Advances in Agronomy, Volume 105

ADVANCES IN AGRONOMY Advisory Board PAUL M. BERTSCH RONALD L. PHILLIPS University of Kentucky University of Minneso...

Author: Donald L. Sparks

62 downloads 1318 Views 4MB 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

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

WARREN A. DICK

MARY C. SAVIN

HARI B. KRISHNAN

APRIL L. ULERY

SALLY D. LOGSDON

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32 Jamestown Road, London, NW1 7BY, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands First edition 2010 Copyright # 2010 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: 978-0-12-381023-6 ISSN: 0065-2113 (series) For information on all Academic Press publications visit our website at elsevierdirect.com Printed and bound in USA 10 11 12 10 9 8 7 6 5 4 3 2 1

CONTRIBUTORS

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

S. Anthony (83) ADAS, Wolverhampton, Woodthorne, Wolverhampton, United Kingdom Jeff Baldock (173) CSIRO Land and Water, PMB2, Glen Osmond, SA, Australia R. Bol (47, 83) Biogeochemistry of Soils and Water group, North Wyke Research, Okehampton, Devon, United Kingdom Bhagirath S. Chauhan (221) Crop and Environmental Sciences Division, International Rice Research Institute, Metro Manila, Philippines H. Cover (117) Vistronix, Inc., Portland, Oregon, USA J. A. Delgado (117) USDA-ARS-Soil Plant Nutrient Research Unit, Fort Collins, Colorado, USA Matthew Forbes (173) Natural Resources Branch, Department of Conservation and Environment, Locked Bag 104, Bentley Delivery Centre, WA, Australia P. Gagliardi (117) USDA-ARS-Soil Plant Nutrient Research Unit, Fort Collins, Colorado, USA S. J. Granger (83) Biogeochemistry of Soils and Water group, North Wyke Research, Okehampton, Devon, United Kingdom C. M. Gross (117) USDA-NRCS, WNTSC, Beltsville, Maryland, USA P. M. Haygarth (83) Centre for Sustainable Water Management, Lancaster Environment Centre, Lancaster University, Lancaster, Lancashire, United Kingdom vii

viii

Contributors

E. Hesketh (117) USDA-NRCS, WNTSC, Amherst, Massachusetts, USA David E. Johnson (221) Crop and Environmental Sciences Division, International Rice Research Institute, Metro Manila, Philippines E. Krull (47) CSIRO Land and Water, PMB2, Glen Osmond, Australia H. Lal (117) USDA-NRCS, WNTSC, Portland, Oregon, USA E. Lopez-Capel (47) The Swan Institute, University of Newcastle, Newcastle upon Tyne, United Kingdom S. P. McKinney (117) USDA-NRCS, WNTSC, Portland, Oregon, USA P. N. Owens (83) University of Northern British Columbia, Prince George, British Columbia, Canada W. A. Payne (1) Assistant Director for Research, Norman E. Borlaug Institute of International Agriculture, and Professor of Crop Physiology, Texas A&M University System, College Station, Texas, USA M. J. Shaffer (117) USDA-ARS (Retired), Fort Collins, Colorado, USA S. P. Sohi (47) School of GeoSciences, University of Edinburgh, Edinburgh, United Kingdom, and Department of Soil Science, Rothamsted Research, Harpenden, Herts, United Kingdom Murray Unkovich (173) School of Agriculture, Food and Wine, The University of Adelaide, PMB 1, Glen Osmond, SA, Australia S. M. White (83) Cranfield University, Cranfield, Bedfordshire, United Kingdom

PREFACE

Volume 105 contains six outstanding reviews dealing with nutrient cycling, soil and water resources, climate change, and crop management. Chapter 1 is a thought provoking commentary on the impacts of biofuels on sustainability of soil and water resources. Chapter 2 discusses the potential effect of biochar on climate change and carbon cycling, crop productivity, and resource management. Chapter 3 is a thorough review on water pollution from intensively managed grasslands. Pollution pathways and ways to minimize contamination from them are also discussed. Chapter 4 is a contemporary review on the use of an innovative GIS Nitrogen Trading Tool for conserving and reducing nitrogen losses in the environment. Chapter 5 discusses the impact of harvest index variability of grain crops on carbon accounting, with application to Australian agriculture. Chapter 6 deals with the role of seed ecology in enhancing weed management in the tropics. I appreciate the excellent reviews of the authors. DONALD L. SPARKS Newark, Delaware, USA

ix

C H A P T E R

O N E

Are Biofuels Antithetic to Long-Term Sustainability of Soil and Water Resources? W. A. Payne*,†

Contents 2 7 7 13 13 14 15 16 16 17 21 22 22 24 29 33 41 43

1. Introduction 2. Some History 2.1. Ethanol as a fuel 2.2. Soil and oil 2.3. Charting our future in the past 3. An Overview of Biofuels 3.1. Ethanol 3.2. Biodiesel 3.3. Cellulosic ethanol 3.4. Biofuel feedstocks and conversion to biofuel 3.5. Bioenergy and biofuel potential on a global scale 4. Sustainability Issues 4.1. Favorable economics? 4.2. Conservation of resources 4.3. Preservation of ecology 4.4. Social justice 5. Summary References

Abstract Sustainability of biofuels is a contentious but old topic that has reemerged with increased use of crops as feedstocks. There are vastly different land requirements for different feedstocks, and disagreement on the energy balance of their conversion to biofuel. To be sustainable, biofuel systems should (1) have favorable economics, (2) conserve natural resources, (3) preserve ecology, and (4) promote social justice. With the possible exception of sugarcane

* Assistant Director for Research, Norman E. Borlaug Institute for International Agriculture, Texas A&M University System, College Station, Texas, USA Professor of Crop Physiology, Texas A&M University System, College Station, Texas, USA

{

Advances in Agronomy, Volume 105 ISSN 0065-2113, DOI: 10.1016/S0065-2113(10)05001-7

#

2010 Elsevier Inc. All rights reserved.

1

2

W. A. Payne

production in Brazil, it seems unlikely that ethanol production from crops will be economically viable without government support. Less is known on cellulosic feedstock economics because there are no commercial-scale plants. Natural resources that may be affected include soil, water, and air. In the United States, agricultural intensification has been associated with greater soil conservation, but this depended on retaining residue that may serve as cellulosic feedstocks. The ‘‘water footprint’’ of bioenergy from crops is much greater than for other forms of energy, although cellulosic feedstocks would have a smaller footprint. Most studies have found that first-generation biofuels reduce greenhouse gas emissions 20–60%, and second generation ones by 70–90%, if effects from land-use change are excluded. But land-use change may incur large carbon losses, and can affect ecological preservation, including biodiversity. Social justice is by far the most contentious sustainability issue. Expanding biofuel production was a major cause of food insecurity and political instability in 2008. There is a large debate on whether biofuels will always contribute to food insecurity, social justice, and environmental degradation in poor countries.

1. Introduction The cacophony of responses to a recent New York Times article (NYT, 2009a,b) in which New Mexico Senator Bingaman suggested further government help for the ailing ethanol industry illustrates what an emotionally and politically charged topic that biofuel has become (Table 1). One can find similar spirited exchanges on biofuel articles at the Christian Science Monitor, The Economist, and other newspapers. Some of the hot button issues that biofuels and especially ethanol raise include patriotism, pro- and antiwar sentiment, terrorism, xenophobia, engine and conversion efficiencies, food for the poor, environmental protection, fair trade, energy independence, urban vs. rural America, big oil companies, and government spending of taxpayers’ dollars. How can scientists possibly make sense of this when, after all, they themselves are not free from partisanship (Clair, 2009; Guston et al., 2009)? There is not even a strong consensus within the scientific community on whether the overall energy output from ethanol and biodiesel production is greater than the input (Liska et al., 2008; Pimentel and Patzek, 2005). Add to that all the other sociopolitical aspects, and one truly has a (metaphorically) volatile mixture. Because of the many biophysical but especially sociopolitical uncertainties and complexities involved, it should come as no surprise that, whether for good-faith or simply politically motivated reasons, there are many contentious views on the topic of biofuels and sustainability. In large part, the topic is linked with that of global climate change, which itself is

3

Are Biofuels Antithetic to Long-Term Sustainability of Soil and Water Resources?

Table 1 Posted reader comments to New York Times article on proposed increased support to the ethanol industry (NYT, 2009a,b)

I think this is a terrible idea, every single subsidized program has been a terrible money draining failure from airlines to welfare. Basically we’re supporting high commodity prices by pushing this plan. This hurts foreign competition and disrupts food markets, we should not be burning food until we can end world hunger. Of course there are also various environmental concerns, the increased fertilizer runoff, by-products from factories and the stuff is less safe than gasoline since it is less stable. The claim that the problems of the ethanol industry are attributable to the recession is dishonest. The ethanol industry is in terrible shape because corn ethanol makes no sense economically or environmentally, and there is no known method for producing cellulosic ethanol on a commercial scale. Please do not prop up corn ethanol. The environmental consequences of growing so much corn conventionally (read mono-crop, petroleum intensive, chemical dependent agriculture) easily cancel out the benefits of ethanol blends. Because we heartlessly treat food as a global free market commodity exposed to the whims of speculation, ethanol production has spiked corn prices and in classic domino effect caused the prices of other staples to ride a roller-coaster as well. This has led to wide spread hunger, food riots and instability. Congressmen, many of whom are deep in the pocket of mega-agribusiness, need to step back for a moment and realize the dangerous consequences of burning food as fuel. Contact your Senators and Representatives and tell them that corn ethanol fuel is a terrible idea both for the economy and the environment. Wow! You mean the government mandated something without making sure it was technologically and economically feasible first? Ethanol uses up as much fuel as it is supposed to save or more, according to recent studies. It makes us more dependent to foreign oil, raises food prices, reduces gas mileage and engine performance, damages the environment. If it wasn’t for. . .lobbyists, congress would have never given those multibillion dollar corporations our tax dollars to subsidize this lunacy. Corn-based ethanol is the ONLY renewable fuel that is available today and is the foundation for the next generation ethanol (cellulosic) of tomorrow. The notion that corn-based ethanol being the culprit for increased food prices has been completely debunked, leaving the GMA and other antiethanol groups with absolutely no credibility. America’s corn growers have just completed one of the largest harvests of corn in our country’s history, with an average of 154 bushels of corn per acre. With continued improvements in agriculture, that yield is expected to double, ON THE SAME AMOUNT OF LAND, over the next decade. This country MUST continue to support corn-based ethanol to get to cellulosic and, more importantly, to reduce our addiction to foreign oil. (continued)

4

W. A. Payne

Table 1

(continued)

In addition to the economic failure of corn ethanol, the environmental costs include using limited water supplies. Ethanol plants are more water efficient than they were, but still have huge water requirements. According to the Feb. 2007 Ethanol Producer Magazine it takes 150–300 million gallons of water to produce 100 million gallons of ethanol. When the water tables are depleted and we cannot get water for food crops, drinking and other activities, where are the tankers of water going to come from? I love how 99% of the people bashing ethanol have never driven a car with ethanol (besides E10), but will quickly attest to how terrible it supposedly is by pointing to bogus studies that use ethanol data in excess of 5 years old. Besides, I’d rather buy my fuel from Farmer Bob down the road than some sheik in the mideast that’s funneling money to terrorist organizations. The price difference makes up for your lost mileage because of very large subsidies and indirect costs that are paid by other consumers and taxpayers. If you want to pay more money to Farmer Bob for ethanol, then by all means do so—but pay him with your own money, not money confiscated from others. And while you’re at it, add on a few bucks per gallon for the environmental damage that you’re inflicting. In short is it not the myth of ‘‘renewable, corn base ethanol’’ that both science and the market place has debunked? Ethanol from corn is not renewable because the energy inputs are roughly the size of what you get out in usable liquid fuels, and the greenhouse gas savings are nil. There is no scientific doubt about these statements, the literature is full of them. There is also no doubt that cellulosic ethanol, if made right, or the kinds of advanced biofuels Berkeley, Stanford and other institutions are working on, MIGHT give true relief on the oil front and the CO2 front. But no responsible scientist, economist or politician (oxymoron) believes cellulosic ethanol or any other biofuel will be cheap, even compared to $100/bbl oil, when all the costs are counted. I challenge you to forego the tax subsidies and shift to a tax on oil, and a tax on carbon, and let the market decide how well ethanol from corn can compete with other fuels, more efficient cars, and less driving. Corn also requires nitrogen fertilizing that is being blamed for increasing dead zones in the Gulf of Mexico and elsewhere. If we want to get more than 10% of our vehicle fuel from corn etc. serious inroads in land and water needed for food crops will have to occur. Biofuels are just recycling carbon dioxide without removing on balance one molecule of that gas already at levels causing major global warming effects. So biofuels really are just a wheel spinning operation going nowhere in getting control of climate change The modern-day definition of agriculture can be said to be ‘‘the process of turning oil into food.’’ Therefore we CANNOT base new generation fuels on conventional modern agriculture.

5


Table 1

(continued)

To be optimistic, the goal of the work in Berkeley is to follow the advice Sean gives above, namely to find a way to turn cellulose into sugar and alcohol, just like termites or other organisms do it, without using large quantities of land or water. Using basic biochemistry there is a good prospect we can do this, but not necessarily cheaply. The late Prof. Alex Farrel was a strong corn ethanol until he and his students here started churning out the disappointing numbers on the low energy yields, the huge land, fertilizer and water impacts of corn, and the lack of any greenhouse gas benefit. While there certainly is promise of higher corn yields per acre, how much of that increase comes at the expense of greater use of oil products for our mechanized agriculture and coal-based electricity for irrigation? One of Alex’ last articles was an Op Ed in the SF Chronicle about a year ago, entitled ‘‘not more biofuels, better biofuels.’’ Until we get there, we should not be subsidizing and earmarking any biofuels, particularly as all of these questions come up about the costly indirect or side effects of plowing so many acres for corn ethanol. If for no other reason, we need to support the sale of ethanol because it replaces foreign oil. I would challenge all of you to read, read, read. Start with Energy Victory by Robert Zubrin. Turn three pages on this book about the Saudis and you will never doubt the need for ethanol and alternative fuels. Forget that it creates jobs, is good for the environment, or supports our agriculture economy, or that you hold ethanol to a standard much higher than gasoline. Forget all that. Read something that sends chills down your spine about the world we live in and the role foreign oil now plays. Every gallon of ethanol produced and consumed are dollars that stay in America. It reduces our dependence on foreign oil, helps our environment, saves American’s money with reduced fuel costs and puts money in the pockets of American farmers instead of Middle Eastern Oil Czars. Farmers produce more corn each year as needed for food and fuel. Ethanol is the most successful biofuel we have at our disposal today. Supporting and using Ethanol today will lead us to the second generation biofuels evolving in the industry. Cut out Ethanol and the farm economy collapses. Our mid-western economy is balanced on the success of the Ethanol industry. Keep it strong and we all succeed. So far the only plants to my knowledge that have closed are the VeraSun plants as a result of some poor commodities buying by their personnel. We have an Ethanol plant in our town Green Plains Renewable Energy that is going strong and profitable and is working on second generation biofuels. One bad apple does not spoil the whole bunch. There should be no regulatory caps on production of Ethanol and standards already set should be kept in place. Higher blends should be encouraged by our government. It’s the right thing to do for our country, the environment, and our economy. (continued)

6

W. A. Payne

Table 1

(continued)

The 9 billion gallons of corn ethanol that were produced last year reduced gasoline and diesel consumption of the United States by all of 0.8%, when the fossil fuel that was used to produce the ethanol is taken into account. Corn ethanol is not an energy program and never was. It is a political largesse program that has hindered the meaningful development of alternative fuels. If you are truly serious about developing alternative transportation fuels, the first thing you would do is eliminate corn ethanol subsidies and mandates. Corn is cheap . . . roughly $4 for 56 lbs. . . We should be embarrassed that 56 lbs of corn can be purchased for $4. The Corn Producers worth their tails off and some people bitch and moan at paying $4 for 56 lbs of corn. . .! We have so much corn we continue to pay Farmers to NOT grow corn. It really boils down to who do you want to support (send you money too) Iran ? Bin Laden and his minions? or keep more of our Money and Pride at home supporting America Farmers, American Producers and American Consumers. . . There are two commercial cellulose ethanol plants under construction in Georgia (Range Fuels) and Florida (Coskata). These use any organic substance to produce ethanol. The producer cost per gallon of ethanol should approach $1.00 a gallon! Your Governmental EPA in 2005 produced a study showing E30 (30% ethanol blend) could produce engine efficiency far superior to plain old gasoline engine efficiency. Yes, that means higher fuel mileage on E30 than that gasoline. Did you know the largest oil reserve in the world in the Mid-East, uses massive amounts of water to get it out of the ground! If you do not think oil (companies) receive tax breaks that amount to billions of dollars, you are dreaming. This amount of amount of money far surpasses what ethanol receives. Remove this and a gallon of gasoline will approach $10 a gallon Well, ‘‘Voice of Reason’’ we need about 0.3/4s of a gallon of that imported petroleum to make 1 gallon of corn ethanol. . .not a good deal at all when you consider additionally that soils and water were used to grow the corn. Additionally the use of the corn kernels for fuel rather than food distorts global food markets. Ethanol has been a blessing for the small independent farmer and all Americans. The government can now subsidize an industry that would not be transferred overseas. People have forgotten the gas shortage in the 1970s and the control OPEC had over our nation. I even question the control the large oil companies have over our nation. Believe me the oil company executives were receiving their multimillion dollar bonuses in 2007. The ethanol industry will take decades to refine production. POET Biorefining in Emmetsburg Iowa is now producing ethanol commercially using corn cobs, but it is a process that needs America’s support to get on its feet. Ethanol may not be the best long-term alternative fuel source but it is


Table 1

7

(continued)

an excellent bridging product. We have the pumps and vehicles already in operation. Also people forget one of the by-products of ethanol production is distiller grain or high protein animal feed. The energy used to produce ethanol also produces refined animal feed. Pulling the rug out from under the growing ethanol business would be a mistake. The agricultural industry has been working smarter producing more crops with less herbicides and fertilizer. Ethanol efficiency is increasing to 3 gallons of ethanol per bushel of corn. Have some faith in our technology. Farmers are continuing to produce more grain on the same amount of acres using fewer inputs, and ethanol production is branching out to cellulosic production.

complex, politically and emotionally charged, and filled with uncertainty and contention (IPCC, 2008—see Key Uncertainties). Before launching into some of the more contentious issues, however, some relevant historical points will be made, followed by an overview of biofuels.

2. Some History 2.1. Ethanol as a fuel The timeline in Table 2 from the US government’s Energy Information Administration (EIA, 2009) allows us to extract some relevant historical highlights:

Ethanol has been used to power internal combustion engines since the 1800s, including that of Henry Ford’s first automobile. The famous model T, first produced in 1908, ran on ethanol, gasoline, or a mixture of the two. In the 1930s, more than 2000 gasoline stations in the US Midwest sold gasohol, which contained 6–12% ethanol. Since the civil war, the economic viability of ethanol has been influenced by government policy, including taxes and subsidies. The model T came into production 2 years after the government repealed a $2 per gallon excise tax on ethanol that had been in place for more than 50 years. The Energy Tax Act of 1978 amounted to a 40 cents per gallon subsidy for every gallon of ethanol blended into gasoline; this was later increased to 50–54 cents. In the 1980s, congress enacted many tax benefits for ethanol producers and blenders. Government loans and price guaranties were also offered, and tariffs were imposed on imported ethanol. Despite all these supports, more than half went out of business by the mid-1980s.

8

W. A. Payne

Table 2

Timeline of ethanol use in the United States

1826 1860

1862

1896 1906

1908

1917–1918 1920s

1930s

1941–1945

1945–1978

1974

1975

Samuel Morey developed an engine that ran on ethanol and turpentine. German engine inventor Nicholas Otto used ethanol as the fuel in one of his engines. Otto is best known for his development of a modern internal combustion engine (the Otto Cycle) in 1876. The Union Congress put a $2 per gallon excise tax on ethanol to help pay for the Civil War. Prior to the Civil War, ethanol was a major illuminating oil in the United States. After the tax was imposed, ethanol cost too much to be used this way. Henry Ford built his first automobile, the quadricycle, to run on pure ethanol. Over 50 years after imposing the tax on ethanol, Congress removed it, making ethanol an alternative to gasoline as a motor fuel. Henry Ford produced the Model T. As a flexible fuel vehicle, it could run on ethanol, gasoline, or a combination of the two. The need for fuel during World War I drove up ethanol demand to 50–60 million gallons per year. Gasoline became the motor fuel of choice. Standard Oil began adding ethanol to gasoline to increase octane and reduce engine knocking. Fuel ethanol gained a market in the Midwest. Over 2000 gasoline stations in the Midwest sold gasohol, which was gasoline blended with between 6% and 12% ethanol. Ethanol production for fuel use increased, due to a massive wartime increase in demand for fuel, but most of the increased demand for ethanol was for nonfuel wartime uses. Once World War II ended, with reduced need for war materials and with the low price of fuel, ethanol use as a fuel was drastically reduced. From the late 1940s until the late 1970s, virtually no commercial fuel ethanol was available anywhere in the United States. The first of many legislative actions to promote ethanol as a fuel, the Solar Energy Research, Development, and Demonstration Act led to research and development of the conversion of cellulose and other organic materials (including wastes) into useful energy or fuels. The United States begins to phase out lead in gasoline. Ethanol becomes more attractive as a possible octane booster for gasoline. The Environmental Protection

9


Table 2 (continued)

1978

1979

1980–1984

1983

Agency (EPA) issued the initial regulations requiring reduced levels of lead in gasoline in early 1973. By 1986 no lead was to be allowed in motor gasoline. The first time gasohol was defined, it was in the Energy Tax Act of 1978. Gasohol was defined as a blend of gasoline with at least 10% alcohol by volume, excluding alcohol made from petroleum, natural gas, or coal. For this reason, all ethanol to be blended into gasoline is produced from renewable biomass feedstocks. The Federal excise tax on gasoline at the time was 4 cents per gallon. This law amounted to a 40 cents per gallon subsidy for every gallon of ethanol blended into gasoline. Marketing of commercial alcohol-blended fuels began. Amoco Oil Company began marketing commercial alcohol-blended fuels, followed by Ashland, Chevron, Beacon, and Texaco. About $1,000,000,000 ($1 billion) eventually went to biomass-related projects from the Interior and Related Agencies Appropriation Act. First US survey of ethanol production was conducted. The survey found fewer than 10 ethanol facilities existed, producing approximately 50 million gallons of ethanol per year. This was a major increase from the late 1950s until the late 1970s, when virtually no fuel ethanol was commercially available. Congress enacted a series of tax benefits to ethanol producers and blenders. These benefits encouraged the growth of ethanol production. The Energy Security Act offered insured loans for small ethanol producers (less than 1 million gallons per year), up to $1 million in loan guarantees per project that could cover up to 90% of construction costs on an ethanol plant, price guarantees for biomass energy projects, and purchase agreements for biomass energy used by federal agencies. Congress placed an import fee (tariff) on foreign-produced ethanol. Previously, foreign producers, such as Brazil, were able to ship less expensive ethanol into the United States. The Gasohol Competition Act banned retaliation against ethanol resellers. The Crude Windfall Tax Act extended the ethanolgasoline blend tax credit. The Surface Transportation Assistance Act increased the ethanol subsidy to 50 cents per gallon. (continued)

10

W. A. Payne

Table 2

1984

(continued) The number of ethanol plants in the United States peaked

at 163.

The Tax Reform Act increased the ethanol subsidy to 60

cents per gallon.

1985

Many ethanol producers went out of business, despite the

subsidies.

Only 74 of the 163 commercial ethanol plants (45%)

1988

1990

1992

remained operating by the end of 1985, producing 595 million gallons of ethanol for the year. Ethanol was first used as an oxygenate in gasoline. Denver, Colorado, mandated oxygenated fuels (i.e., fuels containing oxygen) for winter use to control carbon monoxide emissions. Other oxygenates added to gasoline included MTBE (methyl tertiary butyl ether—made from natural gas and petroleum) and ETBE (ethyl tertiary butyl ether—made from ethanol and petroleum). MTBE dominated the market for oxygenates. Omnibus Budget Reconciliation Act decreased the ethanol subsidy to 54 cents per gallon of ethanol Ethanol plants began switching from coal to natural gas for power generation and adopting other cost-reducing technologies. An expanding market and the high cost of fructose corn syrup encouraged expansion of wet mill plants that produce the syrup as a by-product of the ethanol production process. The Energy Policy Act of 1992 (EPACT) provided for two additional gasoline blends (7.7% and 5.7% ethanol). It defined ethanol blends with at least 85% ethanol as ‘‘alternative transportation fuels.’’ It also required specified car fleets to begin purchasing alternative fuel vehicles, such as vehicles capable of operating on E-85 (a blend of 85% ethanol and 15% gasoline). EPACT also provided tax deductions for purchasing (or converting) a vehicle that could use an alternative fuel such as E-85 and for installing equipment to dispense alternative fuels. The Clean Air Act Amendments mandated the winter-time use of oxygenated fuels in 39 major carbon monoxide nonattainment areas (areas where EPA emissions standards for carbon mioxide had not been met) and required yearround use of oxygenates in nine severe ozone nonattainment areas in 1995. MTBE was still the primary oxygenate used in the United States.

