance in existing varieties and advanced breeding materials, though in low frequency, and high priority to incorporate diverse or durable adult plant resistance by various wheat breeding programs including those of CIMMYT and ICARDA, it is unlikely that stem rust race Ug99 or its descendents will destroy the world wheat crop. However, localized epidemics as observed in Kenya during 2007 cannot be ruled out. Migration of Ug99 is being monitored carefully through field surveys, monitoring nurseries and GIS tools to provide an early warning, which could allow chemical interventions if necessary and guide decision making. Reducing the area currently occupied by susceptible varieties in the primary risk areas of Africa, Arabian Peninsula, Middle East, and West-South Asia with resistant ones should become an immediate priority. It is highly advisable to release in these areas varieties that have durable, adult plant resistance, or have effective race-specific resistance genes in combinations, to prevent further evolution and selection of new virulences leading to ‘‘Boom-and-Bust’’ cycles. This will also allow reduction of inoculum in high-risk areas and thus reduce risks of its spread to secondary risk areas. The emergence of Ug99 race of stem rust pathogen as a global threat to wheat production has highlighted the need and benefits of public-funded

306

Ravi P. Singh et al.

wheat research and improvement where germplasm carrying useful traits and information can be readily shared to ensure a sustainable production of staple food crops at low cost and with the least negative environmental impacts. Response to the alarm raised by Dr. N. E. Borlaug of the global threat from Ug99 has been positive and effective from wheat scientists, research leaders, and importantly the donor community who came forward to support the global effort to mitigate the threat. However, to succeed much needs to be done, especially to ensure that seed of high yielding, stem rust resistant materials is made available to as many farmers as possible to ensure that their wheat crop is not destroyed by stem rust.

ACKNOWLEDGMENTS We wish to express our deepest gratitude to Nobel Laureate Dr. N. E. Borlaug for his tireless and successful efforts to sound the alarm on the threat of stem rust Ug99. We also wish to acknowledge the vital, catalytic role played by Dr. Borlaug’s long-time associate, Mr. C. Dowswell. Finally, we acknowledge the support of our own institutions (CIMMYT, INIFAP, USDA-ARS, and KARI) as well as various other organizations, National Programs, and financial support particularly from USAID, CIDA-Canada, ICAR-India, and USDA-ARS; without which we could not have achieved the progress reported here.

REFERENCES Aquino, P., Carrion, F., and Calvo, R. (2002). Selected wheat statistics. In ‘‘CIMMYT 2000–2001 World Wheat Overview and Outlook: Developing No-Till Packages for Small Scale Farmers’’ ( J. Ekboir, ed.), pp. 52–62. CIMMYT, Mexico, D.F. Bariana, H. S., Hayden, M. J., Ahmed, N. U., Bell, J. A., Sharp, P. J., and McIntosh, R. A. (2001). Mapping of durable adult plant and seedling resistances to stripe rust and stem rust diseases in wheat. Aust. J. Agric. Res. 52, 1247–1255. Bhardwaj, S. C., Nayar, S. K., Prashar, M., Kumar, J., Menon, M. K., and Singh, S. B. (1990). A pathotype of Puccinia graminis f. sp. tritici on Sr24 in India. Cereal Rusts Powdery Mildews Bull 18, 35–37. Biffen, R. H. (1905). Mendel’s laws of inheritance and wheat breeding. J. Agric. Sci. 1, 4–48. Brown, J. K. M., and Hovmller, M. S. (2002). Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science 297, 537–541. CIMMYT (2005). ‘‘Sounding the Alarm on Global Stem Rust.’’ http://www.globalrust. org/uploads/documents/SoundingAlarmGlobalRust.pdf (accessed on November 29, 2007). CIMMYT, Mexico, D.F. Draxler, R. R., and Rolph, G. D. (2003). ‘‘HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory).’’ http://www.arl.noaa.gov/ready/hysplit4.html (accessed on November 29, 2007). NOAA Air Resources Laboratory, Silver, Spring, MD. FAO (Food and Agriculture Organization of the United Nations). (2001). ‘‘Food Outlook.’’ FAO, Rome, Italy. Gold, J., Harder, D., Townley-Smith, F., Aung, T., and Procunier, J. (1999). Development of a molecular marker for rust resistance gene Sr39 and Lr35 in wheat breeding lines. Electron. J. Biotechnol. 2, 35–40.

Stem Rust Threat to Wheat

307

Hayden, M. J., Kuchel, H., and Chalmers, K. J. (2004). Sequence tagged microsatellites for the Xgwm533 locus provide new diagnostic markers to select for the presence of stem rust resistance genes Sr2 in bread wheat (Triticum aestivum L.). Theor. Appl. Genet. 109, 1641–1647. Hodson, D. P., Singh, R. P., and Dixon, J. M. (2005). An initial assessment of the potential impact of stem rust (race Ug99) on wheat producing regions of Africa and Asia using GIS. In ‘‘Abstracts. 7th International Wheat Conference’’, p. 142. November 27–December 2, 2005, Mar del Plata, Argentina. Huerta-Espino, J. (1992). ‘‘Analysis of Wheat Leaf and Stem Rust Virulence on a Worldwide Basis.’’ Ph.D. thesis, University of Minnesota, USA. Hugh-Jones, M. E. (2002). Agricultural bioterrorism. In ‘‘High-Impact Terrorism: Proceedings of a Russian-American Workshop,’’ pp. 219–232. National Academy Press, Washington, DC. Jin, Y. (2007). Resistance to race TTKS of Puccinia graminis f. sp. tritici in Chris and related spring wheat. Phytopathology 97, S162(Abstract). Jin, Y., and Singh, R. P. (2006). Resistance in US wheat to recent eastern African isolates of Puccinia graminis f. sp. tritici with virulence to resistance gene Sr31. Plant Dis. 90, 476–480. Jin, Y., Pretorius, Z. A., and Singh, R. P. (2007a). New virulence within race TTKS (Ug99) of the stem rust pathogen and effective resistance genes. Phytopathology 97, S137(Abstract). Jin, Y., Singh, R. P., Ward, R. W., Wanyera, R., Kinyua, M., Njau, P., Fetch, T., Pretorius, Z. A., and Yahyaoui, A. (2007b). Characterization of seedling infection types and adult plant infection responses of monogenic Sr gene lines to race TTKS of Puccinia graminis f. sp. tritici. Plant Dis. 91, 1096–1099. Jin, Y., Pretorius, Z. A., Singh, R. P., and Fetch, T., Jr (2008). Detection of virulence to resistance gene Sr24 within race TTKS of Puccinia graminis f. sp. tritici. Plant Dis. 92: In press. Khan, R. R., Bariana, H. S., Dholakia, B. B., Naik, S. V., Lagu, M. D., Rathjen, A. J., Bhavani, S., and Gupta, V. S. (2005). Molecular mapping of stem and leaf rust resistance in wheat. Theor. Appl. Genet. 111, 846–850. Knott, D. R. (1980). Mutation of a gene for yellow pigment linked to Lr19 in wheat. Can. J. Genet. Cytol. 22, 651–654. Knott, D. R. (1982). Multigenic inheritance of stem rust resistance in wheat. Crop Sci. 22, 393–399. Knott, D. R. (1988). Using polygenic resistance to breed for stem rust resistance in wheat. In ‘‘Breeding Strategies for Resistance to the Rusts of Wheat’’ (N. W. Simmonds and S. Rajaram, eds.), pp. 39–47. CIMMYT, Mexico, D.F. Kolmer, J. A. (2001). Early research on the genetics of Puccinia graminis stem rust resistance in wheat in Canada and the United States. In ‘‘Stem Rust of Wheat: From Ancient Enemy to Modern Foe’’ (P. D. Peterson, ed.), pp. 51–82. APS Press, St. Paul, MN. Leonard, K. J. (2001). Stem rust—Future enemy? In ‘‘Stem Rust of Wheat: From Ancient Enemy to Modern Foe’’ (P. D. Peterson, ed.), pp. 119–146. APS Press, St. Paul, MN. Le Roux, J., and Rijkenberg, F. H. J. (1987). Pathotypes of Puccinia graminis f. sp. tritici with increased virulence for Sr24. Plant Dis. 71, 1115–1119. Luig, N. H. (1985). Epidemiology in Australia and New Zealand. In ‘‘Cereal Rusts, Vol. II: Diseases, Distribution, Epidemiology, and Control’’ (A. P. Roelfs and W. R. Bushnell, eds.), pp. 301–328. Academic Press, Orlando. Mago, R., Spielmeyer, W., Lawrence, G. J., Lagudah, E. S., Ellis, J. G., and Pryor, A. (2002). Identification and mapping of molecular markers linked to rust resistance genes located on chromosome 1RS of rye using wheat-rye translocation lines. Theor. Appl. Genet. 104, 1317–1324. Mago, R., Spielmeyer, W., Lawrence, G. J., Ellis, J. G., and Prior, A. J. (2004). Resistance genes for rye stem rust (SrR) and barley powdery mildew (Mla) are located in syntenic regions on short arm of chromosome. Genome 47, 112–121.

308

Ravi P. Singh et al.

Mago, R., Bariana, H. S., Dundas, I. A., Spielmeyer, W., Lawrence, G. J., Pryor, A. J., and Ellis, J. G. (2005). Development of PCR markers for the selection of wheat stem rust resistance genes Sr24 and Sr26 in diverse wheat germplasm. Theor. Appl. Genet. 111, 496–504. Mater, Y., Baenzinger, S., Gill, K., Graybosch, R., Whitcher, L., Baker, C., Specht, J., and Dweikat, I. (2004). Linkage mapping of powdery mildew and greenbug resistance genes on recombinant 1RS from ‘Amigo’ and ‘Kaukaz’ wheat-rye translocations of chromosome 1RS.1AL. Genome 47, 292–298. McIntosh, R. A. (1988). The role of specific genes in breeding for durable stem rust resistance in wheat and triticale. In ‘‘Breeding Strategies for Resistance to the Rust of Wheat’’ (N. W. Simmonds and S. Rajaram, eds.), pp. 1–9. CIMMYT, Mexico, D.F. McIntosh, R. A., Luig, N. H., Milne, D. L., and Cusick, J. (1983). Vulnerability of triticales to wheat stem rust. J. Plant Pathol. 5, 61–69. McIntosh, R. A., Wellings, C. R., and Park, R. F. (1995). ‘‘Wheat Rusts: An Atlas of Resistance Genes.’’ CSIRO Publications, Victoria, Australia. Nagarajan, S., and Joshi, L. M. (1985). Epidemiology in the Indian subcontinent. In ‘‘The Cereal Rusts, Vol. II: Diseases, Distribution, Epidemiology, and Control’’ (A. P. Roelfs and W. R. Bushnell, eds.), pp. 371–402. Academic Press, Orlando. Peterson, R. F., Campbell, A. B., and Hannah, A. E. (1948). A diagrammatic scale for estimating rust intensity of leaves and stem of cereals. Can. J. Res. Sect. C 26, 496–500. Pretorius, Z. A., Singh, R. P., Wagoire, W. W., and Payne, T. S. (2000). Detection of virulence to wheat stem rust resistance gene Sr31 in Puccinia graminis f. sp. tritici in Uganda. Plant Dis. 84, 203. Prins, R., Groenewald, J. Z., Marias, G. F., Snape, J. W., and Koebner, R. M. D. (2001). AFLP and STS tagging of Lr19, a gene conferring resistance to leaf rust in wheat. Theor. Appl. Genet. 103, 618–624. Rajaram, S., Singh, R. P., and Torres, E. (1988). In ‘‘Current CIMMYT Approaches in Breeding Wheat for Rust Resistance. Breeding Strategies for Resistance to the Rust of Wheat’’ (N. W. Simmonds and S. Rajaram, eds.), pp. 101–118. CIMMYT, Mexico, D.F. Reynolds, M. P., and Borlaug, N. E. (2006). Applying innovations and new technologies from international collaborative wheat improvement. J. Agric. Sci. 144, 95–110. Roelfs, A. P. (1985). Wheat and rye stem rust. In ‘‘The Cereal Rusts, Vol. II: Diseases, Distribution, Epidemiology, and Control’’ (A. P. Roelfs and W. R. Bushnell, eds.), pp. 3–37. Academic Press, Orlando. Roelfs, A. P., and Martell, L. B. (1984). Uredospore dispersal from a point source within a wheat canopy. Phytopathology 74, 1262–1267. Roelfs, A. P., and Martens, J. W. (1988). An international system of nomenclature for Puccinia graminis f. sp. tritici. Phytopathology 78, 526–533. Roelfs, A. P., Singh, R. P., and Saari, E. E. (1992). ‘‘Rust Diseases of Wheat: Concepts and Methods of Disease Management.’’ CIMMYT, Mexico, D.F. Rotem, J., Wooding, B., and Aylor, D. E. (1985). The role of solar radiation, especially UV, in the mortality of fungal spores. Phytopathology 75, 510–514. Rowell, J. B., and Romig, R. W. (1966). Detection of urediospores of wheat rusts in spring rains. Phytopathology 56, 807–811. Saari, E. E., and Prescott, J. M. (1985). World distribution in relation to economic losses. In ‘‘The Cereal Rusts, Vol. II: Diseases, Distribution, Epidemiology, and Control’’ (A. P. Roelfs and W. R. Bushnell, eds.), pp. 259–298. Academic Press, Orlando. Shank, R. (1994). Wheat stem rust and drought effects on Bale agricultural production and future prospects. Report on February 17–28 assessment. In ‘‘United Nations Emergencies Unit for Ethiopia.’’ http://www.africa.upenn.edu/eue_web/Bale_mar.txtaccessed on November 29, 2007. Addis Ababa, Ethiopia.

Stem Rust Threat to Wheat

309

Singh, R. P. (1991). Pathogenicity variations of Puccinia recondita f. sp. tritici and P. graminis f. sp. tritici in wheat-growing areas of Mexico during 1988 and 1989. Plant Dis. 75, 790–794. Singh, R. P., and McIntosh, R. A. (1986). Genetics of resistance to Puccinia graminis tritici and Puccinia recondita tritici in Kenya plume wheat. Euphytica 35, 245–256. Singh, R. P., and McIntosh, R. A. (1987). Genetics of resistance to Puccinia graminis tritici in ‘Chris’ and ‘W3746’ wheats. Theor. Appl. Genet. 73, 846–855. Singh, R. P., Huerta-Espino, J., Rajaram, S., and Crossa, J. (1998). Agronomic effects from chromosome translocations 7DL.7Ag and 1BL.1RS in spring wheat. Crop Sci. 38, 27–33. Singh, R. P., Huerta-Espino, J., and Rajaram, S. (2000). Achieving near-immunity to leaf and stripe rusts in wheat by combining slow rusting resistance genes. Acta Phytopathlogica Hungarica 35, 133–139. Singh, R. P., Huerta-Espino, J., Pfeiffer, W., and Figueroa-Lopez, P. (2004a). Occurrence and impact of a new leaf rust race on durum wheat in the northwestern Mexico from 2001–2003. Plant Dis. 88, 703–708. Singh, R. P., William, H. M., Huerta-Espino, J., and Rosewarne, G. (2004b). Wheat rust in Asia: Meeting the challenges with old and new technologies. In ‘‘New Directions for a Diverse Planet: Proceedings of the 4th International Crop Science Congress,’’ http://www. cropscience.org.au./icsc2004/symposia/3/7/141_singhrp.htm (accessed on November 29, 2007). September 26–October 1, 2004. Brisbane, Australia. Singh, R. P., Hodson, D. P., Jin, Y., Huerta-Espino, J., Kinyua, M., Wanyera, R., Njau, P., and Ward, R. W. (2006). Current status, likely migration and strategies to mitigate the threat to wheat production from race Ug99 (TTKS) of stem rust pathogen. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources 1, 54. Singh, R. P., Huerta-Espino, J., Sharma, R., Joshi, A. K., and Trethowan, R. M. (2007). High yielding spring bread wheat germplasm for global irrigated agro-ecosystems. Euphytica 157, 351–363. Stakman, E. C. (1957). Problems in preventing plant disease epidemics. Am. J. Bot. 44, 259–267. Stakman, E. C., and Piemeisel, F. J. (1917). A new strain of Puccinia graminis. Phytopathology 7, 73. Steele, K. A., Humphreys, E., Wellings, C. R., and Dickinson, M. J. (2001). Support for a stepwise mutation model for pathogen evolution in Australasian Puccinia striiformis f. sp. tritici by use of molecular markers. Plant Pathol. 50, 174–180. Sunderwirth, S. D., and Roelfs, A. P. (1980). Greenhouse characterization of the adult plant resistance of Sr2 to wheat stem rust. Phytopathology 70, 634–637. The, T. T., Latter, B. D. H., McIntosh, R. A., Ellison, F. W., Brennan, P. S., Fischer, J. A., Hollamby, G. J., Rathgen, A. J., and Wilson, R. E. (1988). Grain yield of near isogenic lines with added genes for stem rust resistance. In ‘‘Proceedings of the 7th International Wheat Genetics Symposium’’ (T. S. Miller and R. M. D. Koebner, eds.), pp. 901–906. Institute of Plant Science Research, Cambridge, UK. Vanderplank, J. E. (1963). ‘‘Plant Diseases: Epidemics and Control.’’ Academic Press, New York and London. Wanyera, R., Kinyua, M. G., Jin, Y., and Singh, R. P. (2006). The spread of stem rust caused by Puccinia graminis f. sp. tritici, with virulence on Sr31 in wheat in Eastern Africa. Plant Dis. 90, 113.

This page intentionally left blank

C H A P T E R

S I X

Genetic Improvement of Forage Species to Reduce the Environmental Impact of Temperate Livestock Grazing Systems M. T. Abberton, A. H. Marshall, M. W. Humphreys, J. H. Macduff, R. P. Collins, and C. L. Marley Contents 1. Introduction 2. Reducing Diffuse Nitrogenous Pollution of Watercourses 2.1. Rationale 2.2. Traits associated with NUE 2.3. Mapping approaches 2.4. The role of forage legumes 2.5. Reducing losses from ensiled red clover 3. Reducing P Pollution of Watercourses 3.1. Rationale 3.2. Breeding for enhanced P use efficiency 3.3. PUE in the rumen 4. Reducing Emissions to Air 4.1. Nitrous oxide 4.2. Ammonia 4.3. Increasing the efficiency of rumen processes 4.4. Manipulating water soluble carbohydrate and protein content 4.5. Exploiting interspecific variation in rumen proteolysis 4.6. Reducing methane emissions 4.7. The role of tanniferous forage species in reducing methane emission from enteric fermentation 4.8. Greenhouse gas emissions from fertilizer production 5. Improving Soil Quality and Reducing Flood Damage 5.1. Flood tolerance and prevention 5.2. Soil porosity and compaction 5.3. Biodiverse mixtures and ecosystem services

312 315 315 316 318 320 321 321 321 323 324 325 325 327 328 328 329 331 332 334 335 335 337 338

Plant Breeding and Genetics Programme Institute of Grassland Environmental Research, Plas Gogerddan, Aberystwyth, Ceredigion SY23 3EB, United Kingdom Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00206-X

#

2008 Elsevier Inc. All rights reserved.

311

312

M. T. Abberton et al.

6. Enhancing Persistency and Resilience 6.1. The role of interspecific hybridization 6.2. Exploiting genetic variation for multiple traits 7. Enhancing C Sequestration in Grasslands 8. Future Prospects Acknowledgments References

339 339 340 341 344 345 345

The livestock agriculture of temperate grasslands is a major provider of meat and milk to the world. These areas also deliver important ecosystem services and are central to tourism, amenity, and leisure in many countries. However, they are also major sources of pollution of waterways and of greenhouse gas emissions. In this review, we focus on how the genetic improvement of some of the major crop species of temperate grasslands can contribute to reduced environmental impacts and climate change mitigation including the enhancement of carbon sequestration. The main species considered are the ryegrasses and fescues and the clovers. With regard to diffuse pollution of waterways, increasing the efficiency with which the plant utilizes nitrogen and phosphorus has significant potential to reduce the amounts available for leaching and overland flow. This also allows reduced fertilizer use, in itself representing a significant saving of greenhouse gas emissions. Changes in the composition of the plant diet have the potential to increase the efficiency of nitrogen use in the rumen and reduce the amount of methane produced by enteric fermentation. Carbon sequestration in temperate grasslands may potentially be enhanced by understanding and altering root architecture and turn over and also litter composition. The role of forage breeding in improving soil quality and flood defense is also considered and wider aspects of the role of persistent perennial species discussed. To fully realize the potential of genetic improvement, breeding programs need to incorporate state-of-the-art genomics and also to be guided by modeling studies. An integrated approach combining the skills of animal scientists, soil scientists, and plant breeders within the context of life cycle analysis is required to find sustainable solutions to the challenge of maintaining the productivity of livestock agriculture while reducing its environmental footprint.

1. Introduction In their entirety, grasslands cover 70% of the world’s agricultural area (Soussana and Luscher, 2007), of which temperate grassland, both natural and anthropogenic in origin, is a major component (Coupland, 1979). Their management in ways that are economically, environmentally, and socially sustainable in the face of frequently conflicting demands is likely to remain a major challenge throughout the 21st century.

Genetic Improvement of Forage Species

313

Historically, grassland management has focussed on food production. However, there is increasing recognition of the need to approach grasslands from the viewpoint of ‘‘multi-functionality,’’ in so far as they are perceived as important not only for livestock production, but also as sources of pollution, while, at the same time, delivering an expanding range of ‘‘ecosystem services’’ as well as underpinning tourism, amenity, and leisure industries in many parts of the world. Recent interest in the potential of grassland to mitigate climate change, for example, through carbon (C) sequestration ( Jones and Donnelly, 2004), is a case in point. The changes in the range of management objectives associated with grassland have been reviewed recently by Kemp and Michalk (2007). The environmental impact of livestock production has become an area of increasing concern (Steinfeld et al., 2006) and the need to mitigate the effects of climate change is widely seen as urgent (Stern, 2007). There are also transnational and national governmental initiatives in place in many parts of the world aimed at reducing pollution of waterways, arising both from environmental concerns and the economic costs of cleaning up water supplies. In Europe, for instance, the Water Framework Directive of the European Union (EU) introduces further legislation concerning water quality. At the same time there are growing concerns over the need to protect soil quality, reflected in the UK, for example, by the development and implementation of soil action plans (Environment Agency, 2007). An important strand to the scientific ‘‘push’’ occurring in parallel with the ‘‘pull’’ from these policy drivers is the development of new crop varieties, underpinned by the recent advances in genetic and genomic understanding. The impact of climate change on European grasslands and their likely role in climate change mitigation have been recently reviewed (Hopkins and Del Prado, 2007; Mannetje, 2007; Soussana and Luscher, 2007). Morgan (2005) considered the global picture in terms of the response of grazing lands to increased atmospheric carbon dioxide, while Hopkins and Del Prado (2007) identified a number of likely responses including increased herbage growth, increased use of forage legumes particularly white and red clover and alfalfa, reduced opportunities for grazing and harvesting on wetter soils, greater incidence of summer drought, and increased leaching from more winter rainfall. However, in terms of C mitigation options, only a reduction in the number of animals can currently be registered as ‘‘a reduction’’ in accordance with The Intergovernmental Panel on Climate Change (IPCC, 1997). Although a wide range of grassland systems and species are found in the temperate zones, this chapter focuses on agriculturally improved grasslands and the role of genetic improvement of forages (i.e., plant breeding underpinned by genetic, physiological, and agronomic knowledge) in reducing the environmental impact of livestock agriculture, particularly grazing, through improvements in resource utilization efficiencies and mitigation of climate change impacts on productive grassland.

314

M. T. Abberton et al.

The key economic species of improved grasslands for grazed livestock throughout substantial parts of Europe, Australasia, and to a lesser extent North and South America are forage grasses in the families Lolium and Festuca and forage legumes in the family Trifolium. Our assessment focuses on the improvement of these species while recognizing the importance of others, in particular alfalfa (Medicago sativa), which is the most important temperate forage legume but is often cut rather than grazed, timothy (Phleum pratense), and cocksfoot (Dactylis glomerata). Much of the work we report is applicable to all temperate grasslands where ryegrasses and clovers predominate. However, many of the examples are drawn from the situations with which we are most familiar, in the UK. The ryegrasses, perennial ryegrass (Lolium perenne) and Italian ryegrass (Lolium multiflorum), are widely considered to offer the optimal combination of forage production and forage quality and provide excellent fodder for livestock agriculture. Their economically productive range is limited by climatic conditions as both species (especially Italian ryegrass) are susceptible to extreme summer and winter conditions (Humphreys et al., 1997). Overall, ryegrasses account for over 60% of all agricultural grass seed used in Europe. Where conditions are suboptimal for ryegrass, alternative more robust species such as timothy, tall fescue (Festuca arundinacea), meadow fescue (Festuca pratensis), cocksfoot, and Kentucky bluegrass (Poa pratensis) are normally employed. The impact of climatic conditions on species utilization across Europe is illustrated by the seed market comparisons between Norway, Poland, France, and the UK. In Norway, growth of L. perenne is confined to areas with the least winter stress and involves only 200 T/year of seeds of L. perenne marketed for 2 year stands. This compares with 1000 T/year of timothy (P. pratense) and 400 T/year of meadow fescue (F. pratensis). In contrast, 75% of the seed marketed annually in the UK is L. perenne, that is, 10,000 T/year (and 13% is L. multiflorum). In Poland, about 1000 T/year of L. perenne is used currently for reseeding meadows and pastures as part of complex seed mixtures comprising between 15% and 50% of L. perenne (information provided by Germinal Holdings (UK), Szelejewo Plant Breeding, Poland, and the Norwegian Crops Research Institute; EU FPV project SAGES http://www.iger.bbsrc.ac.uk/SAGES2/sages2.html). White clover (Trifolium repens L.) is the most widely grown temperate forage legume and is the most common forage legume in pastures grazed by sheep or cattle. Estimates of white clover use include 15 Mha in Australasia and 5 Mha in the USA, with global sowing totaling of 3–4 Mha annually. In comparison, the worldwide area of alfalfa (M. sativa) is thought to be around 30 Mha, with annual sowings about twice that of white clover. However, these figures probably underestimate the total area sown to white clover, due to the contribution of ‘‘indigenous’’ strains of this species and its much longer life than alfalfa. In Europe, the most widespread use of this species is in northern and western parts of the continent, although reliable

Genetic Improvement of Forage Species

315

estimates of the clover content of either recently reseeded or more established pastures on a wider scale are difficult to obtain. Imports of seed produced in Australasia account for a substantial proportion of the total sown in Europe. In comparison, the annual sowing of red clover (Trifolium pratense) is estimated as 1.5 Mha globally. Red clover is grown widely in northern areas of Europe, in particular Scandinavia, Germany, Eastern Europe, and the Baltic States. It is also common in Russia, Japan, and parts of South and North America with around 4–5 Mha of red clover grown in the USA alone. The species described above are the major source of grazed and conserved feed for dairy, beef, and sheep production in many temperate areas. Consequently, plant breeding programs for these and other forage species have traditionally been targeted toward increasing productivity and nutritional quality. In recent years, however, a number of programs have shifted their focus toward increasing resource use efficiency and reducing the environmental footprint of temperate livestock systems as a whole. While this in part reflects the changing national and international policy drivers alluded to above, it has been facilitated by the development and application of a range of molecular tools lending both greater speed and precision to operational breeding. Until now, most systematic approaches to improving the environmental impact of grassland agriculture (e.g., Hatch et al., 2004) have ignored plant breeding-based solutions, relying instead on managing nutrient inputs and outputs in order to reduce emissions or diffuse pollution (e.g., via manure application, or improved guidelines for fertilizer use). However, given the historical successes of plant breeding with respect to improving the yield, quality, and adaptive range of many crops, there seems to be no reason in principle why it should not be equally successful with respect to traits conferring explicit environmental benefits (Humphreys et al., 2006). In the following sections, we discuss recent progress toward increasing the role of genetic improvement aimed at (i) reducing diffuse nitrogenous pollution of waterways, (ii) reducing losses of phosphorus (P) to water, (iii) increasing the efficiency of processes in the rumen to reduce greenhouse gas emissions, (iv) enhancing soil quality and reducing the incidence and severity of flooding, and (v) understanding and enhancing the role of temperate grassland in C sequestration.

2. Reducing Diffuse Nitrogenous Pollution of Watercourses 2.1. Rationale Nitrate pollution is a longstanding major issue of water quality in Europe and a focus of EU legislation, including the Nitrates Directive and the Water Framework Directive. Hence, there is a continuing need to develop

316

M. T. Abberton et al.

measures to promote the more efficient utilization and retention of plantavailable forms of N within agricultural systems. This will not only assist farmers in meeting regulatory requirements but also result in cost savings from reduced fertilizer use. Current nutrient use efficiencies in temperate agroecosystems vary considerably, depending on the transfer components considered. For example, European nitrogen (N) use efficiencies from soil to crop vary between 40% and 80%, but on a whole farm basis range from 10% to 40% for whole dairy farm, compared with 40–80% for whole arable farm (Neeteson et al., 2004). While most nutrient budget information for grassland systems refers to N and P, the available information for K (Alfaro et al., 2003) indicates that K uptake by plants is the main factor controlling K efficiency. Interestingly, nutrient budget studies at the farm-scale have suggested a significant uncoupling/independence between N, P, and K in terms of surpluses and deficits (e.g., Berry et al., 2003). Breeding forage grasses and legumes with higher nutrient use efficiencies in terms of nutrient capture, utilization, and retention offers a ‘‘clean technology’’ route to increased sustainability through lower fertilizer inputs and reduction in the agricultural footprint with respect to pollution and the wider consumption of resources. Its strategic importance is underscored by anticipated trends in global fertilizer consumption and attendant energy costs. While, for example, there has been a gradual decline in average annual NPK fertilizer applications to agricultural land in the UK over the last 20 years (Defra, 2005), global demand for fertilizers continues on an upward trend. According to Frink et al. (1999), 85–90 million metric tonnes (MMt) of N fertilizers are currently applied annually worldwide, compared with 10.2 MMt in 1960. This figure is expected to rise to 240 MMt by 2050 (Tilman, 1999). Life cycle assessment of environmental impacts of European agricultural production systems (Brentrup et al., 2004) indicate that application of N fertilizer at optimal economic rates is compatible on balance with optimal environmental impact, significant under or oversupply with N fertilizer leading to decreasing resource use efficiency primarily, respectively, in terms of land used and eutrophication. The recommendation to maintain optimum yields in order to use land most efficiently with N applied according to crop demand in order to minimize nitrate leaching, provides additional rationale for improving N use efficiency (NUE) on a ‘‘more for less’’ basis.

2.2. Traits associated with NUE NUE has been defined in various ways depending on the scope of the system, whether mass balance or nutrient flux approaches are favored, and on whether uptake, utilization, and nutrient retention within the plant, together with their physiological subcomponents, are considered explicitly

Genetic Improvement of Forage Species

317

(for reviews see Baligar et al., 2001; Garnier and Aronson, 1998; Good et al., 2004; Gourley et al., 1994; Greenwood et al., 2005; Lea and Azevedo, 2006). In an agronomic context, NUE commonly relates yield/productivity to nutrient input, and assumes two components such that NUE ¼ NUpT NUtE, where NUpE is the ‘‘nutrient uptake efficiency’’ (e.g., Shaver and Melillo, 1984), defined as the ratio between the amount of nutrient absorbed by the plant and that supplied/available in the soil, and providing a measure of the efficiency with which roots intercept and capture nutrients. NUtE is the ‘‘utilization efficiency,’’ defined as the ‘‘the unit dry matter produced per unit nutrient in the dry weight’’ (Marschner, 1986) or ‘‘the dry matter flux per unit nutrient flux’’ in a whole stand (Ricklefs, 1990), in units of gram biomass per mole of nutrient. This, essentially, ‘‘black-box’’ definition, with minor variations, has been adhered either explicitly or implicitly in most quantitative trait loci (QTL) studies of NUE, or of its NUpE and NUtE components in crop species. For example, Loudet et al. (2003) defined N use efficiency (NiUE) as the ‘‘quantity of N used to build up a certain amount of biomass.’’ Significant advantages in terms of trait dissection are provided by fluxbased approaches to the analysis of NUE in ecosystems (Berendse and Aerts, 1987; Vitousek, 1982). These offer a semi-mechanistic framework for understanding the physiological processes contributing to the NUtE component, defined as the ‘‘total net primary production per unit nutrient absorbed annually’’ at the whole plant or stand level; formally partitioning this between nutrient productivity and nutrient retention, such that NUtE¼aNPMRT, where aNP is ‘‘mean annual nutrient productivity’’ and MRT is ‘‘mean residence time of the nutrient within the plant.’’ Tradeoffs between attributes conferring high aNP and those conferring high MRT can be assessed by principle component analysis (Garnier and Aronson, 1998). Studies with grasses indicate aNP depends on several factors including the ratio of photosynthesis to leaf N concentration and interorgan N allocation, while MRT depends on N resorption efficiency (REFF) and rate of biomass loss (i.e., organ life span) (Atkin et al., 1996; Garnier et al., 1995). The range of REFF among species appears to be relatively low (e.g., 0.4–0.58), while leaf life span varies sixfold; hence, at the single plant level, the limited available evidence suggests that leaf life span may control the MRT parameter more tightly than REFF. The prerequisite for genetic improvement of NUE is the availability of genetic variation in relevant traits (e.g., Gorny, 1999; Gorny and Sodkiewicz, 2001; Osbourne and Rengel, 2002; Presterl et al., 2003). This has been demonstrated for acquisition (Vance et al., 2003; White et al., 2005; Wissuwa and Ae, 2001) and associated traits such as root surface area and root hair growth (Dunbabin et al., 2003; Gilroy and Jones, 2000), utilization (Hacisalihoglu and Kochian, 2003; Wilkins et al., 1997; Witt et al., 1999), and nutrient retention within the plant (Granstedt, 2000;

318

M. T. Abberton et al.

Johnston, 2000). Genetic variation has also been reported for root distribution in perennial ryegrass and tall fescue (Bonos et al., 2004; Crush et al., 2005) and for root length in rice (Price et al., 1997). Phylogenetic studies also indicate that shoot P and organic N concentrations are species level traits, while shoot C, Ca, and Mg concentrations are influenced by more ancient evolutionary processes, with K intermediate (Broadley et al., 2004). Critical ratios of N:P:K for optimal yield vary between crop species (Greenwood et al., 1980) and genetic variation in shoot P and Kþ concentrations has been demonstrated in a range of forage grasses (Sleper et al., 1977; Vogel et al., 1989), including Italian and perennial ryegrass (Easton et al., 1997; Moseley and Baker, 1991). While nutrient uptake, utilization, and retention have all been targeted explicitly (e.g., Gallais and Hirel, 2004; Mickelson et al., 2003; Sheehy et al., 2005), genetic improvement in NUE has traditionally been pursued through selection for yield at a given, usually high, level of nutrient input (Good et al., 2004), with little attempt to understand its physiological and genetic basis (Hirel et al., 2001). However, recent applications of transgenic, genomic, and quantitative molecular mapping techniques have enabled a more mechanistic approach (e.g., Gallais and Hirel, 2004). Although transgenic studies in model species like Arabidopsis and Nicotiana have generally shown little impact of overexpression of N transporters and nitrate reductase on yield or NiUE, there are positive effects for enzymes associated with remobilization of N (Good et al., 2004). However, in view of positional and background effects (e.g., Opsahl et al., 2002) its stand-alone value remains questionable when addressing complex traits like NUE (Good et al., 2004).

2.3. Mapping approaches Mapping population-based QTL studies of agronomic, physiological, and metabolic traits associated with NUE, usually considered as yield response to nutrient supply, have been reported in a number of model species and crops, although, as yet, few forage species. For NiUE, these include Arabidopsis (Loudet et al., 2003; Rauh et al., 2002), maize (Gallais and Hirel, 2004), rice (Obara et al., 2001), barley (Mickelson et al., 2003), and perennial ryegrass (Van Loo et al., 1997, 2003). In the majority of nutrient-related studies, candidate traits have been scored at ‘‘final harvest’’ after growth at a single (e.g., Harada and Leigh, 2006), or high/low rates of nutrient supply (e.g., Hirel et al., 2001; Loudet et al., 2003). QTLs have been identified for (i) yield and its components (Gallais and Hirel, 2004; Hirel et al., 2001; Van Loo et al., 1997), (ii) root morphology (Beebe et al., 2006), tissue nutrient concentrations (Harada and Leigh, 2006; Hirel et al., 2001; Loudet et al., 2003), (iii) N assimilation enzymes such as total leaf glutamine synthetase (GS) activity (EC 6.3.1.2) and leaf NADH-NR (EC 1.6.6.1)

Genetic Improvement of Forage Species

319

(Hirel et al., 2001), and more rarely (iv) nutrient transfer rates (e.g., Gallais and Hirel, 2004; Mickelson et al., 2003). Significant progress has been made in characterizing some of the QTLs associated with NUE. For example, in Arabidopsis grown under high or low N supply, Loudet et al. (2003) reported heritabilities of around 0.5 for yield and N-related traits, with 4–9 QTLs for each trait, and individual QTL explaining 2–21% of total phenotypic variation. These authors also reported extensive genotypeenvironmental (GE) effects. The position of QTLs not only varies with level of N supply (e.g., Bertin and Gallais, 2000; Loudet et al., 2003) but also with form (i.e., nitrate or ammonium) of supply (e.g., Rauh et al., 2002), implying that the metabolic rate-limiting steps for N utilization vary with supply level and N source, and that the genetic factors controlling plant development are specific to one N environment, with several sets of genes expressed according to the rate of N supply. Interestingly, more QTLs for yield and its components in maize have been detected in plants grown under high N input, whereas more QTLs for N-content have been detected in plants grown under low N input (Gallais and Hirel, 2004). The co-localization of QTLs for N assimilatory enzyme activity, especially cytosolic GS and QTLs for yield variation has also been reported in this species (Hirel et al., 2001). Similarly, in rice, Obara et al. (2001) reported coincidence between a QTL for a yield trait and a structural gene for GS1, suggesting that GS1 might be a key component of NUE and yield. Co-localization of QTLs for different traits can be explained in terms of linkage (i.e., two different closely linked genes influence two different traits independently) or pleiotropy (i.e., the same genetic factor controls both traits) (Lebreton et al., 1995). However, these findings highlight the problem of discriminating between QTLs associated with plant growth/development as opposed to NUE. It has been suggested that QTLs mapped consistently across different N treatments are more likely to be growth/ development related, while QTLs demonstrating large environmental interactions are more likely to be conferred by genes in the N use pathways (Rauh et al., 2002). Perhaps surprisingly, there is still little evidence for which of the three main efficiency components, uptake, utilization, and retention/remobilization offers the most promising target for improving NUE in forage grasses and legumes. However, some indication can be drawn from experience with other species and in particular maize. Genetic variation in NiUE for maize grown under high N input was explained by variation in N uptake, while at low N input it was mainly due to differences in utilization efficiency (Gallais and Hirel, 2004). Likewise, based on correlations between the efficiency of primary N assimilation, N remobilization and yield in maize, Purcino et al. (1998) concluded that increased grain yield during the last two decades derives from more efficient N remobilization rather than enhanced inorganic N metabolism, and specifically from greater

320

M. T. Abberton et al.

leaf longevity improving the ratio between assimilate supply from source leaves and demand in sink leaves during grain filling. The importance of N remobilization in conferring N utilization efficiency has also been highlighted in QTL studies of barley (Mickelson et al., 2003).

2.4. The role of forage legumes In their systems synthesis study of dairy farms, Jarvis et al. (1989) found that use of white clover, especially at relatively low clover contents, was an effective approach to reducing nitrogenous losses. However, there was a cost to production and losses per livestock unit did not differ markedly from those under some alternative management systems. Parsons et al. (1991) showed that 80% of the sheep carrying capacity of a grass sward receiving 420 kg N/ha/year could be maintained with a white clover content of 5%, or less, fixing only 24 kg N/ha/year, leading to a marked reduction in nitrogenous losses. In a similar vein, Jarvis et al. (1989) showed that 66% of the support energy for grassland management on a dairy farm came from fertilizer production and that this could be more than halved by the use of white clover. However, the ultimate fate of biologically fixed N with respect to environmental loss pathways remains a major concern with the use of forage legumes in livestock systems. Davies et al. (1998) considered this in a comparison of ploughed grass and grass/clover swards and Ledgard et al. (1999) studied losses under grazing by dairy cows. Clearly, significant leaching losses can occur under grass/clover swards and in some circumstances they are comparable to those from moderately fertilized grass swards (i.e., 200 kg N/ha/year). Consequently, there is a need to consider the role of germplasm improvement in reducing such losses. The evidence that the use of forage legumes in mixed swards to reduce or replace mineral N fertilizer applications lowers nitrate leaching losses is equivocal (Frame and Laidlaw, 2005). Leaching can occur not only as a result of ploughing up legume rich swards, but also under a standing crop, for example, following cutting of red clover. However, the processes of root and nodule senescence contributing to losses of N are under genetic control and amenable to selection. Assessing both the magnitude and genetic variation in rhizosphere N fluxes associated with root and nodule senescence remains technically challenging. In recent studies at IGER, conducted in Flowing Solution Culture, nodule loss following severe defoliation were compared across a range of red clover varieties and selection lines, and significant genetic variation in propensity for nodule loss was identified. Further work, using small lysimeters, has begun to quantify the relationship between nodule loss and nitrate leaching, and promising lines with reduced nodule loss have been identified, providing the basis for the development of future varieties.

Genetic Improvement of Forage Species

321

2.5. Reducing losses from ensiled red clover New opportunities are also arising within forages to select for specific traits that could reduce postharvest N losses to the environment. A good example of this approach is the emerging research on the enzyme polyphenol oxidase (PPO), which is at a particularly high level of activity in red clover in comparison with other forage species and plays a significant role in protein protection (Owens et al., 2002). This enzyme converts phenols to quinones, which subsequently bind to protein and slow the rate of protein degradation. One of the most promising applications for this kind of inhibition is in the reduction of postharvest losses of N during ensilation. For example, it has been shown that ensiling alfalfa leads to the degradation of 44–87% of forage protein to non forage protein (NPN). In comparison, red clover has up to 90% less proteolysis (Sullivan et al., 2004). Current work at IGER is seeking to enhance the effectiveness of PPO inhibition of proteolysis in ensiled red clover through genetic improvement, thereby reducing the protein available in silo for diffuse pollution of N (e.g., as ammonia). Significant variation for PPO activity in red clover germplasm and differences in activity throughout the year have been shown The targeting of PPO levels for genetic improvement in red clover, as a route to reduced nitrogenous pollution, has been facilitated by recent developments in the cloning of the gene encoding the PPO enzyme (Sullivan et al., 2004). There is also increasing evidence of significant PPO activity within a number of forage grasses (Lee et al., 2006). Winters et al. (2003) reported the existence of an active form of PPO in perennial ryegrass and although its properties were distinct from those of red clover PPO, it also formed high-molecularweight protein complexes in leaf extracts. Of the seven grass species studied by Lee et al. (2006), PPO activity was highest in cocksfoot and at levels that significantly reduced proteolysis.

3. Reducing P Pollution of Watercourses 3.1. Rationale Land managers in the UK are increasingly required to comply with a range of new and existing EU and domestic legislation all of which aims to improve the quality of water (in rivers, lakes, ground water, transitional, and coastal waters). At the EU and domestic level, this legislation includes the Water Framework, Nitrate and related Directives including the OSPAR Convention on Eutrophication. Eutrophication, or nutrient enrichment of surface waters, is highly detrimental to ecosystem health. When it occurs over a relatively short period of time, it is usually due to anthropogenic influences. P is frequently

322

M. T. Abberton et al.

the limiting nutrient in these aquatic systems and P pollution from agriculture is a common cause of eutrophication. Run-off of dissolved nutrients from applied P fertilizers and erosion and delivery of P sorbed to soil particles are both important sources of P pollution of water (Bilotta et al., 2007). At the same time, crop yield is limited by P availability in 30–40% of the arable cropped area (Runge-Metzger, 1995), and global reserves of phosphate fertilizers are estimated at 85 years (USGS, 2002). Furthermore, fertilizer processing is heavily dependent on energy and the costs of oil and gas are likely to rise in the foreseeable future, while the cheapest sources of minerals such as rock phosphate are rapidly being depleted (IFDC, 2005; Orris and Chernoff, 2002; Vance, 2001). Surplus P has accumulated in many UK soils year-on-year over several decades. The majority of UK farms still operate on the basis of annual P surplus (Haygarth and Jarvis, 1999), and these are significantly higher (>20 kg/ha) in intensive livestock production compared with arable (<10 kg/ha) systems (Edwards and Withers, 1998). This is underlined by estimates that livestock systems account for 7700 tonnes out of the 12,000 tonnes of P contributed annually to surface waters by UK agriculture (White and Hammond, 2006). The risks to water quality in the UK (i.e., eutrophication) arising from P transfer by solubilization (leaching), physical detachment (particles and colloids), and incidental mechanisms (direct transfer of fertilizer and slurry) are further increased by the location of significant tracts of improved grassland within major river catchments. Approximately, 60% of river length in England still contains phosphate levels of >0.1 mg P/l, compared with <10% of river length in Scotland and Wales (Defra, 2005). Potential mitigation options for a typical range of farm systems in England and Wales were collated in a Defra-funded program at IGER North Wyke (Haygarth, pers comm.) to assess their potential effectiveness, in reducing mass transfers of P to watercourses, and the potential cost to the farming industry. This was achieved by plotting ‘‘cost curves’’ for prioritising cost and effectiveness, to help planners and scientists gauge where priority areas for further investment and research are required. A simple model framework incorporating a number of assumptions identified 15 methods for mitigating inputs of P to agricultural systems, 19 methods for preventing mobilization of P, and 6 methods for controlling the transport of P to streams. In all cases cost and effectiveness were determined. It was found that the largest reductions in P inputs were to grassland and horticulture, reflecting the preferential accumulation of P at the surface in grassland soils and the use of short season succession cropping in horticultural systems. As a result of this and similar studies, the development of forage varieties capable of reducing the velocity and intensity of overland flow of P has become a new breeding objective at IGER. One of the ways currently being explored to achieve this is through the production of

Genetic Improvement of Forage Species

323

ryegrass/fescue hybrids, combining the agronomic advantages of ryegrass with the erect, broad, and rigid foliage found in several fescue species (e.g., tall fescue). Measures to increase soil porosity and water holding capacity that mitigate against flooding will also reduce overland flow of P (Section 5.1).

3.2. Breeding for enhanced P use efficiency A large number of studies have been carried out, mainly in New Zealand, on the genetics and physiology of the response by white clover to differences in P supply (Caradus, 1994; Caradus and Dunn, 2000; Caradus et al., 1992, 1993). Heritabilities and a positive response to selection for the ability to grow at low P levels have been reported (Crush, 1995). However, it has generally proved difficult to translate knowledge of genotypic variation in response to P supply, or even the ability to grow well at low P levels, into selection criteria yielding the kind of robust and reliable improvements in these traits required in new cultivars of white clover. This is due to a number of factors, not least the difficulty of achieving reliable and effective phenotyping and of relating screens performed in ‘‘artificial’’ conditions (e.g., controlled environment chambers and glasshouse) with performance in the field. A second factor is the complex nature of soil P chemistry (Arai and Sparks, 2007) and the attendant interactions between plant available forms of P, soil conditions, and the ability of the plant to respond through, for example, enzyme release, root hair growth, and changes in metabolism. Further difficulties arise from the genetic heterozygosity and phenotypic plasticity of white clover, and from its ability to form symbioses with vesicular arbuscular mycorrhizal fungi ( VAM). These fungi play a role in the mobilization and absorption of P particularly forms of P that may be otherwise inaccessible to the plant. In return they benefit from C supplied from the plant’s photosynthesis. However, a feature of agricultural soils in temperate regions is that in many cases they have received significant inputs of P in the form of fertilizer and/or manures over many years. It remains an open question as to the extent to which mycorrhizal symbioses are of importance under such circumstances. Preliminary studies on acid upland soils at the IGER site at Bronydd Mawr near the Brecon Beacons in Wales suggest that specific interactions between mycorrhiza and white clover, in terms of the extent of root colonization may be important with respect to yield. Although limited progress has been made in terms of variety development specifically targeted at improving P use efficiency (PUE), understanding of the basic processes involved and their genetic control has developed considerably. This has arisen largely from work on a number of ‘‘model’’ systems which, in addition to conventional models like Arabidopsis thaliana, includes species such as Phaseolus vulgaris and white lupin (Lupinus albus), which are either perceived as agronomically problematic with respect to

324

M. T. Abberton et al.

aspects of P nutrition, or are especially amenable to experimental approaches (e.g., analysis of changes in root architecture). Such work has given considerable insights into changes in gene expression underlying response to changes in availability of P. The challenge is to integrate our growing knowledge of the genomic basis of PUE with our understanding of the mechanisms leading to pollution of watercourses. Clearly, one of the crucial missing links remains the dissection of the key traits involved in the field situation and their conversion to robust and reliable selection criteria. QTL studies of PUE have so far been restricted to relatively few crop species compared with NUE, including rice (Ni et al., 1998; Wissuwa and Ae, 2001) and P. vulgaris (Beebe et al., 2006). A total of 26 QTLs for P accumulation and associated root attributes were identified in P. vulgaris (Beebe et al., 2006) and four associated with differences in P uptake in rice (Wissuwa and Ae, 2001). Fine mapping of genes underlying QTLs associated with PUE is also less advanced compared with NUE. However, genes involved in root development and morphology are regarded as prime candidates for P (Franco-Zorilla et al., 2004). There is also evidence for cross talk between K and P responses (Wang et al., 2002) and for two sequential major transcriptional programs in response to P starvation (Hammond et al., 2003): a generic stress response and a later P-specific response—(i) a transient early stage, preferentially involving genes of general stress response and (ii) a more highly induced later response to Pi starvation stress including genes presumed to play specific role in responding to Pi starvation.

3.3. PUE in the rumen Our understanding of the role of P within the rumen and of the factors determining its utilization efficiency remains relatively limited compared with N. Studies of the fate of P in ruminants are often confounded by interactions with other parameters (e.g., differences in passage rates and particle size among different feeds). However, it is known that P is essential for maximum cellulose digestion by rumen microorganisms (Hall et al., 1961). Hence, it might be expected that increasing the availability of P to rumen microbes in ruminants fed on high forage diets would lead to an increased energy supply, as cellulose breakdown is improved. Nutrient budgeting has become a widely used tool for managing nutrients on farms and as an indicator of the sustainability of farming practices. In livestock enterprises, up to 75% of excreted N can be quickly lost from slurry as ammonia whereas losses of nonvolatiles, such as P, are small. Due to these changes, slurry becomes increasingly P-rich relative to the fertilizer requirements of pastures. This imbalance has a significant implication for farm nutrient budgeting. For example, following N losses from slurry, the land area needed to recycle slurry P is greater than that needed for slurry N.

Genetic Improvement of Forage Species

325

As agricultural manures contribute approximately 30% of P inputs to surface waters in the UK, the impact of diet on P excretion is becoming increasingly important (Powers and Van Horn, 2001). In pursuit of sustainable and economically viable livestock systems which meet consumer demands, landowners are under increasing pressure to maximize the use of home-grown, forage-based diets for their livestock, such as those including forage legumes. Recent research has shown that ensiled legume forages can be used as productive winter feeds for ruminants, with lambs fed on these silages having higher liveweight gains, due to improved food conversion and N utilization efficiency, when compared with lambs fed ensiled ryegrass (Marley et al., 2007). However, despite the potential for these forage legumes to contribute to livestock systems, there is little information on their effects on P budgets at the animal, farm-gate or landscape level. Previous studies indicate that white clover lines may differ in the ratio of inorganic to total P (Caradus et al., 1998). In monogastric animals, phytate P (myo-inositol hexakisphosphate—an organic form of P in plants) is not readily available for absorption by the animal (Nelson, 1967). In ruminants, there is evidence that the rumen microbial population produces sufficient levels of the enzyme, phytase, to hydrolyze phytate P, making it available to the animal (Clark et al., 1986). However, it has been shown more recently that phytate may not be completely hydrolyzed prior to faecal excretion, and that P digestibility is improved (Kincaid et al., 2005) and P excretion reduced (Knowlton et al., 2007) when exogenous phytase is added to ruminant diets. As optimizing P digestibility is a key strategy in reducing P accumulation on livestock farms (Valk et al., 2000), there is a need to determine the ratio of inorganic to total P in forages being developed for livestock systems and their effects on P digestibility when fed to ruminants.

4. Reducing Emissions to Air 4.1. Nitrous oxide Nitrous oxide (N2O) emissions arise directly from N inputs to soil (e.g., animal excreta, fertilizer, manure, crop residues and fixed N) and also indirectly from nitrates. Current mitigation options include the more precise matching of fertilizer and manure N applications to crop demand, refinements in application methods and timing and the use of nitrification inhibitors such as dicyandiamide (e.g., Di and Cameron, 2004). Initiatives targeted at reducing nitrate leaching or ammonia volatilization are also likely to reduce N2O emissions. The soil processes controlling N2O production (i.e., nitrification and denitrification) are affected by a range of

326

M. T. Abberton et al.

Ratio of NH4+ : NO3− uptake

abiotic factors (e.g., aeration, temperature and pH) as well as by fertilizer addition and organic matter content (Hopkins and Del Prado, 2007). Given the dependence of N2O emissions on microbially mediated transformations, one of the breeding strategies currently under investigation at IGER, apart from direct selection for increased NUE (see Section 2.2) is based on enhancing the preferential uptake of NH4þ by forage grasses, with a view to competing more effectively with soil nitrifying microorganisms and thereby reducing the production of NO3 in the soil. Building on work reported by Clarkson et al. (1986) and demonstrating considerable temperature-dependent variation in the ratio of NH4þ: NO 3 uptake by perennial ryegrass cv. S23, studies with genotypes exhibiting contrasting phenotypes for NUE (Fig. 1.) and subsequently with two of IGER’s perennial ryegrass mapping families, grown in flowing solution culture, have revealed promising genetic variation in relative uptake of NH4þ and NO3 under steady-state conditions of supply. A number of QTLs have been identified for high uptake rates, some common to both forms of N and some unique to either NH4þ or NO3. The initial aim is to produce selections exhibiting a preferential NH4þ: NO3 uptake ratio of 10:1, compared with 3:1 commonly found in existing varieties. These will be used to provide ‘‘proof of principle’’ with respect to the plant’s potential for modulating nitrification rates in the soil. The alternative, enhancing direct competition for NO3, will also be tested with selections exhibiting prefer ential uptake of NO3 over NHþ 4 . The development of NO3 specific genotypes would be of additional interest in the context of reducing NO3 leaching. A complementary breeding approach currently being evaluated at IGER is aimed at promoting better soil structure and soil aeration (Section 5.1), which, among other effects, may reduce denitrification. 7 6 5 4 3 2 1 0 (dh ) (45/6) (33/4) (35/1) (42/5) (41/6) (30/8) (31/4) Genotype Figure 1 Cumulative NHþ 4 :NO3 uptake ratios by eight genotypes of perennial ryegrass (n ¼ 72 plants), selected on the basis of contrasting component traits for NUE, during 10 days growth under constant concentrations of 10 mM NH4NO3 maintained in flowing solution culture. (Source: Defra Project LS3617.)

Genetic Improvement of Forage Species

327

4.2. Ammonia Agriculture makes a contribution of 7% to UK greenhouse gas emissions and is estimated to produce 80–90% of European NH3 emissions 80–90% of which comes from urine and faeces produced by livestock (Misselbrook et al., 2005). Nonagricultural emissions arise from a large number of relatively small sources (Sutton et al., 2000), and therefore the greatest reductions in NH3 emissions are likely to be achieved by reducing those from agriculture. The urinary N in livestock is mainly in the form of urea ( Jarvis et al., 1989) and other compounds that are readily hydrolyzed to NH3, giving rise to high concentrations of NH4þ in the soil solution depending on pH Together these labile-N compounds in livestock excreta are referred to as total ammoniacal-N (TAN) and may be regarded as the source of almost all NH3 emissions. This TAN is also the major source, in the season following manure application and grazing, of NO3 leaching and a significant source of N2O, N2, and NO emissions (Yamulki and Jarvis, 1997). The current NH3 emissions ceiling for the UK (set by the Gothenburg protocol) is 297 kilotonnes for the year 2010. This target is being revised and will most likely be reduced. Therefore, the NH3 emissions ceiling target is one potential limit to the livestock sector as a result of livestock and manure management under new climate change scenarios. The rapid breakdown of herbage proteins in the rumen and inefficient incorporation of herbage N by the rumen microbial population are major determinants of N loss in excreta (Dewhurst et al., 1996). A wide range of management strategies for minimizing losses have been recommended (e.g., Peoples et al., 1995) and much of the current breeding effort in this area is centered on increasing the efficiency of rumen processes through forage compositional changes and optimization of the energy/ protein balance, as described below (Sections 4.3–5). More direct approaches are also being pursued at IGER, including the selection of perennial ryegrass genotypes for enhanced uptake of NH4þ, as described above (Section 4.1), and for greater tolerance to the locally high NH4þ concentrations in the environs of dung and urine patches. Selection for enhanced NH4þ uptake also exploits the approximate 1:1 stoichiometry between NH4þ uptake and Hþ release, shifting the NH3: NH4þ equilibrium in the soil solution in favor of NH4þ and thereby assisting the retention and recovery of N applied in manure and slurry to pasture. A similar approach is merited for white clover, particularly as studies in flowing solution culture at IGER have indicated that under steady-state conditions of NH4þ supply its growth is severely inhibited at concentrations an order of magnitude lower than those affecting perennial ryegrass (i.e., 200 mg NH4þ-N/dm3 compared with 2000 mg NH4þ-N/dm3).

328

M. T. Abberton et al.

4.3. Increasing the efficiency of rumen processes In the rumen, scarcity of readily available energy during the time of maximal protein degradation restricts microbial protein synthesis. Ammonia accumulates as a waste product and is absorbed from the rumen and excreted as waste N in urine. This imbalance in nutrient supply therefore reduces the efficiency of conversion of forage protein in the rumen and increases N-pollution. It has been shown that when sheep (MacRae and Ulyatt, 1974) and cattle (Ulyatt et al., 1988) are given fresh forages they can waste 25–40% of forage protein. Where amino acids are used to produce energy, ammonia is produced as a waste product. Alternatively, scarcity of readily available energy (ATP) during time of maximal protein degradation restricts microbial protein synthesis and only a modest proportion of the available N released from the forage protein is incorporated into microbial protein. In the cases outlined above, ammonia accumulates as a waste product and is absorbed from the rumen, subsequently converted to urea in the liver and excreted as waste N in urine. This imbalance in nutrient supply has been variously described in the literature as the ‘‘asynchronous’’ release of N and energy for microbial growth, its main effects being to reduce the efficiency of conversion of forage N in the rumen and increase environmental N-pollution. The N lost from livestock excreta and manures undergoes various fates. Soluble N compounds are quickly converted in soil to nitrate-N (NO3) which enters ground and surface waters following application of manures to land and from grazed pasture Around 29% of leached NO3 is estimated to arise from livestock (Webb et al., 2000). Ammonia (NH3), when deposited to land, increases N eutrophication and soil acidification. Genetic improvement of the forage grasses and legumes in the ruminant diet has the potential to reduce emissions to air. Two possible strategies of increasing the efficiency of conversion of forage-N to microbial-N have been suggested: (i) increase the amount of readily available energy accessible during the early part of the fermentation and (ii) provide a level of protection to the forage proteins, thereby reducing the rate at which their breakdown products are made available to the colonizing microbial population.

4.4. Manipulating water soluble carbohydrate and protein content To compensate for the scarcity of readily available energy during the time of maximal protein degradation in the rumen, new cultivars of forage grasses and legumes providing a better balance between water soluble carbohydrate (WSC) and crude protein (CP) are being developed at IGER through selection for increased WSC content in the grasses and reduced protein

Genetic Improvement of Forage Species

329

content in the legumes. In the latter case, IGER’s unique non-fixing inbred genotypes of white clover (Michaelson-Yeates et al., 1998) were used to demonstrate the principle that material of lower leaf protein content shows much slower protein degradation in the silo. Following this, we established that genotypic variation within elite gene pools of white clover is much greater than that was previously thought. Initial crosses have been made to produce lines for further testing using in vitro rumen simulation and this will lead to the development of improved varieties. Other approaches to developing white clover varieties with a lower CP content involve exploiting the variation that exists elsewhere within the Trifolium genus. Of particular promise in this respect is the interspecific hybridization of white clover and the rhizomatous, drought tolerant species Trifolium ambiguum M. Bieb (Kura or Caucasian clover), the development of which has recently been reviewed by Abberton (2007). Hybrids between these two species have been developed using a number of strategies and have shown lower CP contents and higher WSC contents than the white clover parent. Analysis of the forage quality characteristics of advanced backcross hybrids showed some backcrosses to have a CP content 14.2 g/kg DM lower and a WSC content 6 g/kg DM higher than the parental white clover (Marshall et al., 2004). Modeling studies using IGER’s SIMSDAIRY model indicate that this could lead to a potential 8% reduction in loss of dietary N. Advanced lines of these hybrids are currently progressing to the stage of commercial variety development, and although they were originally developed with the specific objective of introgressing the rhizomatous trait into white clover, as a strategy to improve the persistence and drought tolerance of white clover, their superior agronomic performance coupled with their forage quality characteristics suggests they have much to offer future grassland systems.

4.5. Exploiting interspecific variation in rumen proteolysis Grass forage enters the rumen where it is subjected to anaerobic conditions, a neutral pH of 6.8 and an ambient temperature of 39 C. This induces abiotic stresses to intact ingested cells such as oxygen depletion and elevated temperature, and as a result causes plant-mediated proteolysis which contributes to poor NUE in ruminants. Ryegrass protein has a particularly short half-life under simulated rumen-like conditions, while under similar conditions certain fescue species in particular are much more efficient (Table 1). In experimental lines (Shaw, 2005), the trait in part was transferred successfully from Festuca glaucescens into experimental lines of Italian ryegrass (L. multiflorum) (IRG), and some improvements in the protein stability of Italian ryegrass leaves in rumen-simulated conditions were achieved. An F1 hybrid between L. multiflorum and F. glaucescens had a protein half-life of 6.5 h (compared to 2.06 h for L. multiflorum) and when the F1 hybrid was

330

M. T. Abberton et al.

Table 1 Half-life values of leaf protein in rumen-like anaerobic conditions at 39 C Species

Protein half life (h)

Lolium perenne (perennial ryegrass) Lolium multiflorum (Italian ryegrass) Festuca pratensis (meadow fescue) Festuca glaucescens (glaucous fescue) Festuca mairei (Atlas fescue)

4.33 2.06 5.84 15.83 8.93

backcrossed onto L. multiflorum, backcross 1 (BC1) genotypes were recovered with double the protein stability of the ryegrass. The stability of the protein in the rumen varies with crop phenology (e.g., some large differences were found in samples taken in September and November in IRG, fescue and its hybrids (Shaw, 2005) and its genetic control remains poorly understood. There is considerable evidence that irrespective of fluctuations in protein stability at different times throughout the year, living cells of the fescue species are affected significantly less than ryegrass to exposure to the rumen stresses that induce plant-mediated proteolysis. Furthermore, despite a diminishing stress resistance response over two generations of backcrossing onto ryegrass, the trait for increased protein stability compared to ryegrass in the rumen is retained and is heritable. Some progress in locating the genes responsible of using marker-assisted-selection with AFLP markers and the use of total genomic species-specific DNA probes to identify and localize fescue-derived introgressions, has been achieved. Further research into the genetic control of the trait is required to select the ideal fescue gene complement necessary for generating the optimal protein stability that is possible when transferred into the ryegrass genome. The decrease in net protein in the rumen is the result of the degradation of proteins, especially Rubisco (Beha et al., 2002; Kingston-Smith and Thomas, 2003). However, there is also some evidence to suggest newly synthesized protein is generated in the rumen (Shaw, 2005) but this continued for only 1 h with F. glaucescens and for 4 h with L. multiflorum. During preliminary investigations, the newly synthesized proteins were assessed for a potential protein protection and whether F. arundinacea was more effective than Lolium in synthesizing heat-shock proteins capable of conferring some thermo-tolerance to ingested cells in the rumen. In simulated tests, a heat shock protein HSP-70 was not generated by L. multiflorum until after a 6 h exposure to rumen temperatures (39 C). The slow expression of the heat shock protein may indicate why plantmediated proteolysis and protein degradation proceeds so quickly when L. multiflorum is ingested. On the other hand, F. glaucescens induces expression

Genetic Improvement of Forage Species

331

of HSP-70 even before the onset of the stress conditions, which may derive from its Mediterranean origins (Humphreys et al., 1997) where it tolerates exposure to high summer temperatures.

4.6. Reducing methane emissions On a global scale, agriculture and in particular enteric fermentation in ruminants produces between 21% and 25% of the total anthropogenic emissions of methane. However, in more rural communities, agriculture contributes almost 50% of annual methane emissions with over 90% of this due to rumen fermentation. In the long term, and for many parts of the world where livestock production is important, plant breeding strategies to reduce methane emissions are likely to become of increasing utility. Such strategies may be involved in the development of production systems that result in lower animal numbers without sacrificing economic viability: this is a clear route to reduce methane emissions. That is, emissions per animal would be constant but emissions per unit of production (milk and meat) would decrease. However, another important approach is to build on our (limited) understanding of rumen function and to modify plant composition in such a way as to result in reduced emission per animal without detriment to productivity or health. The two major sources of agricultural methane emission are enteric fermentation in livestock and livestock manures. In this review, we will focus on genetic improvement strategies to reduce the former since this is not only the most important source but also the most amenable to improvement through breeding. However, it should be noted that approaches to alter the composition of livestock diets will also have an effect on manure composition; for example, C:N ratio will affect the decomposition rate. Hopkins and Del Prado (2007) have reviewed the range of strategies to reduce methane emission from enteric fermentation and categorize them into (i) dietary changes, (ii) direct rumen manipulation, and (iii) systematic changes. The latter include considerations of breed, livestock numbers, and intensiveness of production. Increasing production intensity may result in lower methane emissions, but may have other less desirable environmental impacts. This highlights the importance of conducting a rigorous life cycle analysis in which trade-offs between different outcomes can be explicitly considered. Strategies based on direct manipulation of the rumen include reducing the numbers of protozoa parasitizing methanogenic bacteria in the rumen and the addition of ionophores to enhance propionate levels. Both these approaches have drawbacks, however, as reducing protozoa may lead to metabolic disease and the main ionophores used are antibiotics such as monensin where issues of resistance may limit utility. Dietary manipulations include the addition of organic acids (aspartate, malate, and fumarate) and yeast culture. These compounds encourage the production of propionate

332

M. T. Abberton et al.

and butyrate in the rumen which compete for hydrogen and reduce the ability of methanogenic microbes to produce methane. Research into the efficacy of these approaches and the optimum method of delivery of organic acids is on going. There is evidence that using forages with high WSC in animal diets can directly reduce methane emissions (Lovett et al., 2004). It has been demonstrated that increasing the WSC content in perennial ryegrass by 33 g/kg reduces methane production in vitro by 9%. To enhance this effect, changes in the composition of the accompanying forage legume component are being assessed with respect to lignin, fatty acid composition, and overall digestibility. Plant secondary metabolites such as tannins and saponins have also been employed in attempts to reduce methane emissions from enteric fermentation. Although the efficacy of these compounds in reducing methane emissions remains disputed, we believe they represent a promising target for genetic improvement and therefore discuss them in some detail in the following section.

4.7. The role of tanniferous forage species in reducing methane emission from enteric fermentation Forage legumes such as Lotus corniculatus (birdsfoot trefoil) and L. uliginosus (greater trefoil) possess secondary metabolites known as condensed tannins (CTs) that form pH reversible bonds with forage proteins which reduce degradation of protein to ammonia by rumen microorganisms, yet release protein at low pH in the abomasums (Ramirez-Restrepo and Barry, 2005). CTs are flavonoid polymers which complex with soluble proteins and render then insoluble in the rumen; yet release them under the acidic conditions found in the small intestine, reducing bloat and increasing amino acid absorption (Barry et al., 1986; Morris and Robbins, 1997). Methane production from housed sheep fed forage diets that contain L. corniculatus has been shown to be lower than those on ryegrass-based pasture (Ramirez-Restrepo and Barry, 2005) associated, in part, with the presence of condensed tannin (CT) in Lotus foliage. Similar responses have also been found in dairy cows grazing Lotus. The mode of action by which CT reduces methane production is unclear, as is the effect of different levels of CT in Lotus on methane production. Recent work at IGER (Marley et al., 2006) has demonstrated significant differences in CT content between varieties of L. corniculatus as well as intra-varietal variation. Germplasm lines with CT contents ranging from 20 mg/g DM to >100 mg/g DM are now available and are being used at IGER to quantify the relationship between CT content and methane production in combination with other forage species.

Genetic Improvement of Forage Species

333

The reactivity of CT differs between forage species in terms of increasing the net absorption of amino acids from the rumen (Min et al., 2003), with the CTs from L. corniculatus particularly effective in increasing their absorption. Hence, in a parallel approach, rhizomatous lines of L. corniculatus with considerably improved persistence and competitive fitness in mixed swards have been developed at IGER. Progress with the development of rhizomatous varieties has also been made in the USA (Wen et al., 2003). The development of Lotus germplasm with stable CT levels is critical as, depending upon concentration, these compounds can have beneficial or detrimental effects on ruminant livestock. The CT content required to control bloat is considered to be 5 mg CT/g DM (Li et al., 1996), while 20–40 mg CT/g DM may be regarded as being optimal for improved ruminant production (Aerts et al., 1999a). CT Levels greater than 60 mg/g DM can reduce voluntary intake and have been reported to depress digestion efficiency (Barry and Manley, 1986). The CT content of herbage from Lotus is influenced by a number of environmental and developmental factors including elevated CO2, temperature and drought (Carter et al., 1999). There is evidence that the CT content of L. corniculatus is lower than L.uliginosus (Aerts et al., 1999b; Sivakumaran et al., 2006) and that levels in stems are lower than in leaves and flowers and that CT content can vary over the growing season (Gebrehiwot et al., 2002). CT content has been reported as being higher in rhizomatous material and in these experiments, CT levels were influenced by the presence of a companion grass and also fluctuated widely during spring and autumn seasons (Wen et al., 2003). Developing germplasm with a CT content that is relatively stable is therefore desirable. A recent study conducted at IGER (Marshall et al., 2005a) has provided a more detailed assessment of the variation between and within varieties of L. corniculatus and L. uliginosus (Fig. 2). Breeding for CT content has been reported as relatively easy (Miller and Ehlke, 1997) and the development of a high throughput CT assay at IGER (Marshall et al., 2008) should accelerate the selection procedures. Although CTs are absent from the leaves of white and red clover, they are present in the inflorescences of these species. Hence, as an alternative to the inclusion of species such as Lotus, efforts are also being made to breed white clover varieties with a greater density of flowers. Work in New Zealand (Burggraaf et al., 2003) has shown that selection for a greater density of flowers does not alter CT concentrations in flowers. However, because of the greater number of flowers, CT content/ha is increased. Subsequently, Burggraaf et al. (2005) showed that the high flowering density line appeared to reduce protein degradation in the rumen, resulting in lower rumen ammonia concentrations compared with cows grazing white clover cv. Huia. Given that increased flowering must be achieved without reducing plant persistence, the extent to which inflorescences can supply CT is likely to be

334

M. T. Abberton et al.

180

Tannin content (mg/g dry wt.)

160 140 120 Min. Max. Mean

100 80 60 40 20

Le -D o e G w ra ey n Sa No r c n G en ab rie lle Em ly U n ps t G art G ol d A1 ie 05 28 D aw n Te rre O be Em r In pir ia e D ra co H ig hg r St ove ea df as G .M t ak Su u nr is e Al

r

ng

ta

ki

Lo

Vi

ea rg

eo G

In

br

ed

lin

e

0

Variety

Figure 2 Tannin content (milligram per gram dry weight) of the leaf of spaced plants of 19 varieties of Lotus corniculatus and two varieties (Grasslands Maku and Sunrise) of L. uliginosus. Mean, maximum, and minimum tannin contents are derived from 20 plants per variety. Vertical bars represent the S.E.M.

limited. Increased flowering has also been attained by the development of interspecific hybrids between white clover and the annual profuse flowering ball clover (Trifolium nigrescens) and advanced backcrosses have been developed with significantly greater inflorescence production with no loss of persistence (Marshall et al., 2005b).

4.8. Greenhouse gas emissions from fertilizer production Very considerable scope remains for expanding the role of forage legumes in replacing mineral N fertilizer inputs to temperate grassland systems and thereby reducing the greenhouse gas emissions associated with fertilizer manufacture and application. A similar argument can be made in support of breeding programs aimed at increasing NUE and PUE in forage species. Wood and Cowie (2004) carried out a review of studies of greenhouse gas emissions from fertilizer production. N fertilizer manufacture brings with it significant greenhouse gas emissions from the Haber–Bosch process of synthesizing ammonia and from nitric acid production. Nitric acid is used in the manufacturing of ammonium nitrate, calcium nitrate, and potassium nitrate. The oxidation of ammonia to give nitric oxide also produces a tail gas of nitrous oxide, nitric oxide, and nitrogen dioxide. Nitric acid production is the largest industrial source of nitrous oxide, although clearly it has many uses besides fertilizer manufacture, with estimated emission rates of

Genetic Improvement of Forage Species

335

550–5890 CO2 equivalents/kg nitric acid. Urea accounts for almost 50% of world N fertilizer production and is synthesized from ammonia and carbon dioxide at high pressure to produce ammonium carbonate which is then dehydrated by heating to give urea and water. The synthesis of phosphate fertilizers also results in greenhouse gas emissions (reviewed by Wood and Cowie, 2004). Single superphosphate is produced from phosphate rock and sulfuric acid and triple superphosphate from phosphate rock and phosphoric acid. The majority is derived from phosphoric acid which itself is synthesized from phosphate rock and sulfuric acid. Wood and Cowie (2004) state that ‘‘more sulfuric acid is produced than any other chemical in the world and the largest single user is the fertilizer industry.’’ Considerable variation in ‘‘net emissions’’ is seen according to method used and efficiency of plant and in some cases the heat generated in production of sulfuric acid is captured.

5. Improving Soil Quality and Reducing Flood Damage 5.1. Flood tolerance and prevention Flooding and submergence-tolerance are major abiotic stresses that rank alongside drought, salinity, and extreme temperatures as determinants of species distribution and success. Common requirements for flood-tolerance are the development of physiological or anatomical responses such as developed capability to withstand anoxia or development of aerenchyma (Visser et al., 2003). The ability of a crop to withstand flooding will depend on its extent, its frequency, and its timing. Even crops such as rice, considered tolerant to flooded conditions, will perish if seedlings are overexposed to water too deep to prevent successful completion of elongation escape mechanisms. Genes on rice chromosome 9 have been associated with submergence-tolerance (Toojinda et al., 2003) and have been successfully selected during a marker-assisted selection program as an aid to develop a submergence-tolerant variety of Thai jasmine rice (Siangliw et al., 2003). Climate change scenarios for many parts of Europe predict, in addition to water deficits during the summer, the likely occurrence of heavy rainfall leading to flooding during the autumn and winter. This will be exacerbated by heavily compacted soils likely to arise subsequent to soil hardening following summer droughts. The heterogeneity and diverse growth patterns of perennial grassland species are likely to impact on both the soils capabilities for retaining or discharging water and nutrients. Opportunities arise given understanding of the genetic determinants for root growth to design new grass crops to better manage water use given either scenario.

336

M. T. Abberton et al.

As pointed out by Kemp and Michalk (2007), the perennial species that are typically sown in temperate grasslands both efficiently utilize water due to deep roots and long growing season and protect the soil surface from erosion and nutrient loss particularly during storms. A decline in the perennial component (ingress of annual weeds) not only reduces the productivity and quality of the pasture but also contributes to soil erosion and potentially acidification and increased salinity in some environments. This reinforces the need for germplasm improvement to increase the persistence of forage species. As proof of principle, the initiation of submergence-tolerance was achieved successfully in Iris following the transfer by introgression of the relevant alleles from tolerant Iris fulva into dry-adapted I. brevicaulis (Martin et al., 2006) thus demonstrating the potential of introgression procedures as aids to the development of flood tolerance. The success of such an approach will be dependent on species hybridization capabilities, fertility, and the frequency of interspecific chromosome recombination. Similar opportunities are likely to reside within the Lolium–Festuca complex to locate and incorporate novel genetic variants for flooding-tolerance. Among these, one potential source of ‘‘new’’ alleles for use in grass crop development may be the little studied natural hybrid Festulolium loliaceum (derived from hybridization events between L. perenne and F. pratensis) which may be found in flood- and water meadows, where it would appear to be adapted better to the more extreme sites of water-logged soils than either its parental species (Humphreys, unpublished results). In some early investigations into F. loliaceum, Peto (1933), Essad (1966), and Gymer and Whittington (1973) found that they were of three types; diploid (2n ¼ 2x ¼ 14), or triploids (2n ¼ 3x ¼ 21) which were either with Festuca- or Lolium-like morphology. Humphreys and Harper (unpublished results) confirmed recently these results and in detailed cytological investigations using genomic in situ hybridization (GISH) determined that the triploid hybrids contained either two genomes of L. perenne and one of F. pratensis or of two genomes of F. pratensis and one of Lolium. All three hybrid combinations have some pollen fertility, albeit at low frequency (Gymer and Whittington, 1975), providing opportunities for gene introgression into the cultivated parental species. In addition to searching for novel genes as safeguards against environmental perturbations such as flooding, the recent trend for multidisciplinary research has brought new opportunities for crop design. For example, it has been found that plant roots have considerably more impacts on soil water status than that reflected by their differing capabilities for water extraction. Hydrologists have described a pivotal role for vegetation in regulating soil water content and one new component demonstrated recently was that root activity has the capability to initiate biophysical changes in soil hydraulic properties (Whalley et al., 2005). This is a function of the porosity of the soil and the rooting depth of the vegetation. Rooting depth determines

Genetic Improvement of Forage Species

337

the soil volume which plants are able to draw water from and, influenced by various key soil hydraulic properties; it defines the plant available water capacity. The response time of stream flow and volume of discharge from a precipitation event are determined by climate (mainly rainfall), antecedent conditions, physical catchment properties, vegetation characteristics, and management. It is known that forested catchments often produce lower water yields and hence provide a lower risk of flooding compared to grassland dominated catchments (Ruprecht and Schofield, 1991).

5.2. Soil porosity and compaction The impact of grazing animals on soil quality and surface waters has been recently reviewed (Bilotta et al., 2007). Extent of ground cover and tolerance of treading are important factors influencing the extent of direct hoof/ soil contact and of soil compaction, pugging and poaching. The impacts of soil compaction on grass production have received relatively little attention, but substantial variation exists in species’ abilities to grow on compacted soils, which will increase following the onset of substantial soil water deficit, and exacerbate flooding during periods of water surplus. Although the information dates back 30 years, when summer droughts were less prevalent, the data of Luten and Roozeboom (1976) still provide the best quantitative data for variation in tolerance to soil compaction between grass species. Although water stress is the most intensively researched physical stress to root growth, field data show that it may not always be the most critical. Various physical stresses may act in combination to limit root elongation. Hypoxia, water stress, and mechanical impedance to root growth will change with the water content of the soil and their relative importance will depend upon the degree of soil compaction. Plant species differ in their visible effects on soil structure (Drury et al., 1991) and recent studies have demonstrated a change to the soil water release characteristic which tends to be associated with an increased number of larger pores in the rhizosphere, or an increase in water repellence associated with an increased growth of fungi (Macleod et al., 2007). Root activity tends to increase the number of large pores and this is influenced by plant species (Materechera et al., 1994). During a multidisciplinary study entitled SuperGraSS (Macleod et al., 2007) involving soil hydrologists, soil physicists, and grass geneticists and physiologists, the different rooting traits of fescue species and ryegrasses and their impact on soil structure and hydrology are being assessed. Preliminary data from this project supports earlier findings (Durand et al., 2007; Garwood and Sinclair, 1979) that F. arundinacea has deeper and larger roots than Lolium capable of extracting water from greater soil depths. New findings are that the roots of F. arundinacea are stronger than Lolium and more able to penetrate hard compacted soils and also F. arundinacea utilizes water more efficiently both

338

M. T. Abberton et al.

under irrigated and especially under drought stress conditions. Preliminary investigations at Lancaster University, which require further investigations to confirm, would indicate greater soil porosity and by implication better soil water retention in plots of F. arundinacea rather than L. multiflorum (Binley, personal communication). It is anticipated that improving the rooting characteristics of Lolium through incorporation of fescue-derived traits will not only bring benefits in terms of increased water-use-efficiency, but also in terms of better waterholding capabilities and reduced soil erosion during periods of potential flooding, or increased overland flow following heavy rainfall. Stabilization of grassland ecosystems by reducing soil erosion and overland flow would also assist in reducing losses of C (i.e., dissolved organic forms) and thereby assist C sequestration (Kalbitz et al., 2000). The positive effects of forage legumes on soil structure and quality have been investigated in white clover (Holtham et al., 2007; Mytton et al., 1993) and red clover (Papadopoulos et al., 2006). The changes in soil structure brought about by white clover resulted in improvements in water percolation rate (i.e., the soil became more freely drained) and in nutrient uptake by plants (Witty and Mytton, 2001). Holtham et al. (2007) also found evidence of greater local structuring of soil around white clover roots and faster drainage of water through soil cores under white clover compared with perennial ryegrass monocultures. Similar benefits in terms of soil structure were noted for soil cores under red clover monocultures by Papadopoulos et al. (2006), although the effects were transient, and were reversed when a cereal crop was sown the following year. Improved soil structure reduces the risk of soil compaction and water run-off, increases the soil’s biological activity, and facilitates seedling establishment and root penetration. However, it appears likely that legume-driven improvements in soil structure and drainage also directly result in increased leaching of both fixed and applied nitrate in legume monocultures (Holtham et al., 2007).