11


Table 2 (continued)

1995

The excise tax exemption and income tax credits were

extended to ethanol blenders producing ETBE.

The EPA began requiring the use of reformulated gasoline

1995–1996

1997

1998

1999

2000 2001 2002

2003

year round in metropolitan areas with the most smog. With a poor corn crop and the doubling of corn prices in the mid-1990s to $5 a bushel, some States passed subsidies to help the ethanol industry. Major US auto manufacturers began mass production of flexible-fueled vehicle models capable of operating on E-85, gasoline, or both. Despite their ability to use E-85, most of these vehicles used gasoline as their only fuel because of the scarcity of E-85 stations. The ethanol subsidy was extended through 2007 with a gradual reduction from 54 cents per gallon to 51 cents per gallon in 2005. Some States began to pass bans on MTBE use in motor gasoline because traces of it were showing up in drinking water sources, presumably from leaking gasoline storage tanks. Because ethanol and ETBE are the main alternatives to MTBE as an oxygenate in gasoline, these bans increased the need for ethanol as they went into effect. EPA recommended that MTBE should be phased out nationally. A 1998 law reduced the ethanol subsidy to 53 cents per gallon starting January 1, 2001. US automakers continued to produce large numbers of E-85-capable vehicles to meet federal regulations that require a certain percentage of fleet vehicles to be capable of running on alternative fuels. Over 3 million of these vehicles were in use. At the same time, several States were encouraging fueling stations to sell E-85. With only 169 stations in the United States selling E-85, most E-85 capable vehicles are still operating on gasoline instead of E-85. A 1998 law reduced the ethanol subsidy to 52 cents per gallon starting January 1, 2003. As of October 2003, a total of 18 States had passed legislation that would eventually ban MTBE. California began switching from MTBE to ethanol to make reformulated gasoline, resulting in a significant increase in ethanol demand by midyear, even though the California MTBE ban did not officially go into effect until 2004. (continued)

12

W. A. Payne

Table 2

(continued)

2005

2007

2008

The Energy Policy Act of 2005 was responsible for regulations that ensured gasoline sold in the United States contained a minimum volume of renewable fuel called the Renewable Fuels Standard. The regulations aimed to double the use of renewable fuel, mainly ethanol made from corn, by 2012. The Energy Independence and Security Act of 2007 expanded the Renewable Fuels Standard to require that 36 billion gallons of ethanol and other fuels be blended into gasoline, diesel, and jet fuel by 2022. The United States consumed 6.8 billion gallons of ethanol and 0.5 billion gallons of biodiesel in 2007. An Argonne National Laboratory study compared data dealing with water, electricity, and total energy usage from 2001 and 2006. During this period, America’s ethanol industry achieved improvements in efficiency and resource use while it increased production nearly 300%. As of March 2008, US ethanol production capacity was at 7.2 billion gallons, with an additional 6.2 billion gallons of capacity under construction.

From EIA (2009).

A new round of bankruptcies is occurring today (Economist, 2009; NYT, 2009a,b). The demand for ethanol has long been influenced by the supply and demand for gasoline. Demand for fuel and therefore ethanol rose dramatically during World War I, and fell drastically after World War II, when there was reduced need for war materials and plentiful supplies of cheap gasoline. From the late 1940s until the late 1970s, virtually no commercial fuel ethanol was available anywhere in the United States. Until 1972, Americans were accustomed to expanding energy consumption with little concerns about supply or sharp price increases (Hakes, 1998). When gasoline prices have become high due to strong demand or constricted supply, government policies have strongly supported the development and production of ethanol. The legislative actions to promote fuel ethanol in 1974, for example, were in response to turmoil in 1973, which began with electricity brown outs and rapidly rising prices for fuel, food, and other necessities. Then, an oil embargo was imposed in October 1973 on the United States by members of the Organization of Arab Petroleum Exporting Countries, cutting further into the supply of oil and elevating prices to levels previously thought impossible


13

(Hakes, 1998). About $1 billion was invested then by the US government into conversion technologies. Over the past several years, remarkable gains have been made in both efficiency of conversion processes and production capacity.

2.2. Soil and oil Soon after the energy crisis, the soil physicist C.H.M. Bavel (1977) noted that the United States was importing $36 billion worth of oil per year and exporting $23 billion of agricultural products, mostly grain and soybeans. The bounty of our farms, he argued, was supporting and extending the profligacy of our energy consumption. He also wondered whether the production levels at the time of wheat, corn, and soybean could be maintained, particularly in view of very high rates of soil erosion. Van Bavel (1977) cited a 1971 analysis suggesting that unrestricted land use, including expansion of agriculture into marginal lands, could lead to a national soil loss rate of 20 metric tons per ha, which at the time was seen as twice as high as the maximum tolerable rate. In effect, he argued, we were exporting several tons of soil to the Gulf of Mexico for every ton of grain exported to offset our energy demand. This amounted to a bad trade of soil for oil.

2.3. Charting our future in the past In the late 1980s, soon after US legislation began mandating the addition of oxygenates in gasoline, which in effect increased demand for ethanol, then ASA president E.C.A. Runge (1990) pointed out a fundamental dilemma that USDA’s supply control programs and the potential of ethanol from crops presented for agricultural scientists: We are supposed to develop the technology that keeps US farmers competitive, but we aren’t supposed to create the technology that creates a surplus for the secretary of agriculture to deal with. Obviously, you can’t have one without the other. It is this surplus agricultural production which has negatively impacted agriculture and agronomists, in particular, for the past decade. Can we create a demand for this excess agricultural production?

Even then, there were detractors of mixing ethanol into fuels. Runge (1990) cited one author who suggested that ‘‘gasohol’’ would not be economical until crude oil reached $60 per barrel. Runge (1990) argued that whether ethanol production was economically viable depended on the comparison made and the technology assumed, and that if the cost of ethanol production was compared with the cost of USDA’s supply control programs, the ethanol alternative was very economical. He reasoned further that, because a very large percentage of our crops are grown under rainfed conditions, we cannot predict agricultural supply.

14

W. A. Payne

Therefore, government policy must opt for more than enough crop production during average or normal weather years to see us through drought years, when production is below average. That is, excess crop production is the norm rather than the exception, and supply control policies would always be out of phase with need if weather is a variable. Runge (1990) proposed that ethanol be used as a ‘‘sink’’ for any excess crop production rather than utilizing acreage reduction programs, export enhancement programs, etc., to control supply. He calculated that we could have saved nearly a billion dollars by converting the corn that we were exporting in 1987 to ethanol instead. Other positive aspects of such a policy included increased gross domestic product, rural development, improved air quality, CO2 reduction, and revenue enhancement at the local, state, and federal levels. But Runge also stressed that we must have a US agriculture that is not only enhanced by science but in harmony with environmental and human values. He called for policies that would create a favorable climate for investors and companies to design plants to use this excess production, pointing out that there would be little investment if there were no assured supply or unstable prices of the raw materials needed to run their plants. Even 20 years ago, ethanol production technology was changing rapidly, leading Runge (1990) to predict that the positive energy contribution of ethanol produced would increase dramatically in the next few years with state-of-the-art plants. Runge’s (1990) goal of sustainably utilizing our agricultural enterprise to its maximum included a vision of cooperation between US agriculture, the energy industry, the motor fuels industry, and environmentalists to solve our problems of air quality, greenhouse gas emission, energy imports, and agricultural and rural development problems. He believed that the beneficiaries of this cooperative effort would include US agriculture, and agricultural scientists who provide progrowth technologies for US agriculture. But the main beneficiaries would be US citizens because of a revitalized rural economy, increased GDP, and improved air quality.

3. An Overview of Biofuels1 Biofuels contain energy derived from biomass produced through the capture of solar energy through photosynthesis. A wide range of biomass can be used to produce several forms of biofuel. Sources of biomass include waste from food, fiber and wood industrial processes, and any number of 1

Much of this summary relies upon the thorough review by FAO (2008).


15

agricultural and forestry products. Biomass can be used to generate electricity, heat, power, fuel, and other forms of bioenergy. Because the primary source of energy is solar (even if animal products are used), biofuels are seen by many as a form of renewable energy. Biofuels can be in the form of solid, liquid, or gas. They can also be classified as primary (i.e., unprocessed) or secondary (i.e., processed). Primary biofuels are directly combusted, usually for cooking, heating, or electricity production needs in industry.2 Secondary biofuels can be solid (charcoal or wood pellets), liquid (ethanol, biodiesel, or bio-oil), or gaseous (biogas or hydrogen). Secondary biofuels can be used for a wider range of applications, including transport and high-temperature industrial processes. Of course, the strongest growth in recent years has been in secondary liquid biofuels for transport, which are mostly produced using agricultural and food commodities as feedstocks. The most important of these are ethanol and biodiesel (FAO, 2008).

3.1. Ethanol Ethanol produced for biofuel today is based on feedstocks containing either sugar or starch. Common sugar crops used as feedstocks include sugarcane, sugar beet, and sweet sorghum. Feedstocks containing starch or cellulose, which can be converted to sugar, can also be used to produce ethanol. The most common among these include corn, wheat, and cassava. Especially in Brazil and other tropical countries, sugarcane is the most widely used feedstock. In nontropical countries, the starch component of cereals is more commonly used. Ethanol can be blended with gasoline or burned in its pure form in internal combustion engines. One liter of ethanol contains approximately 66% of the energy of 1 l of gasoline, but it has a higher octane level. When mixed in gasoline, it therefore improves performance and fuel combustion in vehicles, thereby reducing emissions of carbon monoxide, unburned hydrocarbons and carcinogens. However, the combustion of ethanol also causes a heightened reaction with nitrogen in the atmosphere, which can result in a marginal increase in nitrogen oxide gases. Ethanol also only contains small amounts of sulfur. When mixed with gasoline, ethanol therefore reduces fuel sulfur content and emissions of sulfur oxide, which contributes to acid rain and is a carcinogen.

2

One recent article (Campbell et al., 2009) suggests that converting biomass to electricity to power batterypowered vehicles is much more land-use efficient, transport-efficient, and emission-offset efficient than ethanol.

16

W. A. Payne

3.2. Biodiesel Biodiesel is produced by combining vegetable oil or animal fat with an alcohol and a catalyst through transesterification. Oil for biodiesel production can be extracted from most oilseed crops. The most popular sources are rapeseed in Europe and soybean in Brazil and the United States. In tropical and subtropical countries, biodiesel is produced from palm, coconut, and jatropha. Small amounts of animal fat, from fish- and animal-processing operations are also used. The production process typically yields additional by-products, such as crushed bean ‘‘cake’’ that can be used as an animal feed, and glycerine. Because biodiesel production can be based on a wide range of oils, the resulting biofuels have a greater range of viscosity and combustibility than ethanol. Biodiesel can be blended with traditional diesel fuel or burned in pure form in compression ignition engines. Its energy content is 88–95% of regular diesel, but it improves lubricity and raises the cetane value, making its fuel economy generally comparable to that of diesel. Its higher oxygen content aids in fuel combustion, thereby reducing emissions of particulate air pollutants, carbon monoxide and hydrocarbons. Similar to ethanol, biodiesel also contains only traces of sulfur. Straight vegetable oil is another potential fuel for diesel engines that can be produced from a variety of sources, including oilseed crops, cooking oil, and animal fat.

3.3. Cellulosic ethanol The starch and sugar components of crops represent only a small fraction of total plant mass, which is mostly composed of cellulose, hemicellulose, and lignin. Cellulose and hemicellulose can be also converted into ethanol after they are first converted into sugar, but the process is more difficult. A second generation of technology—termed recently the ‘‘holy grail’’ of biofuels (CSM, 2009)—promises to make it economically possible to use cellulosic biomass for ethanol production. There is currently much ongoing research and even a few pilot plants devoted to converting cellulosic biomass into ethanol, but little commercial-scale production. As cellulosic biomass is the most abundant biological material on earth, the successful development of commercially viable second-generation cellulose-based biofuels could significantly expand the volume and variety of feedstocks that can be used for production. Cellulosic wastes, including waste products from agriculture (straw, stalks, leaves) and forestry, wastes generated from processing (nut shells, sugarcane bagasse, sawdust) and organic parts of municipal waste, could all be potential sources. Potential crops that could serve as a feedstock source for cellulosic ethanol include short-rotation woody crops, fast-growing trees, and grassy


17

species such as switchgrass. Since the entire crop can be used, an ideal plant species would rapidly produce large amounts of biomass. The use of cellulosic biomass would theoretically permit the production of more fuel per hectare of land. Furthermore, some species are adapted to poor degraded soils, which in theory could provide avenues not only for land rehabilitation but avoid competition for land with food crops.3

3.4. Biofuel feedstocks and conversion to biofuel Because nearly any source of biomass can be used for biofuel, there is a wide array of potential biofuel feedstocks across the world. Currently, for instance, by-products of forest industries are used to produce fuelwood and charcoal, while those of pulp mills provide a major fuel source for bioelectricity generation in many countries. A number of crop and forest residues are also used to produce heat and power. But the largest growth in recent years has been in ethanol and diesel biofuels for transport using agricultural crops as feedstocks. In 2007, 85% of the global production of liquid biofuels was in the form of ethanol (Table 3). Despite the fact that almost any biomass source can be used, most of the world’s ethanol production comes from sugarcane or corn (FAO, 2008). In Brazil, the bulk of ethanol is produced from sugarcane, while in the United States it is produced from corn. Other significant Table 3

2007 ethanol and biodiesel production of the world and selected countries

Country/country grouping

Ethanol (millions of l)

Biodiesel (millions of l)

Brazil Canada China India Indonesia Malaysia United States of America European Union Others World

19,000 1000 1840 400 0 0 26,500 2253 1017 52,009

227 97 114 45 409 330 1688 6109 1186 10,204

From FAO (2008).

3

Cellulosic ethanol from woody crops, fast-growing trees, and grasses is not at all without its strong critics. See, for example, the long and heavily referenced report by the environmental group Global Forest Coalition (2007).

18

W. A. Payne

feedstocks include cassava, rice, sugar beet, and wheat. The two largest ethanol producers, Brazil and the United States, made up nearly 90% of total production in 2007, with major production occurring also in Canada, China, the EU (mostly France and Germany), and India. For biodiesel, the most popular feedstocks are rapeseed in the European Union (EU), and soybean in the United States and Brazil. Palm, coconut, and castor oils are used in tropical and subtropical countries as biodiesel feedstocks, and the use of jatropha has been rapidly increasing. Biodiesel production was principally concentrated in the EU, with much smaller production in the United States. Other significant biodiesel producers include Brazil, China, India, Indonesia, and Malaysia. Because of potentially rapid changes in prices, government policies, land-use patterns, and public perceptions on food security, subsidies, the environment, etc., it is difficult to know how the absolute and relative trends shown in Table 3 for global biofuel production will evolve among countries—changes have been occurring almost weekly (FAO, 2008). Crop yield data in Table 4, taken from FAO (2008), are presented firstly to illustrate what agronomists already know very well—crops vary widely in terms of yield per hectare across regions and production systems. But agronomists also know better than any how crop yields change from place to place and year to year due to a myriad of processes that underlie the complexity of our science. Even the most tranquil agronomist must feel compelled to question single static yield values given for, say, corn and soybean for the entire United States. Agronomists also know that almost any measure of crop quality, which for biofuels includes sugar, starch, oil, and cellulose contents, also changes with complex environmental and genetic processes. Therefore, without even entering into the chemical engineering aspects of conversion efficiencies listed in Table 4, agronomists know that these efficiencies cannot be seen as fixed or static. Agronomists, plant breeders, and other agricultural scientists should critically view such static values not only in terms of how they represent current yields, but how well they represent potential yields that new technologies could bring about. This is part of the centuries-old Malthusian debate (Evans, 1998), which Runge (1990) framed within the context of biofuel and our capacity to sustainably increase agronomic production. This is not meant to criticize the illustrative overview given by FAO (2008). It is rather to illustrate just one source of uncertainty and disagreement from one of many complex scientific disciplines involved in the biofuel debate. Further uncertainties come into play when considering all the energy requirements needed to produce a crop and convert it into biofuel, as illustrated by the range of energy balance calculations shown in Fig. 1. A fossil energy balance of 1.0 implies as much energy is needed to produce 1 l of biofuel as it contains. An energy balance of 2.0 means that 1 l contains

19


Table 4 Static estimates of crop yield, conversion efficiencies, and biofuel yields for the world and selected countries

Crop

Sugar beet Sugarcane Cassava Maize Rice Wheat Sorghum Sugarcane Sugarcane Oil palm Oil palm Maize

Maize Cassava Cassava Soybean

Soybean

Global/ national estimates

Biofuel

Crop yield

Conversion efficiency

Biofuel yield

Global

Ethanol

46.0

110

5060

Global Global Global Global Global Global Brazil India Malaysia Indonesia United States of America China Brazil Nigeria United States of America Brazil

Ethanol Ethanol Ethanol Ethanol Ethanol Ethanol Ethanol Ethanol Biodiesel Biodiesel Ethanol

65.0 12.0 4.9 4.2 2.8 1.3 73.5 60.7 20.6 17.8 9.4

70 180 400 430 340 380 74.5 74.5 230 230 399

4550 2070 1960 1806 952 494 5476 4522 4736 4092 3751

Ethanol Ethanol Ethanol Biodiesel

5.0 13.6 10.8 2.7

399 137 137 205

1995 1863 1480 552

Biodiesel

2.4

205

491

From FAO (2008).

twice the amount required to produce it. The FAO (2008) report used calculations from the Worldwatch Institute (2006), an organization which not all taking part in the lively debate captured in Table 1 would see as neutral. Nonetheless, the figure serves to illustrate that there exist wide variations in energy balances estimated for different feedstocks and fuels. Some of these considerations are mentioned in the contrasting studies of Pimentel and Patzek (2005) and Liska et al. (2008). A simplified list of factors to consider includes the energy associated with land preparation, growing and harvesting the crop, processing the feedstock into biofuel, transport of both feedstock and biofuel, and storage, distribution, and retail of biofuel. For many reasons, including choice of data sources, energy terms that are excluded or included, and methodologies used, energy balance of biofuel is a very contentious subject (FAO, 2008).

20

W. A. Payne

Fuel Petrol Diesel Biodiesel

Ethanol

Feedstock • Crude oil • Crude oil • Soybean • Rapeseed • Waste vegetable oil • Palm oil

0

1

2

3

4

5

6

7

8

9

10

• Sweet sorghum • Maize • Sugar beet • Wheat • Sugarcane • Cellusosic

Figure 1 Estimated ranges for fossil energy balances of biofuel from different feedstocks, based on existing studies. Reproduced from FAO (2008).

Some general conclusions can be drawn nonetheless from the illustrative data in Tables 3 and 4, and Fig. 1:

Per area production of biofuel will vary with biomass yield, which differs among crop species and environments. It also changes due to differences in conversion efficiencies among crops. This implies vastly different land requirements for increased biofuel production, depending on the crop and location. All biofuels appear to make a positive energy contribution, but to widely varying degrees (but see the Pimentel and Patzek (2005) reference, which strongly disagrees). Estimated fossil fuel balances for biodiesel range from around 1 to 4 for rapeseed and soybean feedstocks. Values are much higher for palm oil because other oilseeds must be crushed before oil can be extracted (palm oil is often grown in sensitive rainforest environments) For crop-based ethanol, the estimated balances range from less than 2 for corn to around 2–8 for sugarcane. The favorable fossil energy balance of sugarcane-based ethanol in Brazil is due to high feedstock productivity and the fact that it is processed using biomass residues from the sugarcane as an energy input. The range of estimated fossil fuel balances for cellulosic feedstocks is even wider, reflecting the uncertainty regarding this technology and the diversity of potential feedstocks and production systems. Overall, the FAO (2008) report finds that most studies project that cellulosic ethanol could dramatically reduce greenhouse gas emissions compared to petroleum fuels and first-generation biofuels. It is harder to convert to liquid fuel, but cheaper to handle and easier to store because it resists deterioration. On the other hand, cellulosic biomass can be bulky and would require a developed transportation system, each representing energy requirements.


21

Currently, ethanol production from sugarcane and sugar beet has the highest energy balance values, with sugarcane-based production in Brazil and India near the top of the list. Yields per hectare for biofuel are somewhat lower for corn, but there are marked differences between countries for crop yields. None of this takes into account the cost of producing biofuels from different countries, which among other things would be influenced by government subsidies, transportation infrastructure, and technological capacity.

3.5. Bioenergy and biofuel potential on a global scale It seems appropriate to put the role of biofuels in a global context of energy demand. According to the FAO (2008) report:

4

Bioenergy makes up approximately 10% of total world energy supply. Most consists of traditional, unprocessed biomass such as wood for heating and cooking, but commercial bioenergy is assuming greater importance. Liquid biofuels for transport have received most public attention and have seen a rapid change increase in production and research spending. However, on a global scale their role is only marginal, making up only 1% of total transport fuel consumption and 0.2–0.3% of total energy consumption worldwide. Large-scale production of biofuels will require large areas of land for feedstock production. Among other things, this implies that production of liquid biofuels could dramatically change current land-use practices, but even so they could only potentially displace fossil fuels for transport to a very limited extent.4 Even though liquid biofuels supply only a small share of global energy needs, they still have the potential to have a significant effect on global agriculture and agricultural markets because of the volume of feedstocks required and the relative land areas needed for their production. Second-generation, cellulosic biofuels would increase the quantitative potential for biofuel generation per hectare of land and could also improve the fossil energy and greenhouse gas balances of biofuels. It is not known when such technologies will enter production on a significant commercial scale.

Not all would agree with this. At least one study by Sandia National Laboratories and General Motors Corp. found that plant and forestry waste and dedicated energy crops could sustainably produce 90 billion gallons of ethanol and replace nearly a third of gasoline use by the year 2030 (ScienceDaily, 2009a,b,c). The study assumed 75 billion gallons would be ethanol made from nonfood cellulosic feedstocks and 15 billion gallons from corn-based ethanol.

22

W. A. Payne

Thus far, we have seen that energy balances, greenhouse gas emissions, and potential for global production of biofuels are controversial and uncertain. But attendant issues they raise for sustainability are more uncertain and contentious still.