5.3. Biodiverse mixtures and ecosystem services Recent work by Kirwan et al. (2007) has indicated that the use of multispecies mixtures of legumes and grasses may offer a number of potential advantages over the conventional perennial ryegrass monocultures and binary white clover/perennial ryegrass mixtures sown in UK pastures. These include greater productivity, more efficient use of environmental resources, a positive impact on soil structure, and improved forage quality. Since agriculturally managed grassland constitutes a large proportion of the landscape in many countries, the level of biodiversity maintained in grassland farming will have a significant impact on biodiversity in the countryside in general. However, the value of increased species diversity in terms of

Genetic Improvement of Forage Species

339

more stable swards, improved resource capture, and increased yields with lower inputs has only recently started to be evaluated in an agronomic as opposed to a predominantly ecological context. The same applies with respect to the consequences of any wide-spread ‘‘return’’ to the use of species-rich seed mixtures for germplasm improvement and variety development. The relationship between biodiversity and productivity is currently a central issue in plant ecology and has been the subject of recent investigation using experimental plant communities with different levels of species richness (i.e., numbers) in the multisite BIODEPTH project (Hector et al., 1999). Most research in this area has treated simplified, small scale, systems, focusing on the effects of species and functional group identity, and species richness on community function (Hooper et al., 2005). However, it has become clear that diversity effects in plant communities are also produced by interspecific interactions, the strength of which depends on the relative abundance (evenness) of the species involved (Kirwan et al., 2007). Many studies have described the changes in relative abundance of species that occur as a response to competition in plant communities (e.g., Mulder et al., 2004), but few have used robust experimental designs to vary evenness systematically in multispecies mixtures, and until recently analytical procedures have not incorporated the joint effects of evenness and richness. The novel ‘‘simplex’’ experimental design used by Kirwan et al. (2007) directly addressed these issues in an agronomic context.

6. Enhancing Persistency and Resilience 6.1. The role of interspecific hybridization Increased grassland persistency is considered an important research target to reduce costs of seed, and inputs required for ploughing and resowing and to introduce important environmental safeguards by stabilizing soils, preventing erosion, maintaining soil nutrient status, increasing flora, fauna, and soil microbe diversity, improving soil structure and hydrology, increasing C sequestration, maintaining landscape and multifunctional purpose, and for introducing predictable and consistent high quality forage for livestock consumption. Many genotype environmental factors affect crop persistency including resource-use-efficiency, competitiveness and vigor, flowering time and intensity, and the genome plasticity necessary for adaptations to various climatic and edaphic conditions that may be encountered over years and in various locations. Improved persistency and adaptation may be achieved through increased heterosis following interspecific hybridization, and examples are found both naturally in grass crop evolution, and as new synthetic hybrids developed through the efforts of plant breeders to speedup

340

M. T. Abberton et al.

crop evolution by artificially combining, by various methodologies, the attributes of contrasting plant species (Humphreys et al., 2003). Hybrid formation and consequent persistence and fecundity is an important contributor to speciation and evolution. It frequently enables hybrids to colonize environments previously beyond the range of adaptation of either parent species. A good example was the hexaploid grass species F. arundinacea which arose following the hybridization of winter hardy F. pratensis (2x) with the drought resistant Mediterranean-based species, F. glaucescens (4x) (Humphreys et al., 1995). As a hybrid F. arundinacea, was capable of colonizing climatically diverse European grasslands. For F. arundinacea and many hybrid species, polyploidization subsequent to hybridization was necessary to enhance fertility by encouraging chromosome pairing between homologous sets and thereby facilitating disomic inheritance, normal disjunction, and hybrid stability and fertility (Lewis et al., 1980). Attempts to reproduce synthetic allopolyploids that combine the drought resistance (or winter hardiness) of Festuca species and the forage quality of Lolium have led to the commercialization of several Festulolium allotetraploid cultivars but these have lacked genome stability and over generations have become unbalanced to favor the Lolium genome (Canter et al., 1999; Zwierzykowski et al., 2006). These hybrids have involved combinations of L. perenne or L. multiflorum with F. pratensis, species that are genetically very homogeneous. The genome sizes of L. perenne (1C ¼ 2034 Mb) and F. pratensis (1C ¼ 2181 Mb) are also very similar (RBG Kew Plant DNA C-values database: http://www.rbgkew.org.uk/ cval/homepage.html). New allopolyploid hybrid combinations that involve more distantly related species such as F. glaucescens (Ghesquie`re pers. comm), especially as these are believed to carry genes for preferential homologous chromosome pairing ( Jauhar, 1993; Lewis et al., 1980), may provide better opportunities for stabilizing both the Lolium and the Festuca genomes.

6.2. Exploiting genetic variation for multiple traits Despite increasing understanding of the regulatory control of complex acclimatory gene actions, and the capability to incorporate these in high quality cultivars, the wisdom and procedures employed over decades by plant breeders is still found regularly to offer the best solution for design of grass crops with increased persistency. In many cases, the selection of a beneficial trait for one key research target is likely to benefit several others. For example, the selection and design of large plants with good growth rate and deep and large root systems should provide for excellent forage production, for increased crop persistency and for drought resistance, and in addition provide improved soil structure, hydrology, and stability. The old plant breeders’ maxim of ‘‘crossing the best with the best and hoping for the best’’ remains an excellent strategy for crop improvement. However,

Genetic Improvement of Forage Species

341

an improved understanding of the genetic control of the traits that offer optimal plant size and performance improves the predictability, reliability, and effectiveness of new crop improvement programs. A L. perenne F2 mapping family created at IGER from parents with high and low WSC has proved to be a very valuable source of genetic variation for a range of traits including WSC, NUE, drought resistance, and for root growth (Turner et al., 2008). The QTL for rooting growth are likely to impact on plant drought resistance and can be compared with QTL for drought resistance to be found in the Festuca mapping family. In the Lolium mapping population, three QTL were identified for root density in the top horizon (surface rooting). Two of these overlay, or were located close to, plant size QTL (Turner et al., 2008), in other words ryegrass plants with large shoots unsurprisingly are also likely to carry large roots. Only one tentative QTL for deep rooting was identified, and this was located on chromosome 2 which shares conserved genome regions with rice chromosome 4, a chromosome that contains a range of QTL for root development (Ikeda et al., 2007).

7. Enhancing C Sequestration in Grasslands Increasing C sequestration in the terrestrial biosphere (i.e., the removal of CO2 from the atmosphere and its storage in C pools of varying lifetimes ( Jones and Donnelly, 2004)), is regarded as a win–win strategy with respect to climate change mitigation (Lal et al., 1998) both in terms of acting as a bridge to the future, when ‘‘low’’ or ‘‘zero C’’ fuel sources take effect, and in comparison with the high economic and environmental costs associated with engineering techniques of CO2 capture and injection into geological and oceanic strata. Furthermore, Article 3.4 of the Kyoto Protocol of the UN Framework Convention on Climate Change, allows countries to include C sequestration in the land as a contribution to reducing greenhouse gas emissions (IPCC, 2001). A significant additional impetus for increasing C sequestration in UK soils is provided by the recent finding (Bellamy et al., 2005) that, irrespective of land use, UK soils have lost C over the last 25 years at a mean rate of 0.6% per annum, extrapolating to an annual C loss of 13 million tonnes, equivalent to 8% of annual CO2 emissions. Steady-state simulations of grassland ecosystems (Arah et al., 1997) also predict a halving of system C contents for a þ5 C rise in temperature. A number of management strategies for conserving soil C stocks in the UK have been suggested (e.g., Environment Agency, 2007). Their implementation, together with the development of additional methods to slow and preferably reverse the decline in C content of grassland soils, is essential, given that circa 70% of the UK’s agricultural area is under grassland.

342

M. T. Abberton et al.

The substantial stocks of C sequestered in temperate grassland ecosystems are located largely underground in the roots and soil. The roots, senescent leaves, and stems differ in their rate and process of breakdown in ˚ gren, 2001). However, in a survey of temperate grassthe soil ( Joffre and A land, Jobba´gy and Jackson (2000) found that only 64% of soil-organic-C existed in the top 40 cm of soil which contained 87% of all roots, the remainder of the C is found at greater soil depths probably due to a decreased C turnover at depth in the soil ( Jones and Donnelly, 2004). According to these authors, the principal factors that determine the amount of C sequestration in the soil are (i) the rate of input of organic matter, (ii) the rate of its decomposition, (iii) soil depth, and (iv) the physical protection of aggregates and organo-mineral complexes. Soil C levels under grassland are generally much higher than those under arable crops, and not withstanding the C losses reported for UK soils alluded to above, most temperate grasslands worldwide are considered to be C sinks, with modeling and experimental studies of C balances and fluxes producing widely varying estimates of C sequestration, although typically between 0.2 and 0.6 Mg C/ha/year. (e.g., Jones and Donnelly, 2004). The higher C stocks under grassland compared with arable are explained by a number of factors including the greater flux of C into the soil under grassland, and an increased residence time of C resulting from absence of tilling. It follows that change from arable cropping to perennial grasslands usually leads to substantial increases in soil C. Specific measures for increasing C sequestration by grassland (e.g., Jones and Donnelly, 2004; Sousanna et al., 2004), include modifications to fertilizer, irrigation, and grazing practices. Given that the amount of C sequestered is the difference between net primary production (NPP) and total losses of C from the system, these are classifiable measures designed to (a) increase input rates of organic matter, (b) change the decomposability of C inputs, (c) place organic matter deeper in the soil, and (d) enhance the physical protection of soil C through intra-aggregate or organo-mineral complexes. Not surprisingly, therefore, pasture improvement and the introduction of high yielding forage grasses and legumes has been shown both theoretically (i.e., Sousanna et al., 2004) and practically (e.g., Lal et al., 1998) to substantially increase C stocks and the potential for C sequestration. For example, pasture improvement by sowing with productive forage grasses such as F. arundinacea has been shown to increase the soil C pool by 17.2% (equivalent to C sequestration of circa 3 mg C/ha/year over a 6 year period (Lal et al., 1998). We are unaware of any attempts, as yet, to breed explicitly for increased C sequestration in any of the economically important temperate forage species. This is in large part due to the fact that C sequestration has not until recently been recognized as an environmentally and economically important cofunction of grassland. However, it also reflects unanswered questions regarding (a) the magnitude, (b) genetic control of, and (c) G E

343

Genetic Improvement of Forage Species

modulation of, the various C flux pathways between plant, soil and atmosphere, and consequently, uncertainties about which traits are most likely to confer durable increases in C sequestration. The feasibility of (a) evaluating traits likely to be directly or indirectly linked with C sequestration and (b) of identifying the associated QTLs, thereby enabling a marker-assisted-breeding approach is currently under investigation at IGER. For example, promising variation in litter deposition (independent of herbage yield) was demonstrated within IGER’s amenityforage perennial ryegrass mapping family (Fig. 3A). Likewise, genetic variation in rhizosphere decomposition of organic matter residues was also revealed (Fig. 3B). Provisional QTLs for C return in litter have been associated with loci on chromosomes 1 and 5. While variety development for performance under the elevated temperatures and CO2 levels anticipated in 50 or 100 years time is neither commercially realistic, nor regarded as a priority at this time, elevated CO2 has been shown to increase photosynthesis of L. perenne Such an effect is likely to increase soil C inputs and microbial biomass through increased root exudates and turnover. However, the benefits of increased productivity under elevated CO2 may only be short-term, their sustainability depending on the availability of N in the soil, prospective increases in plant NUE and maintenance of an appropriate balance of soil C and N. Furthermore, recent results (De Boeck et al., 2007) suggest that the effects of global warming on canopy photosynthesis may be less straightforward. Based on a 2-year study of CO2 flux measurements on grassland communities of varying species richness, exposed to either ambient air temperatures, or warmed by 3 C A

B Barley straw residue C:N (g/g)

300

Litter (gC/m2)

250 200 150 100 50 0 0

20

40

Cut herbage (gC/m2)

60

25

20

15

10 50

100 150 200 Ryegrass residue C:N (g/g)

250

Figure 3 (A) Relationship between the total C content of litter in sand-box mini swards of 94 amenity forage ryegrass mapping family genotypes and the C content of the herbage cut taken at the same time. Values are means of two sand boxes per genotype. (B) Relationship between the dry weight C:N ratios of buried perennial ryegrass leaf and barley straw residues under the same 94 genotypes. Values are means of five litter bags per mini sward following decomposition over 21 days (ryegrass) and 36 days (barley straw) under summer conditions. (Source: Defra Project LS3648.)

344

M. T. Abberton et al.

under soils of increasing soil–water deficit, these authors concluded that, while respiratory CO2 outputs were largely unchanged under warming, photosynthetic CO2 inputs were actually lowered, especially in summer when heat and drought stress were at their maximum. It is, however, possible that negative impacts of increased temperature on sward productivity could be mitigated by the selection of appropriate genotypes or by using multispecies communities providing complementary growth patterns and productivity throughout the season.

8. Future Prospects Over the last decade, considerable progress has been made in the genomics of model species that is those that are, due to small genome size, rapid generation times and relatively simple genome organization, particularly tractable to molecular genetics approaches including genetic and genomic analysis and sequencing. The original model was A. thaliana a small dicotyledonous weed, which has now been fully sequenced and which has contributed greatly to our understanding of the genetic control of fundamental processes in plants. More recently, model species have been developed in the major families relevant to grassland agriculture. In the grasses, rice has become a model species as well as a major crop and in the legume family two species have been developed as models, Lotus japonicus and Medicago truncatula. The latter is in the same genus as alfalfa (M. sativa) and closely related to the clovers, white clover (T. repens L.) and red clover (T. pratense). The potential for translating and exploiting the understanding, tools and resources developed in the models to crop species is currently being explored but is likely to be considerable. At the same time significant progress is being made in developing molecular approaches in some of the crop species themselves (Pollock et al., 2005). This is particularly the case for the temperate grasses in the Lolium (ryegrasses) and Festuca (fescues) genera, alfalfa and the clovers. Some of the key resources including genetic maps, bacterial artificial chromosome libraries and databases of expressed sequence tags are in place for many of these species. Conservation of gene order or synteny is a powerful tool in transferring information from models to crops and this has been shown to be extensive both within grasses and legumes (e.g., Armstead et al., 2004). The ability to use genomic insights and to apply molecular approaches to assist plant breeding programs increases the likelihood of successfully developing new varieties of important forages so as to reduce greenhouse gas emissions and enhance C sequestration. Approaches based on plant genetic improvement have the potential to underpin options for reduction together with other approaches, for example, management and animal selection. They can also bring increased

Genetic Improvement of Forage Species

345

understanding: for instance the use of genetic variation to ‘‘perturb’’ systems and bring greater understanding of processes for example, in the rumen. Breeding approaches also have the potential to address multifunctionality and trade-offs for example, maintaining productivity and quality while reducing inputs. It is important that the improvements brought about to individual forage species are seen within the context of the whole system for example, at farm and catchment level and in terms of the balance between different outcomes for example, production, reduced pollution to water and lower emissions to air. To this end the use of modeling approaches is likely to be extremely valuable. Modeling studies have considered the impact of dietary strategies and increasingly consider the full range of economic and environmental attributes important for sustainability for example, the SIMSDAIRY model developed by del Prado and colleagues (del Prado et al., 2006). The key needs for the future are translation from model species, integration of molecular approaches into breeding programs and collaborations between soil scientists, animal scientists, modellers and plant breeders to seek integrated solutions to the challenge of maintaining economically viable livestock production with a reduced environmental footprint.

ACKNOWLEDGMENTS The authors are grateful for the financial support of the UK Biotechnology and Biological Sciences Research Council (BBSRC), the Department for Environment Food and Rural Affairs (Defra), and Germinal Holdings Ltd.

REFERENCES Abberton, M. T. (2007). Interspecific hybridization in the genus Trifolium. Plant Breed. 126, 337–342. Aerts, R. J., Barry, T. N., and McNabb, W. C. (1999a). Polyphenols and agriculture: Beneficial effects of proanthocyanidins in forages. Agric. Ecosyst. Environ. 75, 1–12. Aerts, R. J., McNabb, W. C., Molan, A., Barry, T. N., and Peters, J. S. (1999b). Condensed tannins from Lotus corniculatus and L. pedunculatus exert different effects on the in vitro rumen degradation of ribulose-1,5-biphosphate carboxylase/oxygenase (Rubisco) protein. J. Sci. Food Agric. 79, 79–85. Alfaro, M. A., Jarvis, S. C., and Gregory, P. J. (2003). Potassium budgets in grassland systems as affected by nitrogen and drainage. Soil Use Water Manage. 19, 89–95. Arah, J. R. M., Thornley, J. H. M., Poulton, P. R., and Richter, D. D. (1997). Simulating trends in soil organic carbon in long-term experiments using the ITE (Edinburgh) Forest and Hurley Pasture ecosystem models. Geoderma 81, 61–74. Arai, Y., and Sparks, D. L. (2007). Phosphate reaction dynamics in soils and soil components: A multiscale approach. Adv. Agron. 94, 135–179.

346

M. T. Abberton et al.

Armstead, I. P., Turner, L. G., Farrell, M., Sko¨t, L., Gomez, P., Montoya, T., Donnison, I. S., King, I. P., and Humphreys, M. O. (2004). Synteny between a major heading-date QTL in perennial ryegrass (Lolium perenne L) and the Hd-3 heading-date locus in rice. Theor. Appl. Genet. 108, 822–828. Atkin, O. K., Botman, B., and Lambers, H. (1996). The relationship between the relative growth rate and nitrogen economy of alpine and lowland Poa species. Plant Cell Environ. 19, 1324–1330. Baligar, V. C., Fageria, N. K., and He, Z. L. (2001). Nutrient use efficiency in plants. Commun. Soil Sci. Plant Anal. 32, 921–950. Barry, T. N., and Manley, T. R. (1986). Interrelationships between the concentration of total condensed tannin, free condensed tannin and lignin in Lotus spp. and their possible consequences in ruminant nutrition. J. Sci. Food Agric. 37, 248–254. Barry, T. N., Manley, T. R., and Duncan, S. J. (1986). The role of condensed tannins in the nutritional value of Lotus pedunculatus for sheep: IV. Sites of carbohydrate and protein digestion as influenced by dietary reactive tannin concentration. Br. J. Nutr. 55, 123–137. Beebe, S. E., Rojas-Pierce, M., Yan, X., Blair, M. W., Pedraza, F., Mun˜oz, F., Tohme, J., and Lynch, J. P. (2006). Quantitative trait loci for root architecture traits correlated with phosphorus acquisition in common bean. Crop Sci. 46, 413–423. Beha, E. M., Theodorou, M. K., Thomas, B. J., and Kingston-Smith, A. H. (2002). Grass cells ingested by ruminants undergo autolysis which differs from senescence: Implications for grass breeding and livestock production. Plant Cell Environ. 25, 1299–1312. Bellamy, P. H., Loveland, P. J., Bradley, R. I., Lark, R. M., and Kirk, G. J. D. (2005). Carbon losses from all soils across England and Wales 1978–2003. Nature 437, 245–248. Berendse, F., and Aerts, R. (1987). Nitrogen-use-efficiency: A biologically meaningful definition? Funct. Ecol. 1, 293–296. Berry, P. M., Stockdale, E. A., Sylvester-Bradley, R., Philipps, L., Smith, K. A., Lord, E. I., Watson, C. A., and Fortune, S. (2003). N, P and K budgets for crop rotations on nine organic farms in the UK. Soil Use Manage. 19, 112–118. Bertin, P., and Gallais, A. (2000). Physiological and genetic basis of nitrogen use efficiency in maize: II. QTL detection and coincidences. Maydica 46, 53–68. Bilotta, G. S., Brazier, R. E., and Haygarth, P. M. (2007). The impacts of grazing animals on the quality of soils, vegetation, and surface waters in intensively managed grasslands. Adv. Agron. 94, 237–280. Bonos, S. A., Rush, D., Highnight, K., and Meyer, W. A. (2004). Selection for deep root production in tall fescue and perennial ryegrass. Crop Sci. 44, 1770–1776. Brentrup, F., Ku¨sters, J., Lammel, J., Barraclough, P., and Kuhlmann, H. (2004). Environmental impact assessment of agricultural production systems using the life cycle assessment (LCA) methodology II. The application to N fertilizer use in winter wheat production systems. Eur. J. Agron. 20, 256–279. Broadley, M. R., Bowen, H. C., Cotterill, H. L., Hammond, J. P., Meacham, M. C., Mead, A., and White, P. J. (2004). Phylogenetic variation in the shoot mineral composition of angiosperms. J. Exp. Bot. 55, 321–336. Burggraaf, V. T., Kemp, P. D., Thom, E. R., Waghorn, G. G., Woodfield, D. R., and Woodward, S. L. (2003). Agronomic evaluation of white clover selected for increased floral condensed tannin. Proc. N. Z. Grassland Assoc. 65, 139–145. Burggraaf, V. T., Waghorn, G. C., Woodward, S. L., Woodfield, D. R., Thom, E. R., and Kemp, P. D. (2005). In ‘‘High floral tannin white clover reduces rumen ammonia concentration in dairy cows’’ (F. P. O’Mara, R. J. Wilkins, L. ‘t Mannetje, D. K. Lovett, P. A. M. Rogers, and T. M. Boland, Eds.), p. 244. 26 June–2 July, Dublin, Ireland. Canter, P. H., Pasˇakinskiene, I., Jones, R. N., and Humphreys, M. W. (1999). Chromosome substitutions and recombination in the amphiploid Lolium perenne x F. pratensis cv. Prior (2n ¼ 4x ¼ 28). Theor. Appl. Genet. 98, 809–814.

Genetic Improvement of Forage Species

347

Caradus, J. R. (1994). Selection for improved adaptation of white clover to low phosphorus and acid soils. Euphytica 77, 243–250. Caradus, J. R., and Dunn, A. (2000). Adaptation to low fertility hill country in New Zealand of white clover lines selected for differences in response to phosphorus. N. Z. J. Agric. Res. 43, 63–69. Caradus, J. R., Mackay, A. D., Wewala, S., Dunlop, J., Hart, A., Vandenbosch, J., Lambert, M. G., and Hay, M. J. M. (1992). Inheritance of phosphorus response in white clover (Trifolium repens L.). Plant Soil 146, 199–208. Caradus, J. R., Hay, M. J. M., Mackay, A. D., Thomas, V. J., Dunlop, J., Lambert, M. G., Hart, A. L., Vandenbosch, J., and Wewala, S. (1993). Variation within white clover (Trifolium-repens L.) for phenotypic plasticity of morphological and yield related characters, induced by phosphorus supply. New Phytol. 123, 175–184. Caradus, J. R., Kennedy, L. D., and Dunn, A. (1998). Genetic variation for the ratio of inorganic to total phosphorus in white clover leaves. J. Plant Nutr. 21, 2265–2272. Carter, E., Theodorou, M. K., and Morris, P. (1999). Response of Lotus corniculatus to environmental change. 2. Effect of elevated CO2 and temperature and drought on tissue digestion in relation to condensed tannin and carbohydrate accumulation. J. Sci. Food Agric. 79, 1431–1440. Clark, W. D., Wohlt, J. E., Gilbreath, R. L., and Zajac, P. K. (1986). Phytate phosphorus intake and disappearance in the gastrointestinal-tract of high producing dairy-cows. J. Dairy Sci. 69, 3151–3155. Clarkson, D. T., Hopper, M. J., and Jones, L. H. P. (1986). The effect of root temperature on the uptake of nitrogen and the relative size of the root system in Lolium perenne. 1. Solutions containing both NHþ 4 and NO3 . Plant Cell Environ. 9, 535–545. Coupland, R. T. (1979). The nature of grassland. In ‘‘Grassland Ecosystems of the World’’ (R. T. Coupland, Ed.), pp. 23–31. Cambridge University Press, Cambridge, UK. Crush, J. R. (1995). Phosphorus response of white clover (Trifolium repens L.) genotypes selected for tolerance of aluminium toxicity. N. Z. J. Agric. Res. 38, 451–456. Crush, J. R., Waller, J. E., and Care, D. A. (2005). Root distribution and nitrate interception in eleven temperate forage species. Grass Forage Sci. 60, 385–392. Davies, D. R., Merry, R. J., Williams, A. P., Bakewell, E. L., Leemans, D. K., and Tweed, J. K. S. (1998). Proteolysis during ensilage of forages varying in soluble sugar content. J. Dairy Sci. 81, 444–453. De Boeck, H. J., Lemmens, C. M. H. M., Vicca, S., Van den Berge, J., Van Dongen, S., Janssens, I. A., Cuelmans, R., and Nijs, I. (2007). How do climate warming and species richness affect CO2 fluxes in experimental grasslands? New Phytol. 175, 512–522. Defra, (2005). Defra website, http://statistics.defra.gov.uk/esg/. del Prado, A., Brown, L., Schulte, R., Ryan, M., and Scholefield, D. (2006). Principles of development of a mass balance N cycle model for temperate grasslands: An Irish case study. Nutr. Cycl. Agroecosyst. 74, 115–131. Dewhurst, R. J., Mitton, A. M., Offer, N. W., and Thomas, C. (1996). Effects of the composition of grass silages on milk production and nitrogen utilization by dairy cows. Anim. Sci. 62, 25–34. Di, I. D., and Cameron, K. C. (2004). Effects of temperature and application rate of the nitrification inhibitor, dicyandiamide (DCD), on nitrification rate and microbial biomass in a grazed pasture soil. Aust. J. Soil Res. 42, 927–932. Drury, C. F., Stone, J. A., and Findlay, W. I. (1991). Microbial biomass and soil structure associated with corn, grasses and legumes. Soil Sci. Soc. Am. J. 55, 805–811. Dunbabin, V., Diggle, A., and Rengel, Z. (2003). Is there an optimal root architecture for nitrate capture in leaching environments? Plant Cell Environ. 26, 835–844.

348

M. T. Abberton et al.

Durand, J. L., Bariac, T., Ghesquie`re, M., Biron, P., Richard, P., Humphreys, M., and Zwierzykowski, Z. (2007). Ranking of the depth of water extraction by individual grass plants, using natural 18O isotope abundance. Environ. Exp. Bot. 60, 137–144. Easton, H. S., Mackay, A. D., and Lee, J. (1997). Genetic variation for macro- and micronutrient concentration in perennial ryegrass (Lolium perenne L.). Aust. J. Agric. Res. 48, 657–666. Edwards, A. C., and Withers, P. J. A. (1998). Soil phosphorus management and water quality: A UK perspective. Phosphorus Agric. Water Qual. 14, 124–130. Environment Agency. (2007). ‘‘Soil: A Precious Resource.’’ Environment Agency, Bristol, UK. Essad, S. (1966). Recherches sur l’origine des hybrids naturels Festuca loliacea 2 et 3 par la morphologie et la cytoge´ne´tique comparatives. Ann. Ame´l Plant 16, 5–41. Frame, J., and Laidlaw, A. S. (2005). Prospects for temperate forage legumes. In ‘‘Grasslands: Developments, Opportunities, Perspectives’’ (S. G. Reynolds and J. Frame, Eds.), pp. 1–28. FAO 2005, Science Publishers, Inc., Plymouth, UK. Franco-Zorilla, J. M., Gonza´lez, E., Bustos, R., Linhare, F., Leyva, A., and Paz-Ares, J. (2004). The transcriptional control of plant responses to phosphate limitation. J. Exp. Bot. 55, 285–293. Frink, C., Waggoner, P. E., and Ausubel, J. H. (1999). Nitrogen fertilizer: Retrospect and prospect. Proc. Natl. Acad. Sci. 96, 1175–1180. Gallais, A., and Hirel, B. (2004). An approach to the genetics of nitrogen use efficiency in maize. J. Exp. Bot. 55, 295–306. Garnier, E., and Aronson, J. (1998). Nitrogen-use efficiency from leaf to stand level. In ‘‘Inherent Variation in Plant Growth. Physiological Mechanisms and Ecological Consequences’’ (H. Lambers, H. Poorter, and M. M. I. Van Vuuren, Eds.), pp. 515–538. Backhuys Publishers, Leiden, The Netherlands. Garnier, E., Gobin, O., and Poorter, H. (1995). Nitrogen productivity depends on photosynthetic nitrogen use efficiency and on nitrogen allocation within the plant. Ann. Bot. 76, 667–672. Garwood, E. A., and Sinclair, J. (1979). Use of water by 6 grass species.2. Root distribution and use of soil-water. J. Agric. Sci. 93, 25–35. Gebrehiwot, L., Beuselinck, P. R., and Roberts, C. A. (2002). Seasonal variation in condensed tannin concentration of three Lotus species. Agron. J. 94, 1059–1065. Gilroy, S., and Jones, D. (2000). Through form to function: Root hair development and nutrient uptake. Trends Plant Sci. 5, 56–60. Good, A. G., Shrawat, A. K., and Muench, D. G. (2004). Can less yield more? Is reducing nutrient input into the environment compatible with maintaining crop production? Trends Plant Sci. 9, 597–605. Gorny, A. G. (1999). Inheritance of nitrogen and phosphorus utilization efficiency in spring barley at the vegetative growth stages under high and low nutrition. Plant Breed. 118, 511–516. Gorny, A. G., and Sodkiewicz, T. (2001). Genetic analysis of the nitrogen and phosphorus utilization efficiencies in mature spring barley plants. Plant Breed. 120, 129–132. Gourley, C. J. P., Allan, D. L., and Russelle, M. P. (1994). Plant nutrient efficiency: A comparison of definitions and suggested improvement. Plant Soil 158, 29–37. Granstedt, A. (2000). Increasing the efficiency of plant nutrient recycling within the agricultural system as a way of reducing the load to the environment—Experience from Sweden and Finland. Agric. Ecosyst. Environ. 80, 169–185. Greenwood, D. J., Barnes, A., Liu, K., Hunt, J., Cleaver, T. J., and Loquens, S. M. H. (1980). Relationships between the critical concentrations of nitrogen, phosphorus and potassium in 17 different vegetable crops during growth. J. Sci. Food Agric. 31, 1343–1353.

Genetic Improvement of Forage Species

349

Greenwood, D. J., Stellaccci, A. M., Meacham, M. C., Broadley, M. R., and White, P. J. (2005). Components of P response of different Brassica oleracea genotypes are reproducible in different environments. Crop Sci. 45, 1728–1735. Gymer, P. T., and Whittington, W. J. (1973). Hybrids between Lolium perenne and Festuca pratensis. II. Comparative morphology. New Phytol. 72, 861–865. Gymer, P. T., and Whittington, W. J. (1975). Hybrids between Lolium perenne and Festuca pratensis. IV. Cytological abnormalities. New Phytol. 75, 259–267. Hacisalihoglu, G., and Kochian, L. V. (2003). How do some plants tolerate low levels of soil zinc? Mechanisms of zinc efficiency in crop plants. New Phytol. 159, 341–350. Hall, O. G., Gaddy, D., and Hobbs, C. S. (1961). Influence of phosphorus supplements on cellulose digestion by rumen microorganisms and on ration digestibility by sheep. J. Anim. Sci. 20, 395. Hammond, J. P., Bennett, M. J., Bowen, H. C., Broadley, M. R., Eastwood, D. C., May, S. T., Rahn, C., Swarup, R., Woolaway, K. E., and White, P. J. (2003). Changes in gene expression in Arabidopsis shoots during phosphate starvation and the potential for developing smart plants. Plant Physiol. 132, 1–19. Harada, H., and Leigh, R. A. (2006). Genetic mapping of natural variation in potassium concentrations in shoots of Arabidopsis thaliana. J. Exp. Bot. 57, 953–960. Hatch, D. J., Chadwick, D. R., Jarvis, S. C., and Roker, J. A. (2004). ‘‘Controlling Nitrogen Flows and Losses.’’ Wageningen Academic Publishers, Wageningen. Haygarth, P. M., and Jarvis, S. C. (1999). Transfer of phosphorus from agricultural soil. Adv. Agron. 66, 196–251. Hector, A., Schmid, B., Beierkuhnlein, C., Caldeira, M. C., Diemar, M., Dimitrakopoulos, P. G., Finn, J. A., Freitas, H., Giller, P. S., Good, J., Harris, R., Hogberg, P., et al. (1999). Plant diversity and productivity experiments in European grasslands. Science 286, 1123–1127. Hirel, B., Bertin, P., Quillere, I., Bourdoncle, W., Attagnant, C., Dellay, C., Gouy, A., Cadiou, S., Retailliau, C., Falque, M., and Gallais, A. (2001). Towards a better understanding of the genetic and physiological basis for nitrogen use efficiency in maize. Plant Physiol. 125, 1258–1270. Holtham, D. A. L., Matthews, G. P., and Scholefield, D. S. (2007). Measurement and simulation of void structure and hydraulic changes caused by root-induced soil structuring under white clover compared to ryegrass. Geoderma 142, 142–151. Hooper, D. U., Chapin, F. S., III, Ewel, J. J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J. H., Lodge, D. M., Loreau, M., Naeem, S., Schmid, B., Seta¨la¨, H., et al. (2005). Effects of biodiversity on ecosystem functioning: A consensus of current knowledge. Ecol. Monogr. 75, 3–35. Hopkins, A., and Del Prado, A. (2007). Implications of climate change for grasslands in Europe, impacts, adaptations and mitigation options: A review. Grass Forage Sci. 62, 118–126. Humphreys, M. W., Thomas, H. M., Morgan, W. G., Meredith, M. R., Harper, J. A., Thomas, H., Zwierzykowski, Z., and Ghesquie`re, M. (1995). Discriminating the ancestral progenitors of hexaploid Festuca arundinacea using genomic in situ hybridisation. Heredity 75, 171–174. Humphreys, M. W., Thomas, H. M., Harper, J., Morgan, G., James, A., Ghamari-Zare, A., and Thomas, H. (1997). Dissecting drought- and cold-tolerance traits in the LoliumFestuca complex by introgression mapping. New Phytol. 137, 55–60. Humphreys, M. W., Canter, P. J., and Thomas, H. M. (2003). Advances in introgression technologies for precision breeding within the Lolium-Festuca complex. Ann. Appl. Biol. 143, 1–10. Humphreys, M. W., Yadav, R. S., Cairns, A. J., Turner, L. B., Humphreys, J., and Skt, L. (2006). A changing climate for grassland research. New Phytol. 169, 9–26.

350

M. T. Abberton et al.

IFDC An International Center for Soil Fertility and Agricultural Development. (2005). Factors affecting fertilizer supply in Africa. Draft paper prepared for the World Bank under PASS Contract WB0223B. www.ifdc.org. (accessed November 2005). Ikeda, H., Kamoshita, A., and Manabe, T. (2007). Genetic analysis of rooting ability of transplanted rice (Oryza sativa L.) under different water conditions. J. Exp. Bot. 58, 309–318. IPCC. (1997). IPCC Revised 1996 Guidelines for National Greenhouse Gas Inventories Vol. 3. Greenhouse Gas Inventory Reference Manual. Intergovernmental panel on Climate change (IPCC), Bracknell, UK. IPCC. (2001). Climate Change 2001: The Scientific Basis. Contribution of working group 1 to the third assessment report of the intergovernmental panel on climate change ( J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. Van der Linden, X. Dai, K. Maskell and C. A. Johnson, Eds.), Cambridge University Press, Cambridge, UK. Jarvis, S. C., Hatch, D. J., and Lockyer, D. R. (1989). Ammonia fluxes from grazed grassland annual losses from cattle production systems and their relation to nitrogen inputs. J. Agric. Sci. 113, 99–108. Jauhar, P. (1993). Cytogeneticsofthe Festuca-Lolium complex. Relevance to breeding. In ‘‘Monographs on Theoretical and Applied Genetics, Vol. 18’’. (B. D. Frankel, L. Grossman, and R. Maliga, Eds.), Pubs Springer-Verlag Berlin, Heidelberg, New York. (ISBN 3–540-52113–5). Jobba´gy, E. G., and Jackson, R. B. (2000). The vertical distribution of soil organic carbon and its relation to climate and vegetation. Ecol. Appl. 10, 423–436. Joffre, R., and A˚gren, G. I. (2001). From plant to soil: Litter production and decomposition. In ‘‘Terrestrial Global Productivity’’ (T. A. Mooney, Ed.), pp. 83–89. Academic Press, San Diego, CA, USA. Johnston, A. E. (2000). ‘‘Soil and Plant Phosphate,’’ p. 54. International Fertilizer Industry Association, Paris. ISBN 2–9506299–54. Jones, M. B., and Donnelly, A. (2004). Carbon sequestration in temperate grassland ecosystems and the influence of management, climate and elevated CO2. New Phytol. 164, 423–439. Kalbitz, K., Solinger, S., Park, J. H., Michalzik, B., and Matzner, E. (2000). Controls on the dynamics of dissolved organic matter in soils: A review. Soil Sci. 165, 277–304. Kemp, D. R., and Michalk, D. L. (2007). Towards sustainable grassland and livestock management. J. Agric. Sci. 145, 543–564. Kincaid, R. L., Garikipati, D. K., Nennich, T. D., and Harrison, J. H. (2005). Effect of grain source and exogenous phytase on phosphorus digestibility in dairy cows. J. Dairy Sci. 88, 2893–2902. Kingston-Smith, A. H., and Thomas, H. M. (2003). Strategies of plant breeding for improved rumen function. Ann. Bot. 142, 13–24. Kirwan, L., Luscher, A., Sebastia, M. T., Finn, J. A., Collins, R. P., Porqueddu, C., Helgadottir, A., Baadshaug, O. H., Brophy, C., Coran, C., Dalmansdottir, S., Delgado, I., et al. (2007). Evenness drives consistent diversity effects in an intensive grassland system across 28 European sites. J. Ecol. 95, 530–539. Knowlton, K. F., Taylor, M. S., Hill, S. R., Cobb, C., and Wilson, K. F. (2007). Manure nutrient excretion by lactating cows fed exogenous phytase and cellulose. J. Dairy Sci. 90, 4356–4360. Lal, R., Henderlong, P., and Flowers, M. (1998). Forages and row cropping effects on soil organic carbon and nitrogen contents. In ‘‘Management of Carbon Sequestration in Soil’’ (B. A. Stewart, Ed.), pp. 365–379. CRC Press, Boca Raton, FL, USA. Lea, P. J., and Azevedo, R. A. (2006). Nitrogen use efficiency. 1. Uptake of nitrogen from the soil. Ann. Appl. Biol. 149, 243–247.

Genetic Improvement of Forage Species

351

Lebreton, C., Lazic-Jancic, V., Steed, A., Pekic, S., and Quarrie, S. A. (1995). Identification of QTL for drought responses in maize and their use in testing causal relationships between traits. J. Exp. Bot. 46, 853–865. Ledgard, S. F., Penno, J. W., and Sprosen, M. S. (1999). Nitrogen inputs and losses from clover/grass pastures grazed by dairy cows, as affected by nitrogen fertilizer application. J. Agric. Sci. 132, 215–225. Lee, M. R. F., Colmenero, J.de J. Olmos, Winters, A. L., Scollan, N. D., and Minchin, F. R. (2006). Polyphenol oxidase activity in grass and its effect on plant-mediated lipolysis and proteolysis of Dactylis glomerata (cocksfoot) in a simulated rumen environment. J. Sci. Food Agric. 86, 1503–1511. Lewis, E. J., Humphreys, M. W., and Caton, M. P. (1980). Disomic inheritance in Festuca arundinacea Schreb. Pflanzenzu´´chtg 84, 335–341. Li, Y.-G., Tanner, G., and Larkin, P. (1996). The DMACA-HCL protocol and the threshold proanthocyanidins content for bloat safety in forage legumes. J. Sci. Food Agric. 70, 89–101. Loudet, O., Chaillou, S., Merigout, P., Talbotec, J., and Daniele-Vedele, F. (2003). Quantitative trait loci analysis of nitrogen use efficiency in Arabidopsis. Plant Physiol. 131, 345–358. Lovett, D. K., Bortolozzo, A., Conaghan, P., O’Kiely, P., and O’Mara, F. P. (2004). In vitro total and methane gas production as influenced by rate of nitrogen application, season of harvest and perennial ryegrass cultivar. Grass Forage Sci. 59, 227–232. Luten, W., and Roozeboom, L. (1976). The effect of traffic on the yields of various grasses. Lanbourmechanisatie 27, 561–562. Macleod, C. J. A., Binley, A., Clark, L. J., Hawkins, S. L., Humphreys, M. W., Turner, L. B., Whalley, W. R., and Haygarth, P. M. (2007). Genetically modified hydrographs: What can grass genetics do for temperate catchment hydrology? Hydrol. Processes 21, doi: 10.1002/hyp.6780. MacRae, J. C., and Ulyatt, M. J. (1974). Quantitative digestion of fresh herbage by sheep. J. Agric. Sci. 82, 309–319. Mannetje, L. (2007). Climate change and grasslands through the ages: An overview. Grass Forage Sci. 62, 113–117. Marley, C. L., Fychan, R., and Jones, R. (2006). Yield, persistency and chemical composition of Lotus species and varieties (birdsfoot trefoil and greater birdsfoot trefoil) when harvested for silage in the UK. Grass Forage Sci. 61, 134–145. Marley, C. L., Fraser, M. D., Fisher, W. J., Forbes, A. B., Ones, R., Moorby, J. M., MacRae, J. C., and Theodorou, M. K. (2007). Effects of continuous or rotational grazing of two perennial ryegrass varieties on the chemical composition of the herbage and the performance of finishing lambs. Grass Forage Sci. 62, 255–264. Marschner, H. (1986). ‘‘Mineral Nutrition in Higher Plants.’’ Academic Press, London. Marshall, A. H., Williams, T. A., Abberton, M. T., Michaelson-Yeates, T. P. T., Olyott, P., and Powell, H. G. (2004). Forage quality of white clover (Trifolium repens L.) Caucasian clover (Trifolium ambiguum M. Bieb) hybrids and their grass companion when grown over three harvest years. Grass Forage Sci. 59, 91–99. Marshall, A. H., Ribaimont, F., Collins, R. P., Bryant, D., and Abberton, M. T. (2005a). Variation in tannin content and morphological traits in Lotus corniculatus L. (Birdsfoot trefoil). In ‘‘Proceedings of the XX International Grassland Congress’’ (F. P. O’Mara, R. J. Wilkins, L. ‘t. Mannetje, D. K. Lovett, P. A. M. Rogers, and T. M. Boland, Eds.), p. 245. 26 June–2 July, Dublin, Ireland. Marshall, A. H., Williams, T. A., Olyott, P., Abberton, M. T., and MichaelsonYeates, T. P. T. (2005b). Forage yield and persistency of Trifolium repens Trifolium nigrescens under rotational sheep grazing. Grass Forage Sci. 60, 68–73.

352

M. T. Abberton et al.

Marshall, A. H., Bryant, D., Latypova, G., Olyott, P., Morris, P., and Robbins, M. (2008). A high throughput method for the quantification of proanthocyanidins in forage crops and its application in assessing variation in condensed tannin content within and between varieties of Lotus corniculatus and L. uliginosus. J. Agric. Food Chem. 56, 974–981. Martin, N. H., Bouck, A. C., and Arnold, M. L. (2006). Detecting adaptive trait introgression between Iris fulva and I. brevicaulis in highly selective field conditions. Genetics 153, 965–977. Materechera, S. A., Kirby, J. M., Alston, A. M., and Dexter, A. R. (1994). Modification of soil aggregation by watering regime and roots growing through beds of large aggregates. Plant Soil 160, 57–66. Michaelson-Yeates, T. P. T., Macduff, J. H., Abberton, M. T., and Raistrick, N. (1998). Characterization of novel inbred lines of white clover (Trifolium repens L.). II. Variation in N2 fixation, NO 3 uptake and their interactions. Euphytica 103, 45–54. Mickelson, S., See, D., Meyer, F. D., Garner, J. P., Foster, C. R., Blake, T. K., and Fischer, A. M. (2003). Mapping of QTL associated with nitrogen storage and remobilization in barley (Hordeum vulgare L.) leaves. J. Exp. Bot. 54, 801–812. Miller, P. R., and Ehlke, N. J. (1997). Inheritance of condensed tannins in birdsfoot trefoil. Can. J. Plant Sci. 77, 587–593. Min, B. R., Barry, T. N., Attwood, G. T., and McNabb, W. C. (2003). The effect of condensed tannins on the nutrition and health of ruminants fed fresh temperate forages: A review. Anim. Feed Sci. Technol. 106, 3–19. Misselbrook, T., Powell, J. M., Broderick, G. A., and Grabber, J. H. (2005). Dietary manipulation in dairy cattle: Laboratory experiments to assess the influence on ammonia emissions. J. Dairy Sci. 88, 1765–1777. Morgan, J. (2005). Rising atmospheric CO2 and global climate change: Responses and management implications for grazing lands. In ‘‘Grasslands: Developments, Opportunities, Perspectives’’ (S. G. Reynolds and J. Frame, Eds.), pp. 235–260. FAO 2005. Science Publishers, Inc, Plymouth, UK. Morris, P., and Robbins, M. P. (1997). Manipulating condensed tannins in forage legumes. In ‘‘Biotechnology and the Improvement of Forage Legumes’’ (B. D. McKersie and D. C. W. Brown, Eds.), pp. 147–173. CAB International, Wallingford, CT. Moseley, G., and Baker, D. H. (1991). The efficacy of a high magnesium grass cultivar in controlling hypomagnesemia in grazing animals. Grass Forage Sci. 46, 375–380. Mulder, C. P. H., Bazeley-White, E., Dimitrakopoulos, P. G., Hector, A., SchererLorenzen, M., and Schmid, B. (2004). Species evenness and productivity in experimental plant communities. Oikos 107, 50–63. Mytton, L. R., Cresswell, A., and Colbourn, P. (1993). Improvement in soil structure associated with white clover. Grass Forage Sci. 48, 84–90. Neeteson, J. J., Schro¨der, J. J., and Jakobsson, C. (2004). Drivers towards sustainability: Why change? In ‘‘Controlling Nitrogen Flows and Losses’’ (D. J. Hatch, D. R. Chadwick, S. C. Jarvis, and J. A. Roker, Eds.), pp. 29–38. Wageningen Academic Publishers, The Netherlands. Nelson, T. S. (1967). Utilization of phytate phosphorus by poultry—A review. Poultry Sci. 46, 862–871. Ni, J. J., Wu, P., Senadhira, D., and Huang, N. (1998). Mapping QTLs for phosphorus deficiency tolerance in rice (Oryza sativa L.). Theor. Appl. Genet. 97, 1361–1369. Obara, M., Kajiura, M., Fukuta, Y., Yano, M., Hayashi, M., Yamaya, T., and Sato, T. (2001). Mapping of QTLs associated with cytosolic glutamine synthetase and NADHglutamate synthase in rice (Oryza sativa L.). J. Exp. Bot. 52, 1209–1217. Opsahl, M. L., McClenaghan, M., Springbett, A., Reid, S., Lathe, R., Colman, A., Bruce, C., and Whitelaw, A. (2002). Multiple effects of genetic background on variegated transgene expression in mice. Genetics 160, 1107–1112.

Genetic Improvement of Forage Species

353

Orris, G. J., and Chernoff, C. B. (2002). Data set of world phosphate mines, deposits and occurrences—Part B. Location and mineral economic data. Open file Report 02–156-B. U.S. Department of the Interior and U.S. Geological Survey. Osbourne, L. D., and Rengel, Z. (2002). Screening cereals for genotypic variation in efficiency of phosphorus uptake and utilisation. Aust. J. Agric. Res. 53, 295–303. Owens, V. N., Albrecht, K. A., and Muck, R. E. (2002). Protein degradation and fermentation characteristics of unwilted red clover and alfalfa silage harvested at various times during the day. Grass Forage Sci. 57, 329–341. Papadopoulos, A., Mooney, S. J., and Bird, N. R. A. (2006). Quantification of the effects of contrasting crops in the development of soil structure: An organic conversion. Soil Use Manage. 22, 172–179. Parsons, A. J. P., Orr, R. J., Penning, P. D., Lockyer, D. R., and Ryden, J. C. (1991). Uptake cycling and fate of nitrogen in grass-clover swards continuously grazed by sheep. J. Agric. Sci. 116, 47–62. Peoples, M. B., Freney, J. R., and Mosier, A. R. (1995). Minimizing gaseous losses of nitrogen. In ‘‘Nitrogen Fertilization in the Environment’’ (P. E. Bacon, Ed.), pp. 565–602. Marcel Dekker, Inc, New York. Peto, F. H. (1933). The cytology of certain intergeneric hybrids between Festuca and Lolium. J. Genet. 28, 113–156. Pollock, C. J., Humphreys, M. O., Morris, P., Abberton, M. T., and Humphreys, M. W. (2005). Biotechnological approaches to temperate forage improvement. In ‘‘Grasslands: Developments, Opportunities, Perspectives’’ (S. G. Reynolds and J. Frame, Eds.), pp. 281–302. FAO 2005. Science Publishers Inc., Plymouth, UK. Powers, W. J., and Van Horn, H. H. (2001). Nutritional implications for manure nutrient management planning. Appl. Eng. Agric. 17, 27–39. Presterl, T., Seitz, G., Landbeck, M., Thiemt, E. M., Schmidt, W., and Geiger, H. H. (2003). Improving nitrogen-use efficiency in European Maize: Estimation of quantitative genetic parameters. Crop Sci. 43, 1259–1265. Price, A. H., Tomos, A. D., and Virk, D. F. S. (1997). Genetic dissection of root growth in rice (Oryza sativa L.). I. A hydroponic screen. Theor. Appl. Genet. 95, 132–142. Purcino, A. A. C., Arellano, C., Athwal, G. S., and Huber, S. C. (1998). Nitrate effect on carbon and nitrogen assimilating enzymes of maize hybrids representing seven eras of breeding. Maydica 43, 83–94. Ramirez-Restrepo, C. A., and Barry, T. N. (2005). Alternative temperate forages containing secondary compounds for improving sustainable productivity in grazing ruminants. Anim. Feed Sci. Technol. 120, 179–201. Rauh, B. L., Basten, C., and Buckler, E. S. (2002). Quantitative trait loci analysis of growth response to varying nitrogen sources in Arabidopsis thaliana. Theor. Appl. Genet. 104, 743–750. Ricklefs, R. E. (1990). ‘‘Ecology.’’ WH Freeman & Company, New York. Runge-Metzger, A. (1995). Closing the cycle: Obstacles to efficient P management for improved global security. In ‘‘Phosphorus in the Global Environment: Transfers Cycles and Management’’ (H. Tiessen, Ed.), pp. 27–42. John Wiley & Sons, New York. Ruprecht, J. K., and Schofield, N. J. (1991). Effects of partial deforestation on hydrology and salinity in high salt storage landscapes. J. Hydrol. 129, 19–38. Shaver, G. R., and Melillo, J. M. (1984). Nutrient budgets of marsh plants: Efficiency concepts and relation to availability. Ecology 65, 1491–1510. Shaw, R. K. (2005). Effects of gene-transfer from Festuca to Lolium on plant-mediated proteolysis, Ph.D. Thesis, University of Wales, Aberystwyth. Sheehy, J. E., Mitchell, P. L., Kirk, G. J. D., and Ferrer, A. B. (2005). Can smarter nitrogen fertilisers be designed? Matching nitrogen supply to crop requirements at high yields using a simple model. Field Crop Res. 94, 54–66.

354

M. T. Abberton et al.

Siangliw, M., Toojinda, T., Tragoonrung, S., and Vanavicit, A. (2003). Thai jasmine rice carrying QTLch9 (SubQTL) is submergence tolerant. Ann. Bot. 91, 255–261. Sivakumaran, S., Rumball, W., Lane, G. A., Fraser, K., Foo, L. Y., Yu, M., and Meagher, L. P. (2006). Variation of proanthocyanidins in Lotus species. J. Chem. Ecol. 32, 1797–1816. Sleper, D. A., Garner, G. B., Asay, K. H., Boland, A. R., and Pickett, E. E. (1977). Breeding for Mg, Ca, K and P content in tall fescue. Crop Sci. 17, 433–438. Soussana, J. F., and Luscher, A. (2007). Temperate grasslands and global atmospheric change: A review. Grass Forage Sci. 62, 127–134. Soussana, J. F., Loiseau, P., Vuichard, N., Ceschia, E., Balesdent, J., Chavallier, T., and Arrouays, D. (2004). Carbon cycling and sequestration opportunities in temperate grasslands. Soil Use and Management 20, 219–230. Steinfeld, H., Gerber, P., Wassenaar, T., Castel, V., Rosales, M., and de Haan, C. (2006). ‘‘Livestock’s Long Shadow. Environmental Issues and Options.’’ FAO, 390 pp. Stern, N. (2007). ‘‘The Economics of Climate Change. The Stern Review,’’ 712 pp. Cambridge University Press, Cambridge, UK. Sullivan, M., Thoma, S., Samac, D., and Hatfield, R. (2004). Cloning of red clover and alfalfa polyphenol oxidase genes and expression of active enzymes in transgenic alfalfa. In ‘‘Molecular Breeding of Forage and Turf.’’. Proceedings of the 3rd International Symposium, Dallas, Texas and Ardmore, Oklahoma, May 18–22, 2003, (A. Hopkins, Z. Y. Zang, R. Mian, M. Sledge, and R. E. Barker, Eds.), pp. 189–196. Kluwer Academic Publishers, Dorderecht. Sutton, M. A., Dragosits, U., Tang, Y. S., and Fowler, D. (2000). Ammonia emissions from non-agricultural sources in the UK. Atmos. Environ. 34, 855–869. Tilman, D. (1999). Global environmental impacts of agricultural expansion: The need for sustainable and efficient practices. Proc. Natl. Acad. Sci. 96, 5995–6000. Toojinda, T., Siangliw, M., Tragroonrung, S., and Vanavichit, A. (2003). Molecular genetics of submergence tolerance in rice: QTL analysis of key traits. Ann. Bot. 91, 243–253. Turner, L. B., Cairns, A. J., Armstead, I. P., Thomas, H., Humphreys, M. W., and Humphreys, M. O. (2008). Does fructan have a functional role in physiological traits? Elucidation by QTL mapping. New Phytol. (in press). Ulyatt, M. J., Thomson, D. J., Beever, D. E., Evans, R. T., and Haines, M. J. (1988). The digestion of perennial ryegrass (Lolium perenne cv. Melle) and whiter clover (Trifolium repens cv. Blanca) by grazing cattle. Br. J. Nutr. 60, 137–149. USGS (US Geological Survey) Minerals Information.. (2002). ‘‘Mineral Commodity Summaries.’’ USGS, Reston. http://minerals.usgs.gov/minerals/pubs/msc/. Valk, H., Metcalf, J. A., and Withers, P. J. A. (2000). Prospects for minimizing phosphorus excretion in ruminants by dietary manipulation. J. Environ. Qual. 29, 28–36. Van Loo, E. N., Van Wijk, A. J. P., Dolstra, O., Marvin, H. J. P., and Snijders, C. H. A. (1997). Selection for nitrogen use efficiency in perennial ryegrass using hydroponics. In ‘‘Proceedings of the XVII International Grassland Congress,’’ pp. 4/ 29–4/30. June 8–19, Winnipeg and Saskatoon, Canada. Van Loo, E. N., Dolstra, O., Humphreys, M. O., Wolters, L., Luessink, W., de Reik, J., and Bark, N. (2003). Lower nitrogen losses through marker assisted selection for nitrogen use efficiency and feeding value in ryegrass (NIMGRASS). Vortrage fur PLanzenzuchtung 59, 270–279. Vance, C. P. (2001). Symbiotic nitrogen fixation and phosphorus acquisition. Plant renewable resources. Plant Physiol. 127, 390–397. Vance, C. P., Uhde-Stone, C., and Allan, D. (2003). Phosphorus acquisition and use: Critical adaptations by plants for securing a non renewable resource. New Phytol. 157, 423–447.

Genetic Improvement of Forage Species

355

Visser, E. J. W., Voesenek, L. A. C. J., Vartapetian, B. B., and Jackson, M. B. (2003). Flooding and plant growth. Ann. Bot. 91, 107–109. Vitousek, P. M. (1982). Nutrient cycling and nutrient use efficiency. Am. Nat. 119, 553–572. Vogel, K. P., Mayland, H. F., Reece, P. E., and Lamb, J. F. S. (1989). Genetic variability for mineral element concentration of crested wheatgrass forage. Crop Sci. 29, 1146–1150. Wang, Y. H., Garvin, D. F., and Kochlian, L. V. (2002). Rapid induction of regulatory and transporter genes in response to phosphorus, potassium, and iron deficiencies in tomato roots. Evidence for cross talk and root/rhizosphere-mediated signals. Plant Physiol. 130, 1361–1370. Webb, J., Harrison, R., and Ellis, S. (2000). Nitrogen fluxes in three arable soils in the UK. Eur. J. Agron. 13, 207–223. Wen, L., Roberts, C. A., Williams, J. E., Kallenbach, R. L., Beuselinck, P. R., and McGraw, R. L. (2003). Condensed tannin concentration of rhizomatous and nonrhizomatous birdsfoot trefoil in grazed mixtures and monocultures. Crop Sci. 43, 302–306. Whalley, W. R., Riseley, B., Leeds-Harrison, P. B., Bird, N. R. A., Leech, P. K., and Adderley, W. P. (2005). Structural differences between bulk and rhizosphere soil. Eur. J. Soil Sci. 56(3), 353–360. White, P. J., and Hammond, J. P. (2006). ‘‘Updating the estimate of the sources of phosphorus in UK waters.’’ Final report for Defra funded project WT0701CSF. White, P. J., Broadley, M. R., Greenwood, D. J., and Hammond, J. P. (2005). Genetic modifications to improve phosphorus acquisition by roots. Proceedings No. 568, International fertiliser Society, York, UK. Wilkins, P. W., Macduff, J. H., Raistrick, N., and Collison, M. (1997). Varietal differences in perennial ryegrass for nitrogen use efficiency in leaf growth following defoliation: Performance in flowing solution culture and its relationship to yield under simulated grazing in the field. Euphytica 98, 109–119. Winters, A. L., Minchin, F. R., Merry, R. J., and Morris, P. (2003). Comparison of polyphenol oxidase activity in red clover and perennial ryegrass. Aspects Appl. Biol. 70, 121–128. Wissuwa, M., and Ae, N. (2001). Further characterisation of two QTLs that increase phosphorus uptake in rice (Oryza sativa L.) under phosphorus deficiency. Plant Soil 237, 275–286. Witt, C., Dobermann, A., Abdulrachman, S., Gines, H. C., Guanghuo, W., Nagarajan, R., Satawatananont, S., Son, T. T., Tan, P. S., Tiem, Le V., Simbahan, G. C., and Olk, D. C. (1999). Internal nutrient efficiencies of irrigated lowland rice in tropical and subtropical Asia. Field Crop Res. 63, 113–138. Witty, J., and Mytton, L. R. (2001). Soil quality: Manifestations, mechanisms and measurement. IGER Innovations 2001 5, 4–57. Wood, S., and Cowie, A. (2004). A review of greenhouse gas emission factors for fertiliser production. IEA Bioenergy Task 38. Yamulki, S., and Jarvis, S. C. (1997). Nitrous oxide emissions from excreta from a simulated grazing pattern and fertilizer application to grassland. In: ‘‘Gaseous Nitrogen Emissions from Grasslands’’. (S. C. Javis and B. F. Pain, Eds.), pp. 195–199. CAB International, Oxon, UK. Zwierzykowski, Z., Kosmala, A., Zwierzykowska, E., Jones, N., Joks´, W., and Bocianowski, J. (2006). Genome balance in six successive generations of the allotetraploid Festuca pratensis Lolium perenne. Theor. Appl Genet. 113, 539–547.

This page intentionally left blank

C H A P T E R

S E V E N

Mutagenesis and High-Throughput Functional Genomics in Cereal Crops: Current Status H. S. Balyan,* N. Sreenivasulu,† O. Riera-Lizarazu,‡ P. Azhaguvel,§ and S. F. Kianian** Contents 1. Introduction 2. Insertional Mutagenesis 2.1. T-DNA insertion mutagenesis 2.2. Transposon insertion mutagenesis 2.3. Activation tagging and gene trap systems 2.4. Insertion mutagenesis resources in cereals 2.5. Selected examples of reverse-genetics functional analysis using insertion lines 3. Non-Transgenic TILLING, DEALING, and DeleteageneTM Approaches 3.1. Development of mutagenized population 3.2. Mutation detection technique in TILLING 3.3. TILLING initiatives in cereals 3.4. Eco-TILLING 3.5. DEALING and DeleteageneTM 3.6. Allelic series versus knockout mutations 4. Phenomics Platform for Screening Mutagenized Population 5. Outlook Acknowledgments References

358 361 362 364 373 375 377 380 380 389 393 394 395 397 398 399 401 401

Department of Genetics and Plant Breeding, Ch. Charan Singh University, Meerut 250 004, India Leibniz Institute of Plant Genetics and Crop Plant Research, Corrensstrasse-03, Gatersleben 06466, Germany { Department of Crop and Soil Science, Oregon State University, Corvallis, Oregon 97331 } Texas A&M University Agricultural Research and Extension Center, 6500 Amarillo Blvd West, Amarillo, Texas 79106 ** Department of Plant Sciences, North Dakota State University, Fargo, North Dakota 58105 * {

Advances in Agronomy, Volume 98 ISSN 0065-2113, DOI: 10.1016/S0065-2113(08)00207-1

#

2008 Elsevier Inc. All rights reserved.

357

358

H. S. Balyan et al.

The accumulation of large sequence information in cereal crop plants has opened opportunities for determining gene function using reverse genetics rather than forward-genetics approaches. A number of reverse-genetics approaches have been developed and used in model plant species such as Arabidopsis and rice and recently been extended to other cereal crop species during the last few years to determine gene-function relationship. Insertional and chemical/ physical mutagen induced mutagenesis combined with emerging high-throughput analytical approaches have made it possible to rapidly discover sequencefunction relationship through reverse genetics at genome-wide scale, even in large genome cereal species. In this review, we provide a brief glimpse of the insertional mutagenesis (T-DNA and transposons) based approaches and thereafter discuss the recently discovered non-transgenic technologies such as TILLING (Targeting Induced Local Lesions IN Genomes) and DEALING (Detecting Adduct Lesions In Genomes), and their emerging applications in cereal crop plants using suitable examples.

1. Introduction Cereal crop plants belong to the family Poaceae and contribute about half of the calories in human diet (Paterson et al., 2005). At the same time, these species show varying levels of genome complexity with up to ~30-fold variation in their genome size (e.g., rice = 420 Mb and wheat = 16,000 Mb). The recent availability of complete rice genome sequence (Goff et al., 2002; International Rice Genome Sequencing Project, 2005; Matsumoto et al., 2005; Yu et al., 2002), the genome-wide gene rich region sequencing, and sequencing of a minimum tiling path in maize (Bruggmann et al., 2006; Rabinowicz and Bennetzen, 2006), sorghum (Bedell et al., 2005), and sequencing efforts in wheat and barley (International Wheat Genome Sequencing Consortium, IWGSC, http://www.wheatgenome.org/; Devos et al., 2005; http://www.international.inra.fr./research/some_examples/ sequencing_the_wheat-genome; International Barley Sequencing Consortium, IBSC, http://barleygenome.org; European Triticeae Genomics Initiative, ETGI, http://www.etgi.org) and the ever increasing EST/DNA sequence resources in GenBank, has fuelled the focus on comparative and functional genomics in these cereal crop plants. Functional genomics aims at the development and application of highthroughput genome-wide experimental approaches for the analysis of function of genes discovered and annotated by structural genomics (e.g., genome sequencing projects) and to precisely understand the systems biology (Feng and Mundy, 2006; Hieter and Boguski, 1997). Computational assisted functional annotation based on homology search analysis programs, signal and domain scans, protein localization, and structural element prediction has led to the prediction of the putative functions of genes quickly.

359

Mutagenesis and Functional Genomics in Cereals

Although, the function of a gene of interest can be predicted by its homology to DNA/EST sequence of genes of known function in a sequence repository, such methods proved informative for about 30–40% of sequences of a given genome. Moreover, the sequence homology may exhibit general functional information with numerous inaccuracies and thus specific gene function may not be determined precisely from sequence homology alone (Sessions et al., 2002). Gaining additional information by genome-wide transcriptome profiling studies, protein–protein interaction studies, comprehensive metabolome, and proteomic studies provides mainly correlative data and helps to gain knowledge of the physiological processes that occur during plant development and adaptation to abiotic and biotic stresses (Fig. 1). While an increasing array of genomic tools and resources continue to be developed, the functional analysis of annotated genes/candidate genes for specific traits is still considered a bottleneck. In the past, mutant analysis based functional derivation has been carried out using a forward-genetics approach, where genes responsible for a specific phenotypes are identified by map-based cloning approach. This approach is difficult, laborious, and time consuming Genome sequences and large-scale EST collection in cereals

Genome-wide transcription profiling To identify co-expressed regulators and pathway genes during plant ontogeny or during stress-responses

Transgenic approaches overexpression anti-sense (RNAi)

Gene annotation By implementing bioinformatic approaches

Identification of gene functions by forward and reverse genetic approaches

Insertional mutagenesis T-DNA/transposons

Induced mutagenesis TILLING, DEALING, deleteageneTM

Identification of insertions/mutations in genic regions and their correlation with-phenotypic alterations

Figure 1 Schematic representation showing steps involved in determination of gene function.

360

H. S. Balyan et al.

particularly while dealing with cereal species known to possess large genomes consisting a high proportion (~80%) of repetitive sequences and is inefficient when it comes to performing high-throughput functional genomic analyses. As a part of post-genomic revolution, general and robust high-throughput methods have been developed to determine gene function by identifying mutations in the gene sequence and then comparing the phenotype exhibited by this mutation with wild-type plants (Fig. 1). This approach known as reverse genetics is a powerful strategy for elucidating the biological functions of the gene sequences with unknown function. Mutagenesis-based reverse genetics in the past has relied on insertional mutagenesis using T-DNA (transfer DNA) and mobile genetic elements such as endogenous and introduced transposons leading to the development of large plant populations of Arabidopsis, rice and maize containing insertions dispersed throughout the genomes (Alonso et al., 2003; An et al., 2003; Lawrence et al., 2005; http://www.Arabidposis.org/links/insertion.jsp). The creation of databases containing DNA sequences flanking all insertions in a given population (An et al., 2003) has allowed in silico search of insertions in or adjacent to the target gene and request for critical lines by online ordering. While the insertion mutations can be analyzed with ease, it is not possible to obtain thousands of insertion lines that are required to saturate the whole genome through transformation and regeneration or independent transposon lines by positive/ negative selection for transposition (Krysan et al., 1999; Sundaresan et al., 1995; Tissier et al., 1999). Functional analysis of plant genes has also been done through gene silencing by antisense or sense suppression (Baulcombe, 1996). RNAi and intron-spliced hairpin have also proved quite effective in endogenous gene silencing (Chuang and Meyerowitz, 2000; Smith et al., 2000). However, extending RNAi technique is limited in high-throughput approaches due to the limited application of this approach for single gene or only a few target genes. Besides other limitations of the above gene-silencing methods, the recalcitrant nature of most cereal species to transformation and regeneration is another major bottleneck in their genome-wide applications. In non-transgenic and more broadly applicable approaches, newer and more generic, effective sequence-specific mutation detection (without prior screening of mutant phenotypes) platforms for reverse genetics have become available during the past 5 years (see Waugh et al., 2006). These include TILLING (Targeting Induced Local Lesions IN Genomes), which involves detection of chemically induced single base pair point mutations (McCallum et al., 2000a,b); DEALING (Detecting Adduct Lesions In Genomes), which involves detection of chemically induced deletion mutations ( Jansen et al., 1997, 1999; Kianian et al., personal communications); and DeleteageneTM, which involves detection of fast-neutron-induced deletions (Li et al., 2001, 2002a). As a corollary to the TILLING methodo-logy, the concept of Eco-TILLING has also been proposed for discovering naturally occurring polymorphisms within the gene sequences from

Mutagenesis and Functional Genomics in Cereals

361

different and diverse plant genomes (Comai et al., 2004). During the past 5 years, several reviews have appeared describing the potential of various reverse-genetics approaches in deciphering gene function focusing mainly on the model species Arabidopsis, rice and maize (Comai and Henikoff, 2006; Feng and Mundy, 2006; Gilchrist and Haughn, 2005; Leung and An, 2004; May and Martienssen, 2003; Stemple, 2004; Waugh et al., 2006). Different reverse-genetics approaches have just begun to be applied to a number of cereal species and new initiatives have been launched to harness the potential of this method. In the following sections, we present a general review of the different mutagenesis-based reverse-genetics approaches in the context of ongoing efforts with particular emphasis on their applications and possible limitations with respect to cereal crop species.

2. Insertional Mutagenesis Insertional mutagenesis for high-throughput reverse-genetics functional analysis relies on insertion of T-DNA from Agrobacterium tumefaciens (AzpirozLeehan and Feldmann, 1997; Krysan et al., 1999) and transposable elements (Martienssen, 1998; Parinov and Sundaresan, 2000) into a gene of interest leading to its disruption, which may result into an altered recognizable phenotype. In the course of this process, the mutant alleles are also tagged by inserted piece of DNA, which may lead to cloning of candidate gene. Though the tagged lines can be used for forward-genetic analyses, they have greater potential in performing high-throughput reverse-genetics analyses (Parinov and Sundaresan, 2000). This latter approach is now widely considered as an important route to characterize and identify gene function across the genomes at ease, especially among model species such as Arabidopsis and rice where complete genome sequences are available. The insertional-mutagenesis approach has the following two major advantages: (a) First, it provides a direct means for determining the function of a gene product in planta revealing a direct relationship between the gene sequence and the phenotype (Krysan et al., 1999) and (b) second, it may lead to the isolation of tagged mutant genes through recovery of the DNA sequences flanking the insertion sequence (Alonso et al., 2003; Azpiroz-Leehan and Feldmann, 1997; Krysan et al., 1999), which may, in turn, lead to the isolation of wild-type gene sequence. Initial serious efforts using insertion mutagenesis in the early parts of this decade were extremely successful and led to the creation of large insertion plant populations in Arabidopsis, resulting in a near saturation of the whole genome with independent insertions (Alonso et al., 2003; Sessions et al., 2002; Weigel et al., 2000). However, low transformation and regeneration efficiencies of most cereal species with the possible exception of rice (Hiei et al., 1994; O’Kennedy et al., 2006; Sallaud et al., 2003; Tyagi and Mohanty, 2000;

362

H. S. Balyan et al.

for reviews, see Delseny, 2003; Shrawat and Lorz, 2006) impeded the creation of large populations needed for genome-wide high-throughput functional analysis of genes using reverse-genetics approach. In most of the important cereal species except maize, the creation of large transposon insertion populations is also specifically a challenging task due to either the lack of suitable transposons or the availability of useful methods for activation of endogenous transposons. Nonetheless, in the recent past, increasing major efforts have been undertaken toward creation of large insertion mutagenesis populations for forward- and reverse-genetics analyses in at least three cereal species including rice (Hirochika et al., 2004; Hsing et al., 2007; Jeon et al., 2000; Kim et al., 2003; Kolesnik et al., 2004; Kumar et al., 2005; Sallaud et al., 2004), maize (May et al., 2003), and barley (Ayliffe et al., 2007; Cooper et al., 2004; Koprek et al., 2001; Singh et al., 2006; Zhao et al., 2006). These efforts have been supplemented by the non-transgenic reverse-genetics approaches involving use of endogenous transposons for creating insertional mutagenesis populations (Hirochika et al., 2004; May et al., 2003; McCarty et al., 2005). Reverse-genetics analysis of insertion lines is carried out following two main approaches. In the first approach, gene-specific sequences flanking T-DNA/transposon insertion may be identified in insertion line population via (i) various PCR-based methods including thermal asymmetric interlaced (TAIL) PCR, MuTail PCR, adapter-ligated PCR, a universal biotinylated adapter amplification procedure, a panhandle PCR; (ii) plasmid rescue approach; or (iii) Southern blot analysis (Balzergue et al. 2001; Dilkes and Feldmann, 1998; Krysan et al., 1999; Liu and Whittier, 1995; Sato et al., 1999; Triglia et al., 1988; Walbot, 2000). This will eventually lead to sequence information of the DNA fragment flanking the insertional element (Krysan et al., 1999; Maes et al., 1999; Sato et al., 1999). The co-segregation of the insertion element with the mutant phenotype will help identify the gene function. However, this is a laborious undertaking and therefore, more recently strategies have been developed to circumvent these steps. The alternative approach involves sequencing of sequences of segments flanking the TDNA/transposon and to create directories of Flanking Sequence Tags (FSTs) (Hanley et al., 2000; Parinov et al., 1999). In this case, FSTs of all available insertion mutant lines are used to create a database of knockout genes that can be searched electronically. Eventually, this is likely to help in global functional analysis of higher plant genes with greater ease. A relatively more detailed and separate account of T-DNA and transposon insertion mutagenesis and their applications with particular emphasis on cereal crop species is presented below.

2.1. T-DNA insertion mutagenesis The plant pathogenic bacterium A. tumefaciens has the ability to transfer part of its DNA (T-DNA) into plant nuclear genomes. This transfer of T-DNA depends on the expression of the bacterial vir genes that allows

363

Mutagenesis and Functional Genomics in Cereals

recognization of the right and left borders (RB and LB, respectively) of T-DNA for its transfer. Generally only one copy of T-DNA is integrated, but integration of more than one copy either at the same locus or at different loci has also been noted (reviewed in Springer, 2000). The classical principle of T-DNA/transposon insertional mutagenesis leading to gene knock-out is shown in Fig. 2A. In a large measure, T-DNA characteristically show insertions within the gene or the gene rich regions helping genome-wide saturation mutagenesis (An et al., 2003; Barakat et al., 2000; Chen et al., 2003; Wu et al., 2003). T-DNA insertion mutagenesis has proven very successful in Arabidopsis and rice (http://signal.salk.edu/cgi-bin/tdnaexpress T-DNA/transposon

A Gene knock-out

B Activation

T-DNA/transposon

Enhancer element CaMV

Enhancer TATA

tagging

C Entrapment tagging

(i) Enhancer

Enhancer TATA

TATA Reporter

trap

(ii) Promoter

Enhancer TATA

Reporter

trap

(iii) Gene trap

Enhancer TATA

SD SA

Figure 2 Schematic representation of tagging plant gene. (A) Gene knock out: insertion of T-DNA/transposon in either the exon (black box) or intron (dotted line) can disrupt the expression of tagged gene causing knock-out mutation. (B) Activation tagging: enhancer element of CaMV 35S promoter increases the expression (shown by dark thick line with arrow) of the gene near the insertion position. [C(i)] Enhancer trap: the minimal promoter (TATA) of a reporter gene is activated by an enhancer element resulting in the expression of the reporter gene (shown by gray arrow). [C(ii)] Promoter trap: the promoterless reporter gene can be expressed when insertion occurs in exon resulting into transcriptional fusion indicated by black and gray line with arrow. [C(iii)] Gene trap: the promoterless reporter gene contains splice acceptor (SA) sequences, which allow a transcriptional fusion between tagged gene and the reporter gene by splicing from the splice donor (SD) site to the SA sequence; in this system the reporter can be activated even when it is inserted into an intron. The product in this case is represented by black and gray line with arrow.

364

H. S. Balyan et al.

and http://signal.salk.edu/cgi-bin/RiceGE). A large number of T-DNA knockout mutant lines in rice were produced by Su-May Yu at the Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan (see Table 1). In the past, several mutants with altered phenotypes of vegetative and reproductive organs were identified while working with T-DNA insertion lines of rice (Yin and Wang, 2000). Such mutant lines may be exploited for both the forward and reverse-genetics analyses. The T-DNA insertion approach is not only efficient for identifying knockout mutations, but it has also been used for activation tagging, gene trap and enhancer trap mutagenesis particularly in rice (for details see Section 2.3.; Table 1). However, due to the low copy numbers of T-DNAs, finding insertion in a specific gene at high probability require creation of large mutagenized populations. With smaller size of the gene (<1 kb), the probability of recovering an insertion mutation reduces even in large (360,000 lines) mutagenized populations such as in case of Arabidopsis (http://signal.salk. edu/database/T-DNA/). Lack of detailed knowledge on preferred T-DNA insertion sites, makes T-DNA tagging complicated, especially in large genomes species, which possess huge amounts of repetitive DNA sequences. Multiple insertions may occur both in multiple copies per locus and in multiple loci (Lindsey et al., 1993), frequently impeding the isolation of genomic sequences flanking T-DNA. In contrast to rice, where T-DNA mutagenesis has been extensively utilized, this approach has not been as successful in other cereal species due to above laid reasons and in addition to poor transformation and regeneration efficiencies. Therefore, large collections of usable T-DNA insertion lines with a critically high number of independent insertions are not available so far in these species.

2.2. Transposon insertion mutagenesis Transposons or transposable elements are mobile elements and were first discovered in maize by Barbara McClintock in the last century (McClintock, 1950). These elements occur widely in plants including cereal species (Kumar and Bennetzen, 1999). Transposable elements are categorized into two broad classes according to their known or supposed transposition mechanisms (for a review, see Capy et al., 1998), namely, class I (retrotransposons, which transposes through an RNA intermediate generated by reverse transcription elements) and class II (DNA-based transposons, which transpose through DNA-cut and-paste type mechanisms).The insertion and transposition abilities of the transposons have been widely harnessed in cereal species such as maize and rice. Most transposable elements have a tendency to preferentially transpose into the gene rich regions (Ito et al., 1999). Gene containing the transposon generally results into knockout phenotypes. As a result of these features, transposon tagging has become a powerful tool for functional analysis and isolation of genes (Hamer et al., 2001).

Table 1

Rice mutant resource developed by the international research community

Type (vector/mutagen)

Population size current (target)

Number of FST

Mapping of FSTs on rice pseudochromosomes

Dongjin and Hwayoung

Gene and enhancer trap, activation tags (T-DNA)

80,006

CIRAD-INRAIRDA-CNRS, Genoplante

Nipponbare

Enhancer trap (T-DNAs) Ds-ET and AC lines

50,000 activation lines and 50,000 inactivation lines (100,000) 46,000 (46,000)

Huazhong Agr. Univ. China

Zhonghua 11

Enhancer trap (T-DNA)

Institute of Botany, Academia Sinica, Taiwan

Tainung 67

Insertion and activation tags (T-DNA)

Institution

Variety

Pohang University of Science and Technology

Database

Name of contact/e-mail

78,709

http://postech.ac, kr/life/risd

G. An/genean@ postech.ac.kr

14,137 27,950

13,384

E. Guiderdoni/mail: [email protected]

42,000 (70,000)

–

–

Integrated database: http:// genoplanteinfo. infobiogen.fr/ oryzatagline/ http://orygenesdb. cirad.fr Hirochika et al. (2004) http://www. ricefgchina.org

20,000 (60,000; by the end of 2005)

27,950

7020

http://trim.sinica. edu.tw.