4. Sustainability Issues Even 10 years ago, there were more than 100 published definitions of the term ‘‘sustainability’’ (Payne et al., 2001). The ASA web site (www. agronomy.org) reminds us that there is actually a legal definition of sustainable agriculture in the United States (US Code Title 7, Section 3101): An integrated system of plant and animal production practices having a sitespecific application that will over the long-term:

1. Satisfy human food and fiber needs. 2. Enhance environmental quality and the natural resource base upon which the agriculture economy depends. 3. Make the most efficient use of nonrenewable resources and on-farm resources and integrate, where appropriate, natural biological cycles and controls. 4. Sustain the economic viability of farm operations. 5. Enhance the quality of life for farmers and society as a whole. To become sustainable, Payne et al. (2001) argued that agricultural systems of the world needed to transition toward ones that are characterized by favorable economics, conservation of resources, preservation of ecology, and promotion of social justice. We will consider the current and potential roles of biofuels in these four categories

4.1. Favorable economics? The summary of Table 2’s timeline made the point that demand and government policies for ethanol production had historically been related to demand and supply of petroleum-based fuels. FAO (2008) described a recent stepwise relationship in which the price of oil seemed to raise corn prices as ethanol production expanded. Before mid-2004, oil prices were so low that corn could not compete as an ethanol feedstock, even with available subsidies. As oil prices began to rise in mid-2004, corn prices were still very low, but by early 2005, prices had exceeded $60 per barrel—the threshold value mentioned in Runge’s (1990) paper—and corn was nearly competitive as a feedstock even without subsidies. The US Energy Policy Act of 2005 established the Renewable Fuel Standard


23

starting at 4 billion gallons in 2006, and rising to 7.5 billion by 2012. This prompted a boom in ethanol plant construction, and the demand for corn to produce ethanol expanded rapidly. Corn prices continued to rise steadily throughout 2006, at least partly due to ethanol demand, while oil prices stayed near $60 per barrel. But as corn prices rose, the viability of corn as an ethanol feedstock fell,5 even with subsidies, and many ethanol plants began to operate at a loss. But when oil prices began to rise sharply in mid-2007 to peak at $145 per barrel in the middle of 2008, corn again became economically viable as an ethanol feedstock, at least with subsidies.6 And when the price of oil and other commodities fell after mid 2008, ethanol plants again began to go bankrupt across the country (Economist, 2009; NYT, 2009a,b). As touched on in the Runge (1990) paper, the price of corn itself is influenced by both agricultural and energy policies. A much more thorough review of policy effects on biofuel and other commodity prices is included in the FAO (2008) report, but a major conclusion is that, at least with our current set of technology, production capacity, and policies on subsidies and import tariffs, US production of ethanol from corn will be economically viable only when oil prices are rising but have not yet driven up corn prices. In a more general sense, because energy markets are large relative to agricultural markets, and because agricultural prices themselves are affected by energy costs associated with farming, oil prices will drive agricultural prices. To some extent, government policy can modify this relationship. The FAO (2008) report finds that, in most cases (Brazil’s ethanol production from sugarcane being perhaps an exception), recent policies have been costly and have introduced new distortions to already distorted and protected agricultural markets at both domestic and global levels. Overall, then, economic viability of biofuels is affected by prices of agricultural feedstocks, technology and infrastructure, and government policy. But the real driving force is the price of oil and other energy sources, which plunged by $115 last year due to falling demand associated with a global recession. Today demand for oil is still falling at a sustained rate not seen since the early 1980s, and US inventories are higher than since September 1990. However, in a recent article, The Economist (2009) reported fears of Saudi Arabia’s oil minister, the CEO of Chevron, Britain’s energy minister, and others that, when the global economic crisis comes to an end, the demand for oil will rise and possibly cause another price shock, with prices 5

6

High prices have also challenged the economic viability of traditionally corn-dependent industries, such as beef, pork, and poultry (DiLorenzo and Crawford, 2008; Ellis, 2006), although by-products such as ‘‘distillers grain’’ can be fed to animals. At least some studies suggest that ethanol reduces the price of oil. One Merrill Lynch report found that biofuels kept oil prices $21 a barrel lower than they might otherwise have been when prices peaked in the summer of 2008 (San Francisco Monitor, 2008).

24

W. A. Payne

possibly surpassing $145 per barrel. Many of the factors that drove the price hike last year remain in place: Much of the world’s ‘‘easy’’ oil has already been extracted, or is in the hands of nationalist governments that will not allow foreigners to exploit it. That leaves firms to hunt for new reserves in ever more inhospitable and inaccessible places, such as the deep waters off Africa or the frozen oceans of the Arctic. Such fields take a long time and a lot of expensive technology to develop. Worse, new discoveries tend to be smaller than in the past and to run dry faster.

Additionally, we know that as economies recover, there will be vast new markets in the developing world, including China and India, where oil consumption has been growing fast. As soon as the world economy starts growing again, demand for oil in theory will once again outstrip the capacity to supply it, possibly sending prices soaring again. This would likely usher in another rush to support and promote ethanol production from crops, which, if the recent trends were repeated, would be economically viable until the prices of feedstocks themselves are driven higher. A very recent NYT (2009b) article sums up the current difficulty in forecasting oil prices, and by inference the economic viability of ethanol and other biofuels: ‘‘The extreme volatility that has gripped oil markets for the last 18 months has shown no signs of slowing down, with oil prices more than doubling since the beginning of the year despite an exceptionally weak economy. The instability of oil and gas prices is puzzling government officials and policy analysts, who fear it could jeopardize a global recovery. It is also hobbling businesses and consumers, who are already facing the effects of a stinging recession, as they try in vain to guess where prices will be a year from now—or even next month.’’

4.2. Conservation of resources The natural resources that may be negatively affected by farming in general, and biofuels in particular, include soil, water, and air. Their sustainable use is vital for our very survival. All alternative fuel production technologies could have environmental impacts; greater understanding is needed to guide policy and to avoid or mitigate unintended environmental consequences of biofuel production (Simpson et al., 2008). 4.2.1. Soil resources When van Bavel (1977) wrote about our ability to sustain high yields without unacceptable soil loss, the United States was importing $36 billion of oil per year, exporting $23 billion of agricultural products per year, and experiencing a national soil erosion rate as high as 20 metric tons per ha, or twice as high as the maximum tolerable rate. He realized that high yields did


25

not necessarily cause soil erosion, and that existing research suggested that high production and soil conservation were compatible. How have we done since then? In 2008, we imported 4.7 109 barrels of oil at an average price of about $90 per barrel (EIA, 2009), for a total of about $423 billion, or nearly an astonishing 12 times more than when van Bavel (1977) wrote his letter. Our agricultural productivity has continued to increase as well; the United States exported about $101 billion worth of agricultural goods, or more than four times as much as in 1977, including $33 billion for approximately 70 million tons of grain and $19 billion for 27 million tons of soybean (ERS/FAS, 2008). The bounty of our farms continues therefore to support the profligacy of our energy consumption, but our bounty is not keeping pace with our profligacy—energy consumption has risen much faster than agricultural productivity. And soil erosion? According to a 2003 report cited by NRCS (2007), erosion rates on a per acre basis declined significantly between 1982 and 2003. The report’s major conclusions include:

Water (sheet and rill) erosion on cropland dropped from 4.0 tons per acre per year in 1982 to 2.6 tons per acre per year in 2003. Wind erosion rates dropped from 3.3 to 2.1 tons per acre per year. Declines in soil erosion rates have moderated somewhat since 1997, but the general downward trend in both water and wind erosion continued through 2003. In 2003, 102 million acres (28% of all cropland) were eroding above soil loss tolerance rates, compared to 169 million acres (40% of cropland) in 1982. In 2003, 266 million acres (72% of cropland) were eroding at or below soil loss tolerance rates, compared to 251 million acres (60% of cropland) in 1982. In 2003, highly erodible land (HEL) cropland acreage was about 100 million acres, compared to 124 million acres in 1982. HEL cropland acreage eroding above soil loss tolerance rates declined 35% between 1982 and 2003. Gains in erosion control continue to occur even though the cropland base is continually changing. Significant acreages of cropland are retired or converted to other land uses, while new lands not previously cropped are being converted to cropland.

Thus, at least in the United States, even though above-tolerance rates of erosion continue on a third of our cropland, we have nonetheless seen that continued agricultural intensification has been associated with greater soil conservation, not less. Land-use change and intensification of agricultural production can have significant adverse impacts on soils, but all types of erosion have been

26

W. A. Payne

researched extensively, and many practices and techniques are available for controlling wind and water erosion just about anywhere in the world (Unger et al., 2006a). But some of the regions of the world in which biofuel is viewed as having the potential to expand—Africa, for example—are currently experienced alarming rates of erosion and other forms of land degradation (Payne, 2010). No one practice or technique—or policy, for that matter—is universally effective for all climates, soils, landscapes, and management conditions in all seasons or years, but most effective controls are usually achieved by using some combination of crop residues and soil surface manipulation. But the very crop residues needed to conserve soil are a potential source of cellulosic biofuel feedstock, which worries some soil scientists. The amount of residue produced by a given cropping system, and therefore the amount potentially available as a feedstock while still leaving sufficient amounts for conserving the soil resource, will change with crop species, soil properties, rainfall, and a long list of agronomic practices. In dry areas, there is often not even sufficient moisture to produce enough residue to achieve soil conservation (Unger et al., 2006a), let alone to export for biofuel production. Some dryland crops, such as cotton, tend to produce insufficient residue to protect soil from wind erosion (Baumhardt and SalinasGarcia, 2006), yet cotton residue has been proposed as a source of biofuel. In areas prone to wind erosion, 0.15–0.40 m of standing stubble is recommended for small grains (Unger et al., 2006a), which many dryland crops never attain. Temperature is another consideration, since it strongly influences the rate of residue decomposition (Ladd and Amato, 1985). Recent research suggests that about 25% of corn stover might be sustainably removed from corn cropping systems if they are not on erosion-prone lands (Blanco-Canqui and Lal, 2009). Wilhelm et al. (2004) reported that estimates of the amount of corn stover needed to maintain soil carbon were in the range of 5.25–12.50 Mg ha–1. For wheat-based systems, Lafond et al. (2009) found that potential exists to use crop residues without adversely affecting long-term productivity of medium- to heavytextured soils, provided that 800 C), long vapor residence time

Solid (biochar) (%)

Gas (syngas) (%)

75% (25% water)

12

13

50% (50% water)

25

25

30% (70% water)

35

35

5% tar (5% water)

10

85

Liquid (bio-oil)

similar to fast pyrolysis, but including a limited supply of oxygen). These produce predominantly heavy oil or gas, yielding only small proportions of char. The development of integrated systems that produce energy and char at high efficiency by slow pyrolysis is still, therefore, mainly on the research scale, with technology commercially deployed only at a handful of locations. The energy content of biochar depends on its feedstock, but may reach 30 and 35 MJ kg-1 (Ryu et al., 2007). Therefore char is conventionally used to provide the heat driving the primary pyrolysis through burning or gasification (Demirbas et al., 2006), or to dry incoming feedstock. Maximizing biochar is therefore at the expense of usable energy in gaseous and liquid forms (Demirbas, 2006) and has an opportunity cost. Although a mitigation strategy for the abatement of greenhouse gases may favor maximization of biochar production (Gaunt and Lehmann, 2008), the realistic balance is a function of market and engineering constraints. Process parameters fundamentally affect the properties of the biochar product, modifying its possible value in agriculture, and in the sequestration of carbon. Temperature and residence (heating) time are particularly important, but the feedstock and its interaction with temperature may be equally significant. Human society in general has extensive experience of pyrolysis in the specific context of producing charcoal, mainly as a clean-burning fuel. Currently produced in rural areas for urban markets and industrial smelting of iron (where carbon credits can be obtained for emissions offset from coke), traditional methods of charcoal production predominate and

A Review of Biochar and Its Use and Function in Soil

51

represent the most widely practiced form of pyrolysis at the current time. In charcoal production process heat is generated within the kiln by initiating combustion of some of the feedstock in air, prior to restricting air flow. Although no fuel is required for external heating, the combustion phase and incomplete exclusion of air lowers conversion rates from biomass to charcoal (about 20% in traditional kilns). This type of production is typically small scale and a ‘‘batch’’ process where the full cycle of heating and cooling is applied to a confined and stationary charge. In addition to being energetically inefficient and time-consuming, these processes are highly polluting: potentially useful gases are emitted, together with aerosol (smoke) from partially combusted oil and tar. Unfortunately, in addition to the emission of CO2 and potent trace greenhouse gases in these streams, traditional charcoal manufacture is also implicated in depletion of forest and wood fuel resources, and wasteful in terms of utilization of biomass carbon as such a large fraction is lost back to the atmosphere. Nonetheless, the current definition of biochar does encompass charcoal, ostensibly to recognize its historic value in soil management, and to acknowledge the accessibility and universality of technology used to produce it. Much evidence drawn upon to assess biochar function rests on studies made using charcoal. The difference in the proportion of feedstock carbon retained is the key difference between traditional production of charcoal and slow pyrolysis, by which typically 30–40% of feedstock mass is recovered as char. Although optimised for char production it is thought that 50% retention might be achievable. It is also possible that additional carbon retained in slow pyrolysis is not chemically or physically consistent with traditional charcoal containing, for example, a greater concentration of hydrogen and oxygen (due to less complete pyrolysis), deposited oils and tars, and possibly thermal alteration of these. The formation of secondary char increases with vapor residence time, which is in turn a function of the rate at which gas flows or is propelled from the reactor. The current position of pyrolysis in the context of a range of other biomass conversion processes is shown in Fig. 1. Although feedstock is important in determining the function of biochar in soil, there is no consensus as to optimal feedstock in terms of both soil use and energy production, mainly because commercial pyrolysis plants are scarce, and those that exist are associated with the processing of specific waste streams. A limited amount of research-scale pyrolysis has been conducted using a wider range of feedstock (Gaunt and Lehmann, 2008; Das et al., 2008; Day et al., 2005). Feedstock currently used at commercial and research facilities includes wood chip and wood pellets, tree bark; crop residues including straw, nut shells, and rice hulls; switch grass; organic wastes including paper sludge, sugarcane bagasse, distillers grain, olive waste (Yaman, 2004); chicken litter (Das et al., 2008), dairy manure, and sewage sludge (Shinogi et al., 2002). Research- and pilot-scale pyrolysis has been undertaken at a rate of 28–300 kg h-1 (dry feedstock mass basis), which is

Feedstocks

Process

Product

Uses and applications

Biomass energy crops (corn, cereals, wood pellets, palm oil, oilseed rape)

Fast pyrolysis (anhydrous)

Synthesis gas Bio-oil liquid Biochar solid

– Heat – Fuel (combusted to generate electricity or converted to syngas) – High value bio-chemicals used as food additives or pharmaceuticals – Soil conditioners/fertilisers

Bioenergy residues “cake” Agricultural waste (wheat straw, hazelnut and peanut shells, waste wood, etc.) Compost (green waste) Manure/animal waste (chicken) Kitchen waste plastic, food, etc. Sewage sludge

Figure 1

Slow pyrolysis (low temp. 450– 550 °C, O2-free, some-times steam) Slow pyrolysis (high temp. 600– 900 °C, O2-free) Gasification (high temp., fast heating rate., O2 present)

Syngas Biochar Activated biochar Combustible ethane, methane Char

Fermentation, anaerobic digestion, and mechanical bio-treatment Carbonization (‘brown’ at 300 °C, ‘black’ at 380 °C)

Ethanol Methane and sludge Charcoal

– Soil amendment (neutral/alkaline pH, porosity retains water, cation exchange capacity: robust benefits to plant growth compared to high-temp char) – Fuel (cooking and heat) – Extreme porosity and surface area – Water filtration and adsorption of contaminants (gas, liquid, or solid) – Fuel (low yield, high reactivity) – Contamination of some feedstocks (e.g., metal and plastic in kitchen waste) may preclude use of sludge/char in soil – Fuel (for electricity or cooking) – Bi-products (wood spirits, wood tar) – Substitute for coal-derived coke in metal smelting

Biochar and other products of thermal conversion of biomass according to available technologies and feedstocks.


53

about one-tenth of commercial plants (2–3 t h-1). Comparing the efficiency of pyrolysis plants is difficult as the fate and the quantitative balance for solid, liquid, and gaseous products are rarely fully quantified in individual studies, yet vary considerably. The composition and hence heating value of syngas also varies, particularly with respect to feedstock quality and moisture content. In production of oil from biomass, the feedstock ratio in carbon, oxygen, and hydrogen is considered an indicator of quality (Friedl et al., 2005). Lowmineral and N contents provided by wood and biomass crops include shortrotation willow, high productivity grasses such as Miscanthus spp., and a range of other herbaceous plants. However, abundant and available agricultural by-products, particularly cereal straw, may be suitable. Proportions of hemicellulose, cellulose, and lignin content appear to influence the ratio of volatile carbon in oil and gas and the proportion of carbon stabilized in biochar. Feedstocks with high lignin content generate high yields of biochar when pyrolyzed at temperatures of approximately 500 C (Demirbas et al., 2006; Fushimi et al., 2003). More knowledge is required for feedstock to be selected or processed to achieve a specified balance between all three classes of pyrolysis products (i.e., gas, oil, and char). However, the basic conversion technology—slow pyrolysis or a related process—is the most critical factor. This review concerns biochar produced from the utilization of biomass with simultaneous energy capture. However, much of the current evidence for the impacts on soil properties and soil carbon rests on research using char produced experimentally by more traditional methods.

1.2. Policy context The biochar concept has a strong global context, as it is positioned strongly in the context of climate change (carbon abatement), but intrinsically linked to renewable energy capture (biomass pyrolysis), and food production and land-use change (food and feed production), further extending to the enhancement of environmental quality (control of diffuse pollution) and management of organic wastes (stabilization and use), through management of soil nutrients. Climate change and food production are causally linked as 13.5% of radiative forcing is attributable to greenhouse gases emitted through agricultural activity (Barker et al., 2007). The carbon release associated with switching land previously under natural or unmanaged vegetation to crop production releases typically large amounts of carbon from standing biomass, and also from the soil. The conversion of land use at current rates accounts for another 17.4% of radiative forcing (Barker et al., 2007) mainly through the loss of biomass carbon as CO2, and workable technologies to enhance the productivity of agriculture may be important in stemming these sources as land pressure is a significant factor. Climate change is linked

54

S. P. Sohi et al.

to energy use by the remaining 70% of radiative forcing that results from use of fossil fuels. However, energy and land-use change are also linked. Although the use of replenishable biomass in energy production is considered carbon neutral (and to offset fossil fuel emissions), it is clear that dedicated energy crops produced at a scale that is significant in terms of global energy supply would lead to direct and indirect pressure on natural ecosystems, and a net emission of CO2 in conversion. Losses documented for some existing conversions are extremely large (Fargione et al., 2008), and although the benefit of a long-term annual offset of fossil carbon emissions is important, short-term losses from conversion of natural ecosystems do not favor, even in carbon terms alone, land-use change. This is because the pace of climate change is rapid compared to the timescale for delivery of net benefit from bioenergy crops after conversion. Assessments of the realistic potential for biochar in carbon abatement have converged on a figure of about 1 GtC yr 1 (Lehmann, 2007), which presents a potential ‘‘wedge’’ for climate change mitigation; Pacala and Socolow (2004) proposed that a portfolio of such wedges is required to avert the threat of catastrophic climate change. Currently, abatement potential is the most quantifiable and certain of the many characteristics of biochar. However, at the moment simple stabilization of biomass carbon is not eligible for trading under the Clean Development Mechanism (CDM), the international scheme designed to achieve carbon abatement under the Kyoto protocol. In the absence of eligibility for carbon credits, or simply to supplement a future income stream from carbon stabilization, it is likely that biochar addition to soil will proceed only where sufficient improvements in soil performance and productivity are perceived or assured. In addition to the avoidance of CO2 and methane emissions during normal decomposition of feedstock, a suppression of nitrous oxide and methane emissions from otherwise normally managed soils is frequently claimed. Since the global warming potential (GWP) of these gases is high and the main source globally agricultural, this effect is of potential significance, although the evidence base is still very limited. A role for biochar in control of diffuse pollution from agriculture, including sorption of agrichemicals, has also been proposed, and available evidence is reviewed later in this study.

1.3. Biochar and the global carbon cycle Globally, photosynthesis by plants draws 120 Gt CO2–C from the atmosphere into energy-rich carbohydrate carbon each year (Schlesinger, 1995). Half of this is rapidly returned to the atmosphere through plant respiration, but about 60 Gt CO2–C yr 1 is invested in new plant growth (carbon comprising 45% of plant biomass). After accounting for approximately


55

2.2 Gt CO2–C yr 1 loss through land-use change (Houghton, 2003), and some ‘‘fertilization’’ from increasing atmospheric CO2 concentration, the amount of carbon in living plant biomass (560 GtC; Schlesinger, 1995) appears to be broadly constant. This confirms that the magnitude of cropping and harvest, as well as the deposition of litter and exudates by plants and plant roots into soil, is broadly equal to annual net primary productivity. The pathways by which much of this carbon is returned to the atmosphere are varied and complex, but in a particular year the gross flux to the atmosphere—the product of the progressive degradation of all cohorts of previous plant production—approximates to the amount that is fixed over the same period. Human intervention in the management of forests and agroecosystems means that 20–40% of net primary productivity is in some way associated with active management (Vitousek et al., 1986). A strategy to deploy biochar on a large scale would divert a portion of the existing global carbon flux that resides within managed ecosystems, or to intercept enhanced net primary productivity production in the form of increased harvest or waste biomass. This material would be pyrolyzed and replaced in soil in a stabilized rather than a degradable form, in this way returning much less carbon to the atmosphere from sites of decomposition (soil, landfill, etc.), and simultaneously decreasing associated emission of other greenhouse gases (notably methane). The net anthropogenic addition of carbon to the atmosphere (6.3 GtCO2–C yr 1; Houghton, 2003) is small relative to the scale of the natural atmosphere–plant–soil–atmosphere cycle. If the net primary productivity of managed agricultural and forest ecosystems is between 12 and 24 GtC yr 1, the interception and stabilization of 1 GtC does not appear an extraordinary goal. Assuming that the carbon in biochar is stable, diversion of organic resources and wastes to pyrolysis would result in a permanent offset against future atmospheric CO2, which could be extended on an annual basis, according to other priorities and circumstances for use of biomass. A strategy based around biochar thus differs from the more established proposition for sequestration of carbon in soil, where the objective is to increase the equilibrium level of active soil organic matter, which broadly requires that rate at which carbon—in the form of organic resource or wastes—is forced through the soil system to be permanently increased to a higher rate. Despite the large size of the global soil carbon pool (1500 GtC) the potential for this strategy to accumulate carbon per unit area is limited in absolute quantitative terms. This is because intensively managed lands account for a small part of the global soil carbon pool, and incremental enhancements account for trivial amounts of carbon, and because the capacity for any soil to stabilize labile carbon has fundamental limits. In addition, storage is delivered over a lengthy period of time (usually decades), and this annual rate diminishes as equilibrium is approached. The increased flow of carbon into the soil must be maintained after equilibration to avoid reversal of soil storage, making the diversion of resource into the soil a

56

S. P. Sohi et al.

permanent commitment. Increasing the quantity and turnover of carbon in soil, in the form of organic matter, seems certain to provide crop-related benefits ( Janzen, 2006; Lal et al., 2004). This is therefore a desirable strategy where sufficient organic resource exists, but should be weighed against a more efficient strategy where adopted explicitly for carbon sequestration, especially in fertile soils. The cycling of black carbon produced during wildfire provides a natural analog for a biospheric intervention based on biochar. Wildfire is currently the largest source of black carbon globally, a small proportion of above-ground biomass (about 1%) being incompletely combusted and returned to the soil as char of various forms. The extent and frequency of wildfire in many systems means that this pathway may already provide a terrestrial net sink for about 0.05–0.2 Gt yr 1 atmospheric CO2–C (Kuhlbush, 1998). Increasing recognition for the global significance of this flux arises in part, from development of measurements that discriminate black carbon from other soil carbon. These seem to indicate much larger amounts of black carbon in soil than has been assumed in global stock estimates, or than has been allowed for in soil models. This may affect, among other things, the response of the global soil pool to climate change, black carbon being much more stable than the typical components of soil carbon (Lehmann et al., 2008). Interpretation of black carbon measurements is, however, complicated by some uncertainty over their efficacy and also their capacity to discriminate charcoal from other forms of black carbon, specifically those arising from anthropogenic activity—deliberate vegetation burning, wood fuel, combustion of coal and oil. The possible indirect effects of biomass stabilization on radiative forcing have to be considered. Soot from biomass burning is implicated in an acceleration of polar ice melt, but conversely in facilitating cloud formation and ‘‘global dimming’’ (McConnell et al., 2007; Ramanathan and Carmichael, 2008). The production of biochar under controlled conditions should be clean, but the means to control methods of production are unclear. Biochar in soil also visibly darkens soil color, especially in soils that are already low in organic matter, and a relationship between soil color and occurrence of low temperature wildfire has been demonstrated (Ketterings and Bigham, 2000; Oguntunde et al., 2008). As dark soils absorb more solar energy they may, depending on water content and plant cover, display higher soil temperatures (Krull et al., 2004). This would potentially accelerate cycling of nutrients and beneficially extend growing season in temperate regions, and in Japan it is a traditional farming practice to apply charcoal to accelerate snow melt. The study of Oguntunde et al. (2008) showed a one-third reduction in soil albedo in char-enriched soils from historic charcoal making sites. On a large spatial scale, the application of biochar could potentially reduce the albedo of the Earth’s surface, whereas


57

increasing surface albedo has been proposed as a possible mitigation measure for climate forcing (Crutzen, 2006).