Q. Zhang/ [email protected]. edu.cn Y. C. Hsing/ bohsing@gate. sinica.edu.tw (continued)

Table 1

(continued)

366 Type (vector/mutagen)

Population size current (target)

Number of FST

Mapping of FSTs on rice pseudochromosomes

Tainung 67

Gene trap, gene knockout, and activation tags (T-DNA)

55,000

11,992

11,992

–

Su-MayYu/ sumay@imb. sinica.edu.tw

Nipponbare

Gene trap (Ac/Ds)

20,000 (30,000)

10,878

–

–

R. Srinivasan/sri@tll. org.sg

Nipponbore

Transposons gene traps (En/Spm Ac/Ds)

1200 (40,000)

10,878

10,624

V. Sundaresan/ [email protected]

Dongjin and MGRI079

Gene trap (Ac/Ds)

89,700 and 6200

5271

5271

http://www.plb. au-davis.edu/ Labs/sundar/ index.htm –

Institution

Variety

Institute of Molecular Biology, Academia Sinica, Nankang, Taipei, Taiwan Temasek Life Sciences Laboratory, The National University of Singapore, Singapore University of California Davis, USA Rice Functional Genomics, National Institute of Agricultural Biotechnology, Republic of Korea

Database

Name of contact/e-mail

D. W. Yun/ [email protected]

RDA, Gyeongsang National University, South Korea

Dongjinbyeo

Insertion (Ac/Ds)

30,000 (100,000 each of indica and japonica)

1072

1044

CSIRO

Nipponbare

Gene and enhancer trap (Ac/Ds)

8,000 (5,000/ year)

787

787

European Union

Nipponbare

Enhancer trap and activation tags (Ac/Ds)

10,000 (20,000)

1,380

1,380

www.niab.go.kr

http://www.pi. csiro.au/ fgrttpub/home. htm –

Moo Young/ [email protected]

Chang-deok Han cdhan@nongae. gsnu.ac.kr N. M.Upadhyaya/ narayanaupadhyaya @csiro.au Andy Pereira/

[email protected]/http://orygenesdb.cirad.frNational Institute of Agrobiological ScienceNipponbareInsertion, endogenous retrotransposon (Tos17)50,000 (50,000)18,02417,955http://tos.nias.affrc.go.jpH. Hirochika/[email protected]

367

368

H. S. Balyan et al.

In the past, a number of transposons including Ac/Ds (Activator/Dissociation), En/Spm (Enhance/Suppressor–Mutator), MuDR/M (Mutator), and Tos17 that are endogenous to species like maize, rice, petunia, and snapdragon (Antirhinum majus) have been characterized extensively. The maize Ac/Ds (Baker et al., 1986) and En/(I)Spm (Pereira and Saedler, 1989; Masson and Fedoroff, 1989) transposons have been shown to transpose in a number of heterologous hosts extending their utility for insertion mutagenesis in species that do not have active endogenous transposon (Ito et al., 2004; Shimamoto et al., 1993; Sundaresan, 1996; Upadhyaya et al., 2006). Additionally, the heterologous transposons have advantage over the endogenous transposons for insertional mutagenesis as these may be used after suitable in vitro modification (Pereira, 1998). The application of transposon mutagenesis and tagging was initially restricted to maize. The maize transposable elements Ac/Ds, En/Spm, and MuDR/Mu proved useful for isolation of several maize genes (Fedoroff et al., 1983; Robertson, 1978; Sundaresan, 1996; Walbot, 1992, 2000). For insertion mutagenesis and tagging of genes, strategies involving two component systems, comprising a mobile transposon component (Ds or I/dSpm) and corresponding stable transposase (Ac or En/Spm) source, are used. First, a construct is made by cloning of an Ac element between a promoter and the selectable herbicide resistance gene (such as hygromycin-phosphotransferase, htp) coding region and it is transformed into host plant. Second, the transgenic plants carrying the Ds element are crossed with transgenic plants constitutively expressing Ac transposase, where nonautonomous Ds element has to be transposed in the presence of Ac transposase. Using this approach, transposition of the Ac element was proven by the recovery of hygromycinresistant plants (Izawa et al., 1991; Murai et al., 1991) and newly transposed Ac insertions (Enoki et al., 1999). The nonautonomous Ds element was also transposed in the presence of Ac transposase via direct gene transfer method (Shimamoto et al., 1993; Sugimoto et al., 1994). In rice, the transposition events involving heterologous maize En/Spm element are well distributed (Wisman et al., 1998) thereby making it an effective system for the targeted mutagenesis of closely linked genes. The Ac/Ds elements preferentially insert into closely linked genomic sites, and hence could be used to target specific genomic regions. On the other hand, the Mu elements have no such local bias and can prove effective at genome-wide scale (Walbot, 2000). Transposon insertion mutagenesis has been practically and successfully tried to develop insertion populations in at least three important different cereal crop species including maize, rice and barley, which is briefly reviewed below. A number of public and private organizations have created a number of maize transposon insertion populations by exploiting different transposon systems to identify knockouts in genes of interest (see Table 2; reviewed in Settles, 2005). The most popular transposon systems used include the

Table 2

Maize mutant resources developed by international research community

Transposon

Mutant line/FST

Source

Reference/e-mail/weblink

Activator (Ac)

1234 independent Ac transposants

[email protected]/http:// waksman.rutgers.edu/Waks/Labs/ Dooner/PGRPpage.html

Activator (Ac)

9000 mutants

Activator (Ac)

60 AC elements mapped (200 Ac containing lines expected to be developed) 1000 families (tDs insertion) ~2200 Mu-induced non-photosynthetic maize mutants 700 amplified insertion sequences

Dooner Lab, Waksman Institute of Microbiology Rutgers, The State University of New Jersey, 190 Frelinghuysen Rd Piscataway, NJ 08855, USA Maize Genetic Cooperation Stock Center,M. Freeling, University of California, USA Brutnell Maize Genetics Lab, Boyce Thompson Institute for Plant Research, Tower Roads, Ithaca, NY, USA Plant Genome Database (PlantGDB)

http://www.plantgdb.org/prj/ AcDsTagging/ [email protected]/http:// chloroplast.uoregon.edu/

Mutator (Mu) Mutator (Mu) Mutator (Mu)

Photosynthetic Mutant Library (PML)

Functional genomics Group at Bristol University, UK

Brutnell et al. (2002) http://w3.aces.uiue.edu/maize-crop/ [email protected]/http://bti.cornell. edu/Brutnell_lab2/Projects/ Tagging/BMGG_pro_ currentmap.html

[email protected]:///C:/ Documents%20and%20Settings/ Administrator/My %20Documents/ FST/Maize/ MAIZE% 20MUTAGENESIS/IGF%20Maize. htm, (http://www.cerealsDB.uk.net) (continued)

Table 2

(continued)

Transposon

Mutant line/FST

Source

Reference/e-mail/weblink

Mutator (Mu)

40,000 mutants carrying 1 million insertions (for hundreds of genes) 14,887 flanking genomic loci (RescueMu insertion)

Trait Utility system for corn, Pioneer Hi-Bred International

Brutnell et al. (2002)

Report from the Maize Gene Discovery project, Department of Biological Sciences, Stanford University, Stanford, CA 94305, USA. Plant Genome Research Project, Principal Investigator Thomas Brutnell Maize Targeted Mutagenesis project (MTMdb) of Rob Martienssen, Cold Spring Harbor Laboratory; Michael Freeling, Berkeley and Danny Alexander, Syngenta, USA UniformMu population.

[email protected] Fernandes et al. (2004)/http: //gremlin3.iastate.edu/zmdb/ RescueMuPopulations.html http://www.sciencestorm.com/award/ 0501713.html http://mtm.cshl.edu/

Mutator (Mu)

Mutator (Mu) Mutator (Mu)

10,000 families of Ds insertions 43,776 F2 Mutator (Mu) lines

Mutator (Mu)

1882 nonredundant FSTs

Settles et al. (2007)

Mutagenesis and Functional Genomics in Cereals

371

Mu/MuDR and Ac/Ds systems, which have important practical differences (Walbot, 2000). Rate of forward mutations induced by Ac/Ds system is lower (10–6) than the rate of forward mutations induced by Mutator transposition (10–4). The high rate of forward mutations induced by Mu elements is coupled with their preference for insertion in gene-rich regions of the genome (Cresse et al., 1995; Fernandes et al., 2004; McCarty et al., 2005; Settles et al., 2004). However, the high mutation frequency due to Mu elements is very attractive but copy number of Mu elements result in single plants having multiple germinal insertions, making it difficult to isolate the individual insertion sites and assignment of a mutant phenotype to a particular insertion. Interestingly, efforts have been made to develop ways to design specific transposons in maize, which can be used to help with identification of the insertion location. The RescueMu project has designed a modification of the Mu transposon to contain a lineralized bacterial plasmid, which can be cut from the genome with specific restriction enzymes (Raizada et al., 2001). In this case, adjacent genomic DNA is retained and on circularization of the plasmid, the adjacent DNA is easily sequenced. RescueMu is being used to develop reverse-genetics resources under the Maize Gene Discovery project, which may be screened computationally for mutations in known genes (Table 2; Fernandes, 2004). A new resource for identification of Mu insertions has been developed by Bristol University, UK (http://www.cerealsdb. uk.net/muarray.htm). The MuArray rapidly identifies mutant plants through hybridization with fluorescently labeled gene of interest. Mu flanking sequences are used to construct high density arrays and labeled cDNA of interest is hybridized to the array to identify individual plants with the correct insertions. In a recent attempt, UniformMu maize populations have been generated in an inbred W22 (http://currant.hos.ufl.edu/mutail/) to identify mutations that interfere with endosperm development. Such studies will allow identification of key genes that control endosperm development. Also genomic sequence flanking the Mu insertions may be obtained through MuTail PCR from MuDR-backgrounds. Use of Ac/Ds elements for reverse- and forward-genetics analyses in maize is also underway. With classical genetic techniques and non-transgenic materials, Ahern et al. (2006) are generating a collection of 10,000 families that each harbor a unique Ds insertion distributed throughout the genome. DNA sequences flanking the Ds elements are cloned and sequenced providing unique and specific sequence determining a precise physical position for each insertion in the maize genome. Importantly, each Ds insertion is stable in the absence of Ac, but can be remobilized using stabilized transposase source. As Ds tend to move to closely linked gene rich regions of the genome, each insertion will also serve as a platform for additional rounds of mutagenesis in the linked genes. In addition, Ac/Ds transposons can be used for generation of an allelic series within a single

372

H. S. Balyan et al.

gene (Bai et al., 2007). The Volbrecht and Bruntell labs have developed genetic stocks in W22 inbred to select for Ac transposition events (http:// www.plantgdb.org/prj/AcDsTagging). Coweperthwaite et al. (2002) used the wx-m7 allele tagged with Ac along with the bz1-m2 allele tagged with Ds to select for excision from the wx locus while retaining Ac transposase. In contrast, to this local mutagenesis approach, Kolkman et al. (2005) used Ac dosage effect for Ac transposition events that retained Ac in donor locus. The Ac dosage screen has proven effective as both Ac populations were used to clone seed phenotypes tagged by transposed Ac elements (Singh et al., 2003). To create a large-scale gene tagging resource, maize Ac and Ds elements were used for large-scale creation of gene tagging resources in rice, a heterologous host (Chin et al., 1999; Greeco et al., 2001, 2003; Kim et al., 2004; Kohli et al., 2001; Kolesnik et al., 2004; Nakagawa et al., 2000; Upadhyaya et al., 2002). Originally, the autonomous Ac element cloned between a promoter and the htp gene was introduced in rice by direct transformation (Enoki et al., 1999; Izawa et al., 1991; Murai et al., 1991). Insertional mutagenesis in rice has also been achieved using maize En/Spm transposable elements (Kumar et al., 2005). Both endogenous and heterologous retrotransposons or retroelements were shown to be randomly distributed and used as an efficient tags as well as for gene cloning in rice (Agrawal et al., 2001; Hirochika et al., 1996, 2001; Miyao et al., 2003; Sato et al., 1999; Takano et al., 2001). Some of the rice endogenous retroelements are activated and their transposition is induced when subjected to stresses caused by culturing, pathogen infection, cell culture, and wounding (Chin et al., 1999; Hirochika et al., 1996; Takeda et al., 1998). The low copy number (1–5) of rice retroelement Tos17 could also be increased to ~30 by activation of retrotransposon, which require passage through tissue culture (Hirochika et al., 1996; Parinov et al., 1999). In future, several other transposable elements that were identified earlier in rice may be exploited for reverse-genetics analysis (Fujino et al., 2005; Jiang et al., 2003; Mao et al., 2000; Nakazaki et al., 2003). Compared to rice and maize, transposon insertion mutagenesis approach has been successfully applied on a very limited scale in barley. Efforts to test the potential suitability of this approach in some other cereal species such as wheat and sorghum are underway (see below). The potential application of insertional mutagenesis for gene tagging and gene cloning in barley was demonstrated through production of large number of transgenic insertion lines. Koprek et al. (2000, 2001) generated independent barley transgenic plants carrying Ac and Ds elements, and crossed them to produce F2 population, which was used to identify plants segregating for transposed elements. A number of confirmed independent single-copy Ds containing transgenic lines (Koprek et al., 2003) and insertion lines containing 19 mapped Ds elements were developed in barley (Cooper et al., 2004) and are available for public use. Recently, barley insertion mutagenesis

Mutagenesis and Functional Genomics in Cereals

373

population of about 200 independent lines carrying flanking Ds and TDNA insertion sites were generated by either activating Ds element in existing single-copy lines when they were crossed with Ac containing line or by Agrobacterium-mediated transformation (Zhao et al., 2006). Mapping of at least 100 insertion sites in barley provides evidence that Ds launch pads will be valuable for future targeted gene tagging system to saturate genomewide mutagenesis in barley. Also more than 100 single-copy Ds insertion lines of barley were generated and preferential insertion of Ds elements into genic regions coupled with their high remobilization frequencies validated the ability of Ds for saturation mutagenesis in species such as barley and wheat containing large genomes with huge amounts of repetitive DNA (Singh et al., 2006). Occurrence of functionally redundant duplicated copies of many genes in rice, maize, barley, and other cereal species such as polyploid bread wheat, etc. (Gaut and Doebley, 1997; Goff et al., 2002) impede functional analysis of all the available genes due to the following reasons. First, insertional mutation in such genes may not express, as the loss of function in one copy of the gene may be compensated for by the expression of another copy. Second, many genes function at more than one stage of plant development and are critical for plant’s viability; mutations in such genes may be lethal. Thirdly, some of the insertion mutants are conditional and the recognizable mutant phenotypes in such cases express only under specific conditions (Bouche and Bouchez, 2001). Lastly, mutations with subtle phenotypic alterations may escape detection during phenotypic screening thereby making it impossible to correlate the changed phenotype with the specific disrupted sequence. In order to circumvent the above problems and to identify genes, complementary techniques of activation tagging and gene trap analyses that utilize random integration of reporter gene constructs have been developed (Springer, 2000; Weigel et al., 2000; see below; Fig. 2B and C).

2.3. Activation tagging and gene trap systems Activation tagging generates dominant gain-of-function mutations. This thus allows identification and functional analysis of functionally redundant genes. This approach offers the possibility of expression of gene by insertion of either transposon or T-DNA that contains cauliflower mosaic virus (CaMV) 35S ubiquitous transcriptional enhancers (Odell et al., 1985). The resulted phenotypes can then be tested either by forward or reversegenetics analysis. This method offers an interesting feature to test the gainof-function of a gene and eventually allows correlation with the observed phenotype. Based on T-DNA activation tagging strategy, more than 30 dominant mutant phenotypes, which resulted from over-expression of the

374

H. S. Balyan et al.

genes, were characterized (Weigel et al., 2000: reviewed in Tani et al., 2004). Although this method has successfully been applied to gene discovery programs in Arabidopsis, to date, this technology has mainly been applied to rice among all the cereal crop species (see An et al., 2005b). Obviously, this approach could be a valuable complementary approach for the identification of gene functions involved in reproductive development and seed yield where suppression of key genes involved in reproductive processes results in generating null mutants by transposon insertion mutagenesis approach. Activation tagging approach also has a few limitations. First, activation tagging induces low frequency of mutations; therefore it is highly unlikely to identify every gene by its potential over-expression phenotype. Second, the CaMV 35S enhancers could activate only a subset of adjacent genes due to promoter preference. Thirdly, multiple and complex insertions of T-DNA in the genome and its ubiquitous over-expression may complicate interpretation of expression patterns and subsequent molecular analysis, making it difficult to clone genes deriving the reporter gene expression. Lastly, the T-DNA insertions are generally stable, so that remobilization is not possible, as in case of transposable elements. Along with T-DNA activation tagging, transposon-mediated activation tagging system using the Ac and Ds elements was also developed and used in rice for developing insertion mutant populations (Mori et al., 2000; http://trim.sinica.edu.tw; http://orygenesdb.cirad.fr; Su-May Yu, Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan). Most recently, dominant over-expression phenotypes were generated in barley using a modified Ds element (UbiDs) (Ayliffe et al., 2007). An activation tagging version of Mutator, MightyMu has also been constructed that contains an active promoter designed to cause dominant, ectopic expression, or gain-of-function mutation (Nan and Walbot, 2006). Besides activation tagging, three different gene trap systems (enhancer trap, promoter trap, and gene trap) have also been developed (Fig. 2C). These systems use reporter gene constructs, and respond to cis-acting regulatory sequences at the site of insertion (Springer, 2000). In these systems, for ease of selecting stable transformants, and high-throughput selection of the insertion mutants, T-DNA can be modified to carry a reporter gene (such as GUS, GFP, DsRed, or Luciferase) next to the T-DNA border ( Jefferson et al., 1987; Jeon et al., 2000; Kumar et al., 2005; Pang et al., 1996; Ryu et al., 2004). Details of these systems are given in Fig. 2C. In all the three systems, expression of the tagged genes can be monitored by assaying the reporter gene activity. Transposable elements have also been used as vector for enhancer or gene traps in plant genomes (Fedoroff and Smith, 1993; Klimyuk et al., 1995; Sundaresan et al., 1995). In species that do not have active well-characterized transposable element systems, heterologous elements are utilized (reviewed in Osborne et al., 1991).

Mutagenesis and Functional Genomics in Cereals

375

2.4. Insertion mutagenesis resources in cereals Efforts made over the years have led to the availability of large T-DNA and transposon-based insertion line and FST resources in rice and maize. Such resources have also become available on a much smaller scale in barley and attempts are being currently made to produce similar resources in other cereal species including wheat and sorghum. 2.4.1. Rice In rice, T-DNA insertion mutagenesis was initially performed by Jeon et al. (2000). Subsequently, multinational European Consortium (EU-OSTID) and The International Rice Functional Genomics Consortium (IRFGC) were established to supplement the efforts of individual institutions for the creation of large insertion populations and FSTs for functional analysis of the rice genes (Hirochika et al., 2004; Sallaud et al., 2004; see van Enckevort et al., 2005 and Hsing et al., 2007). The members of the two consortia and other laboratories have together produced more than 600,000 T-DNA and endogenous and heterologous transposon (En/Spm, Ac/Ds and Tos17 ) insertion lines in rice (Table 1). These include knockout lines and lines produced following activation and entrapment tagging strategies. In addition, more than 200,000 mutagenized lines with associated FSTs are also available in the OryGenesDB database (Table 1; Droc et al., 2006). Since the average copy number of T-DNA inserts in each line is reported at 1.4–2.0, the available insertion line populations shall be adequate to find a knockout in a given gene at >90% probability (An et al., 2003, 2005; Barakat et al., 2000; Chen et al., 2003; Wu et al., 2003). As an alternative approach to insertion mutagenesis, Nakamura et al. (2007) implemented Full-length cDNA Over-eXpresser (FOX) gene-hunting system in rice for gain-of-function phenotypes from 13,980 over-expressing independent full-length cDNAs under the control of the maize Ubiquitin-1 promoter. In their collection, they have showed altered morphological characteristics for more than 16.6% of FOX-rice lines including well known example of three super-dwarf mutants resulting due to over-expressed gibberellin 2-oxidase gene. It is hoped that systematic efforts using these collections for reverse genetics based on approaches of both gain and loss of functional analysis will lead to the identification and cloning of newer genes. It is also possible that the available rice resources may prove fruitful to unravel the function of orthologous genes in other cereal species, particularly with large genomes. However, it is now being increasingly realized that rice is not the most perfect model species for triticeae genomics. Therefore, it is also desirable to establish systems for insertional mutagenesis in other related large genome species such as barley and wheat, although this goal may not be easily achievable for the reasons enumerated earlier.

376

H. S. Balyan et al.

2.4.2. Maize A number of different projects and individual laboratories are involved in development of maize reverse-genetics resources (using endogenous transposons and modified vectors), which include mutant lines and FSTs for functional analysis and cloning of maize genes (Table 2). Four major projects including Maize Gene Discovery project (http://gremlin3.iastate.edu), Plant Genome Research Project (http://www.sciencestorm.com), and Maize Targeted Mutagenesis project (MTM, http://mtm,cshl.edu/), and Pioneer HiBred’s Trait Utility System of Corn (Pioneer Hi-Bred’s TUSC) besides several other individual laboratories including Dooner (http://waksman.rutgers.edu/ Waks/Lab/Dooner/) and Bruntell (http://bti.cornell.edu/Brutnell_lab2/ Project/) laboratories are actively involved in accomplishing the above task. The above efforts together helped to produce more than 83,000 mutant lines (Table 2). The Pioneer Hi-Bred’s TUSC and MTM projects also allow users to order seeds homozygous for Mu transposon insertion mutants. Also it is worth to note that screening Photosynthetic Mutant Screen (PMS) to identify Mu-induced alleles and Uniform Mu endosperm mutants created in an NSFfunded project will act as important resources to identify developmental mutants in maize. The details of such mutants are very often available at their websites. Seed of mutants may also be ordered from the Maize Genetics Cooperative Stock Center’s Web site (http://w3.aces.uiuc.edu/maize-coop/ ). Along with insertion lines, more than 15,000 FSTs are also available (Table 2; Fernandes et al., 2004; Settles et al., 2007). 2.4.3. Sorghum, barley, and wheat Because of the recalcitrant nature of sorghum, barley, and wheat to transformation and regeneration, the insertion mutagenesis approach has so far not been as successful as in rice and maize. Nonetheless, recently initiatives have been undertaken in this direction. Among the three crops, due to very low level of gene redundancy, sorghum is considered a very useful system for determining gene function by creating insertion mutant resources. Identification of an active transposable element (Csl ) in sorghum (Chopra et al., 1999) laid the foundation of insertional mutagenesis in sorghum. The low copy number and high transposition frequency makes Csl as an important vehicle for developing mutant lines for reverse-genetics analysis and genes cloning in sorghum. With the recent availability of an improved regeneration protocol following A. tumefaciens-mediated transformation (Nguyen et al., 2007), it should be possible to create large insertion line populations in near future. However, some progress toward development of insertion mutant lines in barley using T-DNA and maize Ds element was recently made and a total of more than 300 insertion lines were produced (Cooper et al., 2004; Koprek et al., 2003; Singh et al., 2006; Zhao et al., 2006). Successful efforts to test heterologous maize Ac/Ds system in wheat

Mutagenesis and Functional Genomics in Cereals

377

for generating insertion mutants were also made (Takumi et al., 1999). The Ac and Ds elements were introduced in wheat using particle bombardment and trans-activation and excision of Ds by Ac transposase gene was demonstrated. The maize Ac/Ds system is being tried with a view to identify genes for pre-harvest sprouting in wheat (Strader et al., 2004).

2.5. Selected examples of reverse-genetics functional analysis using insertion lines Although summarizing the identified functions of genes based on existing reports of reverse-genetics analysis in rice and maize might be a topic of its own, here we are attempting to exemplify well known examples of developmental mutants described in those species. 2.5.1. Functional analysis using T-DNA insertion line Although the flanking sequences to T-DNA insertion events are screened regularly for association with characteristic phenotypes, here we have made an attempt to summarize the benefits realized based on genes with lethal null alleles involved in vegetative and reproductive development in rice. In a reverse-genetics approach, Jung et al. (2003) screened 1995 T-DNA pools for rice chlorophyll-deficient mutants and identified large subunit of Mg-chelatase gene as a key enzyme in the chlorophyll branch of tetrapyrrole biosynthesis pathway. T-DNA knockout alleles, fon1-3 and fon1-4 showed alterations in stem cell proliferation producing fewer tillers and secondary rachis branches, as a result it is concluded that FON1 gene, an ortholog of clavata1 from Arabidopsis, not only affects reproductive organs but also vegetative tissues (Moon et al., 2006). Mutant carrying T-DNA tag in cysteine protease gene OsCP1 showed significant defect in pollen development (Lee et al., 2004). Similarly, Jung et al. (2005) reported T-DNA insertion in tapetum1 gene, which caused male sterility as a consequence of the failure of tapeta differentiation during meiotic stage. To clarify the role of genes expressed during early seed development, Kang et al. (2005) isolated various alleles of floury endosperm-4 from rice mutant collection carrying T-DNA insertion in pyruvate orthophosphate dikinase (PPDK ) and suggested its possible involvement in modulation of carbon metabolism during grain filling. In yet another study in rice aimed at studying the role OsSSIIIa/Flo5 gene during starch synthesis, analysis of null mutants created using T-DNA activation tagging approach revealed that the above gene plays an important role in generating relatively long chains of endosperm starch (Ryoo et al., 2007). Further, transgenic lines carrying T-DNA insertion sites among transcription factor families are being systematically determined by using reverse-genetics approaches in order to identify functions of Myb, WRKY, HD-ZIP, and Zinc-finger families (Sallaud et al., 2004). Interestingly, a T-DNA knockout line of OsMADS3 showed

378

H. S. Balyan et al.

phenotype of homeotic transformation of stamens into lodicules without in any way affecting the normal carpel development (Yamaguchi et al., 2006). 2.5.2. Functional analysis using transposon insertion lines Initial success obtained in the field of transposon mutagenesis in Arabidopsis has been summarized by reviewing function of more than 60 genes involved in plant development (May and Martienssen, 2003). Here we are summarizing initial outcome of transposon mutagenesis in two cereal crop plants, namely rice and maize, by covering examples of genes involved in metabolic processes during development of vegetative and reproductive organs. For example, in rice, mutation caused by insertion of transposon (nDart1) in chlorophyll protease gene (OsClpP5 ) leads to the formation of mutant dwarf allele thl-m with leaf variegation (Tsugane et al., 2006). In another example, insertion of the retrotransposon Tos17 in rice PAIR2 gene results in homologous chromosome pairing during meiosis (Nonomura et al., 2004). Similarly, retrotransposon (Tos17) based insertion mutants created among three different members of cellulose synthase genes (OsCesA4, OsCesA7, and OsCesA9) led to a dramatic decrease of cellulose content resulting into changes in secondary cell wall formation (Tanaka et al., 2003). Insertions have also been found in genes responsible for gametogenesis (Nonomura et al., 2003) and seed set (Tabuchi et al., 2005). In the study by Nonomura et al. (2003), retrotransposon-tagged mutation in rice caused developmental arrest of pollen mother cells, which eventually suggested that MSP1 is essential for restricting the number of cells entering into male and female sporogenesis and to initiate anther wall formation. The putative function of glutathione synthetase 1;1 in rice has been revealed by characterization of Tos17 insertion mutants. Under normal nitrogen conditions, the glutathione synthetase 1;1 mutant plants showed retardation in growth rate and reduced grain filling. Recently, Fujita et al. (2006) characterized important role of starch synthase I for its involvement in amylopectin chain synthesis during grain development in rice using retrotransposon-tagged mutants. In contrast to wide utilization of Ac/Ds, En/Spm, and Tos17 elements in generating rice insertional mutants, genome-wide mutagenesis in maize is accomplished by insertions of Mu family elements. Initial success obtained in the field of transposon mutagenesis in maize has been summarized by reviewing function of approximately 30 genes involved in plant development (May and Martienssen, 2003). In this review, we are providing recent update on application of Mutator transposon insertions in maize to study mutant phenotypes that occur during seed development. Mutator transposon insertions in ZmPRPL35-1 encoding large subunit of plastid ribosomes resulted in alterations of embryo development without affecting endosperm development (Magnard et al., 2004). In these lines, the disturbances in basic events of pattern formation in embryo led to the development of embryo devoid of scutellum. Characterization of several alleles of Mutator transposon insertions

Mutagenesis and Functional Genomics in Cereals

379

in transcription factor IIS depicted cracked mature kernels in maize (da Costa e Silva et al., 2004). Although ABA is well known to influence seed dormancy events and prevent precocious germination during seed maturation, functional characterization of key ABA biosynthetic genes expressed during seed maturation in endosperm and embryo is lacking. Recently, Suzuki et al. (2006) identified allelic mutations in molybdopterin synthase gene (viviparous 15 locus) from Mu lines, which is expressed in both endosperm and embryo tissues during seed maturation. The mutant exhibited viviparous phenotype due to lack of molybdenum cofactor, which is an essential component required for ABA biosynthesis. The utility of Mutator mutant collections for the discovery of genes integral to altered storage products has been revealed in maize. Characterization of starch branching IIa::Mu mutants shows reduced levels of starch branching IIa transcripts with reduced branching activity of glucans in leaves compared to seeds (Blauth et al., 2001). Similarly, mutations in the maize sugary2 gene coding for starch synthase IIa resulted in short glucan chains of amylopectin (Zhang et al., 2004). Reverse-genetics screening of maize Mutator insertion mutant collections turned out to be a useful approach in identifying lines conferring reduced phytic acid content in seeds (Shi et al., 2003, 2005), which is a part of long desired goal of crop improvement. The Mu insertion in maize inositol phosphate kinase (ZmIpk) gene resulted in approximately 30% reduced phytic acid content in seed (Shi et al., 2003). Recently, one of the FSTs from UniformMu populations was shown to co-segregated with the rough endosperm (rgh) seed mutant phenotype that shows highly reduced grain fill in maize. In this case, the insertion disrupts the open reading frame of a predicted 6-phosphogluconate dehydrogenase (6PGDH) gene at the Pgd3 locus, suggesting an important role for 6PGDH enzyme in seed development (Settles et al., 2007). It is clear from the above set of examples that so far insertion resources are available in only a few of the several cereal species, and therefore, T-DNA/transposon-based reverse-genetics analysis is not possible in all the cereal species at the moment. Transposon-based gene tagging in heterologous systems has several drawbacks mainly due to occurrence of multiple copies causing similar difficulties as associated with multiple T-DNA insertions. Transposon insertions are unstable and may result in reversions or footprints in tagged genes leading to mutant phenotypes without a molecular tag in the mutated gene. Furthermore, some transposons such as En/Spm transpose via a replication mode resulting in increasing number of active transposons (Greco et al., 2004). These may cause a multitude of mutant phenotypes, which are very difficult to track specially in cereal species having large genomes with huge proportions of repetitive sequences. Hence, we need to explore the potentialities and possible applications of other more general reverse-genetics technologies such as TILLING and DEALING as alternatives in cereal species.

380

H. S. Balyan et al.

3. Non-Transgenic TILLING, DEALING, and DeleteageneTM Approaches The most direct approach to discover the function of a given gene is to induce knock-out (loss of function) mutations or to create an allelic series in the gene using chemical and physical mutagens and analyze the consequences. Over the past few decades, chemical and physical mutagens were used to create novel genetic variation in various cereal crop plants leading to the development of more than 1008 mutant crop varieties (http:// www-infocris.iaea.org/MVD/) including well characterized and large mutant collections (~10,000) in barley (Von Wettstein-Knowles, 1992; http://www.arsgrin.gov/npgs/index.html; http://barley.ipk-gatersleben.de/ ebdb/). However, with the availability of mutants with altered phenotypes, most often the gene-function relationship could not be established using forward-genetic analysis, as it requires positional cloning of the gene(s) and their characterization ( Wu et al., 2005). Positional or map-based cloning is a slow, tedious, and time-consuming process that requires mapping of the mutated locus with respect to co-segregating molecular marker(s) that can then be used to initiate chromosome walking, sequencing, and eventual cloning of the gene responsible for the altered phenotype. Therefore, mapbased isolation of mutated genes for finding gene-function relationship can not be routinely used. As an alternative to the above and as a result of the accumulation of large amount of DNA/EST sequence data in public domain, high-throughput reverse-genetics strategies coupling induced mutagenesis, molecular detection of mutations (SNPs/deletions), and association of the mutants with altered phenotypes have been developed (Liu et al., 1999b; Nadeau and Frankel, 2000). These new approaches have been termed TILLING (McCallum et al., 2000), DEALING (Kianian, unpublished), and DeleteageneTM (Li et al., 2001, 2002a). Recent developments beginning with the creation of suitable mutagenized populations, the technological advances for the detection of mutations using the above approaches and their successful applications in different cereal crops are reviewed below.

3.1. Development of mutagenized population 3.1.1. Choice of mutagen for TILLING, DEALING, and DeleteageneTM Chemical and physical mutagens are readily applicable to all plant and animal species (Kodym and Afza, 2003). As a result, these mutagens have been successfully applied to induce mutations in a variety of plant species including cereals. Some of the most important and widely used mutagens in plants are summarized in Table 3. Based on their mode of action, the

381

Mutagenesis and Functional Genomics in Cereals

Table 3 Important physical and chemical mutagenic agents and their mode of action

Mutagenic agents

Category

Mode of action

Ultra-Violet (UV) rays

Physical mutagen

X-rays, gamma rays

Physical mutagen

Fast neutrons

Physical mutagen

2-amino purine, 5-bromouracil

Chemical agent

Ethidium bromide, proflavin, acridine orange

Chemical agent

Ethyl methane sulfonate (EMS)

Chemical agent

Pyrimidine dimer formation and error during DNA replication Ionize water and other organic molecules forming radicals causing breaks in DNA strands and alterations in purine and pyrimidine bases Extremely damaging to DNA Base pairing mistake resulting into A/T to G/C transitions Reduce fidelity of DNA replication by intercalating between bases, causing insertions, deletions or additions that frequently induce frameshift mutations Modification of guanine (G) resulting into G/C to A/T transition

Used in TILLING/ DEALING/ DeleteageneTM

NA

DeleteageneTM (gamma rays)

DeleteageneTM NAa

NA

TILLING

(continued)

382 Table 3

a

H. S. Balyan et al.

(continued)

Mutagenic agents

Category

Mode of action

N-ethyl-Nnitrosourea (ENU)

Chemical agent

Nitrous acid

Chemical agent

Hydroxylamine

Chemical agent

Diepoxybutane (DEB)

Chemical agent

Sodium azide

Chemical agent

Trimethylpsoralen

Chemical agent

Modification of thymidine (T) resulting into A/T to T/A transversion Modification of cytosine (C) resulting into A/T to G/C transition Modification of cytosine (C) resulting into G/C to A/T transition Produces inter and intra-strand DNA cross links; cause large deletions and known as deletogen Precise mechanism of action unknown Produces inter and intra-strand DNA cross links in conjunction with UV; cause large deletions and known as deletogen

Used in TILLING/ DEALING/ DeleteageneTM

NA

NA

NA

DEALING

TILLING

DEALING

Not applied.

mutagens could be classified into those inducing point mutations (single base pair change creating single nucleotide polymorphisms i.e., SNP), chromosomal aberrations, and sub-gene to gene level (kb) deletions; deletion causing mutagens are popularly termed deletogens.

Mutagenesis and Functional Genomics in Cereals

383

Among chemical mutagens causing point mutations, ethyl methane sulfonate (EMS) has received a universal acceptance. EMS is considered the most efficient mutagen for small genes (Greene et al., 2003). It induces high density of G/C-to-A/T transitions with little or no chromosomal aneuploidy/aberrations, reduced fertility or dominant lethality (Ashburner, 1990). In a recent reverse-genetics study involving 192 genes in an Arabidopsis population mutagenized by EMS, 99% induced changes were confirmed as G/C-to-A/T transitions (Greene et al., 2003). In barley and rice also, nearly 70% transition from G/C-to-A/T were noticed (Caldwell et al., 2004; Till et al., 2007). Until now, most of the TILLING reverse-genetics resources involving detection of SNP have been created using EMS (for details, see Section 3.1.3.). Only recently, sodium azide (NaN3), an A/T-to-G/C transition inducing chemical mutagen (Olsen et al., 1993) was used either alone for the creation of barley TILLING populations (http://www.intl-pag.org/13/abstracts/ PAG13_P081.html) or in combination with methyl nitrosourea (MNU) in rice (Till et al., 2007). However, a number of other point mutations inducing chemical mutagens are available with the potential of producing different constellations of mutations that lead to different projected codon usage and phenotypes. These include the transversion inducing (A/T-to-T/A) N-ethylN-nitrosourea (ENU) (Stemple, 2004) and the transition inducing nitrous acid (A/T-to-G/C), and hydroxylamine (C/G-to-T/A). Therefore, it would be most appropriate to try and use other mutagenic agents with differing mechanisms besides EMS in the development of TILLING resources to maximize gains. With a view to induce deletion mutations in Caenorhabditis elegans, the first organism used for DEALING studies, Jansen et al. (1997, 1999) and Liu et al. (1999) used different chemical deletogens [ethylmethane sulfonate (EMS), ethylnitrosourea (ENU), diepoxyoctane (DEO) and ultraviolet-activated trimethylpsoralen (UV-TMP)] as deletogens. The EMS and ENU induced deletion mutations at comparable rate while DEO and UV-TMP induced deletion mutations at a higher rate (see Liu et al., 1999). The deleted segments varied from 700 to 2900 bp (average ~1400 bp) when screening was performed in a 3 kb window (Liu et al., 1999). Most of these intragenic deletions represented null or severe mutations resulting in phenotypes that resembled wild-type or dramatic/lethal defects. Although, at present, extensive data on the efficiency and effectiveness of chemical deletogens is lacking in crop plants, DEB and UV-TMP are being increasingly applied in cereal crop plants. For instance, DEB has been used to induce mutations in the Xa21 gene of rice changing disease resistance response against X. oryzae pv. oryzae (Wang et al., 2004) and to create mutagenized population of indica rice cv. IR64 for conducting reverse-genetics analysis at IRRI, Philippines (Wu et al., 2005). DEB and UV-TMP are also being used for developing diploid wheat (Triticum monococcum) and basmati rice reverse-genetics DEALING-resources through a collaborative research project involving one University from the USA (led by

384

H. S. Balyan et al.

S. Kianian, North Dakota State University, Frago) and two institutions from India (H. S. Balyan, Ch. Charan Singh University, Meerut and H. S. Dhaliwal, IIT, Roorkee). For a long time, physical mutagens such as fast neutrons have also been shown to cause deletion mutations in plants. Fast-neutron treatments were shown to induce mutations efficiently in A. thaliana (Koornneef et al., 1982). Subsequent studies related with molecular characterization of Arabidopsis ga1-3 (Sun et al., 1992) and tomato Prf-3 (Salmeron et al., 1996) genes amply demonstrated the ability of fast-neutron irradiation to induce deletion mutations. However, it is only as late as 2001 that the fast-neutron deletion mutagenesis-based reverse-genetics methodology described as DeleteageneTM was reported in model plants like Arabidopsis and rice for the first time (Li et al., 2001, 2002). Fast-neutron and another important physical mutagen, gamma rays, have been recently used to develop reverse-genetics resources in indica rice cv. IR64 at IRRI, Philippines (Wu et al., 2005) and for forward-genetics studies of wheat in Australia (Singh et al., 2006). However, data suggest that these irradiation methods can produce large deletions (range of 100 kb) that may be difficult to detect by simple PCR methods in non-sequenced genomes. We feel that a wider application of fast neutron based reverse-genetics approach for discovering the biological functions of genes in a diverse set of plant species including cereal crops will soon be a reality. 3.1.2. Mutagen treatment and population size Most of the important cereal crop plants including wheat, barley, and rice are self-fertilizing species while only a few species such as maize and rye are cross-fertilizing. The mode of reproduction results in homozygosity and homogeneity in the self-fertilizing species while heterozygosity and heterogeneity in cross-fertilizing species. Thus, the mode of reproduction and its consequences have important bearing on the possible methods of mutagen treatment and post-treatment handling of the population for detection of induced mutations. In seeds, the embryo is already differentiated with one to several genetically effective cells. Therefore, mutagen treatment of the seed gives rise to chimeric M1 plants and segregation the mutations are observed in M2 generation (progeny from self-fertilization of individual M1 plants) (Fig. 3A). In recent TILLING and DeleteageneTM studies involving self-fertilizing species like Arabidopsis, wheat, rice, and barley, a similar strategy was followed for the detection of base substitution and deletion mutations (Caldwell et al., 2004; Li et al., 2001; McCallum et al., 2000a,b; Slade et al., 2005; Till et al., 2003). In this scheme, DNA is isolated from individual M2 plants from a family and pooled for subsequent mutation detection. A part of the M3 seed harvested from M2 plants is stored for future use, and the remaining seed may be used for rearing M3 progenies for identification/confirmation of mutations and subsequent study. A similar

A

B Seed treatment with chemical mutagen

C Pollen treatment with mutagen

Raising of M1 population (Plants will be +/+; +/−)

Pollination of mother variety plants with mutagen treated pollens

Pollen culture

Raising of M2 population

Harvesting M1 seed and raising of M1 plants (Plants will be +/+ and +/−)

Raising of haploid (2n = 1x) plants (Haploids will be + or −) and treatment with colchicine

Isolation of DNA from individual plants

Raising doubled haploid (DH0) plants (2n = 2x) (Plants will be +/+,−/−)

Individual M1 plant progenies (Progenies will be +/+, +/−, −/−)

M1 single seed descent (SSD) population (Plants will be +/+, +/−, −/−)

Preparation of 8-fold pools of DNA for mutation detection by TILLING

Collection and storage of M3 seed; use a portion of harvested seed for raising M3 to identify and confirm mutant progenies

Raising of M2 population; collection and storage of M3 seed; use a portion of harvested seed for raising M3 to identify and confirm mutant progenies

Collection and storage of DH1 seed; use a portion of harvested seed for raising DH1 to identify mutant progenies

Creation of database for data on phenotypes and molecular mutations

In silico association of altered phenotypes with mutations detected by TILLING

Figure 3

Flow chart showing steps involved in three different schemes of creation of mutagenized populations for TILLING.

386

H. S. Balyan et al.

strategy may be used with a cross-fertilizing species provided seed from homozygous inbred lines instead of open pollinated source is used as the starting material for mutagen treatment. Using seed from an open pollinated source as the starting material for mutagenic treatment would complicate matters related to the origin of mutations. For generating TILLING resources in maize, a cross-fertilizing species, Till et al. (2004a) recently used a modified strategy involving mutagen treatment of the pollen rather than seeds (Fig. 3B). The treated pollens were used to pollinate the donor or test-cross parent. The fusion of the mutation-carrying male gamete with the wild-type female gamete would produce M1 seed that would be heterozygous at the concerned locus. The DNA from M1 plants is pooled for TILLING, and the selfed M2/M3 seed is used for subsequent phenotyping of mutations. In the past, pollen mutagenesis coupled with pollen culture was recommended as an in vitro-mutagenesis method for crop improvement (Castillo et al., 2001; Maheshwari et al., 1980; Maluszynski et al., 1995, 2003; Taji et al., 2001; reviewed in Forster and Thomas, 2005, and Szarejko and Forster, 2007). This approach so far has not been utilized in experiments related to reverse-genetics analysis, though the technique, once perfected, has tremendous potential as homozygous mutant doubled haploid (DH) plants may be obtained in a single generation (Fig. 3C), removing the complexity introduced by heterozygosity. As the DHs would represent fixed inbred lines, molecular screening, phenotypic analysis, identification of mutations as well as storage, and maintenance of seed stocks would be a lot easier. However, the DH mutation population has a disadvantage in that the lethal mutations will not be maintained and will be lost. Such mutations hence will not be available to examine the basic biological function of mutated genes conferring lethality. While developing mutagenized populations for reverse-genetics analysis, the determination of the optimum mutation frequency and thus appropriate size of a suitable mutagenized population is crucial. The right balance between the chances of detecting a mutation in the target gene and the size of the population that can be handled with ease within the available logistic limits, for a given mutagen treatment may be empirically determined. Mutagen treatment is usually applied in such a manner that it produces sufficient lethality while allowing sufficient fertility, so that a high frequency of induced mutations may be recovered in the mutagenized population. To assess suitable mutagen dose, mutation frequencies in different plant species were indirectly estimated in the past on the basis of seedling lethality, chlorophyll deficiency, seed set, etc. (Emery, 1960; Gilchrist and Haughn, 2005; Kleinhofs et al., 1978; Li and Redei, 1969; Prina and Favert, 1983; Stemple, 2004). These are all phenotypic measures, which fail to provide direct estimation of mutation frequency at molecular level, which may be more relevant for the reverse-genetics experiments. Although precise estimates of induced global mutation frequency at the nucleotide level of

387

Mutagenesis and Functional Genomics in Cereals

a plant genome may be obtained directly through molecular analysis, routine methods providing comprehensive estimates of mutation frequency are not available at present. Though AFLP analysis (gain or loss of fragments), which in a single 96-well experiment has the potential of screening 160 kb of sequence has been tried to assess EMS, DEB, and gamma-irradiation induced mutation frequency in barley, detailed results have not been reported (see Waugh et al., 2006). Nucleotide level variation is a more desirable guide for the quantitative measure of the mutation frequency to future reverse-genetics studies in plants, but the available data should be treated with some degree of caution as the estimates are based on limited experimental information. Recent reports suggest that density of mutations range from 1/24 kb in common wheat, a polyploid species, to 1/Mb in barley, a diploid species (Table 4). In two other diploid species (maize and rice), density of mutations was reported at 1/500 kb. In an Arabidopsis mutagenized population, mutation density of 1/235 kb or approximately 4 point mutations per eight-fold pool (representing 768 plants = 96 wells pooled DNA of 8 plants per well) was reported (Greene et al., 2003). These data clearly support the long held belief that polyploid species better tolerate mutational load than the diploid species possibly due to their better genetic buffering capacity afforded by multiple homoeoloci (Stadler, 1929). This would also mean that to discover similar number of mutational events in diploids as in the polyploids, one would probably need to analyze a much larger mutagenized populations in diploid species. Recent reports suggest working with relatively smaller M2 populations of 8600 plants in barley (Caldwell et al., 2004). In case of maize, initially two smaller TILLING populations were created and two more populations are recently developed (see Weil and Monde, 2007). These two available populations generated using EMS include one at University of Minnesota, USA with 2370 mutant lines of B73, the inbred line used for maize genome sequencing, and the other at Purdue University, USA that include 1276 line of W22 inbred. The two other populations that have been created also belong to B73 (at Iowa State University, USA) and W22 (at Purdue University, USA) inbred lines. However, current efforts are on to develop Table 4 Density of mutations determined in cereal crop species Species

Mutation density

Reference

Triticum aestivum Triticum durum Zea mays Oryza sativa Oryza sativa Hordeum vulgare

1/24 kb 1/40 kb 1/500 kb 1/500 kb 1/497–1/500 kb 1/Mb

Slade et al. (2005) Slade et al. (2005) Till et al. (2004a,b) Wu et al. (2005) Till et al. (2007) Caldwell et al. (2004)

388

H. S. Balyan et al.

much larger barley-TILLING resources following treatment of 60,000 seeds with NaN3 (S. Salvi, University of Bologna, Italy; Ist International GABI-TILL Workshop, IPK, Gatersleben, Germany) and 80,000 seeds with EMS (Nils Stein, IPK, Gatersleben, Germany; Ist International GABI-TILL Workshop, IPK, Gatersleben, Germany). We believe that the remaining diploid cereal species also require large working M2 populations, which would become available in due course. However, for genome-wide Arabidopsis TILLING Project (ATP), McCallum et al. (2000a) suggested that 10,000 M2 plants would be sufficient for obtaining the desired mutation using just a single primer pair per gene. Considering the low rate of mutation induction by deletogens and fast neutron, the required population size has been estimated to be in excess of 100,000 plants for comprehensive coverage (>95% probability of detecting a mutation in a target gene) if the mutation rate is considered independent of the genome size (Li et al., 2001, 2002). At IRRI, Philippines, 60,000 IR64 mutants have already been generated using both physical and chemical mutagens and 38,000 independent lines have been advanced to M4 generation for evaluation (Wu et al., 2005). To allow availability of mutants, a mutant database has been created in International Rice Information system (IRIS; http://www.iris.irri.org). 3.1.3. Preparation of DNA pools Creation of appropriate size DNA pools for high-throughput detection of induced mutations is crucial for functional analysis of a gene through reverse-genetics methods. Purity of the DNA preparation is also critical for obtaining clean gels free from backgrounds. The average size of good quality DNA should be at least 15 kb and it should be stable under standard storage conditions (Comai and Henikoff, 2006). DNA is isolated from the M2 generation of the seed treated populations and from M1 generation of the pollen treated populations (Fig. 3A–C). DNA from individual plants is pooled at the same concentration to achieve balanced representation. The Seattle TILLING Project (STP) or formerly the ATP, which help establish Maize TILLING Project at Purdue University (http://genome.purdue.edu/ maizetilling) ( Weil and Monde, 2007) routinely uses eight-fold pools that are arranged in the bi-dimensional scheme so that the display of two mutations in the same pools identifies the mutant individual. To improve the high-throughput efficiency of TILLING for the identification of sugar and starch mutants in Arabidopsis, a modified pooling strategy involving 64-fold super pools followed by 8 8 sub-pools is being used (G. Strompen and J. Lunn, Ist International GABI-TILL Workshop, IPK, Gatersleben, Germany, 2006). Although desirable for high-throughput analysis, success in detecting mutations in deeper pools of the large genome cereal crop species is awaited. In C. elegans adduct lesion experiments 96-well plate pools of DNA were used (Liu et al., 1999). In DEALING experiments on diploid wheat (T. monococcum), several pooling strategies have demonstrated

Mutagenesis and Functional Genomics in Cereals

389

effectiveness for screening of 1000-fold pools (S. Kianian, unpublished data) and a poison–primer approach, utilized in C. elegans for detecting small deletions, for screening at 5000-fold pools (Fig. 4, data from Riera-Lizara-zu). It is apparent that the latter strategy of using large DNA pool size will allow handling of large DEALING/ DeleteageneTM resource populations that will compensate for the lower frequency of induced deletion mutations by fast neutron or deletogens. Similarly, large DNA pools can also be applied to DeleteageneTM but strategy for detection may have to be altered to account for larger size deletions generated by agents.

3.2. Mutation detection technique in TILLING For the first time, TILLING, as a reverse-genetics tool was used in Arabidopsis by McCallum et al. (2000a,b) and in Drosophila melanogaster by Bentley et al. (2000). TILLING involves PCR amplification of DNA pools from EMS mutagenized populations (Fig. 3) using gene-specific primers and targeted identification of induced point mutations (base substitutions/ SNPs) in the gene of interest. The induced point mutations can lead to a broader range of effects, including hypomorphic, hypermorphic, and neomorphic effects (Stemple, 2004). The nature of the induced lesion has also implications for the experimental strategy to identify the type and genomic location of the mutation. In the original TILLING experiments, McCallum et al. (2000a,b) made use of denaturing high-performance liquid chromatography (DHPLC) to detect the base pair changes by heteroduplex analysis. However, the need for newer mutation detection methods was felt as DHPLC is unsuitable for high-throughput analysis and fails to provide the location of the detected mutation without sequencing of the whole amplified fragment. To overcome these drawbacks, new TILLING protocol involving CELI enzymatic mismatch cleavage of heteroduplex DNA (Oleykowski et al., 1998) on a LI-COR gel analyzer system (Lincoln, NE; Middendorf et al., 1992) was reported by Colbert et al. (2001). This protocol was first used by the STP for screening EMS mutagenized Arabidopsis populations (Till et al., 2003). The CELI endonuclease, which belong to the S1 nuclease family of single strand-specific (sss) nucleases, is found in celery, Apium graveolens, (celery enzyme is known as CELI) and many other plant species (Oleykowski et al., 1998). CELI cleaves DNA to the 30 side of the mismatches and loops out in heteroduplexes between wildtype and mutant DNA, leaving duplexes intact. In this new method, the PCR products amplified from pooled DNA samples using different IRD700 and IRD800 labeled-primers are first incubated with CELI and then the cleaved products are run and visualized on LI-COR gels. At present, both home-made CELI enzyme preparations and commercial mutation detection kits (SURVEYOR Mutation Detection Kit, Transgenomic) that uses CELI are being used (Qiu et al., 2004). In a few protocols,

390

H. S. Balyan et al.

800 bp InDel

A Deletion

193 bp

Mutant

Wild type e2

e1 403 bp 1000 bp

External

B

A

Internal

D

E

Wild type DNA + + − − Deletion DNA − − + + Competing oligo + − + − 1000 bp 800 bp 600 bp

1:10,000

1:5000

1:1000

1:500

1:100

1:1

Unmixed

1:10

Mutant: wild type ratios

B

A

C

+ + + ++ + + + +

++

+ + +

+ + + ++ + + + +

++

+ + +

−+

− +

+

−+

−+

−+

−+

− Full length primers and poison primer

B

775 bp 300 bp

Internal primers

Figure 4 (A) Schematic representation of the wild-type and mutant barley (Hordeum vulgare) GBSS I (waxy) structural gene. Exons are represented by filled boxes. The primers used for PCR are represented by arrows. (B) Detection of the barley waxy deletion using nested PCR with a poison primer. (A) First amplification step of wildtype, mutant (with the waxy deletion) and various dilutions of mutant to wild-type DNA with the A and C external primers with and without the competing or poison primer B. With no poison primer (only primers A and C), the wild-type DNA produces a 1000-bp product while the deletion or mutant DNA produces an 800-bp fragment. With the poison primer (primers A, B, and C), the wild-type DNA produces a 600-bp product that results from the preferential amplification of DNA by the poison primer B and the external primer C. With the poison primer mutant DNA still produces the expected 800-bp product because there is no binding site for the poison primer B. With the exception of the 1:1 dilution of mutant to wild-type DNA, poison primer B also outcompetes the external A primer in the various dilutions. (B) Second amplification

Mutagenesis and Functional Genomics in Cereals

391

CELI has been replaced by mung bean nuclease. Recently, Endo-1, a genetically engineered endonuclease, manufactured by Serial Genetics, France (www.serialgenetics.com) has also been used as an alternative to CELI for mutation detection following heteroduplex mismatch cleavage in several plant species (A. Bendahmane, Ist International GABI-TILL Workshop, IPK, Gaterlseben, Germany). Endo-1 has the advantage of recognizing all the different types of mismatches including base-substitutions and INDELs with similar efficiency. The CELI/Endo-1 endonuclease-based mutation detection assay has several advantages including simplicity of the assay, home-made/commercial availability of enzymes, and insensitivity to the location and types of different mismatches (see Yeung et al., 2005). Protocols for high-throughput TILLING and detection of mutations have been standardized. For example, a single TILLING run interrogates a total of ~750,000 base pairs (eightfold pooling 1000 bp 96 lanes = 768,000 bp) and detects mutations equivalent to ~3000 sequencing lanes (assuming that a 1 kb fragment requires some four full sequencing lanes with four primers for reliable detection of heterozygotes). The above TILLING protocol including the identification of the mutant in each pool and its sequencing is estimated to be an order of magnitude more economical than full sequencing (Henikoff and Comai, 2003). A summary of the currently available and emerging technologies (modified following Comai and Henikoff, 2006) for high-throughput discovery of mutations and polymorphism is given in Table 5. It may be realized that all these technologies require very expensive equipments where easy access may not be available to all research laboratories. Following discovery of mutation(s) in a pool, DNA from the individual plants of the pool is re-screened, and the mutant plant and the approximate position of the mutation are identified. To complete the analysis, sequencing of the amplified target from the mutant plant is carried out to identify the base pair change (stop codon or mis-sense mutation) and consequent changed phenotype. Freely accessible web-based software to aid TILLING has also been developed. One of these software tools is CODDLE (for Codon Optimized to Detect Deleterious LEsions; http://www.proweb.org/cod/dle/). CODDLE help in choosing the best region of the target gene and the optimal primers to amplify PCR products of up to 1.5 kb that is suitable for CELIbased mismatch–cleavage method for TILLING (Till et al., 2003). Although, CODDLE has served well with small genome model plant Arabidopsis, newer procedures and modified software may need to be developed to carry out the step of DNA from the first PCR step using the D and E internal primers. The wild-type template produces a 775-bp product whether or not the poison primer was used in the first PCR step. Similarly, the mutant template produces a 300-bp product with or without the use of the poison primer. Without the poison primer, we begin to lose the ability to detect the deletion at mutant: wild-type ratios of 1:100. With the poison primer, the deletion can be detected at a ratio of 1:5000.

392

H. S. Balyan et al.

Table 5 Comparison of presently available and emerging technologies for mutation detection through TILLING

Name of technique

DNA resequencing

Method of mutation discovery

Capillary sequencing Chromatographic heteroduplexes analysis

Denaturing high performance liquid chromatography (dHPLC) Denaturing gradient Electrophoretic gel electrophoresis heteroduplexes (DGGE) analysis Single-strand Electrophoretic conformational heteroduplexes polymorphisms analysis (SSCP) Chemical detection Cleavage at the mispaired site using chemical reagents CELI/ENDO 1 Cleavage at the enzymatic mispaired site detection using enzymes Pyrosequencing

Multiplex capillary heteroduplex analysis

Microarray analysis

Sequence determination by measuring the amount of pyrophosphate released and by polony sequencing on 454 Fluorescent-based conformation sensitive heterodulpex analysis Hybridizationbased detection of mismatches

Able to identify location, identity or number of mismatches

References

Yes

–

No

Underhill et al. (1997)

No

Alharbi et al. (2005), Li et al. (2002) Gross et al. (1999)

No

Yes

Smooker and Cotton (1993)

Yes

Oleykowski et al. (1998), Till et al. (2004a,b) Langaee and Ronaghi (2005), Margulies et al. (2005), Shendure et al. (2005)

Yes

Yes

Ganguly et al. (1998), Hoskins et al. (2003)

No

Cutler et al. (2001)

Mutagenesis and Functional Genomics in Cereals

393

functions of CODDLE by taking into account the unique characteristics of the major cereal crops such as their large genome size and deviations from the optimal G + C content. Another freely available web-based tool, PARSESNP (for Project Aligned Related Sequences and Evaluate SNPs; http://www. proweb.org/parsenp/), has been developed to help TILLING community in phenotype analysis. PARSESNP help evaluate mutations and polymorphisms, providing mapping information, predictions of deleterious effects on an encoded protein, changed restriction sites and other information to facilitate phenotype analysis (Taylor and Greene, 2003).

3.3. TILLING initiatives in cereals In the first year of its public operation, the ATP (http://tilling.fhcrc. org:9366), discovered, sequenced, and delivered >1000 mutations in >100 genes ordered by researchers (Till et al., 2003). In terms of its application in functional genomics as well as a non-transgenic method of crop improvement, the success of TILLING in cereal crop plants that have more complex genomes with much less sequence information has just begun. Examples of application of TILLING among cereals have become available in at least wheat, barley, maize, rice, and sorghum. Recently, using an elegant TILLING experiment, the first economically important wheat with near waxy phenotype (waxy starch devoid of amylose) was produced by Slade et al. (2005). Using primers for the different granule bound starch synthase (GBSS) I gene (or waxy gene) homoeoloci in both the hexaploid and tetraploid wheat, they identified ~250 novel alleles with severe to no effect. Ninety four of these mutations were predicted to alter the encoded waxy gene product. Two waxy loci with severe mutations along with a naturally occurring deletion at the third locus were combined to obtain a triple homozygous mutant line with waxy phenotype. In yet another study in wheat using modified TILLNG technique involving agarose gel electrophoresis and unlabeled primers, 100 mutants at two (GBSS) or waxy homoeoloci on chromosomes 7A and 7D involved in amylose synthesis and 20 independent mutations from the two puroindoline genes (Pina and Pinb) responsible for kernel texture were recovered (Sharp et al., 2006). These examples demonstrate that TILLING has the potential for creating novel genetic variability, crop improvement, and also it could be successfully extended to polyploid crop plants with large genome size. Recently, TILLING has also been used in another important cereal crop, barley, (Caldwell et al., 2004). In this case, two structured EMS-mutagenized populations of barley cv. Optic were produced to allow genome-wide forward-genetics and reverse-genetics analysis. These populations were screened for mutations in Hordoindoline-a (Hin-a) and the Hordeum vulgare Floral Organ Regulator-1 (HvFor1) ortholog following cleavage of heteroduplex with CELI. Six induced alleles for HvFor1 and four for Hin-a were identified and

394

H. S. Balyan et al.

confirmed following sequencing. Screening of 750 pollen mutagenized maize plants for 11 different genes in 1 kb window in a TILLING experiment allowed detection of 17 independent mutations (Till et al., 2004b). These included promising allelic series for a chromomethylase gene that has been previously implicated in non-CpG DNA methylation. In yet another effort, Maize TILLING Project (http://purdue.edu/maizetilling) has identified 319 mutations in 62 genes representing 76 kb of sequence. The data of the TILLed genes is put in the public domain (http://genome.purdue.edu/maizetilling; http://maizegdb.org) after 160 days and the sequence is submitted to GenBank (Weil and Monde, 2007). With the creation of a very large collection of IR64 mutants in rice at IRRI, Philippines (http://www.iris. org), screening for 10 genes using TILLING has been carried out ( Wu et al., 2005). Of the 10 genes, independent mutations, which were later confirmed through sequencing, were detected in two genes including pp2A4 encoding serine/threonine protein phosphatase catalytic subunit and cal7 encoding callose synthase. Two pilot scale rice mutant libraries were also recently developed and 57 nucleotide changes in 10 mutagenized genes (Os1433, OsBZIP, OsCALS8R, OsDREB,OSEXTE, OsMAPK, OsPITA, OsRIA, OsRPLDI, and OsTPSI ) were discovered (Till et al., 2007). Efforts for the discovery of induced allelic variation through TILLING in genes of interest have also been initiated in crops like sorghum, which is an important crop for the dry land areas around the world (http://www.ars.usda.gov/ research/publications/publications.htm?SEQ_NO_115 = 180731). For this purpose, EMS-populations have already been developed and alleles for starch synthesis and grain quality genes are being searched using high-throughput capillary electrophoresis (Cross et al., 2006).

3.4. Eco-TILLING Identification of new alleles of the known genes impacting agronomic value of the trait is no less important than the discovery of function of any newly discovered gene. Therefore, as an extension of TILLING strategy, Ecotilling was developed for the detection of polymorphisms (SNP and haplotype structure in a target gene) in natural populations (Comai et al., 2004). The technological details of Ecotilling are similar to TILLING except for using a modified DNA pooling strategy and the detection of several-fold higher level of polymorphism. DNA pools for Ecotilling are made by mixing DNA from the queried individual and the reference sample in 1:1 ratio instead of using several-fold DNA pools in the case of TILLING. Software for the automatic handling of large number of data points and to avoid human error during manual analysis has been developed (Zerr and Henikoff, 2005; http://www.gelbuddy.org). Ecotilling can be used to effectively discover natural polymorphism in a large set of individuals. The discovered polymorphism can be confirmed by

Mutagenesis and Functional Genomics in Cereals

395

sequencing and include base pair changes, small insertions and deletions. Ecotilling allows rapid detection of variation in many individuals and is cost-effective because only one individual for each haplotype needs to be sequenced. A chief advantage of Ecotilling is that to characterize the allelic series and to discover novel alleles, instead of sequencing each individual variant, only unique haplotypes from an Ecotilled germplasm need to be sequenced. Recent successful examples of Ecotilling in cereals include the identification of allelic variation in the powdery mildew resistance genes mlo and Mla of barley (Mejlhede et al., 2006). Ecotilling could also be effectively used to unravel genetic variation even in germplasm consisting of heterogenous (such as land races) and heterozygous individuals/populations without the need of using reference DNA. Also Ecotilling initiative in maize has been undertaken by screening 48 lines, which account for 80% of the diversity in maize (http://genome.purdue.edu/maizetilling/EcoTILLING.htm). Considering its importance in correlating the allelic variation with the phenotypic variation and its utility in the discovery of new and important alleles that may be useful in crop improvement, Ecotilling has vast potential in crop improvement. Currently, more than 1500 national, regional, or international gene banks are engaged in ex situ conservation of a total of 6.1 million (including >3.6 million accessions of 30 different crop plants) accessions from 18,000 plant species (SINGER-The System-wide Information Network for Genetic Resources; http://singer.grinfo.net/). These accessions constitute a rich source of diversity, at both the phenotypic and nucleotide sequence levels and could be the main targets for Ecotilling in future. Currently, a number of initiatives on Ecotilling in various cereal species such as rice, maize, and barley have been launched (Table 6).

3.5. DEALING and DeleteageneTM A year before TILLING was reported, Liu et al. (1999) demonstrated the high-throughput PCR detection of deletions by using gene-specific primers to screen DNA pools of deletion lines of C. elegans. In this study, four chemical mutagens including EMS, ENU, DEO and UV-TMP were used to induce deletion mutations. A majority of deletions (average size = 1400 bp) occurred in exons and led to loss of gene function leading to mutant phenotypes. This approach allowed the establishment of gene-function relationship. A similar approach called DeleteageneTM for detection of deletions in fast-neutron induced mutants in Arabidopsis and rice was reported by Li et al. (2001, 2002). For further details with respect to creation of deletion library, pooling strategy, PCR screening, and DeleteageneTM applications refer to the review by Li and Zhang (2002). Large pools of DNA extracted from mutagenized populations generated by fast neutron or chemical deletogens such as DEB, are screened by using primers flanking the targeted genes and by adjusting the PCR conditions to preferentially amplify deletion alleles

396

H. S. Balyan et al.

Table 6 A summary of institutions undertaking studies on Eco-TILLING of genes in different cereal species Species

Genes

Institute/University

Barley

Drought tolerant genes (Dhn)

Barley

Barley yellow mosaic and barley mild mosaic virus resistant genes (rym4 and rym5)

Barley

Biotic and abiotic resistance genes Drought related genes (DREB2, TPP, and ERF3) and Protein phosphatase (pp2a4), membrane stability (14–3–3), Class I chitinase (cht) Drought response binding protein 1 (dreb1), Trehalose phosphatase (tps), Viviparous14 (vp14) Alk and waxy gene

Institute for Jordbrugsvidenskab, Denmark. Leibniz Institute of Plant Genetics and Crop Plant Research (IPK), Gatersleben, Germany Rothanmsted Research, Harpenden, UK Scottish Crop Research Institute, Invergowrie, UK International Rice Research Institute, Metro Manila, The Philippines Basic Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA

Rice

Rice

Sorghum Biotic and abiotic resistance genes Maize

Biotic and abiotic resistance genes

Texas Agricultural Experiment Station, Beaumont, Texas, USA International Rice Research Institute, Metro Manila, The Philippines Fred Hutchinson Cancer Research Center, Seattle, Washington, USA Purdue University, Seattle, USA

represented by amplification products of smaller size than that expected for the wild-type alleles. Deletion mutants were identified for 84% of the targeted loci from an Arabidopsis population of 51,840 individuals (Li et al., 2001; http://www. bio.net/hlo.ypermail/arab-gen/) and the potential application of DeleteageneTM was demonstrated by the analysis of knockout mutations of Arabidopsis transcription factors TGA2, TGA5, and TGA6 suggesting their

Mutagenesis and Functional Genomics in Cereals

397

redundant but essential roles in systemic acquired resistance (Li et al., 2001; Zhang et al., 2003). In the rice functional genomics project at IRRI, Philippines, deletion mutation detection through PCR in defense response genes was optimized in DNA pools of 100 individuals from a population of 8500 DEB and fast neutron induced mutants (Manosalva et al., 2003). Toward identifying rice lines carrying deletion mutation in defense responsive genes, gene-specific (e.g., the untranslated regions of the genes) or gene-family specific (e.g., to conserved regions of gene family members) primers were used in PCR analysis. These primers allowed identification of mutations in two different PAL (phenylalanine ammonia lyase) family members, which were subsequently, confirmed using nested-PCR and sequence analysis (Wu et al., 2005). Studies on functional genomics aimed at exploiting induced deletions by chemical deletogens such as DEB are in progress in diploid wheat (T. monococcum). A method based on the poison– primer approach (Edgley et al., 2002) was applied to detect mutations in complex pools of cereal DNA (Fig. 4). Result (Riera-Lizarazu, unpublished) clearly indicated that insertion/deletion polymorphisms in the waxy gene can be detected up to a ratio of 1:5000 (Fig. 4B). At a 1:5000 ratio, the DNA from the waxy deletion stock represents only 10 pg of the 50 mg of template DNA used in the first round of PCR. Thus, nested PCR with a poison primer allows the detection of a deletion in a complex mixture where the DNA with a deletion is present in about two genome equivalents (DNA content in one disomic cell barley of T. monococcum). Additional protocols based on various simple PCR methods have also been developed that allow detection easily in pools of 1:1000 DNA. The most desirable component of all these approaches is the use of simple PCR and gel-based assays reducing the need for complex machines and enzymes. The genome-wide oligonucleotide (oligo) chips have been suggested as an expedient way to detect mutations (Borevitz et al., 2003). Detection of genetic lesions in rice deletion mutations by using Syngenta GeneChipÒ containing 24-mer oligos representing 24,000 rice genes was recently attempted (Chang et al., 2003). However, it has been shown that the precision of chip-based mutation detection system may be adversely affected by the large deletions and high number of background mutations that may complicate the interpretation of the hybridization signal.

3.6. Allelic series versus knockout mutations Point-mutations-inducing chemical mutagens yield a large proportion of mis-sense mutations, which are discovered through TILLING. While a large majority of mutations caused by insertional (T-DNA and transposons) and deletion (DEALING and Deleteagene) mutagenesis are knockout mutations. Thus these mis-sense and knockout mutation inducing approaches are complimentary. For example, if a knockout mutation leads

398

H. S. Balyan et al.

to lethality, it is not possible to characterize the phenotypic changes brought about by the induced mutations in the concerned gene. On the other hand, most of the mis-sense mutations should produce an allelic series representing a range of changes in the phenotype, enabling detailed functional study of the concerned gene. If the mis-sense mutations in the allelic series occur in the conserved residues leading to destabilization of proteins, then conditional mutations are produced. Thus, in the event of the induction of lethal knockout mutations by insertion and deletion mutations, TILLING has the potential to provide an allelic series that can be highly informative for functional genomics. In the case of an allelic series, it is important to estimate the mutation’s ability to damage the protein as this information can determine whether the available mutants are adequate or more screening needs to be done. Toward achieving this goal, software tools have been developed that use protein homology information to predict damaging lesions. These include conserved alignments in blocks database of protein families to predict whether a mis-sense mutation is expected to damage protein function (http://www.proweb.org/coddle) and the sorting intolerant from tolerant (SIFT) program uses PSI-BLAST searching of current databases to assess amino acid conservation (Ng and Henikoff, 2001). SIFT is a general web-based program that is applicable to both natural variation and induced mutations (http://www.blocks.fhcrc.org/~pauline/SIFT. html). SIFT predicts damage with ~70% accuracy with experimental data (Ng and Henikoff, 2001, 2002). Therefore, sequence information can be used as a guide to predict damage to protein caused by random mis-sense mutations, and this approach can be applied generally across organisms.

4. Phenomics Platform for Screening Mutagenized Population Implementing high-throughput phenotypic screening system is a key step for systematic documentation of phenotypic mutants for reverse-genetics analysis. Hence systematic phenotyping of large mutagenized populations based on combination of morphological and biochemical techniques will help to categorize the mutants according to trait of interest. Recently a highthroughput phenomics study was performed in rice and provided detailed phenotypic data for more than 20,000 T1 lines (12 plants per line) and furnished the details of T-DNA integration sites and the consequent phenotypes (Chern et al., 2007). Obviously such trait specific categorization of mutants is valuable in limiting the amount of reverse-genetics screening. The growth stages and data collection methodology platform presented in Arabidopsis demonstrates its significance in gathering phenotypic data over the complete life cycle starting from seed imbibition to vegetative phase

Mutagenesis and Functional Genomics in Cereals

399

by plate-based analysis, and other principal growth stages, flower and seed production by soil-based screening techniques (Boyes et al., 2001). In addition, Kaminuma et al. (2003) demonstrated implementation of computational phenomic technology based on 3D-specific traits at rosette leaves in Arabidopsis. Another example of trait specific screening has been demonstrated by implementing iodine-based staining technique for detection of accumulated starch patterns in leaf between wild-type and AGPase mutant detected using the TILLING approach ( John Lunn, MPI Golm, personal communication). Also iodine staining technique has been implemented to stain barley seeds to estimate amylose content in waxy mutants created using ubiDs barley activation tagging system (Ayliffe et al., 2007). We would also like to highlight the elegant screening technique implemented to screen abiotic stress tolerant mutants. Genetic screening for identifying loci associated with abiotic stress responses signalling has been difficult due to the absence of major visible phenotypes and appropriate screening systems. To tackle these problems Zhu and coworkers evolved an approach first by generating transgenic plants by expressing the luciferase under the control of the stress responsive RD29A promoter and eventually treating the seeds of these plants with EMS to generate large number of mutants (see Ishitani et al., 1997). By looking for the alteration in luciferase expression pattern in mutants under various abiotic stress conditions three major groups of mutants (los-low expression of osmotically responsive genes, cos-constitutive expression of osmotically responsive genes and hos-high expression of osmotically responsive genes) were identified and presented genetic analysis of osmotic and cold stress signal transduction pathways mediated by ABA-dependent and ABAindependent pathways (Ishitani et al., 1997). These representative examples demonstrate the role of phenomics in reporting trait specific mutant phenotypes. Such elegant methods, or modifications thereof, need to be implemented for screening mutants in cereal mutagenized populations aimed at determining gene-function relationship through reverse genetics.

5. Outlook At present, determining gene-function relationships by using reversegenetics methods (T-DNA/transposons) rely heavily on transgenic technology. Due to the recalcitrant nature of cereal crop plants to both the transformation and regeneration, application of insertional-mutagenesisbased reverse genetics has serious limitations. Therefore, alternative and newly emerging non-transgenic, induced mutation-based approaches such as TILLING, DEALING, and DeleteageneTM offer a much needed alternative for the functional analysis of genes in cereal crop plants. However, the success of these approaches depends on generating populations with high

400

H. S. Balyan et al.

level of mutational redundancy (>5- to 10-fold mutations of the number of genes in the genome) to ensure detection of mutations in a particular gene with a sufficiently high probability, effective cataloguing, and high-throughput screening of the mutant population (Alonso and Ecker, 2006). Using these approaches, allelic series and knockout mutations can be produced in the target gene conferring subtle to extreme phenotypic alterations and the total loss of function establishing a structure and function relationship. Detection of mutations in induced mutagenesis-based approaches may be hindered by functional redundancy in multigene families, background mutations and the polyploidy in crops such as in wheat and oat. However, encouraging results have been shown by reverse-genetics studies with A. thaliana by knocking out entire gene families uncovering overlapping and specific functions among their members (Okushima et al., 2005; Prigge et al., 2005; To et al., 2004). Second, extensive crossing may help purge background mutations to obtain mutant lines with mutations at only a single locus and also to combine mutant alleles at all the homoeoloci for recovering mutant phenotype in polyploid species. Thirdly, polyploid species such as common wheat also contain major genes that control a particular trait (Friebe et al., 2003). Mutations in such genes may be detected with ease without any complication due to polyploidy (Friebe et al., 2003; Koebner and Hadfield, 2001). Currently, in most of the cereal species, several different cultivars or varieties are being used for reverse-genetics analysis. Although from the stand point of inducing a wide spectrum of mutations, it would perhaps be desirable to use different varieties, but it would be most prudent to use a single variety for which most sequencing information is available. In cereals, the reverse-genetics studies conducted so far have focused on metabolic pathway related genes involved in plant ontogeny. In the future, it would also be interesting to look for mutations in the promoter regions, which may have much larger effect on the phenotypic expression of the trait. Reverse-genetics functional genomics may also help at comparing the functions of closely related genes from related cereal species to address outstanding issues related to evolutionary and developmental biology (Evo-Devo) to understand how the past and present biodiversity arose (Ostergaard and Yanofsky, 2004). Another daunting task that needs to be addressed is the phenotypic characterization of mutagenized population. Thus, the role of phenomics in reporting trait specific mutant phenotypes needs to be considered. Also, systematic forward genetics using reverse-genetics tools that is, simultaneous phenotypic analysis of all the induced mutants needs to be undertaken. This may only be achieved by developing new high-throughput phenomic platforms to examine not only morphological but also biochemical phenotypes. Although induced mutations will continue to be an important component of reverse-genetics analysis, the importance of natural allelic variation

Mutagenesis and Functional Genomics in Cereals

401

to study gene function in plants can not be ignored. Also, future new technologies like sequencing by synthesis and ultra high-throughput sequencing (UHTS) need to be employed for the rapid identification of such alleles (Service, 2006). These and other new emerging sequencing technologies constitute very attractive high-throughput future options for unravelling induced and natural allelic variation for the study of gene-function relationship.

ACKNOWLEDGMENTS The exchange of scientists award by INSA-DFG and the Developing Country Collaboration award by the NSF, USA, to HSB are gratefully acknowledged, during the tenure of which this chapter was conceived and prepared. The DEALING research by ORL and SFK was supported by a grant from the National Science Foundation-Plant Genome Research Program Contract Agreement No. DBI-0321462.

REFERENCES Agrawal, G. K., Yamagachi, M., Kobayashi, M., Hirochika, R., Miyao, A., and Hirochika, H. (2001). Screening of rice viviparous mutant generated by endogenous retrotransposon Tos17 insertion. Tagging of a zeaxanthin epoxidase gene and a novel OsTATC gene. Plant Physol. 125, 1248–1257. Ahern, K., Deewatthanawong, P., Conrad, L., Schnable, J., Dong, Q., Sigmon, B., Hall, B., Schares, J., Brendel, V., Vollbrecht, E., and Brutnell, T. (2006). A two component Activator/dissociation platform for reverse and forward genetic analysis in maize. Maize Genet. Conf. Abstr. 48, 155. Alharbi, K. K., Aldahmesh, M. A., Spanakis, E., Haddad, L., Whittall, R. A., Chen, X.-H., Rassoulian, H., Smith, M. J., Sillibourne, J., Ball, N. J., Graham, N. J., Briggs, P. J., et al. (2005). Mutation scanning by meltMADGE: Validations using BRCA1 and LDLR, and demonstration of the potential to identify severe, moderate, silent, rare, and paucimorphic mutations in the general population. Genome Res. 15, 967–977. Alonso, J. M., and Ecker, J. R. (2006). Moving forward in reverse: Genetic technologies to enable genome-wide phenomic screens in Arabidopsis. Nat. Rev. Genet. 7, 524–536. Alonso, J. M., Stepanova, A. N., Leisse, T. J., Kim, C. J., Chen, H. M., Shinn, P., Stevenson, D. K., Zimmerman, J., Barajas, P., Cheuk, R., Gadrinab, C., Heller, C., et al. (2003). Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 301, 653–657. An, G. H., Jeong, D. H., Jung, K. H., and Lee, S. (2005a). Reverse genetic approaches for functional genomics of rice. Plant Mol. Biol. 59, 111–123. An, G., Lee, S., Kim, S.-H., and Kim, S.-R. (2005b). Molecular genetics using T-DNA in rice. Plant Cell Physiol. 46, 14–22. An, S. Y., Park, S., Jeong, D. H., Lee, D. Y., Kang, H. G., Yu, J. H., Hur, J., Kim, S. R., Kim, Y. H., Lee, M., Han, S. K., Kim, S. J., et al. (2003). Generation and analysis of end sequence database for T-DNA tagging lines in rice. Plant Physiol. 133, 2040–2047. Ashburner, M. (1990). ‘‘Drosophila: A Laboratory Handbook.’’ Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Ayliffe, M. A., Pallotta, M., Landgridge, P., and Pryor, A. J. (2007). A barley activation tagging system. Plant Mol. Biol. 64, 329–347.

402

H. S. Balyan et al.

Azpiroz-Leehan, R., and Feldmann, K. A. (1997). T-DNA insertion mutagenesis in Arabidopsis: Going back and forth. Trends Genet. 13, 152–156. Bai, L., Singh, M., Lauren, P. L., Sweeney, M., and Brutnell, T. P. (2007). Generating novel allelic variation through activator (Ac) insertional mutagenesis in maize. Genetics 175, 981–992. Balzergue, S., Dubreucq, B., Chuvin, S., Le-Clainche, I., Le Boulaire, F., de Rose, R., Samson, F., Baiudet, V., Lecharny, L., Cruaud, C., Werissenbach, M., Caboche, M., et al. (2001). Improved PCR-walking for large scale isolation of plant T-DNA borders. Biotechniques 30, 496–504. Baker, B., Schell, J., Lorz, H., and Fedoroff, N. (1986). Transposition of the maize controlling element ‘Activator’ in tobacco. Proc. Natl. Acad. Sci. USA 83, 4844–4848. Barakat, A., Gallois, P., Taynal, M., Mestre-Ortega, D., Sallaud, C., Guiderloni, E., Delseny, M., and Bermnardi, G. (2000). The distribution of T-DNA in the genomes of transgenic Arabidopsis and rice. FEBS Lett. 417, 161–164. Baulcombe, D. C. (1996). RNA as a target and an initiator of post-transcriptional gene silencing in transgenic plants. Plant Mol. Biol. 32, 79–88. Bedell, J. A., Budiman, M. A., Nunberg, A., Citek, R. W., Robbins, D., Jones, J., Flick, E., Rohlfing, T., Fries, J., Bradford, K., McMenamy, J., Smith, M., et al. (2005). Sorghum genome sequencing by methylation filtration. PLoS Biol. 3, 103–115. Bentley, A., MacLennan, B., Calvo, J., and Dearolf, C. R. (2000). Targeted recovery of mutations in Drosophila. Genetics 156, 1169–1173. Blauth, S. L., Yao, Y., Klucinec, J. D., Shannon, J. C., Thompson, D. B., and Guilitinan, M. J. (2001). Identification of mutator insertional mutant of starch-branching enzyme 2a in corn. Plant Physiol. 125, 1396–1405. Borevitz, J., Liang, D., Plouffe, D., Chang, H., Zhu, T., Weigel, D., Berry, C., Winzeler, E., and Chory, J. (2003). Large-scale identification of single-feature polymorphisms in complex genomes. Genome Res. 13, 513–523. Bouche, N., and Bouchez, D. (2001). Arabidopsis gene knockout; phenotypes wanted. Curr. Opin. Plant Biol. 4, 111–117. Boyes, D. C., Zayed, A. M., Ascenzi, R., McCaskill, A. J., Hoffman, N. E., Davis, K. R., and Gorlach, J. (2001). Growth stage-based phenotypic analysis of Arabidopsis: A model for high throughput functional genomics in plants. Plant Cell 13, 1499–1510. Bruggmann, R., Bharti, A. K., Gundlach, H., Lai, J. S., Young, S., Pontaroli, A. C., Wei, F. S., Haberer, G., Fuks, G., Du, C. G., Raymond, C., Estep, M. C., et al. (2006). Uneven chromosome contraction and expansion in the maize genome. Genome Res. 16, 1241–1251. Brutnell, T. P. (2002). Transposon tagging in maize. Funct. Integr. Genomics 2, 4–12. Caldwell, D. G., Mccallum, N., Shaw, P., Muehlbauer, G. J., Marshall, D. F., and Waugh, R. (2004). A structured mutant population for forward and reverse genetics in barley (Hordeum vulgare L.). Plant J. 40, 143–150. Capy, P., Bazin, C., Higuet, D., and Langin, T. (1998). ‘‘Dynamics and Evolution of Transposable Elements.’’ Landes Bioscience, Austin, Texas. Castillo, A. M. W., Cistue, L. W., Valles, M. P. W., Sanz, J. M. W., Romagossa, I., and Molina-cano, J. L. (2001). Efficient production of androgenic doubled-haploid mutants in barley by the application of sodium azide to anther and microspore culture. Plant Cell Rep. 20, 105–111. Chang, H. S., Wu, C., Zeng, L., Dunn, M., Wang, G. L., Leung, H., Goff, S., Wang, X., Zhu, T., and Leach, J. E. (2003). Detection of deleted genes in rice mutants using the Rice GeneChip genome array. In ‘‘Abstract Papers of the Plant and Animal Genome XI Conference,’’ p. 100. Chen, S., Jin, W., Wang, M., Zhang, F., Zhou, J., Jia, Q., Wu, Y., Liu, F., and Wu, P. (2003). Distribution and characterization of over 1000 T-DNA tags in rice genome. Plant J. 36, 105–113.