1.4. Scenarios for the production and deployment of biochar Producing charcoal using traditional kilns liberates greenhouse gases, particularly methane and nitrous oxide, and conserves relatively small propositions of carbon in the feedstock (FAO, 1985) and wastes the heat energy product. Apart from being associated with deforestation, sequestration of carbon into charcoal using unmodified, traditional methods may therefore not, depending on the source and ordinary fate of feedstock, provide climate change mitigation. Controlled pyrolysis stabilizes some carbon in solid form but also captures energy-rich liquids and gases which can be used to drive the pyrolysis reactions or used elsewhere. Although energy is retained in solid char the amount of energy liberated from the pyrolyzed feedstock may be higher, per mass of feedstock carbon, than in combustion. Pyrolysis could therefore be more efficient in terms of carbon emissions (CO2 MJ 1), and production of biochar carry greater abatement potential than biomass combustion, provided there is an overall adequate supply of feedstock, and storage for the biochar product is available. Although it has previously been proposed that entire valleys might be dedicated to provide storage for carbon stabilized as biochar (Seifritz, 1993), applying biochar to agricultural soil is proposed for three reasons: (1) only the soil seems to have a capacity sufficient to accommodate biochar at the scale relevant to the long-term mitigation of climate change, (2) there is potential for biochar to enhance soil function for agricultural productivity and thus offset the opportunity cost associated with its residual energy value, and (3) the possible suppression of methane and nitrous oxide release would increase the value of biochar as a means to offset agricultural greenhouse gas emissions. The impact of biochar on existing and future levels of nonbiochar soil carbon should also be considered in this context. Ideally biochar will provide reliable agronomic benefits and command a value in crop production that precludes combustion for energy, with or without a value placed on sequestration of the carbon that it contains. In this evaluation of value, allowance has to be made for the cost of acquisition and incorporation of biochar into soil. A value can also be assigned by producers and upstream food processors to the marketing potential of low-carbon or ‘‘carbon-neutral’’ food products produced in systems that deploy biochar. It is expected that a growing understanding of the relationship between feedstock, the manipulation of the pyrolysis process, and the function of biochar in soil will ultimately enable biochar to be ‘‘engineered’’ to provide

58

S. P. Sohi et al.

the balance of benefits most appropriate to a particular system. The value of the energy captured in pyrolysis must also exceed the price for the alternative use of the feedstock, unless it is genuinely a waste, in which case the normal cost of disposal can be added to the value of the energy. It should be recognized that the price of feedstock depends on demand, however, and from a market perspective wastes may cease to be wastes once demand as novel feedstock exceeds their rate of production within a relevant catchment area. In a closed-loop biochar system, the biochar product is returned to the same land that provided the feedstock. There are opportunity costs in diverting straw to pyrolysis, and there is an additional direct cost of applying biochar. There are also possible non-monetary costs associated with the collection of straw for pyrolysis, in terms of disruption of schedules and soil disturbance. Currently, there are no socioeconomic studies that have addressed such matters. A key advantage of a biochar strategy is that, assuming that the provision of key functions is limited only by the longevity of the biochar, its stability would dictate that the application frequency required to deliver benefits is single or occasional, rather than annual. Scenarios using non-waste feedstock for co-production of energy and biochar may impact commodity prices and feedback into feedstock costs. The proximity of a pyrolysis facility to an adequate catchment for feedstock must be economically and logistically viable, and can potentially affect the CO2-equivalent savings. This is the case for biomass and bioenergy facilities generally. However, for biochar the proximity of suitable locations for biochar application to soil is important as well. If the gathering of feedstock and the distribution of biochar occur over the same area the logistical and cost impacts may not be greatly affected. Possible off-farm sources of pyrolysis feedstock include municipal green waste (from gardens and parks), composted urban waste, digested sewage sludge and mixed municipal waste. In addition, in the future, by-products of other bioenergy or biofuel systems are likely to become available. Utilizing off-farm wastes is attractive in cost terms where it results in avoided landfill and other disposal costs. In addition, compared to typical or existing disposal methods, there may be a lower emission of CH4 and N2O greenhouse gases than placement in soil, enhancing the net gain in carbon equivalents through avoided emissions of gases with much higher GWP. However, many such wastes have a high water content which will incur increased emissions (and cost) associated with higher requirement for process energy in pyrolysis. In the ‘‘closed-loop’’ scenario, biochar is incorporated into the same land, the same enterprise, or groups of enterprises from which the pyrolysis feedstock originates. A typical scenario would involve the use of cereal crop straw that, in intensive arable areas, is often available at a relatively low cost. Although there is no published laboratory work to support the use of


59

biochar produced from wheat straw, there is limited existing information on the relative stability of biochar from rice husk, sugarcane bagasse, and straw from maize. Using literature evidence Gaunt and Lehmann (2008) compared the carbon-equivalent gain to be derived from pyrolysis of maize straw versus a dedicated biomass crop. Biochar produced in the latter scenario has implications for land use as indicated earlier, and it was assumed that biochar was incorporated into different land from that producing the feedstock offering greater potential for agronomic gain. In general, the defined spatial boundaries are important and indirect as well as direct land-use impacts should be considered in establishing the overall net greenhouse gas benefit (Searchinger et al., 2008). In the combined energy and bio-oil case study considered by Ogawa et al. (2006), biochar was returned as a by-product to adjacent arable land. Most scenarios considered to date have focused on conventionally managed arable land, where biochar could be added to soil as part of an existing tillage regime. Biochar could be incorporated during conversion of land to no-till, but strategic integration of application into no-till and grazed grassland systems has not been widely considered, although amendment of slurry and manure already spread presents clear opportunities.

1.5. Trading and acceptability issues for biochar carbon Biochar products used in soil will ultimately have specified agronomic value, and if associated with a carbon credit for the storage of a fraction of the carbon that they contain, a high and predictable level of stability. The ability to verify actual rates of degradation at a research level is important (Matthews, 2008; Ogawa et al., 2006), and identifying the specific characteristics that govern stability will enable such properties to be quantified and optimized, and to adapt production process accordingly. Until research provides the tools required to optimize the agronomic function of biochar, the viability of biochar-based soil management may depend on a claim for carbon credits. However, knowledge and awareness of bioenergy and carbon markets is lacking, and in the absence of markets where credit for carbon stabilized in biochar can be claimed, it is not possible for management of carbon to influence decision making of individual land users. The absence of robust figures for the costs and benefits to crop production arising from the use of biochar in soil is a major barrier to deployment. The economic case for pyrolysis of biomass rather than complete combustion is sensitive to the prevailing value of heat and power generated using other fuels, as well as the value of biochar in soil. It can be further affected by subsidy for renewable energy, which may have the effect of further inflating the value of the energy in biochar, as a renewable source.

60

S. P. Sohi et al.

Compared to schemes intended to store carbon by increasing standing plant biomass or stocks of soil organic matter, accounting for the impact of biochar should be relatively straightforward. However, a framework within which biochar carbon could be certified as an offset has not yet been established and, moreover, land-based offsets are not allowed under the CDM. Methodologies do exist for estimating the avoided emission of methane in the pyrolytic stabilization of crop residues (UNFCCC, 2007), as well as for carbon storage enhanced by adoption of no-tillage in the Voluntary Carbon Market which currently trades carbon worth USD 30 bn yr 1. Rather than dealing with carbon stabilized into biochar, a suitable accounting methodology might deal with CO2 avoided in decomposition, based on an assumption of the proportion of feedstock carbon released as CO2 during pyrolysis, and the proportion of the residual char that is entirely stable. Key elements of a suitable system based on this principle would be limited to: (1) documentation showing delivery of crop-derived feedstock to a pyrolysis facility, (2) a guarantee from the processor for the stability of the pyrolyzed product, and (3) a simple system to verify the presence of biochar in amended soils. Establishing the total carbon-equivalent gain from applications of biochar to soil, taking into account the avoidance of any non-CO2 greenhouse gases is more complex, and unlikely to emerge for some time. The mechanisms for these effects are not understood or proven, and the timescales over which they occur is much more uncertain. At the current time, a number of Annex II nations are seeking inclusion of biochar into the successor to the Kyoto Protocol, through the United Nations Convention to Combat Desertification (UNCCD), which tackles the productivity of dryland as well as desertified and desertifying areas. The CDM specifies the offsets by Annex II nations (broadly, developing and newly industrialized countries) that are available to Annex I nations (broadly, industrialized countries). Various governments have implemented national carbon-trading schemes to meet their emission reduction commitments under the Kyoto protocol. In Europe this has led to the EU Greenhouse Gas Emission Trading Scheme, and subsidiary schemes such as the UK Emissions Trading Scheme. More recently, Australia has announced plans for a Carbon Pollution Reduction Scheme (Anon, 2007). The inclusion of biochar into some national trading schemes is anticipated. In terms of potential risk to human health posed by biochar, attention has focused on two classes of toxic compounds associated with combustion processes, namely PAHs and dioxins. Dioxins predominantly form at temperatures in excess of 1000 C, but there are no published studies to confirm their absence in biochar (Garcia-Perez, 2008). The proliferation of PAH in secondary pyrolytic reactions above 700 C is well established (Ledesma et al., 2002), but may form at low concentration in the operating temperature range of pyrolysis reactors (Garcia-Perez, 2008), and PAH profiles have


61

even been considered to provide a reliable thermal history for environmental samples (Brown et al., 2006). Unpublished analyses of several biochar samples found PAH content similar to those of rural UK soils (Manning, pers. comm, 2009); a single published study examined the full PAH profile (40 individual PAH compounds) in a number of synthetic char samples, manufactured at relatively high heating rates (Brown et al., 2006). In the latter study, total PAH concentration was 3–16 mg g 1 depending on peak temperature, compared to 28 mg g 1 in char from a prescribed burn in pine forest. However, empirical relationships to relate these results to process parameters—and which could then be used to predict their formation—have not been established. It has been noted by Ahmed et al. (1989) that while biochar comprises entire systems of PAH molecules, the existing evidence suggests no leachable PAH are present. Thus although the timescale over which PAH from biochar are altered in the soil, and most importantly, the rate at which they become bioavailable, it appears that degradation predominates. In one study in temperate soils, a mean residence time for PAH has been estimated at less than ten years (Paterson et al., 2003), which is low relative to most estimates for the stability of biochar.

2. Characterization of Biochar The key challenge in quantification is to distinguish biochar from soil organic matter and from other forms of black carbon present in bulk soil samples. A variable and unpredictable level of interference from the mineral matrix in soil presents a major challenge in the application of many potential techniques, and many of the techniques depend on spectroscopic characteristics rather than physical separation or isolation. Some of the techniques that most effectively distinguish different types of biochar can also be used to characterize individual biochar fragments (or collections of fragments) recovered from soil. Examination of pure samples removes the matrix effects, but where function of a recalcitrant substrate depends on its surface characteristics or those of accessible pores, separation of active and inactive components presents a significant challenge.

2.1. Quantification of char in soil The categorization of soil organic carbon in general presents a major challenge. Quantifying black carbon is particularly difficult on account of its chemical complexity and diversity, yet inherently unreactive nature. Due to its recalcitrance, biochar cannot meaningfully be extracted from soil using chemicals, though potential biomarkers may be. Results from studies using

62

S. P. Sohi et al.

the physical location of char within a soil matrix (Brodowski et al., 2006; Glaser et al., 2000; Liang et al., 2009a; Murage et al., 2007; Shindo et al., 2004) suggest that efficacy of physical separations using density or means other than hand picking (which is limited to very small samples) are sensitive to site factors. Until recently, the most practical approaches have sought to remove non-black carbon fractions (i.e., soil organic matter and mineral carbonates) with subsequent evaluation of the residue. However, for quantifying biochar specifically this type of quantification may be affected by the presence of the more recalcitrant black carbon forms, as well as by the presence of highly resistant organic compounds—such as those stabilized on clay—not incompletely removed, and which in some cases are estimated separately. Different techniques discriminate components of increasing minimum stability: partially charred biomass, char, charcoal, soot, and graphitic black carbon. Leading methods in this category include removal of non-black carbon by oxidation—chemically (e.g., sodium chlorite, potassium dichromate), using ultraviolet radiation, or by a thermal approach (De la Rosa et al., 2008). Hydrogen pyrolysis (HyPy) is alternative approach to removal of non-black carbon (Ascough et al., 2009), while evolved gas analysis seeks to infer source from the character of the diverse gaseous products of thermal decomposition. A combined chemothermal oxidation method, with a temperature threshold of 375 C (Gustafsson et al., 2001), forms the basis of a standard procedure for the determination of fixed carbon, which comprised the most stable fraction of black carbon, and has the more stable component of biochar. Virtual separations have traditionally relied on spectroscopic techniques in combination with pretreatment (or other allowance) for mineral interference, for example using hydrofluoric acid (Simpson and Hatcher, 2004)—pyrolysis gas chromatography mass spectroscopy (PyCG/MS), and solid-state 13C nuclear magnetic resonance (NMR) spectroscopy with cross-polarization, Bloch decay, and combined with magic angle spinning (MAS) (Skjemstad et al., 1999; Smernik et al., 2002)—or chemically extracted and purified biomarkers, particularly benzene polycarboxylic acids (BPCA) (Brodowski et al., 2005), and levoglucosan (Kuo et al., 2008). A further approach considered in this category is matrix-assisted laser desorption ionization (MALDI–TOF) (Bourke et al., 2007). The applicability of these methods depends on the purpose of the analysis and the specific nature of the target fraction, so although all have been evaluated for a set of 12 environmental and black carbon samples in a ring trial (Hammes et al., 2007, 2008), there remains relatively little consensus as to a universal standard. Since most depend on progressive exclusion based on increasing recalcitrance, the methods cannot readily exclude both recalcitrant soil organic matter and graphitic and soot fractions, and directly reveal the content of relatively less condensed (stable) charcoal or char fractions. Nonetheless, UV or chemical oxidation with elemental and 13C


63

NMR analysis of residues, thermal analysis (De la Rosa et al., 2008; Hammes et al., 2007), and HyPy (Ascough et al., 2009) were identified as the most promising techniques. A relatively new development in the quantification of black carbon has been the application of correlative techniques based on mid-infrared (MIR) spectroscopy. Initially evaluated for the estimation of organic carbon content of bulk soil samples, among other key properties, algorithms have been developed for relating the MIR response spectrum to black carbon, using a calibration set assessed using a UV-oxidation method ( Janik et al., 2007). The method has been applied to evaluate charcoal content in regional evaluation using archived soils (Lehmann et al., 2008) and holds potential for similar assessments in the global context provided the algorithms can be shown to hold for soils of contrasting organic carbon contents.

2.2. Chemical composition Some of the quantification techniques may also be relevant to the characterization and comparison of various samples of pure biochar, that is, ex situ. The purpose here is to assess variation in properties of black carbon between samples, and to document the process of aging in contrasting soils and environments. Elemental ratios of O:C, O:H, and C:H have been found to provide a reliable measure of both the extent of pyrolysis and the level of oxidative alteration of biochar in the soil, and are relatively straightforward to determine. Diffuse reflectance infrared Fourier transform spectroscopy (FTIR), X–ray photoelectron spectroscopy (XPS), energy dispersive X–ray spectroscopy (EDX), near-edge X-ray absorption fine structure (NEXAFS) spectroscopy (Baldock and Smernik, 2002; Fernandes and Brooks, 2003; Lehmann et al., 2006) have been used to examine surface chemistry of biochar in more detail. These analyses provide qualitative information that may enable the mechanisms behind aging and functionalization of biochar to be elucidated. Biogeochemical characterization may also help understand the agronomic function of biochar products at the soil process level and facilitate production of biochar that offers specified benefits. To develop the predictive capacity for the longevity and interaction of biochar in soil, determining its value as a carbon sink and soil conditioner, the nature of its interventions in typical soil processes must be established. From a practical point of view it is important that the devised methods enable biochar characteristics to be determined sufficiently rapidly and inexpensively as to permit widespread application and use. A preliminary set of seven key properties for the evaluation of biochar have been defined: pH, content of volatile compounds, content of ash, water-holding capacity, bulk density, pore volume, and specific surface area (Okimori, et al., 2003). Feedstock is a key factor governing the status of such physicochemical properties. Pyrolysis temperature is the most significant

S. P. Sohi et al.

64

process parameter, carbon content of biochar inversely related to biochar yield, increasing from 56% to 93% between 300 and 800 C in one study, while yield of biochar decreased from 67% to 26% (Okimori, et al., 2003). Beyond a certain temperature threshold, biochar yield may continue to decrease with no further increase in the concentration of carbon within it. However, since ash is broadly conserved, the ash content of biochar increases with temperature. In the study described earlier, the ash content of remaining biochar rose from 0.67% to 1.26% as peak formation temperature was increased from 300 to 800 C

2.3. Physical characterization Scanning electron microscopy (SEM) is often used to describe the physical structure of biochar. The macroporous structure (pores of approximately 1 mm diameter) of biochar produced from cellulosic plant material inherits the architecture of the feedstock, and is potentially important to waterholding and adsorption capacity of soil (Day et al., 2005; Ogawa et al., 2006; Yu et al., 2006). Surface area measured by gas adsorption, however, is influenced by micropores (nm scale) that are not relevant to plant roots, microbes, or to the mobile soil solution. Process temperature is the main factor governing surface area, increasing in one study from 120 m2 g 1 at 400 C to 460 m2 g 1 at 900 C (Day et al., 2005). The importance of temperature leads to the suggestion that biochar created at low temperature may be suitable for controlling release of fertilizer nutrients (Day et al., 2005), while high temperatures would lead to a material analogous to activated carbon (Ogawa et al., 2006). It is also noted that the surfaces of low temperature biochar can be hydrophobic, and this may limit the capacity to store water in soil. The form and size of the feedstock and pyrolysis product may affect the quality and potential uses of biochar. Initially, the ratio of exposed to total surface area of biochar will be affected by its particle size. However, although low temperature biochar is stronger than high temperature products, it is brittle and prone to abrade into fine fractions once incorporated into the mineral soil. It may be proposed that the surface area over the long term, that is, of weathered biochar, is not greatly affected by this parameter.

3. Biochar Application in Agriculture 3.1. Historic usage The fertile terra preta of the central Amazon are anthropogenic dark earths, in a landscape characterized by soils of generally low fertility. Archaeological evidence and carbon dating indicates that these soils were created over a


65

period of millennia from about 9000 ybp, through the activity of dispersed but relatively large and settled communities eliminated, presumably, by western disease, approximately 1000 ybp. These soils subsequently recolonized by natural forest were uncovered relatively recently, and are locally popular for the production of cash crops such as papaya and mango, which anecdotal evidence suggests grow three times faster on this land compared to the surrounding soil. The terra preta are distributed patchily in areas of historic habitation, averaging 20 ha in area, but with individual sites of up to 350 ha reported so far (Smith et al., 2009). The fertility of terra preta has been attributed to a high char content (Glaser et al., 2001), which largely determines their dark color. The source of char is considered to have been incompletely combusted biomass from both domestic fires and burning in-field, but the extent of the deposits suggests that the applications were increasingly deliberate, presumably as a management strategy to address low soil fertility. Residually, terra preta display elevated soil organic matter content, and enhanced nitrogen, phosphorus, potassium, and calcium status. Similar soils have been documented elsewhere within the region, namely Ecuador and Peru, in West Africa (Benin, Liberia), and the savanna of South Africa (Lehmann et al., 2003). Use of charring in traditional soil management in the past (Young, 1804) or at the current time (Lehmann and Joseph, 2009) has also been reported in other countries. It seems probable that these practices have been ubiquitous globally through history and that further examples will emerge in the future. Currently, Japan has the largest commercial production of charcoal for soil application, with approximately 15,000 t traded annually (Okimori et al., 2003). Growing recognition for the potential of the terra preta as a model for modern management of soil fertility using biproducts of bioenergy is now well established, and has spurred a slew of research effort, published outputs of which are reviewed later. A larger number of current experiments are yielding data, which are not yet in press.

3.2. Impact on crop productivity Glaser et al., 2001 reviewed a number of early studies conducted during the 1980s and 1990s. These tended to show marked impacts of low charcoal additions (0.5 t ha 1) on various crop species, but inhibition at higher rates. Since then, data have been published only for approximate field experiments (Asai et al., 2009; Blackwell et al., 2007; Kimetu et al., 2008; Rondon et al., 2007; Steiner et al., 2007, 2008a; Yamato et al., 2006), where the short-term response of staple grain crops to biochar application, in terms of plant biomass or crop yield, has been assessed. Universally, these studies have used charcoal, produced commercially, traditionally, or under conditions designed to simulate wildfire and in all cases from wood, one from short-rotation forestry crops (Blackwell et al.,

66

S. P. Sohi et al.

2007). An additional two studies have examined the agronomic value of biochar produced under zero-oxygen conditions, although these also used contrasting feedstock—poultry litter (Chan et al., 2008) and ‘‘green waste’’ (Chan et al., 2007)—and indicator plants (radish) in pot experiments making comparisons against the function of the charcoal difficult. Seven of the eight studies in total have tested moderate rates of addition, broadly 5–15 t ha 1 (or up to 0.5% by soil mass). Four included rate of addition as a test variable, either to high (60–300 t ha 1, or 2–10% by mass), or even higher rates (in pot experiments). The charcoal in these experiments was produced either at a documented lower temperature (approximately 350–450 C) or at a likely similar range of temperature in a traditional kiln, and was generally alkaline. Biochar in these experiments was added to acidic, tropical soils, though collectively encompassed a textural range. The mean duration for the experiments was less than eight months with measurements—in addition to yield—related in some way to nutrient dynamics. The precise regime of nutrient management was the most common second variable included in these studies. Positive yield effects from biochar addition were reported by Kimetu et al. (2008), who were able to establish that the impact were in part due to non-nutrient improvement to soil function. Improved fertilizer use efficiency was pin-pointed as an explanation for biochar maintaining crop yields after forest clearance in Amazonia, in essentially a recreation of terra preta (Steiner et al., 2008a). Biochar-amended plots receiving NPK sustained higher crop yield compared to control plots where yield declined rapidly. Results from semi-arid soils in Australia have shown positive response to biochar in combination with fertilizer in pot trials (Chan et al., 2007), and in Indonesia maize and peanut yields were enhanced where bark charcoal was applied in combination with N fertilizer in the field (Yamato et al., 2006). The view that nutrient management and pre-existing soil nutrient status determine crop response to biochar was supported by a study in rice (Asai et al., 2009), where statistically higher first-season yield was observed only when biochar (at a low rate) was applied together with fertilizer N and in a low-yielding crop variety; yield was lower than the control in an equivalent treatment using a high-yielding (and thus N-demanding) variety. However, some studies show no significant yield response, for example at low rates of application in an Australian study in wheat (Blackwell et al., 2007). A pot study of maize showed higher biological nitrogen fixation with biochar addition due to nutrient effects (Rondon et al., 2007); higher yield and N uptake reported in pot trials using radish (Chan et al., 2007, 2008). A key consideration highlighted in several studies is the potential for biochar to immobilize previously plant available N. This could be from the mineralization of labile, high C–to–N fractions of biochar drawing N into microbial biomass, sorption of ammonium, or sequestration of soil solution into fine pores.


67

The mechanisms of crop response are covered later and in the absence of long-term data (other than terra preta), development of predictive certainty for the longevity and durability of yield and other effects, particularly in relation to specific crop and soil types, is critical to guide selection of feedstock, production method, and application rate. Predictability and certainty are required to assign a financial value to the agronomic value of biochar and to open the possibility for large-scale deployment.