Mutagenesis and Functional Genomics in Cereals

403

Chern, C.-G., Fan, M.-J., Yu, S.-M., Hour, A.-L., Lu, P.-C., Lin, Y.-C., Wei, F.-J., Huang, S.-C., Chen, S., Lai, M. H., Tseng, C.-S., Yen, H.-M., et al. (2007). A rice phenomics study-phenotype scoring and seed propagation of a T-DNA insertioninduced rice mutant population. Plant Mol. Biol. 65, 427–438. Chin, H. G., Choe, H. C., Lee, S. H., Park, S. H., Koo, J. C., Kim, N. Y., Lee, J. J., Oh, B. G., Yi, G. H., Kim, S. C., Choi, H. C., Cho, M. J., et al. (1999). Molecular analysis of rice plants harboring an AC/Ds transposable element-mediated gene trapping system. Plant J. 19, 615–623. Chopra, S., Brendel, V., Zhang, J. B., Axtell, J. D., and Peterson, T. (1999). Molecular characterization of a mutable pigmentation phenotype and isolation of the first active transposable element from Sorghum bicolor. Proc. Natl. Acad. Sci. USA 96, 15330–15335. Chuang, C. F., and Meyerowitz, E. M. (2000). Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 97, 4985–4990. Colbert, T., Till, B. J., Tompa, R., Reynolds, S., Steine, M. N., Yeung, A. T., McCallum, C. M., Comai, L., and Henikoff, S. (2001). High-throughput screening for the induced point mutations. Plant Physiol. 126, 480–484. Comai, L., and Henikoff, S. (2006). TILLING: Practical single-nucleotide mutation discovery. Plant J. 45, 684–694. Comai, L., Young, K., Till, B. J., Reynolds, S. H., Greene, E. A., Codomo, C. A., Enns, L. C., Johnson, J. E., Burtner, C., Odden, A. R., and Henikoff, S. (2004). Efficient discovery of DNA polymorphisms in natural populations by Ecotilling. Plant J. 37, 778–786. Cooper, L. D., Marquez-Cedillo, L., Singh, J., Sturbaum, A. K., Zhang, S., Edwards, V., Johnson, K., Kleinhofs, A., Rangel, S., Carollo, V., Bregitzer, P., Lemaux, P. G., et al. (2004). Mapping Ds insertions in barley using a sequence-based approach. Theor. Appl. Genet. 272, 181–193. Coweperthwaite, M., Park, W., Xu, Z., Yan, X. M., Mauris, S. C., and Dooner, H. K. (2002). Use of transposon Ac as a gene-searching engine in the maize genome. Plant Cell 14, 713–726. Cresse, A. D., Hulbert, S. H., Brown, W. E., Lucas, J. R., and Bennetzen, J. L. (1995). MuIrelated transposable elements of maize preferentially insert into low copy number DNA. Genetics 140, 315–324. Cross, M. J., Lee, S. L., Rice, N. F., and Henry, R. J. (2006). Targeted mutagenesis in sorghum using high-throughput screening platform—A reverse genetics strategy to complement the sorghum genomics effort. In ‘‘Abstracts International Plant and Animal Genome Conference,’’ January 14–18, 2006, San Diego, CA, USA, p. 371. Cutler, D. J., Zwick, M. E., Carrasquillo, M. M., Yohn, C. T., Tobin, K. P., Kashuk, C., Mathews, D. J., Shah, N. A., Eichler, E. E., Warrington, J. A., and Chakravarti, A. (2001). High throughput variation detection and genotyping using microarray. Genome Res. 11, 1913–1925. da Costa e Silva, O., Lorbiecke, R., Garg, P., Mu¨ller, L., Waßmann, M., Lauert, P., Scanlon, M., Hsia, A.-P., Schnable, P. S., Krupinska, K., and Wienand, U. (2004). The Etched1 gene of Zea mays (L.) encodes a zinc ribbon protein that belongs to the transcriptionally active chromosome (TAC) of plastids and is similar to the transcription factor TFIIS. Plant J. 38, 923–939. Delseny, M. (2003). Towards an accurate sequence of the rice genome. Curr. Opin. Plant Biol. 6, 101–105. Devos, K. M., Ma, J. X., Pontaroli, A. C., Pratt, L. H., and Bennetzen, J. L. (2005). Analysis and mapping of randomly chosen bacterial artificial chromosome clones from hexaploid bread wheat. Proc. Natl. Acad. Sci. USA 102, 19243–19248. Dilkes, B. P., and Feldmann, K. A. (1998). Cloning genes from T-DNA tagged mutants. Methods Mol. Biol. 82, 339–351.

404

H. S. Balyan et al.

Droc, G., Ruiz, M., Larmande, P., Pereira, A., Piffanelli, P., Morel, J. B., Dievart, A., Courtois, B., Guiderdoni, E., and Perin, C. (2006). OrygenesDB: A database for rice reverse genetics. Nucleic Acids Res. 34, D736–D740. Edgley, M., D’Souza, A., Moulder, G., McKay, S., Shen, B., Gilchrist, E., Moerman, D., and Barstead, R. (2002). Improved detection of small deletions in complex pools of DNA. Nucleic Acids Res. 30, e52. Emery, G. (1960). Biological effects of a chemical mutagen, Diepoxybutane, on tomato. Science 131, 1732–1733. Enoki, H., Izawa, T., Kawahara, M., Komatsu, M., koh, S., kyozuka, J., and Shimamoto, K. (1999). Ac as a tool for the functional genomics of rice. Plant J. 19, 605–613. Fedoroff, D. L., and Smith, D. L (1993). A Versatile system for detecting transposition in Arabidopsis. Plant J. 3, 273–289. Fedoroff, N. V., Wessler, S., and Shure, M. (1983). Isolation of the transposable maize controlling elements Ac and Ds. Cell 35, 235–243. Feng, C. P., and Mundy, J. (2006). Gene discovery and functional analysis in model plant Arabidopsis. J. Integr. Plant Biol. 48, 5–14. Fernandes, J., Dong, Q., Schneider, B., Morrow, D. J., Nan, G.-L., Brendel, V., and Walbot, V. (2004). Genome-wide mutagenesis of Zea mays L. using RescueMu transposons. Genome Biol. 5, R82. Forster, B. P., and Thomas, W. T. B. (2005). Doubled haploids in genetics and plant breeding. Plant Breed. Rev. 25, 57–88. Friebe, B., Zhang, P., Nasuda, S., and Gill, B. (2003). Characterization of a knockout mutation at the Gc2 locus in wheat. Theor. Appl. Genet. 111, 509–517. Fujino, K., Sekiguchi, H., and Kiguchi, T. (2005). Identification of an active transposon in intact rice plants. Mol. Genet. Genomics 273, 150–157. Fujita, N., Yoshida, M., Asakura, N., Ohdan, T., Miyao, A., Hirochika, H., and Nakamura, Y. (2006). Function and characterization of starch synthase I using mutants in rice. Plant Physiol. 140, 1070–1084. Ganguly, T., Dhulipala, R., Godmilow, L., and Ganguly, A. (1998). High throughput fluorescence based conformation-sensitive gel electrophoresis (F-CSGE) identifies six unique BRCA2 mutations and an overall low incidence of BRCA2 mutations in highrisk BRCA1-negative breast cancer families. Hum. Genet. 102, 549–556. Gaut, B. S., and Doebley, J. F. (1997). DNA sequence evidence for the segmental allotetraploid origin of maize. Proc. Natl. Acad. Sci. USA 94, 6809–6814. Gilchrist, E. J., and Haughn, G. W. (2005). TILLING without a plough: A new method with applications for reverse genetics. Curr. Opin. Plant Biol. 8, 211–215. Goff, S., Ricke, D., lan, T., Presting, G., Wang, R., Dunn, M., Galezebrook, J., Session, A., Oeller, P., Varma, H., Hadley, D., Hutchinson, D., et al. (2002). A draft of the rice genomes (Oryza sativa L. spp. japonica). Science 296, 92–100. Greco, R., Ouwerkerk, P. B. F., Sallaud, C., Kohli, A., Colombo, L., Puigdomenech, P., Guiderdoni, E., Christou, P., Hoge, J. H. C., and Pereira, A. (2001). Transposon insertional mutagenesis in rice. Plant Physiol. 125, 1175–1177. Greco, R., Ouwerkerk, P. B. F., de Kam, R. J., Sallaud, C., Favalli, C., Colombo, L., guiderloni, E., Meijer, A. H., Hoge, J. H. C., and Pereira, A. (2003). Transpositional behaviour of an Ac/Ds system for reverse genetics in rice. Thoer. Appl. Genet. 108, 10–24. Greco, R., Ouwerkerk, P. B., Taal, A. J. C., Sallaud, C., Guiderdoni, E., Meijer, A. H., Hoge, J. H. C., and Pereira, A. (2004). Transcription and somatic transposition of the maize En/Spm transposon system in rice. Mol. Genet. Genomics 270, 514–523. Greene, E. A., Codomo, C. A., Taylor, N. E., Henikoff, J. G., Till, B. J., Reynolds, S. H., Enns, L. C., Burtner, C., Johnson, J. E., Odden, A. R., Comai, L., and Henikoff, S. (2003). Spectrum of chemically induced mutations from a large-scale reverse-genetic screen in Arabidopsis. Genetics 164, 731–740.

Mutagenesis and Functional Genomics in Cereals

405

Gross, E., Arnold, N., Goette, J., Schwarz-Boeger, U., and Kiechle, M. (1999). A comparison of BRCA1 mutation analysis by direct sequencing, SSCP and DHPLC. Hum. Genet. 105, 72–78. Hamer, L., Dezwaan, T. M., Montenegro-Chamorro, M. V., Frank, S. A., and Hamer, J. E. (2001). Recent advances in large-scale transposon mutagenesis. Curr. Opin. Chem. Biol. 5, 67–73. Hanley, S., Edwards, D., Srevenson, D., Haines, S., Hegarty, M., Schuch, W., and Edwards, K. J. (2000). Identification of transposon-tagged genes by random sequencing of mutator-tagged DNA fragments from Zea mays. Plant J. 23, 557–566. Henikoff, S., and Comai, L. (2003). Single-nucleotide mutations for plant functional genomics. Annu. Rev. Plant Biol. 54, 375–401. Hiei, Y., Ohta, S., Komari, T., and Kumashiro, T. (1994). Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of TDNA. Plant J. 6, 271–282. Hieter, P., and Boguski, M. (1997). Functional genomics: Its all how you read it. Science 278, 601–602. Hirochika, H. (2001). Contribution of the Tos17 retrotransposon to rice functional genomics. Curr. Opin. Plant Biol. 4, 118–122. Hirochika, H., Sugimoto, K., Otsuki, C., Tsugawa, H., and Kanda, M. (1996). Retrotransposons of rice involved in mutation induced by tissue culture. Proc. Natl. Acad. Sci. USA 93, 7783–7788. Hirochika, H., Miyao, A., Yamazaki, M., Takeda, S., Abe, K., Hirochika, R., Agarwal, G. K., Watanabe, T., Sugimoto, K., Sasaki, T., Murata, K., and Tonaka, K. (2001). Retrotransposon of rice as a tool for functional analysis of genes. In ‘‘Rice Genetics IV Proceedings of the Fourth International Rice Genetics Symposium,’’ 22–27 October 2000 (G. S. Khus, D. S. Brar, and B. Hardy, Eds.), pp. 279–292. Science Publishers Inc., New Delhi (India) and International Rice Research Institute, Las Banos (Philippines). Hirochika, H., Guiderdoni, E., An, G., Hsing, Y.-I., Eun, M. Y., Han, C.-D., Upadhyaya, N., Ramchandran, S., Zhang, Q., Pereira, A., Sundresan, S., and Leug, H. (2004). Rice mutant resources for gene discovery. Plant Mol. Biol. 54, 325–334. Hoskins, B. E., Thorn, A., Scambler, P. J., and Beales, P. L. (2003). Evaluation of multiplex capillary heteroduplex analysis: A rapid and sensitive mutation screening technique. Hum. Mutat. 22, 151–157. Hsing, Y.-I., Chern, C.-G., Fan, M.-J., Lu, P. C., Chen, K. T., Lo, S.-F., Sun, P.-K., Ho, S.-L., Lee, K.-W., Wang, Y.-C., Huang, W.-L., Ko, S.-S., et al. (2007). A rice gene activation/knockout mutant resource for high throughput functional genomics. Plant Mol. Biol. 63, 351–364. International Rice Genome Sequencing Project. (2005). The map-based sequence of the rice genome. Nature 436, 793–800. Ishitani, M., Xiong, L., Stevenson, B., and Zhu, J. K. (1997). Genetic analysis of osmotic and cold stress signal transduction in Arabidopsis: Interaction and convergence of abscisic acid-dependent and abscisic acid-independent pathways. Plant Cell 9, 1935–1949. Ito, T., Seki, M., Hayashida, N., Shibata, D., and Shinozaki, K. (1999). Regional insertional mutagenesis of genes on Arabidopsis thaliana chromosome V using the Ac/Ds transposon in combination with a cDNA scanning method. Plant J. 17, 433–444. Ito, Y., Eiguchi, M., and Kurata, N. (2004). Establishment of an enhancer trap system with Ds and GUS for functional genomics in rice. Mol. Genet. Genomics 271, 639–650. Izawa, T., Miyazaki, C., Yamamoto, M., Terada, R., Iida, S., and Shimamoto, K. (1991). Introduction and transposition of the maize transposable element Ac in rice (Oryza sativa L.). Mol. Gen. Genet. 227, 391–396. Jansen, G., Hazendonk, E., Thijssen, K. L., and Plasterk, R. H. A. (1997). Reverse genetics by chemical mutagenesis in Caenorhabditis elegans. Nat. Genet. 17, 119–121.

406

H. S. Balyan et al.

Jansen, G., Thijssen, K. L., Werner, P., van der Horst, M., Hazendonk, E., and Plasterk, R. H. A. (1999). The complete family of genes encoding G proteins of Caenorhabditis elegans. Nat. Genet. 21, 414–419. Jefferson, R. A., Kavanagh, T. A., and Beven, M. W. (1987). GUS fusions: ß-Glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901–3907. Jeon, J. S., Lee, S., Jung, K. H., Jun, S. H., Jeong, D. H., Lee, J., Kim, C., Jang, S., Yang, K., Nam, J., An, K., Han, M. J., et al. (2000). T-DNA insertional mutagenesis for functional genomics in rice. Plant J. 22, 561–570. Jiang, N., Bao, Z., Zhang, X., Hirochika, H., Eddy, S. R., and Wessler, S. R. (2003). An active DNA transposon family in rice. Nature 421, 163–167. Jung, K. H., Hur, J., Ryu, C. H., Choi, Y., Chung, Y. Y., Miyao, A., Hirochika, H., and An, G. H. (2003). Characterization of a rice chlorophyll-deficient mutant using the T-DNA gene-trap system. Plant Cell Physiol. 44, 463–472. Jung, K. H., Han, M. J., Lee, Y. S., Kim, Y. W., Hwang, I. W., Kim, M. J., Kim, Y. K., Nahm, B. H., and An, G. H. (2005). Rice Undeveloped Tapetum1 is a major regulator of early tapetum development. Plant Cell 17, 2705–2722. Kaminuma, E., Heida, N., Tsumoto, Y., Matsui, M., Toyoda, T., and Konagaya, A. (2003). Quantifying the spiral leaf trait of Arabidopsis from the 3D shape model towards computational phenomic. Genome Inform. 14, 627–628. Kang, H. G., Park, S., Matsuoka, M., and An, G. (2005). White-core endosperm floury endosperm-4 in rice is generated by knockout mutations in the C4-type pyruvate orthophosphate dikinase gene (OsPPDKB). Plant J. 42, 901–911. Kim, S. R., Lee, J., Jun, S. H., Park, S., Kang, H. G., Kwon, S., and An, G. (2003). Transgene structures in T-DNA-inserted rice plants. Plant Mol. Biol. 52, 761–773. Kim, C. M., Piao, H. L., Park, S. J., et al. (2004). Rapid, larger-scale generation of Ds transposant lines and analysis of the Ds insertion sites in rice. Plant J. 39, 252–263. Kleinhofs, A., Warner, R. L., Muehlbauer, F. J., and Nilan, R. A. (1978). Induction and selection of specific gene mutations in Hordeum and Pisum. Mutat. Res. 51, 29–35. Klimyuk, V. I., Nussaume, L., Harrison, K., and Jones, J. D. G. (1995). Novel GUS expression patterns following transposition of an enhancer trap Ds element in Arabidopsis. Mol. Gen. Genet. 249, 357–365. Kodym, A., and Afza, R. (2003). Physical and chemical mutagenesis. Methods Mol. Biol. 236, 189–204. Koebner, R., and Hadfield, J. (2001). Large-scale mutagenesis directed at specific chromosomes in wheat. Genome 44, 45–49. Kohli, A., Xiong, J., Greco, R., Christou, P., and Pareira, A. (2001). Tagged Transcriptome Display (TTD) in indica rice using Ac transposition. Mol. Genet. Genomics 266, 1–11. Kolesnik, T., Szeverenyi, I., Bachmann, D., Kumar, C. S., Jiang, S., Ramamoorthy, R., Cai, M., Ma, Z. G., Sundaresan, V., and Ramachandran, S. (2004). Establishing an efficient Ac/Ds tagging system in rice: Large-scale analysis of Ds flanking sequences. Plant J. 37, 301–314. Kolkman, J. M., Conrad, L. J., Farmer, P. R., Hardeman, K., Ahern, K. R., Lewis, P. E., Sawers, R. J. H., Lebejko, S., Chomet, P., and Brutnell, T. P. (2005). Distribution of Activator (Ac) throughout the Maize genome for use in regional mutagenesis. Genetics 169, 981–995. Koornneef, M., Dellart, L. W. M., and van der Veen, J. H. (1982). EMS-and radiationinduced mutation frequencies at individual loci in Arabidopsis thaliana (L.) Heynh. Mutat. Res. 93, 109–123. Koprek, T., McElroy, D., Louwerse, J., Williams-Carrier, R., and Lemaux, P. G. (2000). An efficient method for dispersing Ds elements in the barley genome as a tool for determining gene function. Plant J. 24, 253–263.

Mutagenesis and Functional Genomics in Cereals

407

Koprek, T., Rangel, S., McElroy, D., Louwerse, J. D., Williams-Carrier, R. E., and Lemaux, P. G. (2001). Transposon-mediated single-copy gene delivery leads to increased transgene expression stability in barley. Plant Physiol. 125, 1354–1362. Koprek, T., Zhao, T., Zimmermann, D., and Schulze-Lefert, P. (2003). SystematicAc/Ds transposon mutagenesis in barley. In ‘‘From Biodiversity to Genomics: Breeding Strategies for Small Grain Cereals in the Third Millennium.’’ EUCARPIA Proceedings, pp. 453–456. Krysan, P. J., Young, J. C., and Sussman, M. R. (1999). T-DNA as an insertional mutagen in Arabidopsis. Plant Cell 11, 2283–2290. Kumar, A., and Bennetzen, J. L. (1999). Plant retrotransposons. Annu. Rev. Genet. 33, 479–532. Kumar, C. H., Wing, R. A., and Sundaresan, V. (2005). Efficient insertional mutagenesis in rice using the maize En/Spm elements. Plant J. 44, 879–892. Langaee, T., and Ronaghi, M. (2005). Genetic variation analysis by pyrosequencing. Mutat. Res. 573, 96–102. Lee, H., Suh, S. S., Park, E., Cho, E., Ahn, J. H., Kim, S. G., Lee, J. S., and Kwon, Y. M.Lee, I. (2000). The AGAMOUS_LIKE 20 MADS domain protein integrates floral inductive pathways in Arabidopsis. Genes Dev. 14, 2366–2376. Lee, S., Jung, K. H., An, G. H., and Chung, Y. Y. (2004). Isolation and characterization of a rice cysteine protease gene, OSCP1, using T-DNA gene-trap system. Plant Mol. Biol. 54, 755–765. Leung, H., and An, G. (2004). Rice functional genomics: Large-scale gene discovery and applications to crop improvement. Adv. Agron. 82, 55–111. Li, S. L., and Redei, G. P. (1969). Estimation of mutation rate in autogamous diploids. Radiat. Bot. 9, 125. Li, X., and Zhang, Y. (2002). Reverse genetics by fast neutron mutagenesis in higher plants. Funct. Integr. Genomics 2, 254–258. Li, X., Song, Y., Century, K., Straight, S., Ronald, P., Xinnian, D., Lassner, M., and Zhang, Y. (2001). A fast neutron deletion mutagenesis-based reverse genetics system for plants. Plant J. 27, 235–242. Li, X., Lassner, M., and Zhang, Y. (2002a). Deleteagene: A fast neutron deletion mutagenesis-based gene knockout system for plants. Comp. Funct. Genomics 3, 158–160. Li, Q., Liu, Z., Monroe, H., and Culiat, C. T. (2002b). Integrated platform for detection of DNA sequence variants using capillary array electrophoresis. Electrophoresis 23, 1499–1511. Lindsey, K., Wei, W., Clarke, M. C., McArdle, H. F., Rooke, L. M., and Topping, J. F. (1993). Tagging genomic sequences that direct transgene expression by activation of a promoter trap in plants. Transgenic Res. 2, 33–47. Liu, Y.-G., and Whittier, R. (1995). Thermal asymmetric interlaced PCR: Automatable amplification and sequencing of insert-end fragments from PI and YAC clones for chromosome walking. Genomics 25, 674–681. Liu, D., Zhang, S., Fauque, T. C., and Crawford, N. M. (1999a). The Arabidopsis transposon Tag1 is active in rice, undergoing germinal transposition and restricted, late somatic excision. Mol. Gen. Genet. 262, 413–420. Liu, L. X., Spoerke, J. M., Mulligan, E. L., Chen, J., Reardon, B., Westlund, B., Sun, S., Abel, K., Armstrong, B., Hardiman, G., King, J., McCague, L., et al. (1999b). Highthroughput isolation of Caenorhabditis elegans deletion mutants. Genome Res. 9, 859–867. Maes, T., De Keukeleire, P., and Gerats, T. (1999). Plant technology. Trends Plant Sci. 3, 90–96. Magnard, J. L., Heckel, T., Massonneau, A., Wisniewski, J. P., Cordelier, S., Lassagne, H., Perez, P., Dumas, C., and Rogowsky, P. M. (2004). Morphogenesis of maize embryos requires ZmPRPL35-1 encoding a plastid ribosomal protein. Plant Physiol. 134, 649–663.

408

H. S. Balyan et al.

Maheshwari, S. C., Tyagi, A. K., Malhotra, K., and Sopory, S. K. (1980). Induction of haploidy from pollen grains in angiosperms-the current status. Theor. Appl. Genet. 58, 193–206. Maluszynski, M., Ahloowalia, B. S., and Sigurbjornsson, B. (1995). Application of in vivo and in vitro mutation techniques for crop improvement. Euphytica 85, 303–315. Maluszynski, M., Kasha, K. J., Forster, B. P., and Szarekjo, I. (2003). ‘‘Doubled Haploid Production in Crop Plants: A Manual.’’, Kluwer Academic Publisher, Dordrecht, p. 428. Manosalva, P., Ryba-White, M., Wu, C., Lei, C., Baraoidan, M., Leung, H., and Leach, J. (2003). A PCR-based screening strategy for detecting deletions in defence response genes in rice. Phytopathology 93, S57. Mao, L., Wood, T. C., yu, Y., Budiman, M. A., Tomkins, J., Woo, S., Sasinowski, M., Presting, G., Frisch, D., Goff, S., Dean, R. A., and Wing, R. A. (2000). Rice transposable elements: A survey of 73,000 sequence-tagged connectors. Genome Res. 10, 982–990. Margulies, M., Egholm, M., Altman, W. E., Attiya, S., Bader, J. S., Bemben, L. A., Berka, J., Braverman, M. S., Chen, Y. J., Chen, Z. T., Dewell, S. B., Du, L., et al. (2005). Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380. Masson, P., and Fedoroff, N. (1989). Mobility of the maize Suppressor-mutator element in transgenic tobacco cells. Proc. Natl. Acad. Sci. USA 86, 2219–2223. Martienssen, R. A. (1998). Functional genomics: Probing plant gene function and expression with transposon. Proc. Natl. Acad. Sci. USA 95, 2021–2026. Matsumoto, T., Wu, J. Z., Kanamori, H., Katayose, Y., Fujisawa, M., Namik, N., Mizuno, H., Yamamoto, K., Antonio, B. A., et al. (2005). The map-based sequence of the rice genome. Nature 436, 793–800. May, B. P., and Martienssen, R. A. (2003). Transposon mutagenesis in the study of plant development. Crit. Rev. Plant Sci. 22, 1–35. May, B. P., Liu, H., Vollbrecht, E., Senior, L., Rabinowicz, P. D., Roh, D., Pan, X., Stein, L., Freeling, M., Alexander, D., and Martiessen, A. (2003). Maize-targeted mutagenesis: A knockout resource for maize. Proc. Natl. Acad. Sci. USA 100, 11541–11546. McCallum, C. M., Comai, L., Greene, E. A., and Henikoff, S. (2000a). Targeted screening for induced mutations. Nat. Biotechnol. 18, 455–457. McCallum, C. M., Comai, L., Greene, E. A., and Henikoff, S. (2000b). Targeting induced local lesions in genomes (TILLING) for plant functional genomics. Plant Physiol. 123, 439–442. McCarty, D. R., Settles, A. M., Suzuki, M., Tan, B. C., Latshaw, S., Porch, T., Robin, K., Baier, J., Avigne, W., Lai, J. S., Messing, J., Koch, K. E., et al. (2005). Steady-state transposon mutagenesis in inbred maize. Plant J. 44, 52–61. McClintock, B. (1950). The origin and behavior of mutable loci in maize. Proc. Natl. Acad. Sci. USA. 36, 344–355. Mejlhede, N., Kyjovska, Z., Backes, G., Burhenne, K., Rasmussen, S. K., and Jahoor, A. (2006). EcoTILLING for the identification of allelic variation in the powdery mildew resistance genes mlo and Mla of barley. Plant Breed. 125, 461–467. Middendorf, L. R., Bruce, J. C., Bruce, R. C., Eckles, R. D., Grone, D. L., Roemer, S. C., Sloniker, G. D., Steffens, D. L., Sutter, S. L., and Brumbaugh, J. A. (1992). Continuous on-line DNA sequencing using a versatile infrared laser scanner/electrophoresis apparatus. Electrophoresis 13, 487–494. Miyao, A., Tanaka, K., Murata, K., Sawaki, H., Takeda, S., Abe, K., Shinozuka, Y., Onosato, K., and Hirochika, H. (2003). Target site specificity of the Tos17 retrotransposon shows a preference for insertion within genes and against insertion in retrotransposon-rich regions of the genome. Plant Cell 15, 1771–1780. Moon, S., Jung, K. H., Lee, D. E., Lee, D. Y., Lee, J., An, K., Kang, H. G., and An, G. (2006). The rice FON1 gene controls vegetative and reproductive development by regulating shoot apical meristem size. Mol. Cells 21, 147–152.

Mutagenesis and Functional Genomics in Cereals

409

Mori, A., Sugimoto, K., Satoh, K., Okabe, K., Toki, S., Ugaki, M., Hirochika, H., and Kikuchi, S. (2000). Toward activation tagging in rice. In ‘‘Proceedings of the Sixth International Congress of Plant Mol. Biol,’’ pp. S1–S57. Murai, N., Li, Z., Kawagoe, Y., and Hayashimamoto, A. (1991). Transposition of the maize activator element in transgenic rice plants. Nucleic Acids Res. 19, 617–622. Nan, G. L., and Walbot, V. (2006). Constructing MightyMu-tagged lines for gene and enhancer trapping. Maize Genet. Conf. Abstr. 48, 243. Nadeau, J. H., and Frankel, W. N. (2000). The roads from phenotypic variation to gene discovery: Mutagenesis versus QTLs. Nat. Genet. 25, 381–384. Nakazaki, T., Okumoto, Y., Horibata, A., Yamahira, S., Teraishi, M., Nishida, H., Inoue, H., and Tanisaka, T. (2003). Mobilization of a transposon in the rice genome. Nature 421, 170–172. Nakagawa, Y., Machida, C., Machida, Y., and Toriyama, K. (2000). Frequency and pattern of transposition of maize transposable element Ds in transgenic rice plants. Plant Cell Physiol. 41, 733–742. Nakamura, H., Hakata, M., Amano, K., Miyao, A., Toki, N., Kajikawa, M., Pang, J., Higashi, N., Ando, S., Toki, S., Fujita, M., Enju, A., et al. (2007). A genome-wide gainof-function analysis of rice genes using the FOX-hunting system. Plant Mol. Biol. 65, 357–371. Ng, P. C., and Henikoff, S. (2001). Predicting deleterious aminoacid substitutions. Genome Res. 11, 863–874. Ng, P. C., and Henikoff, S. (2002). Accounting for human polymorphisms predicted to affect protein function. Genome Res. 12, 436–446. Nguyen, T.-V., Thu, T. T., Claeys, M., and Angenon, G. (2007). Agrobacterium-mediated transformation of sorghum (Sorghum bicolor (L.) Moench) using an improved in vitro regeneration system. Plant Cell Tissue Organ Cult. 99, 155–164. Nonomura, K. I., Miyoshi, K., Eiguchi, M., Suzuki, T., Miyao, A., Hirochika, H., and Kurata, N. (2003). The MSP1 gene is necessary to restrict the number of cells entering into male and female sporogenesis and to initiate anther wall formation in rice. Plant Cell 15, 1728–1739. Nonomura, K. I., Nakano, M., Murata, K., Miyoshi, K., Eiguchi, M., Miyao, A., Hirochika, H., and Kurata, N. (2004). An insertional mutation in the rice PAIR2 gene, the ortholog of Arabidopsis ASY1, results in a defect in homologous chromosome pairing during meiosis. Mol. Genet. Genomics 271, 121–129. Odell, J. T., Nagy, F., and Chua, N. (1985). Identification of DNA sequences required for activity of the cauliflower mosaic virus 35S promoter. Nature 313, 810–812. O’Kennedy, M. M., Grootboom, A., and Shewry, P. R. (2006). Harnessing sorghum and millet biotechnology for food and health. J. Cereal Sci. 44, 224–235. Okushima, Y., Overvoorde, P. J., Arima, K., Alonso, J. M., Chan, A., Chang, C., Ecker, J. R., Hughes, B., Lui, A., Nguyen, D., Onodera, C., Quach, H., et al. (2005). Functional genomic analysis of the AUXIN RESPONSE FACTOR gene family members in Arabidopsis thaliana: Unique and overlapping functions of AFR7 and AFR19. Plant Cell 17, 444–463. Olsen, O., Wang, X., and von Wttestein, D. (1993). Sodium azide mutagenesis: Preferential generation of A.T G.C transitions in the barley Ant18 gene. Proc. Natl. Acad. Sci. USA 90, 8043–8047. Oleykowski, C. A., Bronson Mullins, C. R., Godwin, A. K., and Yeung, A. T. (1998). Mutation detection using a novel plant endonuclease. Nucleic Acids Res. 26, 4597–4602. Osborne, B. I., Corr, C. A., Prince, J. P., Hel, R., Tanksley, S. D., McCormick, S., and Baker., B. (1991). AC transposition from a T-DNA can generate linked and unlinked clusters of insertions in the tomato genome. Genetics. 129, 833–844. Ostergaard, L., and Yanofsky, M. F. (2004). Establishing gene function by mutagenesis in Arabidopsis thaliana. Plant J. 39, 682–696.

410

H. S. Balyan et al.

Pang, S. Z., DeBoer, D. L., Wan, Y., ye, G., Layton, J. G., Neher, M. K., Armstrong, C. L., Fry, J. E., Hinchee, M. A., and Fromm, M. E. (1996). An improved green fluorescent protein gene as a vital marker in plants. Plant Physiol. 112, 893–900. Parinov, S., and Sundaresan, V. (2000). Functional genomics in Arabidopsis: Large scale mutagenesis complements the genome sequencing project. Curr. Opin. Biotechnol. 11, 157–161. Parinov, S., Sevugan, M., De, Y., Yang, W., Kumaran, M., and Sundaresan, V. (1999). Analysis of flanking sequence from dissociation insertion lines: A database for reverse genetics in Arabidopsis. Plant Cell 11, 2263–2270. Paterson, A. H., Freeling, M., and Sasaki, T. (2005). Gains of knowledge: Genomics of model cereals. Genome Res. 15, 1643–1650. Pereira, A. (1998). Heterologous transposon tagging system. In ‘‘Transgenic Plant Research’’ (K. Lindsey, Ed.), pp. 91–108. Harwood Academic Publishers, UK. Pereira, A., and Saedler, H. (1989). Transposition of the maize En/Spm element in transgenic tobacco. EMBO J. 8, 1315–1321. Prigge, M. J., Otsuga, D., Alonso, J. M., Ecker, J. R., Drews, G. N., and Clark, S. E. (2005). Class III homoeodomain-leucine zipper gene family members have overlapping antagonistic, and distinct roles in Arabidopsis development. Plant Cell 17, 61–76. Prina, A. R., and Favert, E. A. (1983). Parabolic effect is sodium azide mutagenesis in barley. Hereditas 98, 89–94. Qiu, P., Shandilya, H., D’Alessio, J. M., O’ Connor, K., Durocher, J., and Gerard, G. F. (2004). Mutation detection using Surveyor nuclease. Biotechniques 36, 702–707. Rabinowicz, P. D., and Bennetzen, J. L. (2006). The maize genome as a model for efficient sequence analysis of large plant genomes. Curr. Opin. Plant Biol. 9, 149–156. Raizada, M. N., Nan, G., and Walbot, V. (2001). Somatic and germinal mobility of the RescueMu transposon in transgenic maize. Plant Cell 13, 1587–1608. Robertson, D. S. (1978). Characterization of Mutator system in maize. Mutat. Res. 51, 21–28. Ryoo, N., Yu, C., Park, C.-S., Baik, M.-Y., Park, I. M., Cho, M.-H., Bhoo, S. H., An, G., Hahn, T.-R., and Jeon, J. S. (2007). Knockout of starch synthase gene OsSSIIIa/Flo5 causes white-core floury endosperm in rice (Oryza sativa L.). Plant Cell Rep. 26, 1083–1095. Ryu, C.-H., You, J.-H., Kang, H.-G., Hur, J., Kim, Y.-H., han, M.-J., An, K., Chung, B.-C., Lee, C.-H., and An, G. (2004). Generation of T-DNA tagging lines with a bidirectional gene trap vector and the establishment of an insertion-site database. Plant Mol. Biol. 54, 489–502. Sallaud, C., Meynard, D., van Boxtel, J., Gay, C., Be`s, M., Brizard, J. P., Larmande, P., Ortega, D., Raynal, M., Portefaix, M., Ouwerkerk, P. B. F., Rueb, S., et al. (2003). Highly efficient production and characterization of T-DNA plants for rice (Oryza sativa L.) functional genomics. Theor. Appl. Genet. 106, 1369–1408. Sallaud, C., Gay, C., Larmande, P., Bes, M., Piffanelli, P., Piegu, B., Droc, G., Regad, F., Bourgeois, E., Meynard, D., Perin, C., Sabau, X., et al. (2004). High throughput T-DNA insertion mutagenesis in rice: A first step towards in silico reverse genetics. Plant J. 39, 450–464. Salmeron, J. M., Oldroyd, G. E. D., Rommens, C. M. T., Scofield, S. R., Kim, H., Lavelle, D. T., dahlbeck, D., and Staskawwicz, B. J. (1996). Tomato Prf is the member of the leucine-rich class of plant disease resistance genes and lies embedded within the Pto kinase gene cluster. Cell 86, 123–133. Sato, Y., Sentoku, N., Miura, Y., Hirochika, H., Kitano, H., and Matsuoka, M. (1999). Loss-of-function mutations in the rice homeobox gene OSH15 affect the architecture of internodes resulting in dwarf plants. EMBO J. 18, 992–1002. Service, R. F. (2006). The race for the $ 1000 genome. Science 311, 1544–1546.

Mutagenesis and Functional Genomics in Cereals

411

Sessions, A., Burke, E., Presting, G., Aux, G., McElver, J., Patton, D., Dietrich, B., Ho, P., Bacwaden, J., Ko, C., Clarke, J. D., Cotton, D., et al. (2002). A high-throughput Arabidopsis reverse genetics system. Plant Cell 14, 2985–2994. Settles, A. M. (2005). Maize community resources for forward and reverse genetics. Maydica 50, 405–414. Settles, A. M., Latshaw, S., and McCarty, D. R. (2004). Molecular analysis of high-copy insertion sites in maize. Nucleic Acids Res. 32, e54. Settles, A. M., Holding, D. R., Tan, B. C., Latshaw, S. P., Liu, J., Suzuki, M., Li, L., O’Brien, B. A., Fajardo, D. S., Wroclawska, E., Tseung, C.-W., Lai, J., et al. (2007). Sequence-indexed mutations in maize using the UniformMU transposon-tagging population. BMC Genomics 8, 116–127. Sharp, P., Dong, C., Dalton-Morgan, J., and Vincent, K. (2006). Identification of GBSS and puroindoline allelic variation by TILLING. In ‘‘Abstracts 2006 ITMI Workshop & Australian Centre for Plant Functional Genomics Symposium held at McCracken Country Club,’’ Victor harbor, South Australia. (27th—31st August, 2006), p. 49. Shendure, J., Porreca, G. J., Repass, N. B., Lin, X., McCutcheon, J. P., Rosenbaum, A. M., Wang, M. D., Zhang, K., Mitra, R. D., and Church, G. M. (2005). Accurate multiplex polony sequencing of an evolved bacterial genome. Science 309, 1728–1732. Shi, J., Wang, H., Wu, Y., Hazebroek, J., Meeley, R. B., and Ertl, D. S. (2003). The maize low phytic acid mutant lpa2 is caused by mutation in an inositol phosphate kinase gene. Plant Physiol. 131, 1–9. Shi, J., Wang, H., Hazebroek, J., Ertl, D. S., and Harp, T. (2005). The maize low-phytic acid 3 encodes a myo-inositol kinase that plays a role in phytic acid biosynthesis in developing seeds. Plant J. 42, 708–719. Shimamoto, K., Miyazaki, C., Hashimoto, H., Izawa, T., Itoh, K., Terada, R., Inagaki, Y., and Iida, S. (1993). Trans-activation and stable integration of the maize transposable element Ds cotransfected with the Ac transposase gene in transgenic rice plants. Mol. Gen. Genet. 239, 354–360. Shrawat, A. K., and Lorz, H. (2006). Agrobacterium-mediated transformation of cereals: A promising approach crossing barriers. Plant Biotechnol. J. 4, 575–603. Singh, M., Lewis, P. E., Hardeman, K., Bai, L., Rose, J. K., Mazourek, M., Chomet, P., and Brutnell, T. P. (2003). Activator mutagenesis of the pink scutellum 1/viviparous7 locus of maize. Plant Cell 15, 874–884. Singh, R. P., Spielmeyer, W., Apples, R. A., Huberta-Espino, Mc Fadden, H., Luo, M. C., Cloutier, S., Chalhoub B., Dann, D., and Lagudah., E. S. (2006). Mutational and comparative genomic analysis of a multi-pathogenic disease resistance locus in wheat. In Abstacts International Triticeae Mapping Initiative Workshop. Victor. Singh, J., Zhang, S., Chen, C., Cooper, L., Bregitzer, P., Sturbaum, A., Hayes, P. M., and Lemaux, P. G. (2006). High-frequency Ds remobilization over multiple generations facilitates gene tagging in large genome cereals. Plant Mol. Biol. 62, 937–950. Slade, A. J., Fuerstenberg, S. I., Loeffler, D., Steine, M. N., and Facciotti, D. (2005). A reverse genetic, non-transgenic approach to wheat crop improvement by TILLING. Nat. Biotechnol. 23, 75–81. Smith, N. A., Singh, S. P., Wang, M. B., Stoutjesdijk, P. A., Green, A. G., and Waterhouse, P. M. (2000). Total silencing by intron-spliced hairpin RNAs. Nature 407, 319–320. Smooker, P. M., and Cotton, R. G. (1993). The use of chemical reagents in the detection of DNA mutations. Mutat. Res. 288, 65–77. Springer, P. S. (2000). Gene traps: Tools for plant development and genomics. Plant Cell 12, 1007–1020. Stadler, I. J. (1929). Chromosome number and the mutation rate in Avena and Triticum. Proc. Natl. Acad. Sci. USA 15, 876–881.

412

H. S. Balyan et al.

Stemple, D. L. (2004). TILLING-a high throughput harvest for functional genomics. Nat. Rev. Genet. 5, 1–7. Strader, C. L., Zale, Z. M., and Steber, C. M. (2004). Mutation- and transposon-based approaches for the identification of genes for pre-harvest sprouting in wheat. In Vitro Cell. Dev. Biol. Plant 40, 256–259. Sugimoto, K., Otsuki, Y., Saji, S., and Hirochika, H. (1994). Transposition of the maize Ds element from viral vector to the rice genome. Plant J. 5, 863–871. Sun, T. P., Goodman, H. M., and Ausubel, F. M. (1992). Cloning the Arabidopsis GA1 locus by genomic subtraction. Plant Cell 4, 119–128. Sundaresan, V. (1996). Horizontal spread of transposon mutagenesis: New uses of old elements. Trends Plant Sci. 1, 184–191. Sundaresan, V., Springer, P., Volpe, T., Haward, S., Jones, J. D., Dean, C., Ma, H., and Martienssen, R. (1995). Patterns of gene action in plant development revealed by enhancer trap and gene trap transposable elements. Genes Dev. 9, 1797–1810. Suzuki, M., Settles, A. M., Tseung, C.-W., Li, O.-B., Latshaw, S., Wu, S., Porch, T. G., Schmelz, E. A., James, M. G., and McCarty, D. R. (2006). The maize viviparous15 locus encodes the molybdopterin synthase small subunit. Plant J. 45, 264–274. Szarejko, I., and Forster, B. P. (2007). Doubled haploidy and induced mutation. Euphytica 158, 359–370. Tabuchi, M., Sugiyama, K., Ishiyama, K., Inoue, E., Sato, T., Takahashi, H., and Yamaya, T. (2005). Severe reduction in growth rate and grain filling of rice mutants lacking OsGS1;1, a cytosolic glutamine synthetase1;1. Plant J. 42, 641–651. Taji, A., Kumar, P., and Lakshmanan, P. (2001). ‘‘In Vitro Plant Breeding.’’ The Howerth Press, p. 161. Takano, M., Kanegae, H., Shinomura, T., Miyao, A., Hirochika, H., and Furuya, M. (2001). Isolation and characterization of rice phytochrome A mutants. Plant Cell 13, 521–534. Takeda, S., Sugimoto, K., Otsuki, H., and Hirochika, H. (1998). Transcriptional activation of the tobacco retrotransposon Tto1 by wounding and methyl jasmonate. Plant Mol. Biol. 36, 365–376. Takumi, S., Murai, K., Mori, N., and Nakamura, C. (1999). Trans-activation of maize Ds transposable element in transgenic wheat plants expressing the Ac transposase gene. Theor. Appl. Genet. 98, 947–953. Tanaka, K., Murata, K., Yamazaki, M., Onosato, K., Miyao, A., and Hirochika, H. (2003). Three distinct rice cellulose synthase catalytic subunit genes required for cellulose synthesis in the secondary wall. Plant Physiol. 133, 73–83. Tani, H., Chen, X., Nurmberg, P., Grant, J. J., hini, A., Gilroy, E., SantaMaria, M. C., Birch, P. R., and Llyod, D. B. (2004). Activation tagging in plants: A tool for gene discovery. Funct. Integr. Genomics 4, 258–266. Taylor, N., and Greene, E. A. (2003). PARSESNP: A tool for the analysis of nucleotide polymorphisms. Nucleic Acids Res. 31, 3808–3811. Till, B. J., Reynolds, S. H., Greene, E. A., Codomo, C. A., Enns, L. C., Johnson, J. E., Burtner, C., Odden, A. R., Young, K., Taylor, N. E., Henikoff, J. G., Comai, L., et al. (2003). Large-scale discovery of induced point mutations with high-throughput TILLING. Genome Res. 13, 524–530. Till, B., Burtner, C., Comai, L., and Henikoff, S. (2004a). Mismatch cleavage by singlestrand specific nucleases. Nucleic Acids Res. USA 32, 2632–2641. Till, B. J., Reynolds, S. H., Weil, C., Springer, N., Burtner, C., Young, K., Bowers, E., Codomo, C. A., Enns, L. C., Odden, A. R., Greene, E. A., Comai, L., et al. (2004b). Discovery of induced point mutations in maize genes by TILLING. BMC Plant Biol. 4, 12–23.

Mutagenesis and Functional Genomics in Cereals

413

Till, B. J., Cooper, J., Tai, T. H., Colowit, P., Greene, E. A., Henikoff, S., and Comai, L. (2007). Discovery of chemically induced mutations in rice by TILLING. BMC Plant Biol. 7, 19–30. Tissier, A. F., Marillonnet, S., Klimyuk, V., Patel, K., Torres, M. A., Murphy, G., and Jones, J. D. G. (1999). Multiple independent defective Suppressor-mutator transposons insertions in Arabidopsis: A tool for functional genomics. Plant Cell 11, 1841–1852. To, J. P. C., Haberer, G., Ferreira, F. J., Derue`re, J., Mason, M. G., Schaller, G. E., Alonso, J. M., Ecker, J. R., and Kieber, J. J. (2004). Type-A Arabidopsis response regulators are partially redundant negative regulators of cytokinin signaling. Plant Cell 16, 658–671. Triglia, T., Peterson, M. G., and Kemp, D. J. (1988). A procedure for in vitro amplification of DNA segments that lie outside the boundaries of known sequences. Nucleic Acids Res. 16, 8186. Tsugane, K., Maekawa, M., Takagi, K., Takahara, H., Qian, Q., Eun, C. H., and Iida, S. (2006). An active DNA transposon nDart causing leaf variegation and mutable dwarfism and its related elements in rice. Plant J. 45, 46–57. Tyagi, A. K., and Mohanty, A. (2000). Rice transformation for crop improvement and functional genomics. Plant Sci. 158, 1–18. Underhill, P. A., Jin, L., Lin, A. A., Mehdi, S. Q., Jenkins, T., Vollrath, D., Davis, R. W., Cavalli-Sfroza, L. L., and Oefner, P. J. (1997). Detection of numerous Y chromosome biallelic polymorphisms by denaturing high-performance liquid chromatography. Genome Res. 7, 996–1005. Upadhyaya, N. M., Zhou, X. R., Zhu, Q. H., et al. (2002). An iAc/Ds gene and enhancer trapping system for insertional mutagenesis in rice. Funct. Plant Biol. 29, 547–559. Upadhyaya, N. M., Zhu, Q. H., Zhou, X. R., Eamens, A. L., Hoque, M. S., Ramm, K., Shivakkumar, R., Smith, K. F., Pan, S. T., Li, S. Z., Peng, K. F., Kim, S. J., et al. (2006). Dissociation (Ds) constructs, mapped Ds launch pads and a transiently-expressed transposase system suitable for localized insertional mutagenesis in rice. Theor. Appl. Genet. 112, 1326–1341. van Enckevort, L. J., Droc, G., Piffanelli, P., Greco, R., Gagneur, C., Weber, C., Gonza´lez, V. M., Cabot, P., Fornara, F., Berri, S., Miro, B., Lan, P., et al. (2005). EU-OSTID: A collection of transposon insertional mutants for functional genomics in rice. Plant Mol. Biol. 59, 99–110. Von Wettstein-Knowles, P. (1992). Cloned and mapped genes: Current status. In ‘‘Barley: Genetics, Biochemistry, Molecular Biology and Biotechnology’’ (P. Shewry, Ed.), pp. 73–98. CAB Int., Wallingford, Oxon. Walbot, V. (1992). Strategies for mutagenesis and gene cloning using transposon tagging and T-DNA insertional mutagenesis. Annu. Rev. Plant. Physiol. Plant Mol. Biol. 43, 49–82. Walbot, V. (2000). Saturation mutagenesis using maize transposons. Curr. Opin. Plant Biol. 3, 103–107. Wang, G. L., Wu, C., Zeng, L., He, C., Baraoidan, M., da Silva, F. D. G., Williams, C. E., Ronald, P. C., and Leung, H. (2004). Isolation and characterization of rice mutants compromised in Xa21-mediated resistance to X. oryzae pv. oryzae. Theor. Appl. Genet. 108, 379–384. Waugh, R., Leader, D. J., McCallum, N., and Caldwell, D. (2006). Harvesting the potential of induced biological diversity. Trends Plant Sci. 11, 71–78. Weigel, D., Ahn, J. H., Blazquez, M. A., Borevitz, J. O., Christensen, S. K., Fankhauser, C., Ferrandiz, C., Kardailsky, I., Malancharuvil, E. J., Neff, M. M., Nguyen, J. T., Sato, S., et al. (2000). Activation tagging in Arabidopsis. Plant Physiol. 122, 1003–1013. Weil, C. F., and Monde, R.-A. (2007). Getting the point-mutations in maize. Crop Sci. 47, S-60–S-67.

414

H. S. Balyan et al.

Wisman, E., Cardon, G. H., Fransz, P., and Saedler, H. (1998). The behaviour of the autonomous maize transposable element En/Spm in Arabidopsis thaliana allows efficient mutagenesis. Plant Mol. Biol. 37, 989–999. Wu, C., Li, X., Yuan, W., Chen, G., Kilian, A., Li, J., Xu, C., Li, X., Zahou, D. X., Wang, S., and Zhang, Q. (2003). Development of enhancer trap lines for function analysis of the rice genome. Plant J. 35, 418–427. Wu, J. L., Wu, C. J., Lei, C. L., Baraoidan, M., Bordeos, A., Madamba, M. R. S., RamosPamplona, M., Mauleon, R., Portugal, A., Ulat, V. J., Bruskiewich, R., Wang, G. L., et al. (2005). Chemical- and irradiation-induced mutants of indica rice IR64 for forward and reverse genetics. Plant Mol. Biol. 59, 85–97. Yamaguchi, T., Lee, D. Y., Miyao, A., Hirochika, H., An, G. H., and Hirano, H. Y. (2006). Functional diversification of the two C-class MADS box genes OSMADS3 and OSMADS58 in Oryza sativa. Plant Cell 18, 15–28. Yeung, A. T., Hattangadi, D., Blakesley, L., and Nicolas, E. (2005). Enzymatic mutation detection technologies. Biotechniques 38, 749–758. Yin, Z., and Wang, G. L. (2000). Evidence of multiple complex patterns of T-DNA integration into the rice genome. Theor. Appl. Genet. 100, 461–470. Yu, J., Hu, S., Wang, J., Wong, G., Li, S., Liu, B., Deng, Y., Dai, L., Zhou, Y., Zhang, X., Cao, M., Liu, J., et al. (2002). A draft sequence of the rice genome (Oryza sativa L. spp. indica). Science 296, 79–92. Zhang, Y., Tessaro, M. J., Lassner, M., and Li, X. (2003). Knockout analysis of Arabidopsis transcription factors TGA2, TGA5, and TGA6 reveals their redundant and essential roles in systemic acquired resistance. Plant Cell 15, 2647–2653. Zhang, X. L., Colleoni, C., Ratushna, V., Sirghle-Colleoni, M., James, M. G., and Myers, A. M. (2004). Molecular characterization demonstrates that the Zea mays gene sugary2 codes for the starch synthase isoform SSIIa. Plant Mol. Biol. 54, 865–879. Zhao, T., Palotta, M., Langridge, P., Prasad, M., Graner, A., Schulze-Lefert, P., and Koprek, T. (2006). Mapped Ds/T-DNA launch pads for functional genomics in barley. Plant J. 47, 811–826. Zerr, T., and Henikoff, S. (2005). Automated band mapping in electrophoretic gel images using background information. Nucleic Acids Res. 33, 2806–2812.

Index

A Activation tagging, 364–365. See also Cereals and gene trap systems, 375–376 null mutants created by T-DNA, 379 waxy mutants by ubiDs barley, for amylose, 400 Active measurement device, 221–222. See also Ammonia sampling and measurement Aegilops tauschii, 291 Agricultural ammonia emission, 203 Agricultural nonpoint source pollution (AGNPS), 23 Agrobacterium tumefaciens, 362 Agroenvironmental sustainability, 3 Agronomic efficiency of fertilizer N (AEN), 150–151 Agropyron elongatum, 278 Air pollution and flooded/non-flooded systems. See also Emissions ammonia (NH3), 327 greenhouse gas emission control, 334–335 methane emission control, 331–334 nitrous oxide (N2O), 325–326 rumen processes, efficiency of, 328 rumen proteolysis, interspecific variation in, 329–331 WSC and protein content, 328–329 Alcaligenes, 69 Allelic series mutation and knock-out mutation, 399–400 Ammoniacal copper arsenate (ACA), 49 Ammoniacal copper zinc arsenate (ACAA), 49 Ammonia (NH3). See also Tanniferous forage species emission control, in atmosphere, 327 oxidation of, 334 volatilization, 9, 325 Ammonia quick test (AQT), 227 Ammonia sampling and measurement air sampling, 206–207 location and time of, 208–210 methods and devices for, 211–218 sampling method selection, 218–221 volume of, 210 concentration data ammonia measurement standards, 255 data collection, 243–244 error reduction, 245–254 precision and bias, 244–245

concentration measurement features of techniques for, 221–224 techniques selection of, 224–225 determination of, 205–206 emission, pollution and health hazards, 203–204 measurement techniques, 204 ChemcassetteÒ detection system, 239–240 chemiluminescence analyzer, 235–238 electrochemical sensor, 238–239 fourier transform infrared spectroscopy, 230–231 gas detection tubes, 228–230 infrared gas analyzer, 231–234 measuring devices, detection of, 241–243 solid-state/electronic sensor, 240–241 ultraviolet differential optical absorption spectroscopy, 234–235 wet methods, 225–228 Antirhinum majus, 369 Apium graveolens, 392 Arabidopsis spp., 361–365, 375, 382, 385–386, 392, 397, 400 Arabidopsis TILLING Project (ATP), 389–390, 393 ArsC gene, 104 Arsenate (AsO4). See also Arsenic (As) adsorption on amorphous iron and aluminum oxides, FTIR investigation, 53 on oxides and clays, pH dependent, 56 biotransformation, 76 breakthrough curves, 74 diffusion, to reaction sites within soil matrix, 60–61 Freundlich equation and distribution coefficient (Kd) for, 66 and glucose-fermenting microorganism, 71 isotherms of desorption, 63 Langmuir adsorption coefficient (KL) for, 61 occluded within short range ordered materials, 55 and phosphate equally adsorbed on goethite, 58 sorbed vs. time during adsorption–desorption, 95 time-dependent sorption isotherm, 61 Arsenical pesticide, 50

415

416 Arsenic (As) biogeochemistry of, 52 binding mechanisms in soils, 54 desorption, 62–64 heterogeneous oxidation, 66–69 pH dependency, 56–57 precipitation and arsenic retention, 55 reduction and oxidation, microbialmediated, 69–72 solution composition, effect of, 57–59 sorption kinetics, 60–62 sulfides, reaction with, 64–66 contaminated soils, remediation of, 101–104 capping of contaminated soils, 102 phytoremediation, 104 PRB and MNA, 103 solidification/stabilization and soil flushing, 102 sorption and precipitation, 103 empirical equilibrium models, 84–86 equilibrium thermodynamic models, 81–84 geochemical models, application of, 99 GLUE methodology, 99, 101 kinetic models kinetic dissolution, 95–96 kinetic reduction–oxidation, 96–97 kinetic retention, 90–95 in soils, 48–52 compounds containing and in poultry industry, 50 leaching and disposal, 51 surface complexation models, 86–89 toxicity of, 47–48 transport of, 73–78 factor affecting mobility, 75–76 mechanisms, 73–78 mobility and field conditions, 78–80 models, 97–99 Arsenicosis, 48 Arsenite (AsO3). See also Arsenic (As) adsorption capacity on minerals and soils, 57 on metal oxides and hydroxides, 53 biotransformation, 76 heterogeneous oxidation on mineral surface, 96–97 kinetics of oxidation in aerated soil and, 69 NOM and sorption kinetic, 59 oxidation kinetics of, 70 simulation competition on Fe and Al oxides and, 88 in soil solution under flooded conditions, 71 sorption on iron sulfides and, 55 toxicity on binding to sulfhydryl groups, 47 weathering process and oxidation, 64 Arsine gas (AsH3), 47 AsO4 and AsO3 sorption edges, 58 Avena sativa, 123

Index B Bacillus benzoevorans, 76 Barley crop, insertional mutagenesis on, 374, 378 Biofuels, 2 Biological N2 fixation (BNF), 159–160 Biomass production, 3, 8 Biomethylation, 52, 72 Black rust. See Stem rust, of wheat Brassica napus, 123 B. vulgaris, 274 C Caenorhabditis elegans, 385, 390, 397 Carbonate anions in soil, 58 Cauliflower mosaic virus (CaMV), 35S enhancer element, 365, 375 Cereals density of mutations determined in, 389 induce mutations in a variety of plant species, 382 insertion mutagenesis resources in, 376–378 mutagenesis and high-throughput functional genomics in insertional mutagenesis, 362–380 non-transgenic approaches, 381–399 phenomics platform for screening population, 400–401 TILING initiatives in, 393–396 ChemcassetteÒ detection system, 239–240 Chemical deletogens, 385, 397–398 Chemical weathering, 64. See also Arsenic Chemiluminescence (CL) analyzer, 218, 235–238 Chillgard refrigerant leak detection system, 233, 249 Chromated copper chromate (CCA), 49, 80 Citric acid (CA), 59, 228 Closed sampling method, 211–215. See also Ammonia sampling and measurement C:N ratio, 144, 158, 176, 331, 343 Codon Optimizedto Detect Deleterious LEsions (CODDLE) software, 393 Composting, off-field residue management, 131–132 Condensed tannins (CTs), 332 Conservation management, 3, 9, 16, 22, 28 Contaminant of concern (COC), 47 Crop residues, rice-based cropping systems bioenergy production of, 181 flooded soil and, 120 open-field burning of, 119 residue and straw removal of, 121 Crude protein (CP), 328–329 D Dactylis glomerata, 314 Data quality indicators (DQIs), 244

417

Index

DeleteageneTM, 360, 362, 382–383, 386, 390. See also Cereals dealing and, 397–399 Denuder device, 228 Detecting adduct lesions in genomes (DEALING), 360, 362, 381–382, 385, 390, 399 Diepoxybutane (DEB), 384–385, 388, 397–398 Digital elevation models (DEMs), 23–24 Dimethylarsinic acid (DMA), 71–72 Dissolved organic carbon (DOC), 59 Dissolved oxygen (DO), 64–66, 96 DNA pools for Ecotilling, 396 preparation of, 390 to screen deletion lines of C. elegans, 397 Doubled haploid (DH) mutation, 388 Dra¨ger sensor, in ammonia measurement, 238–239 Drosophila melanogaster, 390 Dry methods, 221. See also Ammonia sampling and measurement Dynamic flows modeling, 10. See also Geographic information systems (GIS) E Eco-TILLING, 396–398. See also Cereals Electrochemical (EC) ammonia sensors, 238–239 Electrophoretic mobility (EM), 53 Emissions black C in China, 154 greenhouse gas and enhance C sequestration, 341 for non-flooded crop, 175–176, 179 for rice, 154, 157–158 reducing to air, 325–335 of trace gases, spatial variability in, 15 Environmental protection agency (EPA), 46, 203 Erosion rates, 2 Escherichia coli, 69 Ethyl methane sulfonate (EMS), 382–383, 385, 388–389, 392, 395–397, 401 N-Ethyl-N-nitrosourea (ENU), 384–385 European Triticeae genomics initiative (ETGI), 358 Eutrophication, 316, 321–322, 328 Evapotranspiration, 2, 24 Extended X-ray absorption fine structure (EXAFS), 51, 53, 55–56, 62, 67 F Ferrihydrite, 54, 56–58, 60–61, 63, 71, 89, 92 Fertilizer, non-flooded crop residue management, 167, 172 Festuca arundinacea, 314 Festuca glaucescens, 329

Festuca pratensis, 314 Festulolium loliaceum, 336 Field level flows. See also Precision conservation to reduce N2O emissions, management for, 16–17 variable erosion and transport, 12–16 erosion patterns, 137Cs modeling, 14 predicted NO3-N leaching, spatial distribution of, 16 sand, spatial distribution of, 15 Flood damage control biodiverse mixtures, role of, 338–339 prevention of, 335–337 soil porosity and compaction for, 337–338 Fourier transform infrared spectroscopy (FTIR), 230–231 Full-length cDNA Over-eXpresser (FOX) gene, 377 Fulvic acid (FA), 59 G Gain-of-function mutations, 375–376 Gas detection tubes for ammonia, 228–230 Gasification, 181 Gas manufacturers intermediate standards (GMIS), 250 Gene function determination forward genetics approach, 361 reverse genetics approach, 359–361, 363–364 steps in, 359 Gene trap systems, 365, 376 Geographic information systems (GIS), 10, 12, 14, 20, 23–24, 27, 29 Geospatial technologies. See also Precision conservation desktop mapping, 6–8 surface modeling, 5 gibberellin 2-oxidase gene, 377 Global positioning systems (GPS), 4, 25–26, 28, 39 Global rust initiative, 289. See also Stem rust, of wheat Global warming, 2 Granule bound starch synthase (GBSS) I gene, 395 Grassland enhancing C sequestration in, 341–344 management and food production, 313–314 nutrient budget information for, 316 persistency and resilience genetic variation, implementation of, 340–341 interspecific hybridization, role of, 339–340 role of forage legumes, 334 stabilization of, 338

418

Index

Greenhouse gas emission from fertilizer production, 334–335 and non-flooded crop, 175–179 rice and, 154–158 Ground covering rice production system (GCRPS), 131 H Happy Seeder machine, 175 Humic acid (HA), 55 Hybrid single-particle lagrangian integrated trajectory (HYSPLIT), 284 Hydrous ferric oxide (HFO), 58, 72, 89 I In-field residue management, in rice cropping systems, 127 Infrared gas analyzer, 231–234 Insertion mutagenesis, 362–364. See also Cereals activation tagging, 375 gene trap systems, 376 T-DNA and, 364–369, 378–379 and transposon elements, 369–375, 379–381 Intergovernmental panel on climate change (IPCC), 313 International barley sequencing consortium (IBSC), 358 International rice functional genomics consortium (IRFGC), 376 International rice genome sequencing project, 358 International wheat genome sequencing consortium (IWGSC), 358 Iris fulva, 336 Italian ryegrass (IRG), 314, 329–330 K Knock-out mutation, 364–365. See also Cereals; Mutation T-DNA insertion mutation, 364–365 vs. allelic mutation, 399–400 L Leaf protein, half-life values, 330 Lolium multiflorum, 314, 329–330 Lolium perenne, 314, 330 Long-term experiments, on rice. See also Rice crop with continuous cropping of flooded rice in Philippines, 159 with removal of all above-ground biomass, 160 on rice–rice cropping systems, 152 and soil organic C, 180 Lotus corniculatus, 334 Lotus japonicus, 344

Lowland rice CH4 and N2O emissions from, 154–156 ecosystems in Asia, 124 effect of residue incorporation on yield of, 144 mulch as weed suppressant, importance of, 135, 151 variety under non-flooded conditions, 131 yeild, 144 Lr19 gene, 290 Lr24 gene, 289 Lupinus albus, 323 M Maize crop insertional mutagenesis on, 377–378 mutator transposon insertions in, 380 transposable elements in, 369–370 transposon insertion populations on, 371–372 Maize gene discovery project, 370, 377 Maize genetics cooperative stock center’s Web site, 377 Maize targeted mutagenesis (MTM) project, 377 Maximum contaminant level (MCL), 46 Mean annual nutrient productivity (aNP), 317 Mean residence time (MRT), 317 Medicago sativa, 314 Medicago truncatula, 344 Metal oxides, 53, 56–57, 60, 103, 105 Methane (CH4) contributor to climate change, 154 reduction of emissions, 331–332 tanniferous forage species, role of, 332–334 uptake and, 15–16 Micrometeorological technique for ammonia sampling, 215 Miran 203 infrared analyzer, 207, 234 Molybdopterin synthase gene, 380 Monitored natural attenuation (MNA), 103–104 Monomethylarsonic acid (MMA), 71–72 Mulching, rice-based cropping systems biomass transfer in, 134–135 conventional tillage and, 134 direct drilling and, 133 Happy Seeder approach in, 134 soil puddling and, 130 Multimedia mapping, 10, 12 Multi-point sampling system (MPSS), 215–216 Mutagenesis-based reverse genetics, 360–362 Mutagens for development of mutagenized population, 382–386 point-mutations-inducing chemical, 399 screening mutagenized population, 400–401 (see Cereals) treatment and population size, 386–390 Mutation. See also Mutagens allelic series vs. knockout, 399–400

419

Index

analysis, identification of, 388 deletion mutations in Caenorhabditis elegans, 385 density determined in cereal crop species, 389 detection technique in TILING, 390–393 expression of tagged gene causing knock-out, 365 induced by Mu elements, 370 insertion of nDart1 in OsClpP5 in rice, 379 knockout mutations of Arabidopsis transcription factors, 397 toward more complex virulence, 282 waxy loci with severe, 395 Mu transposon, 369–370, 377 N National air emission monitoring study (NAEMS), 203 National institute of standards and technology (NIST), 249–250 Natural organic matter (NOM), 59 N fertilizer, 17, 150, 172, 316, 320, 334–335 Nitrate leaching, 3, 316, 320, 325 Nitrogen pollution, in watercourses causes of, 315–316 forage legumes, role of, 320 mapping techniques, 318–320 NUE, characteristics of, 316–318 red clover, losses from, 321 Nitrogen trading, 27 3-Nitro-4-hydroxyphenylarsonic acid (Roxarsone), 50 Nitrous oxide (N2O), 9, 154 and CH4 emission, 158, 177–179 emissions control, in atmosphere, 325–326 in oxidation of ammonia, 334 Nondispersive infrared analyzers (NDIR), 231 Non-flooded crop, residue management bioenergy implications for biopower options in, 181 straw characteristics, 182–183 fertilizer efficiency and, 167, 172 grain yield for mulching crop residues effect of, 168–172 N immobilization in, 161 residual effect of rice, 162–163 residue incorporation effects of, 164, 167 rice-wheat systems in, 165–166 Happy Seeder approach in, 175–176 N2O and CH4 emission and, 177–179 P and K management of, 179–180 pest and disease pressure for, 173–174 SOM in, 180–181 water use efficiency for, 173 NO3-N leaching, 14, 20, 29, 31. See also Field level flows; Off-site transport, connection of field with predicted spatial distribution, 16

N use efficiency (NUE), 316–318 Nutrient cycling, crop residue management, 120–121 Nutrient uptake efficiency (NUpE), 317 Nutrient utilization efficiency (NUtE), 317 O Occupational safety and health administration (OSHA), 203 Off-site transport, connection of field with flows from field to nonfarm areas, 17–20 effective erosion buffers, 19 pollutants in vadose zone, GIS software and models for, 18 PSMs, for capturing runoff phosphorus, 20 transport of chemicals in shallow underground tile, 18 ideal buffer width and riparian zones, 21–22 RUSLE and VFSMOD, to determine locations, 21 On-field residue management, rice cropping systems, 126 Open-field burning, rice-based cropping systems, 119 banning of, 122 residue incorporation, 129 Open-path Fourier transform infrared system (OP-FTIR), 218 Open-path sampling, 217–218. See also Ammonia sampling and measurement Opsis AR-500 UV open-path monitor, 235 OryGenesDB database, 376 Oryza sativa, 118 P Passive measurement device, 221–222. See also Ammonia sampling and measurement Path-weighted average (PWA), 217 Permeable reactive barrier (PRB), 103 Phenotype screening system, for mutagen population, 400–401 Phleum pratense, 314 Phosphate (PO4) in soils, 57 Phosphorus pollution control, in watercourses causes of, 321–323 PUE, in rumen, 324–325 P use efficiency, 323–324 Photoacoustic spectrophotometer (PAS), 232–233, 248, 252 Photosynthetic mutant screen (PMS), 377 Phyllosilicates, 57 Pioneer Hi-Bred’s trait utility system of corn, 377 Poa pratensis, 314 Point sampling method, 215–217. See also Ammonia sampling and measurement Point-zero-charge (PZC), 53 Polyphenol oxidase (PPO), 321

420

Index

Population growth, 2 Potential conservation practices, 30–38 Precision agricultural-landscape modeling system (PALMS), 24 Precision conservation buffers and riparian zones, 20–21 different degrees of, 4 environmental impacts and production systems sustainability, 24 on field scale, 13 to generate maps for use in analysis in field of, 10 GIS mapping approach and map analysis, 6–8 hydrologically sensitive area and, 22 to identify hot spots on farm and watershed, 28 to increase for soil and water, conservation practices, 30–38 integration of information and locations for riparian buffers, 29 management and conservation, integration and maps for, 9 for management of flows, 16–17 manure management, technology for, 17 modeling approach to, 23 patterns and relationships, identification of, 9–12 GIS research for, 10, 12 Map analysis procedures, 10–11 multimedia mapping and Cartesian coordinate system, 10 static coincidence analysis vs. dynamic three-dimensional flows, 10, 12 and potential for site-specific applications, 6 to reduce the transport of nutrients, 20 site-specific and three-dimensional scale approach, 4, 13 at watershed scale, 24–27 for animal management and soil and water conservation, 26 Precision conservation management zones (PCMZ), 16 Precision farming, 3 Project aligned related sequences and evaluate SNPs (PARSESNP), 393 Pteris vittata, 104 Puccinia graminis, 273 Puccinia graminis tritici, 272 Pyrolysis, 181 Q Quality assurance and quality control (QAQC), 244, 250–251, 256 R Race Ug99, 273, 278–279. See also Stem rust gene Sr, immunity to, 291 and long-term control, 288–305

markers associated to stem rust resistance genes, 293–294 pandemic, prediction of, 287–288 threat to wheat production, 281–288 Red clover, 313, 315, 321, 333, 338 Remote sensing (RS), 4, 24–25, 28, 39 RescueMu project. See Mu transposon Retrotransposon-tagged mutation, 380 Reverse genetics approach, gene function determination, 359–361 T-DNA insertion line in, 363–364, 378–379 transposon insertion lines in, 379–381 Rice-based cropping systems in Asia, residue management decision tree, 184–185 in-field residue management practices, 127 monocropping systems in, 128 non-flooded crop following rice biomass, transfer of, 134–135 incorporation, 132–133 mulching, 133–134 non-flooded crop, rice following rice composting, 131–132 incorporation, 129–130 mulching, 130–131 nutrient cycling in, 120–121 on-field residue management practices, 126 production area and grain yield in, 123–124 productivity, profitability and environmental impact, 122 residue production and area for, 124 soil puddling in, 125 Rice crop, evaluation for residue management options biological N2 fixation and, 159–160 grain yield effect of incorporation of upland crop residue on, 141–143 effect of rice residue incorporation on, 135–140 incorporation, profitability, 152–153 insertional mutagenesis on, 376–377 monocropping systems, 128, 151–152 residue incorporation effects of, 136–140 T-DNA knockout mutant lines in, 366–368 water use efficiency, 151 yield and residual effect relationship, 145 Rice yield in a barley–rice rotation, 144 with and without fertilizer application, 149 Riparian ecosystem management model (REMM), 21 Riparian zones denitrification of N in, 27 precision conservation and, 3–4 RNAi technique, 360–361. See also Cereals Rosemount gas analyzer, 233–234

421

Index S Seattle TILLING Project (STP). See Arabidopsis TILLING Project (ATP) Secale cereale, 281 Semidwarf wheat varieties, 278–279 Septoria nodorum, 273 Septoria tritici, 273 Silicic acid, 59 Site-specific management zones (SSMZ), 16 Smart dust, 9 Smoke particles, 153–154 Soil aggregation, 12 Soil and water assessment tool (SWAT), 23 Soil erosion, 2, 18, 336, 338 Soil organic matter (SOM), 55, 178, 180–181 Soil puddling, 125 mulching and, 130 Soil quality, improvement of biodiverse mixtures, 338–339 flood tolerance and prevention, 335–337 soil porosity and compaction, 337–338 Solid-state/electronic ammonia sensor, 240–241 Sorghum crop. See Cereals Sorting intolerant from tolerant (SIFT) program, 400 Spatial patterns and relationships. See also Precision conservation GIS research and changes in geo-referencing and, 12 map analysis procedures, 10 maps of surface flow, 11 Sr gene, 278, 280, 290–291. See also Stem rust Staphylococcus aureus, 69 Staphylococcus xylsis, 69 Static coincidence modeling, 10. See also Geographic information systems (GIS) Stem rust, of wheat breeding for resistance, 277–281, 288–289 future perspectives of, 305–306 high-yielding wheat, 296–300 race-specific resistance genes for Ug99, 290–292 race-specific resistance genes in wheat, 292–295 resistant wheat varieties development, 304–305 Sr24 gene breakdown, 289–290 Ug99 resistance of plant, 295–296 world wheat area reduction, 300–304 occurrence of, 273–274 pathogens and epidemiology of, 274–277 race Ug99 of avirulent and virulent genes in, 281–282 epidemic prediction of, 287–288 geographical distribution of, 282–283 migration of, 284–285

wheat germplasm resistance/susceptibility, 285–287 resistance genes, PCR-based markers, 293–294 Surface mulching, 132 CH4 production and, 179 residue incorporation and, 158 and rice residue incorporation in no-till sown wheat, 176 Surfurosprillum barnesii, 76 Syngenta GeneChipÒ , 399 T Tanniferous forage species, 332–334. See also Methane (CH4) Targeting induced local lesions in genomes (TILLING). See also Cereals; Mutagen; Arabidopsis TILLING Project (ATP); Eco-TILLING creation of mutagenized populations, different schemes of, 387 DNA pool preparation, 390 mutagen agents for, 382–386 mutagen treatment, 386–390 mutation detection technique in CELI enzymatic mismatch cleavage of DNA, 392–393 PCR amplification of DNA pools, 390–391 software’s used and emerging techniques, 393–394 T-DNA insertion mutagenesis. See also Cereals; Mutation; Mutagens gene knock-out mutation, 364–365 insertion lines in, 378–379 tagging mutation, 365–369 Thinopyrum elongatum, 289 Thinopyrum ponticum, 278 Tilletia indica, 174 Time weighted average (TWA), 215 Total ammoniacal-N (TAN), 327 Transposon-based gene tagging, 381 Transposon insertion mutagenesis. See also Maize crop Ac/Ds transposons, uses of, 373 barley transposons, 374 insertion lines in, 379–381 maize transposon insertion populations, 371–372t Mu transposons, 370 transposable elements, 369 Trifolium ambiguum, 329 Trifolium nigrescens, 334 Trifolium pratense. See Red clover Trifolium repens L. See White clover Trimethylarsine oxide (TMAO), 72 Triticum aestivum, 119, 272

422

Index

Triticum monococcum, 385, 390, 399 Triticum turgidum, 291 Triticum ventricosum, 278, 281 U Ultraviolet differential optical absorption spectrometer (UV-DOAS), 217–218, 234–235 Universal soil loss equation (USLE), 23 Upland (non-flooded) crop residue incorporation, 141–143 V VECHTA air sampling system, 210 Vegetative filter strip model (VFSMOD), 21 Vesicular arbuscular mycorrhizal fungi (VAM), 323 W Water and tillage erosion model, 13–14 Water erosion prediction project (WEPP), 24 Watershed scale. See also Precision conservation models and tools, 22–24 DEMs, AGNPS and SWAT models, 23–24 WEPP and PALMS, 24 variable hydrology, 22 Water soluble carbohydrate (WSC), 328–329 waxy gene. See Granule bound starch synthase (GBSS) I gene

Wet methods, 221, 225–228. See also Ammonia sampling and measurement Wheat. See also Stem rust, of wheat breeding for rust resistance, 277–281, 288–289 future perspectives of, 305–306 high-yielding wheat for, 296–300 race-specific resistance genes for Ug99, 290–292 resistant wheat varieties development, 304–305 Sr24 gene breakdown, 289–290 Ug99 resistance of plant, 295–296 wheat improvement strategies, race-specific resistance genes in, 292–295 world wheat area reduction, 300–304 crop, insertional mutagenesis on, 378 (see also Cereals) fungal diseases in, 273 production of, 272–273 rust pathogens, dispersal modes of, 275–277 stem rust (see Stem rust, of wheat) White clover, 314, 320, 323, 325, 329, 338 X X-ray absorption near-edge spectroscopy (XANES), 51, 65, 67, 69 Z Zea mays, 119

Precision conservation Precision Ag Wind erosion

Chemicals

Soil erosion Runoff Leaching

Leaching

Terrain

Leaching

Soils Yield Potassium

3-dimensional Flows Cycles

Coincidence

CIR image

2-dimensional Interconnected perspective

Isolated perspective

Jorge A. Delgado and Joseph K. Berry, Figure 1 The site-specific approach can be expanded to a three-dimensional scale approach that assesses inflows and outflows from fields to watershed and region scales. (From Berry et al., 2003.)

Surface modeling

Point samples are spatially interpolated into a continuous surface

53.2 ppm

4.2 ppm

Field sample locations Phosphorus surface

Discrete data spikes

Min = 4.2 Max = 53.2 Avg = 13.4 SDev = 5.2

Spatial data mining 32c,62r

45c,18r

Map surfaces are clustered to identify data pattern groups

P 53.2

Relatively low responses in P, K, and N Relatively high responses in P, K, and N

11.0

Cluster 2 Cluster 1

N

K 412.0

177.0

27.9

32.9

N K P Geographic space

Data space

Clustered data zones

Jorge A. Delgado and Joseph K. Berry, Figure 2 Surface modeling is used to derive map surfaces that utilize spatial data mining techniques to investigate the numerical relationships in mapped data. (From Berry et al., 2005.)

Map analysis Desktop mapping Field data Standard normal curve fit to the data

Spatially interpolated data

34.1% 34.1%

68.3% +/−1 standard deviation

Average = 22.0 StDev = 18.7

22.0

28.2

Discrete spatial object (generalized)

80 60 40 20 0 −20 −40 −60

High = 50

80 60 40 20 Average = 22.0

0 −20 −40 −60

N

Continuous spatial distribution (detailed)

Jorge A. Delgado and Joseph K. Berry, Figure 3 Desktop mapping uses aggregated, nonspatial statistics to summarize spatial objects (points, lines, and polygons), whereas map analysis uses continuous spatial statistics to characterize gradients in geographic space (surfaces).

Inclination of a fitted plane to a location and its eight surrounding elevation values

2418

2404

2393

2409

2395

2341

2383

2373

2354

Slope(47,64) = 33.23%

35% 30% 25% 20% 15% 10% 5% 1% 0%

Steep

Moderate Gentle flat

Slope map draped on elevation Slope map

Elevation surface

Flow(28,46) = 451 paths

537 Paths Heavy 256 Paths 123 Paths 64 Paths 32 Paths 16 Paths Moderate 8 Paths 4 Paths Light 2 Paths 1 Paths minimal

Total number of the steepest downhill paths flowing into each location Flow map draped on elevation Slope map

Jorge A. Delgado and Joseph K. Berry, Figure 4 Maps of surface flow confluence and slope are calculated by considering relative elevation differences throughout a project area. (From Berry et al., 2005.)

Tillage erosion

Water erosion

Tillage–water erosion

Total erosion (cesium-137 measurements)

−1 −1 Mg ha yr

−33

Soil loss net erosion

−22

Accelerated erosion

−11 0

Soil loss T value = −11 Mg ha−1 yr−1

11 22 33 120

Soil gain net deposition

Slope % map and cesium137 sampling sites Slope % 8 7 6 5 4 3 2 1 0 Elevation contour lines are overlaid on all maps elevation labels are shown only on total erosion map

Jorge A. Delgado and Joseph K. Berry, Figure 6 Erosion patterns developed from tillage, water, tillage-water, and total erosion (137Cs) modeling of the research field are displayed. Cesium sampling sites are also displayed on a contour map of slope percentage for the field. (From Schumacher et al., 2005.)

A 100

100

90

80

70

70

60

Sand (%)

90

80

60

50 50

200

200

150

150

100

100

50 0

50

kg NO3-N/ha 0–1.5m

B

0

Jorge A. Delgado and Joseph K. Berry, Figure 7 Spatial distribution of sand content in the top 1.5 m of soil across different productivity zones (A). Spatial distribution of observed residual soil NO3 -N in the top 1.5 m of soil for study one across the different productivity zones during the 2000 growing season (B). (From Delgado and Bausch, 2005.)

250

200

150

150

100 100 50

50

0

kg NO3-N/ha 0–1.5m

250

200

0

Jorge A. Delgado and Joseph K. Berry, Figure 8 Spatial distribution of predicted NO3-N leaching from the root zone of corn (1.5 m depth) in study one across the different productivity zones during the 2000 growing season. (From Delgado and Bausch, 2005.)

Erosion potential

Slopemap

Reclassify

Overlay

Reclassify

3 steep 2 moderate 1 gentle

Flow/slope

Slope_classes Reclassify

33 heavy flow: steep 33 heavy flow: moderate 33 heavy flow: gentle 23 moderate flow: steep 22 moderate flow: moderate 21 moderate flow: gentle 13 light flow: steep 12 light flow: moderate 11 light flow: gentle

Flowmap

Erosion_potential High Moderate Low

Flow_classes 3 heavy 2 moderate 1 light

Effective erosion buffers Effective erosion potential distance

Erosion_potential

Far

Distance

Close

Erosion buffers Streams

Jorge A. Delgado and Joseph K. Berry, Figure 9 Effective erosion buffers around a stream expand and contract depending on the erosion potential of the intervening terrain. (From Berry et al., 2005.)