3.3. Impact on soil performance and resource implications Both the mineral and the organic components of soil influence waterholding capacity. Although higher levels of soil organic matter increase water-holding capacity and can be deliberately managed, changes will be temporary unless a regime is maintained. Glaser et al. (2002) reported that water retention in terra preta was 18% higher than in adjacent soils where charcoal was low or absent, and likely a combined consequence of higher biochar content and higher levels of organic matter that appear to be associated with charcoal in these soils. As biochar is broadly stable in soil, it has potential to provide a direct and long-term modification to soil waterholding capacity through its often macroporous nature (predominantly mmsized pores), reflecting cellular structures in the feedstock from which it is typically produced. The direct impact of particle size distribution in biochar added to the soil may have a direct impact on soil texture at the macroscale, but this effect must be short-lived as physically biochar appears to divide rapidly in soil to particles of silt size or less (Brodowski et al., 2007), presumably by abrasion and the effects of shrink–swell or freeze-thaw, etc. This suggests that in the longer term the effect of biochar on available moisture will be positive in sandy soils ordinarily dominated by much larger pores than present in biochar, rather neutral in medium-textured soils, and potentially detrimental to moisture retention in clay soils—though since preferential flow is important in cracking clays, the impact of biochar on the nature of soil cracking might be important. The usual measure for pore size distribution in soil is the moisture release curve, which shows how quickly moisture is drawn from a soil under increasing tension. Although this method is well suited to discriminating soils of contrasting texture, it is not well able to discriminate the effects of subtle differences in soil management at a particular location: levels of replication have not been sufficient to attach statistical significance to differences in mean characteristics of amended and non-amended soils. In recent work, moisture release curves were determined for a loamy sand field soil to which up to 88 t ha 1 biochar had been applied (Gaskin et al., 2007). For soils where biochar has been added at rates up to 22 t ha 1 there was no difference compared to nonamended soil, but at the highest rate there was a significant effect at water potentials in the range 0.01–0.20 MPa. At the highest potential the

68

S. P. Sohi et al.

volumetric water content was double that of soil without biochar added. Soils of lower bulk density are generally associated with higher soil organic matter, and bulk density provides a crude indicator for how organic matter modifies soil structure and pore-size distribution. Many studies where the effect of biochar on crop yield has been assessed have cited moisture retention as a key factor in the results. Soil temperature, soil cover, evaporation, and evapotranspiration affect soil water availability, so comparison of volumetric water content between biochar-amended and control soils in field experiments may be confounded by indirect effects, that is, on plant growth and soil thermal properties. In addition to the chemical stabilization of nutrients, modification of the physical structure of the bulk soil may result in biochar not simply increasing the capacity of soil to retain water, but also nutrients in soil solution. There are several reasons why biochar might be expected to decrease the potential for nutrient leaching in soils, and thus enhance nutrient cycling and also protect against leaching loss. In field studies where positive yield response to biochar application has been observed, enhanced nutrient dynamics has been frequently cited as an explanation. However, the underlying processes have not been demonstrated directly, and no empirical or mechanistic description has been established. In general, both mineral and organic fractions of soil contribute to cation exchange capacity (CEC) in soil, although not in a summative manner. The CEC largely controls the flush of positively charged ammonium ions after fertilizer or manure application, and rapid mineralization of soil organic matter under favorable environmental conditions. These relatively loose associations do not automatically preclude acquisition by the plant, but have an important effect on mitigating losses of nitrate by leaching, and consequently on agronomy and avoided eutrophication of aquatic and marine environments. Only certain inorganic components of the soil contribute significant CEC due to mineralogy, abundance, and particle size and surface area, with certain types of clay being most important. On a mass basis the exchange capacity of soil organic matter is up to 50 times greater than for any mineral, but is a small proportion of soil mass in most agricultural situations, particularly under tropical conditions. In heavy textured soils in climates favoring organic matter about one-third of total CEC may derive from organic matter (Stevenson, 1982). Since the mineralization of organic matter is also a major source of ammonium in soil, increasing organic matter inputs to increase soil organic matter can potentially increase rather than decrease leaching losses. Available evidence suggests that the specific CEC of biochar is consistently higher than that of whole soil, clay minerals, or soil organic matter and analogy can be drawn to the very high CEC associated with activated carbon that defines its function as a sorption medium for decolorizing and purifying solutions. Since secondary thermal treatment of charcoal is one method for activating charcoal substrate, it is expected that of the


69

process parameters that appear to affect the CEC of biochar, temperature should be the most critical (Gaskin et al., 2007). This function of biochar arises from specific surface area, which increases with temperature through the formation of micropores (Bird et al., 2008), and the abundance of carboxyl groups on those surfaces. The apparent proliferation of carboxyl groups on char surfaces over time, within or outside the soil environment, suggests either partial oxidation of accessible surfaces by biotic and abiotic processes (Cheng et al., 2006) or, alternatively, chemisorption. To develop understanding of this process and the rate at which it proceeds, it may be necessary to perfect methods for recovery of larger samples of intact and increasingly aged biochar from field soils. Although information on the CEC of fresh pyrolysis products relates to limited feedstock and production conditions, and it appears that CEC can substantially develop prior to biochar application to soil. The inherent stability of biochar creates a distinction between the CEC that it provides, and CEC associated with soil organic matter. Importantly, there is no obvious constraint on the level of benefit that that could be attained with repeated addition, by incremental enhancement of CEC. Provided that biochar is biologically stable, the benefits of higher CEC could be achieved but without causing seasonal flushes of nitrate. It is possible that if biochar were proven to significantly impact retention of nutrients and benefit water dynamics at application rates feasible for strategic deployment in vulnerable catchments, it could assist in the mitigation of diffuse pollution from agriculture. It could also be possible to utilize its sorptive capacity to remove contamination in water treatment processes. Studies that demonstrate the capacity for biochar to remove nitrate (Mizuta et al., 2004) and phosphate (Beaton et al., 1960) have been cited in this context. However, although biochar may loosely hold nutrient elements in a plant-available form, it also has an affinity for organic compounds and may sorb toxic by-products from the wastewater treatment process (Yu et al., 2006). Using post-treatment biochar products on land would be subject to regulation, and the economic and overall carbon and environmental gain achieved from centralized rather than distributed production has not been assessed. A centralized system using activated carbon for the removal of chlorine and organic chemicals such as phenols, polychlorinated biphenyls, trihalomethanes, pesticides and halogenated hydrocarbons, heavy metals, and organic contaminants (Boateng, 2007) has been discussed. However, it is not clear whether the use of biochar derived from agricultural crop wastes would provide qualitative of quantitative differences in efficacy through contrasting surface area and sorptive capacity (Zanzi et al., 2002). Indirect effects of biochar on soil chemistry appear to arise from modification of soil pH, although no dynamic studies of biochar in agronomic trials have controlled for this effect. Intensively studied terra preta sites are higher in pH than surrounding soils, as well as exhibiting higher phosphorus

70

S. P. Sohi et al.

status. Biochar also contains ash which may be in a soluble or a more accessible form in biochar than in the unpyrolyzed feedstock. The indirect effect of biochar on soil phosphorus availability, plus the accessibility of mineral ash (containing phosphorus, potassium and other potentially important trace elements) in the biochar matrix may be important in explaining some short-term impacts of biochar on crop growth, especially since phosphorus availability cannot, unlike ammonium, be improved simply by increasing soil organic matter status (Lehmann, 2007; Steiner et al., 2007). However, there has been no systematic work published to show the relative availability of mineral nutrients in biochar under different process parameters, and appropriate extraction protocols are not yet defined. Much speculation concerns the effects of biochar on microbial activity in soil, which in the context of terra preta has been reviewed in detail by Steiner et al. (2003). Assuming that plant inputs and hence microbial substrate are not changed by biochar, enhanced microbial activity will diminish soil organic matter. However, this is contrary to the observation in terra preta that nonblack carbon is generally higher than surrounding soil (Liang et al., 2006) and that the rate of stabilization of substrates seems to be increased (Liang et al., 2009b). It is possible that the contrasting balance in microbial activity between different functional groups impacts crop nutrition, specifically enhancement of mycorrhizal fungi (Ishii and Kadoya, 1994), with soil organic matter maintained through a positive feedback from increased net primary productivity of plants and hence carbon input to the soil. There is relatively extensive literature documenting stimulation of indigenous arbuscular mycorrhizal fungi by biochar, and this has been associated with enhanced plant growth (e.g., Nishio, 1996; Rondon et al., 2007). This literature has been reviewed in some detail by Warnock et al. (2007) who proposed four mechanistic explanations of which combined effects on nutrient availability, water storage, and CEC were considered most probable. Assessing the impact of fresh biochar addition to soils that previously contained negligible quantities is a different proposition to evaluation of terra preta soils where biochar is abundant but heavily aged and appears to contain a distinct microbial community structure (Kim et al., 2007). Microbial communities may respond in the short term to labile components of biochar when added to the soil, especially at higher rates, and traces of pyrolysis condensates seem to promote microbial activity (Steiner et al., 2008b).

3.4. Additional impacts on greenhouse gas balance There may be additional and potentially important affects of biochar addition on the emission of other greenhouse gases from soil and on indirect emissions. Key indirect emissions savings would include those associated with manufacture of fertilizer not required to produce equivalent crop yield where positive effects on crop nutrient use efficiency have been achieved.


71

This extends to avoided emissions of N2O during the manufacture of the fertilizer and from the use of the fertilizer via the soil. It is estimated that currently the production and use of 1 t of fertilizer nitrogen results in a carbon-equivalent emission of 1.9 CO2 by these pathways (Mortimer et al., 2003). More speculatively, emission of carbon from above- and belowground stocks may be avoided if agricultural productivity is enhanced to a level where reduced pressure on natural ecosystems avoided conversion of forest or savannah land to agriculture (Searchinger et al., 2008). Reduced irrigation costs from improved water-holding capacity could be important in certain cropping systems, and potentially reduced energy requirement in tillage in soil physical properties are significantly altered. Surprisingly little information has been collected on the impact of biochar on such parameters, though in a tropical environment topsoil bulk density was found to be approx. 10% lower at old kiln sites due to the presence of aged biochar than in nearby soils (Oguntunde et al., 2008). Observations of the terra preta suggest that biochar in soil can lead to a net stabilization of other organic matter (Liang et al., 2009b). If this were proven in soils of modern agroecosystems, the overall net carbon gain from biochar-based soil management strategies would be considerably enhanced. This is a particularly important prospect, since it would provide a means to benefit from higher soil organic matter without depending on the capacity of clay surfaces which is finite and fixed for a particular soil (Verheijen et al., 2005). In addition to representing a carbon store of its own, biochar would enhance the intrinsic soil organic carbon storage capacity of soil itself, by affecting the turnover of indigenous carbon. However, apparently contradictory data have been published (Wardle et al., 2008) which suggested accelerated decomposition of organic matter by charcoal. This appeared to be a short-term effect, possibly resulting from the perturbation of nutrient and pH status of essentially plant litter (Lehmann and Sohi, 2008). The mechanisms and predictive description for these effects remain to be determined and defined. Although no peer-reviewed studies document suppression of nitrous oxide or methane emission in the field from biochar application in the field, data have been presented in conference proceedings suggesting drastic reduction of these fluxes (Renner, 2007), and some limited laboratory evidence is published (Yanai et al., 2007). Nitrous oxide has a GWP of 310 (IPCC, 1996) and is emitted mainly by heterotrophic denitrifying bacteria, which under anaerobic conditions reduce nitrate rather than oxygen (NO3 to N2O via nitrite and nitric oxide). More continuous, low-rate production of N2O may occur at aerobic as well as anaerobic positions in the soil, from the activity of the chemotrophic bacteria that convert ammonium from mineralization processes to soluble nitrate (Bateman and Baggs, 2005). Higher soil organic matter is associated with greater nitrification rates, but the greatest immediate impact on soil nitrate

S. P. Sohi et al.

72

concentrations and hence denitrification rate is the application of inorganic nitrogen fertilizers and also manure or slurry. Elimination of reduction of trace gas fluxes from soil would significantly impact carbon-equivalent impacts of agriculture (Gaunt and Lehmann, 2008). Proposed mechanisms for the suppression of N2O from biochar revolve around modification of soil water dynamics, that is, drawing soil solution (and dissolved nitrate) into pores inaccessible by microbes and maintaining aerobic conditions inside inhabited pore space. Increase of soil pH which under anaerobic conditions also favors completion of nitrate reduction to N2 (from N2O) or the adsorption of ammonium that prevents nitrification and denitrification. The effect of water addition cannot be evaluated completely under constant conditions, but increasing water filled pore space from partial to nearcomplete saturation has been seen to reverse 90% suppression in small laboratory incubations with biowaste charcoal, at a high rate equivalent to 180 t ha 1(Yanai et al., 2007). Application of ash separately from charcoal in the same experiment did not equally suppress emission, suggesting that pH was not a factor in the result, though simultaneous monitoring of N2 would be required to confirm N2O reduction. In conference proceedings, research has shown nitrate to accumulate where N2O has been suppressed.

4. Research Priorities and Future Challenges 4.1. Mechanistic understanding The fundamental mechanisms by which biochar affects the function of soil and the wider agroecosystem are poorly defined and consequently current capacity to predict the effects on biochar are inadequate. In short-term experiments of months to a few years biochar may be seen to generally enhance plant growth and soil nutrient status, and decrease N2O emissions. However, the explanation for these benefits is not fully described, and neither the quantitative variability in response nor the durability of the effects is specified. Consequently soil–biochar dynamics need to be strategically investigated to deliver understanding in several areas. Functional interaction with soil microbial communities. Biochar may modify symbiotic relationships between plants and microbes in close proximity to the root (the rhizosphere). At the moment the net effect of physical protection provided to microbial colonies, and evidence for adequate access of those same colonies to labile and soluble carbon substrates has not been established. The component of biochar stability provided by association of individual biochar particles and fine mineral particles has not been established, and the role of microbial and rhizosphere secretions in promoting them will be important. Fundamentally, the apparent conflict between high stability, soil organic matter accumulation, and apparent enhancement of


73

soil microbial activity needs to be resolved. Useful methods will separate indirect effects of increased water-holding capacity or altered water release characteristics, pH effects, and allow for their potentially transient nature. Surface interactions. It seems that the CEC of biochar surfaces develops over time, but the role of feedstock and production parameters in determining initial and ultimate activity of surfaces needs to be established, as well as the trajectory of development. Once the relative importance of biotic and abiotic processes is known as is the net effect of simultaneous change in the relative abundance of external and internal surface area following physical disintegration with soil movement, the effect of climate might become predictable. The nature of interactions between biochar and nutrient anions, most importantly phosphate, needs to be established and the extent to which nutrient effects are derived from finite supply of mineral ash within the biochar matrix, or supplied from the wider soil, must be determined. Nutrient use efficiency. Understanding the link between biochar function and its interactions with nutrients and crop roots may enable fertilizer use efficiency to be improved, with concomitant benefit to diffuse pollution in watercourses and wetlands. Soil physical effects. Quantitative description for the function of biochar with respect to water infiltration, water retention, macroaggregation, and soil stability would enable prediction of beneficial properties, and selection of biochar with particular properties for use in specific environmental and agricultural contexts. Methods for spreading and incorporation of biochar in soil are required to mitigate lateral movement, and the potential for surface flow and loss at depth need to be specified. Fate of biochar. The stability of biochar carbon is intrinsic to fulfilling its role as a significant CO2 sink, but to perform an agronomic role it must also remain within the soil to which it is applied. The environmental role or impact of biochar once it has moved through a soil profile, or into watercourses, is yet to be assessed. Information on the extent to which physical breakdown of biochar changes the balance in its properties, particularly with respect to soil water dynamics, exchange capacity and soil micro- and macroaggregation is lacking. Methods are urgently required to assess the long-term biological stability of specific biochar samples, possibly extrapolating from the dynamics of atypically high initial rates of loss in soil. Impacts on soil nitrous oxide and methane emission. Published data for the effect on trace gas emission are extremely limited, but this could have an important effect in determining the net greenhouse gas balance, especially over longer timeframes if the suppression is continuous. It is unlikely that these effects will be tradable in carbon accounting for biochar projects without a predictive model, and at the moment a clear underlying explanation for the effect is lacking.

74

S. P. Sohi et al.

Impact on physiological plant response. Given that biochar seems to have some complex effects on plant growth beyond simple supply of nutrients, this is an area of fundamental research that needs to be targeted for future work.

4.2. Properties, qualities, and environmental risks The role of biochar in mitigation of climate change hinges on the stability of the bulk of the carbon that it contains. Transferable methods for rapid assessment of relative long-term levels of stability for specific biochar samples in soil are required, rather than relying on direct linear extrapolations from short-term losses. A critical, experimental analysis of the risks arising from deployment of biochar is required, particularly in the context of pyrolysis bioenergy, since most charcoal used in traditional ways is also produced using traditional kilns. This assessment would consider exposure to solid, liquid, and gaseous products of pyrolysis at all stages of production, in distribution, and in the physical application of biochar to soil, as well as the impacts on the health of the soil and plants grown on it. This assessment is critical as biochar is irretrievable once added to soil, and given its apparent permanency, the scale and speed at which it would have to be added in order to address climate change is large and rapid. In addition, the position regarding responsibility and liability around large-scale deployment of biochar will become much clearer through such assessment. Moreover, deriving minimum standards will remove a key impediment to those seeking to invest biochar in soil that is also used for producing food. Although the economic case may be otherwise favorable, producing biochar from sewage sludge and municipal waste streams requires different assessments of risk, as considering only the contaminants that might form during pyrolysis is insufficient. In these cases it is important to consider whether contaminants present in the pyrolysis feedstock are eliminated, or modified to become more or less available in the biochar product. The environmental impact and function of biochar in subsoil and in watercourses needs to be assessed. Methodologies for validation and audit of biochar deployment will emerge, and it will be necessary to be able to retrospectively determine the source of biochar applied at a particular location. Databases should play an important role in enabling optimal biochar products to be selected for use at a particular location.

4.3. Modeling capacity for the soil–biochar system Two types of carbon modeling are required: static spreadsheet models to compare alternative scenarios for their relative carbon-equivalent gain, and mechanistic soil simulation models that capture information


75

from short-term experiments to predict longer-term impacts on soil function. In addition, socioeconomic models that incorporate a spatial dimension are required to assess the workability of particular scenarios. So far only generic, theoretical analyses have been published. Full assessment spreadsheet models based on improved experimental evidence are required to conduct scenario comparisons for strategies based around specific feedstock streams and pyrolysis technologies, taking the spatial dimension of feedstock supply and biochar use into account. Economic models must define carefully selected conceptual and geographic boundaries, and account for the entire supply chain. Socioeconomic constraints relevant to the application of biochar must be recognized. Simulation modeling for the carbon and nitrogen cycles in soil with and without biochar is essential to understanding the functional behavior of biochar, and the impact on soil-based greenhouse gas emissions. Modeling of soil carbon currently relies on conceptual pools and essentially ignores black carbon from a mechanistic perspective. Progress in this area is dependent on improved quantification methods for biochar in soil. Development correlative techniques based on spectral analysis may lead to an approach that is both rapid and low cost, relying on the MIR wavelength ( Janik et al., 2007). It is likely that sites used for parameterization of soil models contain atypically low levels of char. Adjustments have previously been required to simulate plots with documented history of burning (Coleman et al., 1997), and reallocating carbon between soil pools according to direct measurement can benefit model prediction (Skjemstad et al., 2004). In the absence of long-term experiments using biochar, modeling will be central to attempts to predict long-term fate. Globally it appears that char in soil may be more abundant than previously assumed (Preston and Schmidt, 2006; Schmidt et al., 1999), and that greater amounts of stable carbon will lessen the rate at which soil carbon is lost in feedback effects from climate change (Lehmann et al., 2008).

4.4. Barriers and limitations to biochar systems A market for carbon credits in which land managers can engage in deploying biochar does not exist. In general, there also remains a lack of knowledge and awareness of bioenergy and carbon markets, how to access these markets, and particularly a way to accurately evaluate costs and benefits associated with the use of biochar in soil. Crucially, as well as being multiple, the relevant markets are interdependent, extending to the supply of feedstock for pyrolysis. However, no framework exists within which the carbon sequestered in biochar can be certified as a tradable commodity, and this extends beyond trading under the CDM to the voluntary carbon markets. To date the methodology required to recognize the stabilization

76

S. P. Sohi et al.

of degradable organic matter as an avoided CO2 emission has not been presented. There are significant organizational and institutional obstacles to the use of biochar in soil. Since biochar could be used on a large scale and cannot be removed from soil once applied, there is a need to carefully assess the potentially negative impacts on occupational health, environmental pollution, water quality, and food safety. This requires a concerted effort to evaluate potential products, and ideally define product standards. Support for the use of biochar in meeting policy objectives will draw upon life-cycle analysis with full greenhouse accounting, backed by a body of experimental data. Where biochar is designated as a regulated waste material, pending defined standards land users in many countries may be subject to a complex and expensive approval process. The lack of mechanistic understanding as to the function of biochar, and its interaction with already complex soil processes, means predicting the return to an investment in biochar between locations in terms of extent, predictability, and durability of benefits does not yet exist. Providing a measure of certainty to the many possible benefits is a key challenge to be addressed by further research.

ACKNOWLEDGMENTS This original review on which this work is based was undertaken with support from CSIRO Land and Water in Australia (Sohi et al., 2009). The contributions of Peter Brownsort (University of Edinburgh) and Keith Goulding (Rothamsted Research) to the preparation of this manuscript is a acknowledged. Rothamsted Research is an institute of the UK Biotechnology and Biological Sciences Research Council in the UK.

REFERENCES Ahmed, A., Pakdel, H., Roy, C., and Kaliaguine, S. (1989). Characterization of the solid residues of vacuum pyrolysis of Papulus tremuloides. J. Anal. Appl. Pyrolysis 14, 281–294. Anon. (2007). ‘‘Australia’s climate change policy: our economy, our environment, our future.’’ The Australian Government, Barton, Australia. Asai, H., Samson, B. K., Stephan, H. M., Songyikhangsuthor, K., Homma, K., Kiyono, Y., Inoue, Y., Shiraiwa, T., and Horie, T. (2009). Biochar amendment techniques for upland rice production in northern Laos. Field Crops Res. 111, 81–84. Ascough, P. L., Bird, M. I., Brock, F., Higham, T. F. G., Meredith, W., Snape, C. E., and Vane, C. H. (2009). Hydropyrolysis as a new tool for radiocarbon pretreatment and the quantification of black carbon. Quat. Geochronol. 4, 140–147. Baldock, J. A., and Smernik, R. J. (2002). Chemical composition and bioavailability of thermally altered Pinus resinosa (Red pine) wood. Org. Geochem. 33, 1093–1109. Barker, T., Bashmakov, I., Bernstein, L., Bogner, J. E., Bosch, P. R., Dave, R., Davidson, O. R., Fisher, B. S., Gupta, S., Halsnæs, K., Heij, G. J., Kahn Ribeiro, S., et al. (2007). Climate change 2007: Mitigation. Contribution of working group III to the


77

fourth assessment. In ‘‘Report of the Intergovernmental Panel on Climate Change’’ (B. Metz, O. R. Davidson, P. R. Bosch, R. Dave, and L. A. Meyer, Eds.), Cambridge University Press, Cambridge. Bateman, E. J., and Baggs, E. M. (2005). Contributions of nitrification and denitrification to N2O emissions from soils at different water-filled pore space. Biol. Fert. Soils 41, 379– 388. Beaton, J. D., Peterson, H. B., and Bauer, N. (1960). Some aspects of phosphate adsorption by charcoal. Soil Sci. Soc. Am. Proc. 24, 340–346. Bird, M. I., Ascough, P. L., Young, I. M., Wood, C. V., and Scott, A. C. (2008). X-ray microtomographic imaging of charcoal. J. Archaeol. Sci. 35, 2698–2706. Blackwell, P., Shea, S., Storer, P., Solaiman, Z., Kerkmans, M., and Stanley, I. (2007). Improving wheat production with deep banded oil mallee charcoal in Western Australia. In ‘‘The First Asia-Pacific Biochar Conference’’ Terrigal, Australia. Boateng, A. A. (2007). Characterization and thermal conversion of charcoal derived from fluidized-bed fast pyrolysis oil production of switchgrass. Ind. Eng. Chem. Res. 46, 8857–8862. Bourke, J., Manley-Harris, M., Fushimi, C., Dowaki, K., Nunoura, T., and Antal, M. J. (2007). Do all carbonized charcoals have the same chemical structure? 2. A model of the chemical structure of carbonized charcoal. Ind. Eng. Chem. Res. 46, 5954–5967. Brodowski, S., Rodionov, A., Haumaier, L., Glaser, B., and Amelung, W. (2005). Revised black carbon assessment using benzene polycarboxylic acids. Org. Geochem. 36, 1299–1310. Brodowski, S., John, B., Flessa, H., and Amelung, W. (2006). Aggregate-occluded black carbon in soil. Eur. J. Soil Sci. 57, 539–546. Brodowski, S., Amelung, W., Haumaier, L., and Zech, W. (2007). Black carbon contribution to stable humus in German arable soils. Geoderma 139, 220–228. Brown, R. A., Kercher, A. K., Nguyen, T. H., Nagle, D. C., and Ball, W. P. (2006). Production and characterization of synthetic wood chars for use as surrogates for natural sorbents. Org. Geochem. 37, 321–333. Chan, K. Y., Van Zwieten, L., Meszaros, I., Downie, A., and Joseph, S. (2007). Agronomic values of greenwaste biochar as a soil amendment. Aus. J. Soil Res. 45, 629–634. Chan, K. Y., Van Zwieten, L., Meszaros, I., Downie, A., and Joseph, S. (2008). Using poultry litter biochars as soil amendments. Aus. J. Soil Res. 46, 437–444. Cheng, C.-H., Lehmann, J., Thies, J. E., Burton, S. D., and Engelhard, M. H. (2006). Oxidation of black carbon by biotic and abiotic processes. Org. Geochem. 37, 1477–1488. Coleman, K., Jenkinson, D. S., Crocker, G. J., Grace, P. R., Klir, J., Korschens, M., Poulton, P. R., and Richter, D. D. (1997). Simulating trends in soil organic carbon in long-term experiments using RothC-26.3. Geoderma 81, 29–44. Crutzen, P. J. (2006). Albedo enhancement by stratospheric sulfur injections: A contribution to resolve a policy dilemma? Climatic Change 77, 211–219. Das, K. C., Garcia-Perez, M., Bibens, B., and Melear, N. (2008). Slow pyrolysis of poultry litter and pine woody biomass: Impact of chars and bio-oils on microbial growth. J. Environ. Sci. Health: Part A 43, 714–724. Day, D., Evans, R. J., Lee, J. W., and Reicosky, D. (2005). Economical CO2, SOx, and NOx capture from fossil–fuel utilization with combined renewable hydrogen production and large-scale carbon sequestration. Energy 30, 2558–2579. De la Rosa, J. M., Knicker, H., Lopez-Capel, E., Manning, D. A. C., Gonzalez-Perez, J. A., and Gonzalez-Vila, F. J. (2008). Direct detection of black carbon in soils by Py-GC/MS, carbon-13 NMR spectroscopy and thermogravimetric techniques. Soil Sci. Soc. Am. J. 72, 258–267. Demirbas, A. (2006). Production and characterization of bio-chars from biomass via pyrolysis. Energy Sources Part A 28, 413–422.

78

S. P. Sohi et al.

Demirbas, A., Pehlivan, E., and Altun, T. (2006). Potential evolution of Turkish agricultural residues as bio-gas, bio-char and bio-oil sources. Int. J. Hydrogen Energy 31, 613–620. FAO. (1985). ‘‘Industrial Charcoal Making’’ FAO Forestry Department, Rome. Fargione, J., Hill, J., Tilman, D., Polasky, S., and Hawthorne, P. (2008). Land clearing and the biofuel carbon debt. Science 319, 1235–1238. Fernandes, M. B., and Brooks, P. (2003). Characterization of carbonaceous combustion residues: II. Nonpolar organic compounds. Chemosphere 53, 447–458. Friedl, A., Padouvas, E., Rotter, H., and Varmuza, K. (2005). Prediction of heating values of biomass fuel from elemental composition. Analyt. Chimica Acta 544, 191–198. Fushimi, C., Araki, K., Yamaguchi, Y., and Tsutsumi, A. (2003). Effect of heating rate on steam gasification of biomass 2. Thermogravimetric-mass spectrometric (TG-MS) analysis of gas evolution. Ind. Eng. Chem. Res. 42, 3929–3936. Garcia-Perez, M. (2008). ‘‘The Formation of Polyaromatic Hydrocarbons and Dioxins During Pyrolysis’’ Washington State University, Pullman. Gaskin, J. W., Speir, A., Morris, L. M., Ogden, L., Harris, K., Lee, D., and Das, K. C. (2007). Potential for pyrolysis char to affect soil moisture and nutrient status of loamy sand soil. In ‘‘Georgia Water Resources Conference’’, University of Georgia. Gaunt, J. L., and Lehmann, J. (2008). Energy balance and emissions associated with biochar sequestration and pyrolysis bioenergy production. Environ. Sci. Technol. 42, 4152–4158. Glaser, B., Balashov, E., Haumaier, L., Guggenberger, G., and Zech, W. (2000). Black carbon in density fractions of anthropogenic soils of the Brazilian Amazon region. Org. Geochem. 31, 669–678. Glaser, B., Haumaier, L., Guggenberger, G., and Zech, W. (2001). The ‘Terra Preta’ phenomenon: A model for sustainable agriculture in the humid tropics. Naturwissenschaften 88, 37–41. Glaser, B., Lehmann, J., and Zech, W. (2002). Ameliorating physical and chemical properties of highly weathered soils in the tropics with charcoal—A review. Biol. Fert. Soils 35, 219–230. Gustafsson, O., Bucheli, T. D., Kukulska, Z., Andersson, M., Largeau, C., Rouzaud, J.-N., Reddy, C. M., and Eglinton, T. I. (2001). Evaluation of a protocol for the quantification of black carbon in sediments. Global Biogeochem. Cycles 15, 881–890. Hammes, K., Schmidt, M. W. I., Smernik, R. J., Currie, L. A., Ball, W. P., Nguyen, T. H., Louchouarn, P., Houel, S., Gustafsson, O., Elmquist, M., Cornelissen, G., Skjemstad, J. O., et al. (2007). Comparison of quantification methods to measure firederived (black/elemental) carbon in soils and sediments using reference materials from soil, water, sediment and the atmosphere. Global Biogeochem. Cycles 21, 1–18. Hammes, K., Smernik, R. J., Skjemstad, J. O., and Schmidt, M. W. I. (2008). Characterisation and evaluation of reference materials for black carbon analysis using elemental composition, colour, BET surface area and C NMR spectroscopy. Appl. Geochem. 23, 2113–2122. Houghton, R. A. (2003). Why are estimates of the terrestrial carbon balance so different? Global Change Biol. 9, 500–509. IEA. (2007). ‘‘IEA Bioenergy Annual Report 2006’’ International Energy Agency, Paris. IPCC. (1996). ‘‘IPCC Second Assessment: Climate Change 1995’’ World Meteorological Organization, Geneva. Ishii, T., and Kadoya, K. (1994). Effects of charcoal as a soil conditioner on citrus growth and vesicular-arbuscular mycorrhizal development. J. Jap. Soc. Horticult. Sci. 63, 529–535. Janik, L. J., Skjemstad, J. O., Shepherd, K. D., and Spouncer, L. R. (2007). The prediction of soil carbon fractions using mid-infrared-partial least square analysis. Aus. J. Soil Res. 45, 73–81. Janzen, H. H. (2006). The soil carbon dilemma: Shall we hoard it or use it? Soil Biol. Biochem. 38, 419–424.


79

Ketterings, Q. M., and Bigham, M. (2000). Soil color as an Indicator of slash-and-burn fire severity and soil fertility in Sumatra, Indonesia. Soil Sci. Soc. Am. J. 64, 1826–1833. Kim, J., Sparovek, G., Longoc, R., Jose De Meloc, W., and Crowley, D. (2007). Bacterial diversity of terra preta and pristine forest soil from the Western Amazon. Soil Biol. Biochem. 39, 684–690. Kimetu, J. M., Lehmann, J., Ngoze, S. O., Mugendi, D. N., Kinyangi, J. M., Riha, S., Verchot, L., Recha, J. W., and Pell, A. N. (2008). Reversibility of soil productivity decline with organic matter of differing quality along a degradation gradient. Ecosystems 11, 726–739. Krull, E. S., Skjemstad, J. O., and Baldock, J. A. (2004). Functions of Soil Organic Matter and the Effect on Soil Properties. Grains Research and Development Corporation, Barton. Kuhlbush, T. A. J. (1998). Black carbon and the carbon cycle. Science 280, 1903–1904. Kuo, L. J., Herbert, B. E., and Louchouarn, P. (2008). Can levoglucosan be used to characterize and quantify char/charcoal black carbon in environmental media? Org. Geochem. 39, 1466–1478. Lal, R., Griffin, M., Apt, J., Lave, L., and Granger Morgan, M. (2004). Managing soil carbon. Science 304, 393. Ledesma, E. B., Marsh, N. D., Sandrowitz, A. K., and Wornat, M. J. (2002). Global kinetic rate parameters for the formation of polycyclic aromatic hydrocarbons from the pyrolysis of catechol, a model compound representative of solid fuel moieties. Energ. Fuels 16, 1331–1336. Lehmann, J. (2007). A handful of carbon. Nature 447, 143–144. Lehmann, C. J., and Joseph, S. (2009). Biochar systems. In ‘‘Biochar for Environmental Management: Science and Technology’’ (C. J. Lehmann and S. Joseph, Eds.), Earthscan, London. Lehmann, C. J., and Sohi, S. P. (2008). Comment on ‘‘Fire-derived charcoal causes loss of forest humus". Science 321, 1295c. Lehmann, J., Kern, D. C., Glaser, B., and Woods, W. I. (Eds.) (2003). In ‘‘Amazonia Dark Earths: Origin, Properties, Management’’, Kluwer Academic Publishers, Dordecht. Lehmann, J., Gaunt, J., and Rondon, M. (2006). Bio-char sequestration in terrestrial ecosystems—a review. Mitigat. Adaptat. Strateg. Global Change 11, 403–427. Lehmann, J., Skjemstad, J., Sohi, S., Carter, J., Barson, M., Falloon, P., Coleman, K., Woodbury, P., and Krull, E. (2008). Australian climate–carbon cycle feedback reduced by soil black carbon. Nat. Geosci. 1, 832–835. Liang, B., Lehmann, J., Solomon, D., Kinyangi, J., Grossman, J., O’Neill, B., Skjemstad, J. O., Thies, J., Luizao, F. J., Petersen, J., and Neves, E. G. (2006). Black carbon increases cation exchange capacity in soils. Soil Sci. Soc. Am. J. 70, 1719–1730. Liang, B., Lehmann, J., Solomon, D., Sohi, S., Thies, J. E., Skjemstad, J. O., Luizaõd, F. J., Engelharde, M. H., Nevesf, E. G., and Wirick, S. (2009a). Stability of biomass-derived black carbon in soils. Geochim. Cosmochim. Acta 72, 6069–6078. Liang, B., Lehmann, L., Sohi, S. P., Thies, J. E., O’Neill, B., Trujilo, L., Gaunt, J., Solomon, D., Grossman, J., Neves, E., and Luizaõ, F. J. (2009b). Black carbon affects the cycling of non-black carbon in soil. Org. Geochem. (in press). Matthews, J. A. (2008). Carbon negative biofuels. Energ. Policy 36, 940–945. McConnell, J. R., Edwards, R., Kok, G. L., Flanner, M. G., Zender, C. S., Saltzman, E. S., Banta, J. R., Pasteris, D. R., Carter, M. M., and Kahl, J. D. W. (2007). 20th-century industrial black carbon emissions altered arctic climate forcing. Science 317, 1381–1384. Mizuta, K., Matsumoto, T., Hatate, Y., Nishihara, K., and Nakanishi, T. (2004). Removal of nitrate nitrogen from drinking water using bamboo powder charcoal. Bioresource Technol. 95, 255–257.

80

S. P. Sohi et al.

Mortimer, N. D., Cormack, P., Elsayed, M. A., and Horne, R. E. (2003). Evaluation of the comparative energy, global warming and socio-economic costs and benefits of biodiesel. Defra, London. Murage, E. W., Voroney, P., and Beyaert, R. P. (2007). Turnover of carbon in the free light fraction with and without charcoal as determined using the 13C natural abundance method. Geoderma 138, 133–143. Nishio, M. (1996). Microbial fertilizers in Japan. National Institute of Agro-Environmental Sciences, Ibaraki, Japan. Ogawa, M., Okimori, Y., and Takahashi, F. (2006). Carbon sequestration by carbonization of biomass and forestation: Three case studies. Mitigat. Adaptat. Strateg. Global Change 11, 429–444. Oguntunde, P. G., Abiodun, B. J., Ajayi, A. E., and van de Giesen, N. (2008). Effects of charcoal production on soil physical properties in Ghana. J. Plant Nutrit. Soil Sci. 171, 591–596. Okimori, Y., Ogawa, M., and Takahashi, F. (2003). Potential of CO2 emission reductions by carbonizing biomass waste from industrial tree plantation in south Sumatra, Indonesia. Mitigat. Adaptat. Strateg. Global Change 8, 261–280. Pacala, S., and Socolow, R. (2004). Stabilization Wedges: Solving the climate problem for the next 50 years with current technologies. Science 305, 968–972. Paterson, E., Towers, W., Bacon, J. R., and Jones, M. (2003). Background Levels of Contaminants in Scottish Soils.Craigiebuckler, Aberdeen. Preston, C. M., and Schmidt, M. W. I. (2006). Black (pyrogenic) carbon: A synthesis of current knowledge and uncertainties with special consideration of boreal regions. Biogeosci. Discuss. 3, 397–420. Ramanathan, V., and Carmichael, G. (2008). Global and regional climate changes due to black carbon. Nat. Geosci. 1, 221–227. Renner, R. (2007). Rethinking biochar. Environ. Sci. Technol. 41, 5932–5933. Rondon, M. A., Lehmann, J., Ramirez, J., and Hurtado, M. (2007). Biological nitrogen fixation by common beans (Phaseolus vulgaris L.) increases with bio-char additions. Biol. Fert. Soils 43, 699–708. Ryu, C., Sharifi, V. N., and Swithenbank, J. (2007). Waste pyrolysis and generation of storable char. Int. J. Energ. Res. 31, 177–191. Schlesinger, W. H. (1995). An overview of the global carbon cycle. In ‘‘Soils and Global Change’’ (R. Lal, J. Kimble, E. Levine, and B. A. Stewart, Eds.), pp. 9–25. Lewis Publishers, Boca Raton, Florida. Schmidt, M. W. I., Skjemstad, J. O., Gehrt, E., and Kogel-Knabner, I. (1999). Charred organic carbon in German chernozemic soils. Eur. J. Soil Sci. 50, 351–365. Searchinger, T., Heimlich, R., Houghton, R. A., Dong, F., Elobeid, A., Fabiosa, J., Tokgoz, S., Hayes, D., and Yu, T.-H.r.v. (2008). Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 213, 1238–1240. Seifritz, W. (1993). Should we store carbon in charcoal? Int. J. Hydrogen Energy 18, 405–407. Shindo, H., Honna, T., Yamamoto, S., and Honma, H. (2004). Contribution of charred plant fragments to soil organic carbon in Japanese volcanic ash soils containing black humic acids. Org. Geochem. 35, 235–241. Shinogi, Y., Yoshida, H., Koizumi, T., Yamaoka, M., and Saito, T. (2002). Basic characteristics of low-temperature carbon products from waste sludge. Adv. Environ. Res. 7, 661–665. Simpson, M. J., and Hatcher, P. G. (2004). Determination of black carbon in natural organic matter by chemical oxidation and solid-state 13C nuclear magnetic resonance spectroscopy. Org. Geochem. 35, 923–935.


81

Skjemstad, J. O., Taylor, J. A., and Smernik, R. J. (1999). Estimation of charcoal (char) in soils. Commun. Soil Sci. Plant Anal. 30, 2283–2298. Skjemstad, J. O., Spouncer, L. R., Cowie, B., and Swift, R. S. (2004). Calibration of the Rothamsted organic carbon turnover model (RothC ver. 26.3), using measurable soil organic carbon pools. Aus. J. Soil Res. 42, 79–88. Smernik, R. J., Baldock, J. A., Oades, J. M., and Whittaker, A. K. (2002). Determination of T1ph relaxation rates in charred and uncharred wood and consequences for NMR quantitation. Solid State Nucl. Magn. Reson. 22, 50–70. Smith, K., Fowler, G., Pullket, S., and Graham, N. (2009). Sewage sludge-based adsorbents: A review of their production, properties and use in water treatment applications. Water Res. 43, 2569–2594. Sohi, S., Lopez-Capel, E., Krull, E., and Bol, R. (2009). ‘‘Biochar, climate change and soil: A review to guide future research’’. Rep. No. 05/09. CSIRO. Steiner, C., Teixeira, W., Lehmann, J., and Zech, W. (2003). Microbial response to charcoal amendments of highly weathered soils and Amazonian Dark Earths in Central Amazonia—Preliminary results. In ‘‘Amazonian Dark Earths: Origin, Properties, Management’’ ( J. Lehmann, D. C. Kern, B. Glaser, and W. I. Woods, Eds.), Kluwer Academic Publishers, Dordrecht. Steiner, C., Teixeira, W. G., Lehmann, J., Nehls, T., MaceˆDo, J. L. V., Blum, W. E. H., and Zech, W. (2007). Long term effects of manure, charcoal and mineral fertilization on crop production and fertility on a highly weathered Central Amazonian upland soil. Plant Soil 291, 275–290. Steiner, C., Glaser, B., Teixeira, W. G., Lehmann, J., Blum, W. E. H., and Zech, W. (2008a). Nitrogen retention and plant uptake on a highly weathered central amazonian ferralsol amended with compost and charcoal. J. Plant Nutr. Soil Sci. 171, 893–899. Steiner, C., Das, K. C., Garcia, M., Forster, B., and Zech, W. (2008b). Charcoal and smoke extract stimulate the soil microbial community in a highly weathered xanthic Ferralsol. Pedobiologia 51, 359–366. Stevenson, F. J. (1982). Humus Chemistry, Genesis, Composition, Reactions.Wiley and Sons, New York. UNFCCC, F. J. (2007). Indicative simplified baseline and monitoring methodologies for selected small-scale CDM project activity categories. In ‘‘Avoidance of Methane Production from Biomass Decay Through Controlled Pyrolysis’’ (C. E. Board, Ed.). United Nations, Bonn. Verheijen, F. G. A., Bellamy, P. H., Kibblewhite, M. G., and Gaunt, J. (2005). Organic carbon ranges in arable soils of England and Wales. Soil Use Manag. 21, 2–9. Vitousek, P. M., Ehrlich, P. R., Ehrlich, A. H., and Matson, P. A. (1986). Human appropriation of the products of photosynthesis. BioScience 36, 368–373. Wardle, D. A., Nilsson, M.-C., and Zackrisson, O. (2008). Fire-derived charcoal causes loss of forest humus. Science 320, 629. Warnock, D. D., Lehmann, J., Kuyper, T. W., and Rillig, M. C. (2007). Mycorrhizal responses to biochar in soil—Concepts and mechanisms. Plant Soil 300, 9–20. Yaman, S. (2004). Pyrolysis of biomass to produce fuels and chemical feedstocks. Energ. Conver. Manag. 45, 651–671. Yamato, M., Okimori, Y., Wibowo, I. F., Anshori, S., and Ogawa, M. (2006). Effects of the application of charred bark of Acacia mangium on the yield of maize, cowpea and peanut, and soil chemical properties in South Sumatra, Indonesia. Soil Sci. Plant Nutr. 52, 489–495. Yanai, Y., Toyota, K., and Okazaki, M. (2007). Effects of charcoal addition on N2O emissions from soil resulting from rewetting air-dried soil in short-term laboratory experiments. Soil Sci. Plant Nutr. 53, 181–188.

82

S. P. Sohi et al.

Young, A. (1804). Paring and burning. The Farmers Calendar, pp. 171–182. Richard Phillips, London. Yu, X.-Y., Ying, G.-G., and Kookana, R. S. (2006). Sorption and desorption behaviors of diuron in soils amended with charcoal. J. Agric. Food Chem. 54, 8545–8550. Zanzi, R., Sjostrom, K., and Bjornborn, E. (2002). Rapid pyrolysis of agricultural residues at high temperature. Biomass Bioenergy 23, 357–366.

C H A P T E R

T H R E E

Towards a Holistic Classification of Diffuse Agricultural Water Pollution from Intensively Managed Grasslands on Heavy Soils S. J. Granger,* R. Bol,* S. Anthony,† P. N. Owens,‡ S. M. White,§ and P. M. Haygarth} Contents 84 86 87 88 90 92 94 94 96 97 98 98 100 101 103 106 110 110

1. Introduction 2. Defining a Conceptual Framework 2.1. Source aspects 2.2. Mobilization aspects 2.3. Delivery aspects 2.4. Pollutant transport through the framework 3. The Assignment of Pollutants in the Conceptual Framework 3.1. Fine-grained sediment 3.2. Ammonium 3.3. Nitrate 3.4. Nitrite 3.5. Phosphorus 3.6. Faecal pathogens 4. Identifying Potential Pollutant SMD Scenarios 4.1. A concentration and discharge case study 5. Summary and Future Research Acknowledgments References

* Biogeochemistry of Soils and Water group, North Wyke Research, Okehampton, Devon, United Kingdom ADAS, Wolverhampton, Woodthorne, Wolverhampton, United Kingdom University of Northern British Columbia, Prince George, British Columbia, Canada } Cranfield University, Cranfield, Bedfordshire, United Kingdom } Centre for Sustainable Water Management, Lancaster Environment Centre, Lancaster University, Lancaster, Lancashire, United Kingdom { {


#


83

S. J. Granger et al.

84

Abstract With the increasing demand for food security comes an increasing pressure on the environment. Contamination of surface water by diffuse agricultural pollutants is widely recognized as an area of concern; however, this has still led to a fragmented approach to scientific research. Pollutants tend to be treated in isolation and only infrequently in the context of an environment where other pollutants may be an issue. This is an important concept, as to achieve costeffective mitigation the effect of any method implemented must take into account the positive as well as the negative effects on other pollutants which exist in the environment in which a method has been implemented. In this chapter, we synthesize the current state of understanding relating to a suite of typical aquatic diffuse pollutants associated with agricultural systems, more specifically those that may originate from intensively managed grassland systems on heavy, clay-rich soil types. This chapter is necessarily wide ranging but tries to draw together the information on each pollutant and present it within a single framework. This is only possible by characterizing the pollutants using shared characteristics along a source–mobilization–delivery (SMD)-continuum. Through this process, we highlight five possible SMD scenarios which can lead to contamination of water bodies. Further information on the nature of these SMD-scenarios can be gained by assessing the relationship between pollutant concentration and discharge of multiple pollutants. In this regard, we highlight the lack of literature available detailing multiple pollutant dynamics and also draw attention to areas of research that we feel need to be addressed if a more holistic approach to diffuse pollution mitigation is to be achieved.

1. Introduction Historically, research into diffuse agricultural pollution has been carried out on specific individual pollutants, focusing on their respective chemistries and processes. This has led to many extremely detailed bodies of research being generated on some pollutants. In grassland systems, it is widely recognized that the pollutants of greatest concern are the nutrients phosphorus (P) (Hawkins and Scholefield, 1996; Haygarth et al., 1998; Sharpley and Syers, 1979) and nitrate (NO3) (Ryden et al., 1984; Scholefield et al., 1993) which can lead to the eutrophication of surface waters (Conley et al., 2009). Both nutrients are added to grasslands via inorganic fertilizers and livestock manures to boost herbage yields; however, NO 3 is highly soluble and readily lost from soils while P, typically being perceived as the limiting nutrient in most inland aquatic ecosystems (Grimm et al., 2003), can cause eutrophication with only minor losses. Phosphorus and NO3 aside, other pollutants of potential concern have

85

Diffuse Pollution from Agricultural Grasslands

received far less attention. Increasingly, the contribution that grasslands make to pools of other pollutants is being questioned (Brazier et al., 2007). Recently, attempts to understand their role in the supply of finegrained sediment and pathogenic organisms to water bodies have been made (Bilotta et al., 2007; Oliver et al., 2005a), while the other forms of pollutant such as ammonium (NH4þ) (Hatch et al., 2004) and NO2 (Burns et al., 1995; Smith et al., 1995a) have received, by comparison, little attention despite having serious adverse affects on aquatic ecosystems (Table 1). When research does exist, it tends to occurs in isolation, without consideration of the other pollutants that often occur contemporaneously in the environment. As concern about water quality has grown, so the legislation governing the quality of waters has developed. Since the United Kingdom joined the European Economic Community (EEC), legislation has been driven by a series of European directives, which have latterly been transposed into national legislation. These have typically been targeted at specifically defined, localized environments, or at specific pollutants such as NO3 (EEC, 1991), NH4þ, NO2, fine-grained sediment (EEC, 1978), and faecal pathogens (EEC, 1976, 1979). However, recent legislation (EEC, 2000) aims to tackle catchments, and pollutants, more holistically with emphasis on ecological quality rather than chemical parameters. Therefore, researchers are now often trying to combine the available information on individual pollutants to create a more holistic approach to diffuse pollution, treating it as a single issue. This is an extremely important concept in terms of understanding pollutant behavior and mitigation. Typically, any mitigation method implemented against any given pollutant has in the past been largely ‘pollutant centric’, a reaction to an environmental pressure attributed to a single pollutant. However, any implemented mitigation method exists in a ‘real world’ system where other pollutants will be present and therefore may impact positively or adversely on those other pollutants. Clearly, the full environmental and economic impacts of a mitigation method cannot be fully appreciated until its effects on all pollutants are Table 1 bodies

The diffuse pollutants covered within the chapter and their effects on water

Pollutant

Ammonium Nitrate Nitrite Phosphorus Pathogens Sediment

Eutrophi- Oxygen cation depletion

X X X X

AcidifiPhysical Pathogenic cation Toxicity effects contamination

X

X

X

X

X

X X

X

X


86

understood. A more integrated framework would potentially take into account the effects of a measure on all pollutants, not just a single pollutant, making diffuse pollution control more ‘method centric’. Such a multipollutant approach has the potential to become complicated and cumbersome. It is therefore necessary for researchers to understand the broader nature of agricultural diffuse pollutants, in terms of both their environmental impacts and pathways. This requires a fundamental understanding of the similar behaviors and characteristics of different pollutants to categorize them. This chapter draws upon the available literature to identify the current state of our knowledge with regard to diffuse agricultural water pollution from intensively managed temperate grasslands, focusing on finegrained sediment, inorganic nitrogen (NO3, NO2, and NHþ 4 ), P, and faecal pathogens. It also explores the possibility of a more holistic classification of diffuse agricultural pollution by using generic pollutant characteristics. This is an area of emerging importance if the increasing food requirements of a growing population and demand for food security are to be met without compromising environmental quality. More specifically we will focus on grasslands on heavy clay soils, generically described within the HOST classification system as Models H-J (Boorman et al., 1995) which account for about 40% of the soils in the United Kingdom, and which are normally intensively managed grasslands (Granger et al., 2009). This specific characterization was chosen as the processes of water movement through this soil type are considered to be more complex and less well understood when compared to lighter sandy, more free-draining soils (Beven and Germann, 1982; Youngs and Leeds-Harrison, 1990). In structured clay soils, the interaction between macropores and micropores, the potential rapid transfer of water to depth, and the installation of lateral drainage systems all lead to a more complex hydrological system (Gooday et al., 2008; Matthews et al., 2000). Within the United Kingdom a broad management and geographic division can be made between these soil types whereby the heavy soils tend to dominate the western maritime regions which are subject to high amounts of precipitation, are surface-water dominated environments, and as such are less suitable for arable cropping. The lighter, more free-draining soils however, are more prevalent in the east of the United Kingdom, where lower levels of precipitation and groundwater-dominated environments are often more suited to arable systems.

2. Defining a Conceptual Framework To better understand diffuse pollutant behavior from intensively managed, seasonally waterlogged, grasslands and to provide an overarching framework within which many pollutants can be described, it is necessary

87


to step back from the specific detailed environmental behavior of potential pollutants and try to find common characteristics. Conceptually, some authors (Haygarth et al., 2005a; Heathwaite et al., 2003) have described the entrainment of diffuse agricultural pollutants into surface waters in terms of a transfer continuum along which pollutant transfer may be reduced. These involve ‘controls’ targeted at the source of the pollutant, at the processes by which the pollutant becomes mobilized, and at the pathways by which pollutants are delivered to water bodies. Each of these locations along the pollutant source–mobilization–delivery (SMD)-continuum can be further conceptually subdivided into aspects which can be used to describe the characteristics of individual diffuse pollutants. Within this chapter the term ‘manure’ is used to describe both solid and liquid animal wastes, with ‘slurry’ being applied specifically to liquid manures.

2.1. Source aspects Sources of agricultural pollutants can be divided into three aspects: external, cycled, and internal sources. Applying mitigation methods to these sources will reduce the pollutant source strength and potentially represent the most effective way of reducing diffuse pollution from agriculture. These sources are illustrated in Fig. 1 and are perhaps best visualized by taking a unit area

External sources

Inorganic fertilizers

Feedstuffs

Cycled sources Manures

Internal sources

Losses to water

Soil system Atmospheric inputs

Figure 1 Examples of the source aspects and their source materials which may lead to diffuse water pollution.

88


of grassland and considering where potentially polluting contaminants are being sourced from. Initially, this grassland is unfarmed and so potential pollutants being sourced from it are considered to be the internal sources. These sources are largely contained within the soil system, but would also include inputs arriving through atmospheric deposition. Once an agricultural system has been set up a second category of sources is created, termed cycled sources. These are generated as a function of the agricultural process and are materials that have been cycled through the farm system. In grassland systems these are dominated by the presence of livestock and include manures, silage effluent, and excreta from grazing animals. Modern farm systems, with high livestock numbers, also need additional supplements to sustain their stocking densities. These inputs represent the third level of source materials which are termed external sources. These sources are imported into the farm system and include inorganic fertilizers and some feedstuffs. It is often the case that many sources may be contributing to aquatic pollution and that these different sources are indistinguishable from one another at the point at which they leave the farm system. It is also perfectly possible for pollutants to pass through any number of these source aspects before arriving at a water body. For example, inorganic P fertilizer brought onto a farm (external) may be spread onto the land and become incorporated into the soil P reservoir (internal). This soil P may then be taken up by vegetation (internal) before being eaten by livestock and excreted (cycled) directly into a water body. Although the P has passed through many different sources aspects, conceptually it is the final source aspect that is considered to be the ultimate source of that specific pollutant, in this case a cycled source. However, it is important to note that in terms of pollutant mitigation, any mitigation method that could have reduced the pollutant source strength is relevant and should be taken into account. So within the example given, although it is a cycled source that leads to the pollution, any suitable mitigation method applied to any source aspect that the pollutant passed through before reaching this point would have had an effect, such as reducing the amount of P fertilizer brought onto the farm. Therefore, it is important not only to know which source aspect the pollution is coming from, but also be aware that mitigation at other source aspects may have an effect on the strength of the source at the point at which it is mobilized.

2.2. Mobilization aspects The term ‘mobilization’ can be used to encompass those processes that enable potential pollutants to separate from their sources and become entrained within hydrological pathways. Pollutant mobilization operates at the soil profile scale and includes physical, chemical, and biological


89

processes. Pollutants that are mobilized from the bulk soil mass can become mobile by a variety of processes, but these processes have been grouped into solubilization or detachment mechanisms by Haygarth et al. (2005a). Solubilization is used to describe the variety of processes that lead to the chemical and biological release of potential pollutants from the bulk soil mass into the soil solution. These are typically low-energy processes which include desorption of material from soil particles, the mineralization of organic matter, and the release of organic and inorganic particles through their decay. Detachment describes the high-energy processes that cause the removal of particles away from the bulk mass of the soil. These particles may be individual pollutants in their own right, such as individual pathogenic organisms, but typically they are soil aggregates composed of organic and inorganic soil material, as well as potentially pathogenic microorganisms and other adsorbed pollutants (Droppo, 2001). Detachment can occur either through the hydraulic action of raindrop impact on the soil surface or by the shear stresses generated by any hydrological flows created on or within the soil. The boundary between the two processes of solubilization and detachment can be defined by the physical size of the material, and this definition is especially strong with reference to the literature surrounding P. Historically within the study of P, material that is >0.45 mm is termed ‘particulate’, while that which is 0.45 mm (Oliver et al., 2005a) and often associated as clumps of cells or with organic material and inorganic soil aggregates (Abu-Ashour et al., 1998b), they can be considered as particles in their own right. When they are mobilized from the internal soil source, they are mobilized by detachment processes in a similar fashion to that of soil aggregates of which they are often a part (Tyrrel and Quinton, 2003). The same processes that cause the detachment of aggregates and microbes from the soil also breakdown and mobilize material from fresh excreta and managed manures on the soil surface leading to the incidental mobilization of faecal pathogens. The pathways that both detached internal soil source material and incidental


101

cycled manure sources follow to their delivery to water bodies are highenergy pathways such as overland flow and macropore/drain flow (AbuAshour and Lee, 2000; Abu-Ashour et al., 1998a; Deeks et al., 2005; Oliver et al., 2005b), while the low-energy pathways yield less pathogens as the soil acts as a filter, straining out microbial particles ( Jamieson et al., 2002).

4. Identifying Potential Pollutant SMD Scenarios Having now ascribed specific pollutants to various aspects within the SMD framework, it is now possible to determine which of the five SMD scenarios contribute which pollutants to water bodies. The five potential SMD scenarios through the conceptual framework by which each pollutant may enter surface water bodies are highlighted in Figure 5. From this, shared pollutant SMD scenarios to water bodies can be determined (Table 3). It is possible to establish that some pollutants appear to have very specific routes, such as NO2, sediment, and PP, whereas other pollutants, such as SP, have many potential routes through the framework. Under ‘normal’ diffuse losses, with just the internal sources contributing, the literature suggests that during periods of low rainfall and base flow the suite of pollutants being delivered to water bodies are all soluble (SP, NO3, and NO 2 ). During periods of rainfall we might expect to see a reduction in the concentration of soluble pollutants delivered to water bodies (due to dilution) and an increase in the concentration of particulate pollutants (PP, paths, sed), with the exception of SP which may also exhibit increased concentrations. During periods of rainfall we may also expect to see ‘extreme’ diffuse losses through the incidental mobilization of cycled and external source materials occurring alongside normal diffuse losses. Unlike the internal source material, which may be considered to be broadly characteristic of a large area, these extreme diffuse losses are highly localized and unpredictable, and the pollutants yielded will be extremely dependent upon the nature of the local source material and its source strength. For example, while SMD scenario 5, which deals with external source materials, may potentially deliver SP, NH4þ, and NO3 (Table 3), it is highly unlikely that all three will be delivered simultaneously as the inorganic fertilizer applied may only contain one or two of these pollutants. This variability will be more predictable with SMD scenario 4 as cycled sources tend to be less variable when compared to the external sources (i.e., manures are likely to contain NHþ 4 , SP, and pathogens); however, the strength of each may vary depending on the nature of the source material. The highly specific nature of extreme diffuse losses means that they are distinctly different when compared to normal diffuse losses. During


102

NO−3

Sed

External

Incidental

High energy

Cycled

Internal

External

Incidental Detachment Solubilization

High energy

High energy 3

High energy

Incidental

Low energy

NO−2 External

Incidental

High energy

High energy 5

Cycled

Internal


High energy

High energy

High energy

Low energy 1

+

NH4 Cycled

Internal

External


High energy

High energy

High energy

Incidental

Low energy 1

PP

High energy 5

Cycled

Internal


High energy 4

High energy

High energy

Low energy

SP

External

Incidental

High energy

Cycled

Internal

External


High energy

High energy 3

High energy

Incidental

Low energy

High energy 5

Cycled

Internal


High energy 4

High energy

High energy 2

Low energy 1

Paths External

Incidental

High energy

Cycled

Internal


High energy 4

High energy 3

High energy

Low energy

Figure 5 The potential SMD scenarios by which diffuse pollutants may become entrained into drainage waters. The SMD scenarios available to specific pollutants are identified in white.

103


Table 3 The source, mobilization, and delivery aspects of normal and extreme diffuse pollution losses and the pollutants at risk of transport Normal loss scenarios Internal

Source aspect Mobilization aspect

Extreme loss scenarios

Solubilization

Delivery aspect Low energy

Cp/Q response Type 1 or 2a SMD scenario 1 PP Sed Paths SP X NHþ 4 NO X 3 NO X 2

Cycled Detachment

External

Incidental

High energy

Type 2b 2

3 X X X

X

Type 2b or 3 4 5

X X X

X X X

precipitation events it might be expected that PP, sediment, and pathogens (and possibly SP) would exhibit a Type 2b response (where Cp and Q are positively related), while NO 3 and NO2 would be negatively related or show no change with Q (Type 1 or 2a). However, should an incidental loss of a cycled source occur it might be expected that a significant increase in SP, pathogens, and especially NHþ 4 might be observed. Depending on the nature of this loss and how quickly the source is exhausted the Cp/Q relationship of these pollutants would either be a Type 2b or 3 response. Similarly with losses of external sources, although we may not be able to predict the nature of the source material, what is reasonably certain is that a Type 2/3 response will be observed in a soluble pollutant phase. This in itself is characteristic of SMD scenario 5 as soluble pollutants under the conditions of normal diffuse losses would typically exhibit a Type 1 or 2a response. A further point of note is that SMD scenarios 4 and 5 do not necessarily rely on energy supplied in the form of precipitation, and therefore it is perfectly possible for extreme diffuse losses to occur alongside the normal loss scenarios of SMD scenario 1. Under these circumstances both SMD scenarios 4 and 5 would exhibit a Type 3 response.

4.1. A concentration and discharge case study Some examples of Cp/Q response types are presented in Fig. 6 and data were taken from a multipollutant study during an intense storm event monitored within a 48-ha grassland catchment in Devon, UK (Granger


104

A

0

100

80

0.5

60

1.0

40

1.5

20

2.0

0 02/06/2008

Rainfall (mm)

Catchment outlet Q (l s−1)

Rainfall Q

2.5 04/06/2008

03/06/2008 Time

C

50

60

0

NO3−

3

2 50 1

0

0

0

D

E 100

NH4+

16

100

8000 SP

50

8

0

0

50

0

4000

SP (mg l −1)

Median catchment Q (l s−1)

100

120

NO3−-N (mg l −1)

NO−2

NO2−-N (mg l −1)

100

NH4+-N (mg l −1)

Median catchment Q (l s−1)

B

0

Figure 6 The hydrograph generated by a stream draining a 48-ha grassland catchment in Devon during a summer storm (A), and the Cp/Q relationship for NO2 (B), NO3 (C), NH4þ (D), and SP (E). With kind permission from Springer ScienceþBusiness Media: Granger et al. (2009), Figs. 2 and 3.

et al., 2009). The hydrograph was generated by an intense storm event which occurred over a 2-day period in June 2008. Measurements of Q were made on a 1-min time step and the response is typical of these heavy clay soil systems with a rapid increase in Q followed by a slower exponential decrease back to pre-event levels (Fig. 6A). However, due to the frequency of Q measurements more detail was revealed in this hydrograph than would normally be expected from a dataset with less frequent measurements. A small subsidiary hydrograph occurs on the rising limb of the main hydrograph which has been resolved sufficiently for its falling limb to be observed before the main rising limb dominates. The interpretation of this hydrological


105

response was that within the catchment a small area of farm buildings and hard standings had some form of direct connection to the stream and that the intense nature of the rainfall had caused such rapid runoff from this area that it had produced its own hydrograph before the more buffered response of the grassland which makes up the majority of the catchment. In effect two hydrographs exist although one dominates over the other. From Fig. 5 it is predicted that the loss of NO 2 can only be via SMD scenario 1, as it can only have an internal source, with a Type 1/2a Cp/Q response. This is indeed what was observed across the two hydrographs, with concentrations rapidly dropping when Q increased and only increasing again when Q subsides (Fig. 6B). The response of NO3 concentration and Q (Fig. 6C) is virtually identical to NO2, and given that the mechanisms of NO3 loss from such systems are understood (i.e., rainwater dilution of soil water NO3), these processes can also be applied to NO2. The response of NH4þ, however, is completely different with very low concentrations before and during the main hydrograph interrupted by a large spike in concentration on the rising limb of the hydrograph (Fig. 6D). This spike in concentration is associated with the peak of the subsidiary hydrograph and concentrations rapidly return to pre-event levels once the rising limb of the main hydrograph begins to dominate. It is clear from this pollutant and its response that the two hydrographs have distinctly different NHþ 4 Cp/Q relationships. For the main hydrograph no change in NHþ 4 concentration occurs, exhibiting a Type 1 Cp/Q relationship characteristic of a pollutant without a source. For the subsidiary hydrograph the Cp/Q relationship is a Type 2b, positively related, response. Given that there are no internal sources of NH4þ this means that we are observing some form of extreme diffuse loss through incidental mobilization (Fig. 3) of either an external or a cycled source. Given that no fertilizers or manures had been applied to the catchment recently and that the subsidiary hydrograph is believed to represent runoff from the animal housing and storage area it can be inferred that this NH4þ spike is a cycled source. Interestingly, without the resolution of the subsidiary hydrograph, the NH4þ Cp/Q response would have been classified as a Type 3, which would still indicate an extreme diffuse loss, although the mechanisms for that would have been less clear with one pollutant response and one hydrograph, as opposed to two Cp/Q relationships. The nature of the Cp/Q response for NO2/NO3 and NHþ 4 appears to indicate that different spatially separate source areas within the catchment were contributing distinctly different diffuse pollutants. However, other diffuse pollutants appear to be derived from both contributing areas leading to a combination of Cp/Q response types. For SP, two chemistry responses have been resolved which are associated with the two hydrological peaks (both Type 2b): (i) a large spike in concentrations associated with the subsidiary hydrograph, similar to that of the NH4þ response and (ii) a smaller broader peak associated with the main hydrograph peak.


106

Given the conclusions drawn about the sources and response types of the forms of inorganic N, it would seem to be logical that similar conclusions be drawn here, namely that the subsidiary hydrograph and associated pollutant spikes are incidental losses from hard standings within the catchment and that the main hydrograph and its chemical responses are representative of the grassland which make up the majority of the land area. It is clear from this case study that to understand better the sources and mechanisms of diffuse pollutant entrainment in water bodies, that high-resolution monitoring of multiple pollutants must be undertaken. Without frequent sampling the significant incidental loss of certain pollutants would have been missed and without high-resolution measurement of Q the small subsidiary hydrograph would not have been recorded. Furthermore, the analysis of different pollutants in itself can allow for a more detailed explanation of diffuse pollutant delivery than from a single pollutant alone. While extensive research has been carried out on some pollutants, such as sediment, P, and NO3, other pollutants have received, by comparison, less attention (i.e., NHþ 4 , NO2 , pathogens). Of these better-studied pollutants, a large part of the work undertaken has focused upon arable systems on light sandy soils and has not dealt with grasslands on heavy soils. Furthermore, there is even less information on the entrainment of multiple pollutants from all systems. If multiple pollutant delivery is to be expected as predicted by the framework then it should be possible to review the literature on pollutant losses to assess whether the broad categorization proposed here is supported by preexisting studies. A review of the literature is presented in Table 4 which focuses on the entrainment of multiple pollutants in drainage from intensively managed agricultural grassland systems on heavy soil types. Within this table, where possible, the various SMD aspects that have been studied have been inferred.

5. Summary and Future Research In this chapter, we have built a (conceptual) framework for the classification of diffuse agricultural pollutants from intensive grassland on heavy clay-rich soils. By focusing on a more generic functional classification we have been able to categorize these pollutants within an SMD-continuum. Through this process we have identified five specific SMD scenarios by which one or more potential pollutants may become entrained in water draining agricultural grasslands. The applicability of this framework was examined using an existing multipollutant dataset which enabled a more comprehensive understanding of a spatially complex mixture of pollutant sources within a grassland catchment. However, for such systems to be better understood, multipollutant research needs to be undertaken and this

Table 4 Review of the literature detailing multiple pollutants in drainage from intensively managed grassland systems Framework classification (? ¼ inferred) Land use

Soil type

Pollutant

Delivery

Mobilization

Source

Comments

References

Heavily grazed permanent grassland, UK

Clay loam (HOST 17)

Sediment PP NHþ 4

High energy (surface flow)

Detachment?

Internal?

Heathwaite et al. (1990)

Incidental

Cycled

Long-term grassland, Wales

Silty clay loam (HOST 17)

NHþ 4 (Total)P NO 3

High energy (drain flow)

Incidental Incidental? Solubilization

Cycled Cycled Internal?

Permanent grassland, UK

Silty clay loam (HOST 17, 21, 24)

NHþ 4 NO 3 SP


Incidental

External

Sediment PP


Detachment

Internal

Pathogen NO 3


Incidental

Cycled

Less than 20% was SP with 80% PP being organic, particulate or adsorbed NO 3 levels not greatly increased while TP and NHþ 4 increase to levels of environmental significance Losses of applied fertilizers were low compared to amount applied but high with regard to water quality Indication that SP and ‘‘dissolved solids’’ may also be linked Unclear as to the source of the NO 3 as manure should contain little

Grazed grassland Sandy loam or clay loam catchment, (HOST 24) UK Silt loam Recently cut grassland, USA

Williams and Nicholson (1995)

Scholefield and Stone (1995)

Dils and Heathwaite (2000) Fajardo et al. (2001)

(continued)

108

Table 4

(continued) Framework classification (? ¼ inferred)

Land use

Soil type

Pollutant

Delivery

Mobilization

Source

Comments


Incidental?

Cycled?

Low energy? High energy (drain flow)

Solubilization Solubilization Incidental?

Internal Internal Cycled?

Drained pastures, Slowly permeable NHþ 4 New Zealand silt loam Pathogen (Total)P


Incidental?

Cycled?

Vegetated runoff Silt loam boxes, USA


Incidental

Cycled

Soils were drained but HelnonenTanski and drainage could not be Uusicollected. NO 3 was Ka¨mppa¨ increased in surface (2001) and flow after slurry Uusiapplication Ka¨mppa¨ and HeinonenTanski (2001) Ga¨chter et al. The source of SP not (2004) determined; however, slurry was applied. NO 3 inversely related to flow Nonrainfall-driven link Monaghan and Smith between applied waste (2004) effluent and pollutants in drains (not NO 3 and sediment) Stout et al. Unclear as to what (2005) extent P is PP or SP or whether soil P is contributing

Drained grassland Clay/silty clay soil NHþ 4 plots, Finland

Predominantly grassland catchments, Switzerland

Relatively impermeable till

NO 3 SP

(Total)P Pathogen

References

Grazed pasture, New Zealand

Silt loam

Sediment PP SP NHþ 4 Pathogen


Cycled Incidental Detachment? Internal? Solubilization?

Grassland catchments, Ireland

Dense clay till

Sediment PP SP?

High energy

Detachment

Grazed and fertilized permanent grassland, UK Farmyards, Scotland

Clayey (HOST 24)

Sediment PP

High energy (surface/drain flow)

Detachment

Internal

Incidental?

Cycled?

na

NHþ 4 SP Pathogen


Incidental

Cycled

Intensively managed grassland catchment, UK

Clay soil (HOST 24)

NO2 NO 3 NHþ 4 Sediment PP SP

Low energy

Solubilization

Internal

High energy

Incidental Detachment Incidental

Cycled Internal Cycled

Internal

Solubilization?

Manure is the probable McDowell et al. (2006) source of these pollutants; however, the effect of manure on the soil has not been considered Douglas et al. Sed/PP link holds for (2007) nonstorm periods perhaps indicating that SP is dominant P form and is becoming rapidly adsorbed onto sediment Bilotta et al. Sediment and TP (2008) losses greater from undrained land

109

Edwards and Variation between Hooda farms was high. NO 3 (2008) and was present but not a Edwards et al. significant form of N (2008) Granger et al. NO2 concentrations (2009) were found to exceed EU limits in base flow, while NO 3 was below limits. During storm event farm building was found to cause an incidental spike in certain pollutant losses

110


fundamental need for multipollutant research has largely been overlooked by the scientific community. This process has also identified research gaps within the literature which we now highlight:

In many studies the sources of pollutant are inferred. For example, while manure may have been applied to soil, and an increase in NO 3 and SP observed in drainage, the link between the pollutant and the source has not been directly established. Have the NO 3 and an SP come from the manure, or from the soil as a consequence of the manure application and what is each source contributing to the total loss from the system? Recent evidence suggests that different sources may still contribute to ‘base flow’ concentrations long after the elevated concentrations have apparently disappeared. There remains a great deal of uncertainty surrounding the mobilization and delivery of P forms. Early studies indicated that the grasslands are dominated by SP losses and that PP losses are generally dismissed with grasslands being considered nonerosive environments. More recently, however, the contribution of PP has been shown to be far more important than previously believed, yet it is still uncertain as to whether this actually represents PP mobilization or SP mobilizations which becomes adsorbed onto particles during delivery. Incidental losses remain an enigma. While there is no doubt that they may be locally significant, is their cumulative effect important at a larger scale or are they too dispersed? Moreover, if incidental losses of manures can lead to increases in particulate contaminants such as pathogens might, do they also significantly contribute to fine-grained sediment concentrations? Furthermore, while incidental losses of P (both fertilizer and manure) have been investigated, are incidental losses of other pollutants (i.e., NHþ 4 and NO3 ) important?

ACKNOWLEDGMENTS The authors thank David Chadwick and Les Firbank of North Wyke Research for their comments on the chapter. We acknowledge UK Department for Environment, Food and Rural Affairs (Defra) Project ES0121 for funding that resulted in this work and the input of Brian Chambers, David Harris, and Ken Smith of the Agricultural Development and Advisory Service (ADAS). North Wyke Research is supported by core funding from the UK Biotechnology and Biological Sciences Research Council (BBSRC).

REFERENCES Abu-Ashour, J., and Lee, H. (2000). Transport of bacterial on sloping soil surfaces by runoff. Environ. Toxicol. 15, 149–153.


111

Abu-Ashour, J., Joy, D. M., Lee, H., Whiteley, H. R., and Zelin, S. (1998a). Movement of bacteria in unsaturated soil columns with macropores. Am. Soc. Agric. Eng. 41, 1043–1050. Abu-Ashour, J., Joy, D. M., Lee, H., Whitley, H. R., and Zelin, S. (1998b). Movement of bacteria in unsaturated soil columns with macropores. Trans. ASAE 41, 1043–1050. Armstrong, A. C., Leeds-Harrison, P. B., Harris, G. L., and Catt, J. A. (1999). Measurement of solute fluxes in macroporous soils: Techniques, problems and precision. Soil Use Manage. 15, 240–246. Barraclough, D., Hyden, M. J., and Davies, G. P. (1983). Fate of fertilizer nitrogen applied to grassland. I. Field leaching results. J. Soil Sci. 34, 483–497. Beven, K., and Germann, P. (1982). Macropores and water flow in soils. Water Resour. Res. 18, 1311–1325. Beven, K., Heathwaite, L., Haygarth, P., Walling, D., Brazier, R., and Withers, P. (2005). On the concept of delivery of sediment and nutrients to stream channels. Hydrol. Processes 19, 551–556. 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. Bilotta, G. S., Brazier, R. E., Haygarth, P. M., Macleod, C. J. A., Butler, P., Granger, S., Krueger, T., Freer, J., and Quinton, J. (2008). Rethinking the contribution of drained and undrained grasslands to sediment-related water quality problems. J. Environ. Qual. 37, 906–914. Boorman, D. B., Hollis, J. M., and Lilly, A. (1995). Hydrology of soil types: A hydrologically based classification of the soils of the United Kingdom. Institute of Hydrology report No. 126. Wallingford, Oxfordshire. Bradley, R. I., Milne, R., Bell, J., Lilly, A., Jordan, C., and Higgins, A. (2005). A soil carbon and land use database for the United Kingdom. Soil Use Manage. 21, 363–369. Brazier, R. E., Bilotta, G. S., and Haygarth, P. M. (2007). A perspective on the role of lowland, agricultural grasslands in contributing to erosion and water quality problems in the UK. Earth Surf. Processes Landforms 32, 964–967. Burns, L. C., Stevens, R. J., Smith, R. V., and Cooper, J. E. (1995). The occurrence and possible sources of nitrite in a grazed, fertilized, grassland soil. Soil Biol. Biochem. 27, 47–59. Chadwick, D. R., and Chen, S. (2002). Manures. In ‘‘Agriculture, Hydrology and Water Quality’’ (P. M. Haygarth and S. C. Jarvis, Eds.), pp. 57–82. CABI Publishing, Wallingford, Oxon, UK. Conley, D. J., Paerl, H. W., Howarth, R. W., Boesch, D. F., Seitzinger, S. P., Havens, K. E., Lancelot, C., and Likens, G. E. (2009). Controlling eutrophication: Nitrogen and phosphorus. Ecology 323, 1014–1015. Deeks, L. K., McHugh, M., and Owens, P. N. (2005). Faecal contamination of watercourses from farm waste disposal for three sites in the UK with contrasting soil types. Soil Use Manage. 21, 212–220. Dils, R. M., and Heathwaite, A. L. (2000). Tracing phosphorus movement in agricultural soils. In ‘‘Tracers in Geomorphology’’ (I. D. L. Foster, Ed.), pp. 259–276. John Wiley & Sons Ltd., Chichester UK. Douglas, R. W., Menary, W., and Jordan, P. (2007). Phosphorus and sediment transfers in a grassland river catchment. Nutr. Cycl. Agroecosyst. 77, 199–212. Droppo, I. G. (2001). Rethinking what constitutes suspended sediment. Hydrol. Processes 15, 1551–1564. Edwards, A. C., and Hooda, P. S. (2008). Farmyard point discharges and their influence on nutrient and labile carbon dynamics in a second order stream draining through a dairy unit. J. Environ. Manage. 87, 591–599.

112


Edwards, A. C., Kay, D., McDonald, A. T., Francis, C., Watkins, J., Wilkinson, J. R., and Wyer, M. D. (2008). Farmyards, an overlooked source for highly contaminated runoff. J. Environ. Manage. 87, 551–559. EEC. (1976). Council directive of 8 December 1975 concerning the quality of bathing water (76/160/EEC). Official J. Eur. Communities 19, 1–7. EEC. (1978). Council directive of 18 July 1978 on the quality of fresh waters needing protection or improvement in order to support fish life (78/659/EEC). Official J. Eur. Communities 21, 1–10. EEC. (1979). Council directive of 30 October 1979 on the quality required of shellfish waters (79/923/EEC). Official J. Eur. Communities 22, 47–52. EEC. (1991). Council directive of 12 December 1991 concerning the protection of waters against pollution caused by nitrates from agricultural sources (91/676/EEC). Official J. Eur. Communities L375, 1–8. EEC. (2000). Directive 2000/60/EC of the European parliament and the council of 23 October 2000 establishing a framework for Community action in the field of water policy. Official J. Eur. Communities L327, 1–72. Fajardo, J. J., Bauder, J. W., and Cash, S. D. (2001). Managing nitrate and bacteria in runoff from livestock confinement areas with vegetative filter strips. J. Soil Water Conserv. 56, 185–191. Foster, I. D. L., Chapman, A. S., Hodgkinson, R. M., Jones, A. R., Lees, J. A., Turner, S. E., and Scott, M. (2003). Changing suspended sediment and particulate phosphorus loads and pathways in underdrained lowland agricultural catchments; Herefordshire and Worcestershire, UK. Hydrobiologica 494, 119–126. Ga¨chter, R., Steingruber, S. M., Reinhardt, M., and Wehrli, B. (2004). Nutrient transfer from soil to surface waters: Differences between nitrate and phosphate. Aquat. Sci. 66, 117–122. Gessel, P. D., Hansen, N. C., Goyal, S. M., Johnson, L. J., and Webb, J. (2004). Persistence of zoonotic pathogens in surface soil treated with different rates of liquid manure. Appl. Soil Ecol. 25, 237–243. Gooday, R., Anthony, S., and Fawcett, L. (2008). A field scale model of soil drainage and nitrate leaching for application in nitrate vulnerable zones. Environ. Model. Softw. 23, 1045–1055. Goulding, K. W. T., Bailey, N. J., and Bradbury, N. J. (1998). A modelling study of nitrogen deposited to arable land from the atmosphere and its contribution to nitrate leaching. Soil Use Manage. 14, 70–77. Granger, S. J., Hawkins, J. M. B., Bol, R., White, S. M., Naden, P., Old, G., Bilotta, G. S., Brazier, R. E., Macleod, C. J. A., and Haygarth, P. M. (2009). High temporal resolution monitoring of multiple pollutant responses in drainage from an intensively managed grassland catchment caused by a summer storm. Water Air Soil Pollut. 205, 377–393. Grimm, N. B., Gergel, S. E., McDowell, W. H., Boyer, E. W., Dent, C. L., Groffman, P., Hart, S. C., Harvey, J., Johnson, C., Mayorga, E., McClain, M. E., and Pinay, G. (2003). Merging aquatic and terrestrial perspectives of nutrient biogeochemistry. Oecologia 442, 485–501. Haigh, R. A., and White, R. E. (1986). Nitrate leaching from a small, underdrained, grassland, clay catchment. Soil Use Manage. 2, 65–70. Hatch, D. J., Chadwick, DR., Smith, K. A., and Chambers, B. J. (2004). Ammonium-N losses from agriculture. In ‘‘Proceedings of the SAC and SEPA Biennial Conference’’ (D. Lewis and L. Gairns, Eds.), pp. 165–173. Edinburgh, 24–25 March 2004. Hawkins, J. M. B., and Scholefield, D. (1996). Molybdate-reactive phosphorus losses in surface and drainage waters from permanent grassland. J. Environ. Qual. 25, 727–732. Haygarth, P. M., and Jarvis, S. C. (1999). Transfer of phosphorus from agricultural soils. Adv. Agron. 66, 195–249. Haygarth, P. M., Hepworth, L., and Jarvis, S. C. (1998). Forms of phosphorus transfer in hydrological pathways from soil under grazed land. Eur. J. Soil Sci. 49, 65–72.


113

Haygarth, P., Dils, R., and Leaf, S. (2000). Phosphorus. In ‘‘Diffuse Pollution Impacts’’ (B. J. D’Arcy, J. B. Ellis, R. C. Ferrier, A. Jenkins, and R. Dils, Eds.), pp. 73–84. Terence Dalton Publishers, Lavenham Press, Lavenham, Suffolk, UK. Haygarth, P. M., Turner, B. L., Fraser, A., Jarvis, S., Harrod, T., Nash, D., Halliwell, D., Page, T., and Beven, K. (2004). Temporal variability in phosphorus transfers: Classifying concentration-discharge event dynamics. Hydrol. Earth Syst. Sci. 8, 88–97. Haygarth, P. M., Condron, L. M., Heathwaite, A. L., Turner, B. L., and Harris, G. P. (2005a). The phosphorus transfer continuum: Linking source to impact with an interdisciplinary and multi-scaled approach. Sci. Total Environ. 344, 5–14. Haygarth, P., Chadwick, D., Granger, S., Chambers, B., Anthony, S., Smith, K., and Harris, D. (2005b). Characteristic behaviour and potential mitigation of some diffuse pollutants in England and Wales: A review of ammonium, nitrite, some potential pathogens and biological oxygen demand. A review prepared as part fulfilment of Defra project ES0121 COST-DP: Cost Effective Diffuse Pollution Management. Haygarth, P. M., Bilotta, G. S., Bol, R., Brazier, R. E., Butler, P. J., Freer, J., Gimbert, L. J., Granger, S. J., Krueger, T., Macleod, C. J. A., Naden, P., Old, G., et al. (2006). Processes affecting transfer of sediment and colloids, with associated phosphorus, from intensively farmed grasslands: An overview of key issues. Hydrol. Processes 20, 4407–4413. Heathwaite, A. L., Burt, T. P., and Trudgill, S. T. (1990). The effect of land use on nitrogen, phosphorus and suspended sediment delivery to streams in a small catchment in southwest England. In ‘‘Vegetation and Erosion’’ ( J. B. Thornes, Ed.). John Wiley & Sons Ltd., Chichester, UK. Heathwaite, A. L., Fraser, A. I., Johnes, P. J., Hutchins, M., Lord, E., and Butterfield, D. (2003). The phosphorus indicators tool: A simple model of diffuse P loss from agricultural land to water. Soil Use Manage. 19, 1–11. Helnonen-Tanski, H., and Uusi-Ka¨mppa¨, J. (2001). Runoff of faecal microorganisms and nutrients from perennial grass ley after application of slurry and mineral fertilizer. Water Sci. Technol. 43, 143–146. Hutchison, M. L., Walters, L. D., Avery, S. M., Synge, B. A., and Moore, A. (2004). Levels of zoonotic agents in British livestock manures. Lett. Appl. Microbiol. 39, 207–214. Jamieson, R. C., Gordon, R. J., Sharples, K. E., Stratton, G. W., and Madani, A. (2002). Movement and persistence of faecal bacteria in agricultural soils and subsurface drainage water: A review. Canadian Biosyst. Eng. 44, 1.1–1.9. Jawson, M. D., Elliott, L. F., Saxton, K. E., and Fortier, D. H. (1982). The effect of cattle grazing on indicator bacteria in a runoff from a Pacific Northwest watershed. J. Environ. Qual. 11, 621–627. Lennox, S. D., Foy, R. H., Smith, R. V., and Jordan, C. (1997). Estimating the contribution from agriculture to the phosphorus load in surface water. In ‘‘Phosphorus Loss from Soil to Water’’ (H. Tunney, O. T. Carton, P. C. Brookes, and A. E. Johnson, Eds.), pp. 55–75. CAB International, Oxon, UK. Lewis, W. M., and Morris, D. P. (1986). Toxicity of nitrite to fish: A review. Trans. Am. Fish. Soc. 115, 183–195. Matthews, A. M., Portwood, A. M., Armstrong, A. C., Leeds-Harrison, P. B., Harris, G. L., Catt, J. A., and Addiscott, T. M. (2000). CRACK-NP, development of a model for predicting pollutant transport in cracking clay soils. Soil Use Manage. 16, 279–284. McDowell, R. W., Muirhead, R. W., and Monaghan, R. M. (2006). Nutrient, sediment, and bacterial losses in overland flow from pasture and cropping soils following dung deposition. Commun. Soil Sci. Plant Anal. 37, 93–108. McDowell, R. W., Nash, D. M., and Robertson, F. (2007). Sources of phosphorus lost from a grazed pasture receiving simulated rainfall. J. Environ. Qual. 36, 1281–1288.

114


Misselbrook, T. H., Chadwick, D. R., Chambers, B. J., Smith, K. A., Webb, J., Demmers, T., and Sneath, R. W. (2003). Inventory of ammonia emissions from UK agriculture. Rep. No. Defra Contract AM0127. Monaghan, R. M., and Smith, L. C. (2004). Minimising surface water pollution resulting from farm-dairy effluent application to mole-pipe drained soils. II. The contribution of preferential flow of effluent to whole-farm pollutant losses in subsurface drainage from a West Otago dairy farm. N.Z. J. Agric. Res. 47, 417–428. Morgan, R. P. C. (2005). Soil Erosion and Conservation.Wiley-Blackwell Publishing, UK. Nicholson, F. A., Groves, S. J., and Chambers, B. J. (2005). Pathogen survival during livestock manure storage and following land application. Bioresour. Technol. 96, 135–143. Oliver, D. M., Clegg, C. D., Haygarth, P. M., and Heathwaite, A. L. (2005a). Assessing the potential for pathogen transfer from grassland soils to surface waters. Adv. Agron. 85, 125–180. Oliver, D. M., Heathwaite, L., Haygarth, P. M., and Clegg, C. D. (2005b). Transfer of Escherichia coli to water from drained and undrained grassland after grazing. J. Environ. Qual. 34, 918–925. Parr, W., Andrews, K., Mainstone, C. P., and Clarke, S. J. (1998). Diffuse pollution: Sources of N and P. WRc report number FR/D 0031 to the Department of the Environment, Transport and the Regions. WRc Medmenham, UK. Preedy, N., McTiernan, K., Matthews, R., Heathwaite, L., and Haygarth, P. (2001). Rapid incidental phosphorus transfers from grassland. J. Environ. Qual. 30, 2105–2112. Reddy, K. R., Khaleel, R., and Overcash, M. R. (1981). Behavior and transport of microbial pathogens and indicator organisms in soils treated with organic wastes. J. Environ. Qual. 10, 225–266. Ryden, J. C., Ball, P. R., and Garwood, E. A. (1984). Nitrate leaching from grassland. Nature 311, 50–53. Scholefield, D., and Stone, A. C. (1995). Nutrient losses in runoff water following application of different fertilizers to grassland cut for silage. Agric. Ecosyst. Environ. 55, 181–191. Scholefield, D., Tyson, K. C., Garwood, E. A., Armstrong, A. C., Hawkins, J., and Stone, A. C. (1993). Nitrate leaching from grazed lysimeters: Effects of fertilizer input, field drainage, age of sward and patterns of weather. J. Soil Sci. 44, 601–613. Sharpley, A. N., and Syers, J. K. (1979). Phosphorus inputs into a stream draining an agricultural watershed: II. Amounts contributed and relative significance of runoff types. Water Air Soil Pollut. 11, 417–428. Smith, L. C., and Monaghan, R. M. (2003). Nitrogen and phosphorus losses in overland flow from a cattle-grazed pasture in Southland. N.Z. J. Agric. Res. 46, 225–237. Smith, R. V., Foy, R. H., Lennox, S. D., Jordan, C., Burns, L. C., Cooper, J. E., and Stevens, R. J. (1995a). Occurrence of nitrite in the Lough Neagh river system. J. Environ. Qual. 24, 952–959. Smith, R. V., Lennox, S. D., Jordan, C., Foy, R. H., and McHale, E. (1995b). Increase in soluble phosphorus transported in drainflow from a grassland catchment in response to soil phosphorus accumulation. Soil Use Manage. 11, 204–209. Smith, K. A., Chalmers, A. G., Chambers, B. J., and Christie, P. (1998). Organic manure phosphorus accumulation, mobility and management. Soil Use Manage. 14, 154–159. Steeń, I. (1997). A European fertilizer industry view on phosphorus retention and loss from agricultural soils. In ‘‘Phosphorus Loss from Soil to Water’’ (H. Tunney, O. T. Carton, P. C. Brookes, and A. E. Johnson, Eds.), pp. 311–328. CAB International, Oxon, UK. Stephenson, G. R., and Street, L. V. (1978). Bacterial variations in streams from a southwest Idaho rangeland watershed. J. Environ. Qual. 7, 150–157. Stout, W. L., Pachepsky, Y. A., Shelton, D. R., Sadeghi, A. M., Saporito, L. S., and Sharpley, A. N. (2005). Runoff transport of faecal coliforms and phosphorus released from manure in grass buffer conditions. Lett. Appl. Microbiol. 41, 230–234.


115

Trehan, S. P. (1994). Immobilization of 15NH4 by cattle slurry decomposing in soil. Soil Biol. Biochem. 26, 743–745. Tyrrel, S. F., and Quinton, J. N. (2003). Overland flow transport of pathogens from agricultural land receiving faecal wastes. J. Appl. Microbiol. 94, 87S–93S. Uusi-Ka¨mppa¨, J., and Heinonen-Tanski, H. (2001). Runoff of nutrients and faecal microorganisms from grassland after slurry application. In ‘‘Animal Husbandry. DIAS Rep. 21’’ (H. B. Rom and C. S. Sorensen, Eds.), pp. 144–151. Agricultural Research Centre of Finland, Jokioinen, Finland. Vinten, A. J. A., Douglas, J. T., Lewis, D. R., Aitken, M. N., and Fenlon, D. R. (2004). Relative risk of surface water pollution by E. coli derived from faeces of grazing animals compared to slurry application. Soil Use Manage. 20, 13–22. Walling, D. E. (2005). Tracing suspended sediment sources in catchment and river systems. Sci. Total Environ. 344, 159–184. Williams, J. R., and Nicholson, R. J. (1995). Low rate irrigation of dirty water and its effect on drainage water quality. Agric. Eng. 50(2), 24–28. Williams, J. R., Smith, K. A., Chambers, B. J., and Cross, R. B. (2002). Nitrogen and phosphorus losses in drainage water following pig slurry applications to a drained clay soil. In ‘‘RAMIRAN 2002’’ ( J. Venglovsky and G. Greserova, Eds.), pp. 109–113. Strbske Pleso, High Tatras, Slovak Republic. Withers, P. J. A., Edwards, A. C., and Foy, R. H. (2001). Phosphorus cycling in UK agriculture and implications for phosphorus loss from soil. Soil Use Manage. 17, 139–149. Withers, P. J. A., Ulen, B., Stamm, C., and Bechmenn, M. (2003). Incidental phosphorus losses—Are they significant and can they be predicted? J. Plant Nutr. Soil Sci. 166, 459–468. Youngs, E. G., and Leeds-Harrison, P. B. (1990). Aspects of transport processes in aggregated soils. J. Soil Sci. 41, 665–675.

C H A P T E R

F O U R

A New GIS Nitrogen Trading Tool Concept for Conservation and Reduction of Reactive Nitrogen Losses to the Environment J. A. Delgado,* C. M. Gross,† H. Lal,‡ H. Cover,§ P. Gagliardi,* S. P. McKinney,‡ E. Hesketh,# and M. J. Shaffer** Contents 1. Introduction 2. Understanding the Nitrogen Cycle with Respect to Nitrogen Management and Trading 2.1. Understanding the relationships between the soil–crop–hydrologic cycle and nitrogen trading 2.2. Inputs 2.3. Transformations and pathways for reactive and total nitrogen losses 2.4. Nitrogen management and long-term effects on nitrogen pools 2.5. Relationships: Carbon and nitrogen sequestration and emissions of N2O 3. New Technologies 3.1. Tier one spreadsheet approaches 3.2. New prototypes: Web-based and stand-alone modeling approaches 4. Case Scenarios: GIS Trading Tool Concept Evaluations 4.1. Irrigated systems from dry western US 4.2. No-till systems from north atlantic region 4.3. Manure operations from midwest region

118 123 127 132 145 152 153 154 154 155 156 157 157 158

* USDA-ARS-Soil Plant Nutrient Research Unit, Fort Collins, Colorado, USA USDA-NRCS, WNTSC, Beltsville, Maryland, USA USDA-NRCS, WNTSC, Portland, Oregon, USA } Vistronix, Inc., Portland, Oregon, USA # USDA-NRCS, WNTSC, Amherst, Massachusetts, USA ** USDA-ARS (Retired), Fort Collins, Colorado, USA { {


#


117

J. A. Delgado et al.

118

5. Current Applications and Trends 5.1. Water quality markets 5.2. Air quality markets 6. Summary and Conclusions References

158 158 159 160 160

Abstract Nitrogen (N) inputs to agricultural systems are important for their sustainability. However, when N inputs are unnecessarily high, the excess can contribute to greater agricultural N losses that impact air, surface water, and groundwater quality. It is paramount to reduce off-site transport of N by using sound management practices. These practices could potentially be integrated with water and air quality markets, and new tools will be necessary to calculate potential nitrogen savings available for trade. The USDA-NRCS and USDA-ARS Soil Plant Nutrient Research Unit developed a web-based and stand-alone Nitrogen Trading Tool (NTT) prototype. These prototypes have an easy-to-use interface where nitrogen management practices are selected for a given state and the NTT calculates the nitrogen trading potential compared to a given baseline. The stand-alone prototype can also be used to calculate potential savings in direct and indirect carbon sequestration equivalents from practices that reduce N losses. These tools are powerful, versatile, and can run with the USA soil databases from NRCS (SSURGO) and NRCS climate databases. The NTT uses the NLEAP model, which is accurate at the field level and has GIS capabilities. Results indicate that the NTT was able to evaluate management practices for Ohio, Colorado, and Virginia, and that it could be used to quickly conduct assessments of nitrogen savings that can potentially be traded for direct and indirect carbon sequestration equivalents in national and international water and air quality markets. These prototypes could facilitate determining ideal areas to implement management practices that will mitigate N losses in hot spots and provide benefits in trading.

1. Introduction The use of nitrogen (N) inputs in agricultural systems has heavily influenced the sustainability and economical viability of agricultural systems worldwide. These N inputs help maximize yields, which is necessary to supply food to the ever-growing world population. However, when these N inputs are higher than necessary, the excessive N can contribute to greater agricultural N losses that impact air, surface water, and groundwater quality (Fig. 1). One of the reasons that excessive N can lead to increased losses is that it is a very mobile and dynamic nutrient. Fortunately, best management practices for N can be used to synchronize N inputs with crop N uptake sinks in a way that minimizes N losses to the environment.

Aeolian erosion and transport of soil, organic matter and nutrients

Anthropogenic emissions of C, N and S

Sea spray contributes to transport of nutrients

Wet and dry nutrient SO2 and NH3 can also be absorbed or precipitation and emitted directly from plants acid rain N2 fixation, N-P-K and other H2SO4 HNO3

Photosynthesis CO2

symbiotic (e.g. rhizobium) and NH3 Nonsymbiotic (e.g. blue-green algae)

fertilizers

N P K Water erosion sediments, organic matter and surface transport of nutrients to water bodies

N2 N2O NOx CO2

Soil organic matter storage: C, N, P, S and others

Under anaerobic conditions H2S CH4 and Higher N2/N2O ratios SO4 and NO3

Leaching Exchange products from animal manure and crop litter: C, N, P, S, K, and others Release of N, NO3, P, K, S, SO4 and other nutrients Chelates: Mn, Zn, Fe, Cu and others

Microbial biomass nutrient cycling

Weathering Minerals Adsorption

Leaching Unconfined aquifer

Solution ions: K, Mg, Ca, and others

Nutrients in underground water are recycled back with irrigation water (NO3, S, Ca, Mg, K, and others)

Leaching of bases, anions and carbonates such as K, Ca, Mg NO3 Cl, SO4 and HCO3

Irrigation water has some quantities of essential elements such as NO3 Cl, S, Ca, Mg, and K

Figure 1 Nutrient cycles of essential elements for crop production, showing their fate and transport in the environment (from Delgado and Follett, 2002).

120


Most agricultural systems are naturally deficient in N, which makes N inputs necessary to maximize yields, crop quality, and economic returns required to sustain viable operations. This is especially true for intensive irrigated systems with higher average yields than nonirrigated systems, particularly during times when crops are growing faster and have greater N uptake. Nitrogen inputs to agricultural systems are very important for the sustainability of these systems. A key positive feature of N inputs is their contribution to crop yields and crop quality, which ensure higher economic returns for farmers. Another positive feature of N inputs is that they reduce the need to cultivate low-productivity agricultural land, allowing those areas to be left alone and allowing farmers to cultivate areas more suitable for agricultural production. Nitrogen inputs also contribute to higher water use efficiencies (kg mm-1 ha-1), which are increasingly necessary for global sustainability as water resources in some regions become depleted. However, across any landscape system combination, any N application in excess of what is needed can increase the risk of negative effects on the environment. It is paramount to reduce the off-site transport of N from fields with sound management practices. In order to continue the efforts to minimize agriculture’s negative impacts on the environment, we need to continue developing and implementing best management practices for N at a field level. Even after N has left the boundaries of a field, there are other conservation efforts that can help identify areas of higher N transport (hot spots). Specifically, precision conservation techniques around fields and across water pathways and off-site management practices such as buffers, filter strips, riparian zones, sediment ponds, denitrification traps, irrigation and drainage ditches, and other management of natural areas within a watershed can help reduce reactive N transport across the landscape. For example, some researchers have proposed that we can even harvest N and reduce its transport across water bodies by using information about N dynamics to determine the best strategic placement of wetlands as a practice that can increase denitrification and removal of nitrates (NO3-N) from surface waters (Hey, 2002; Hey et al., 2005). We suggest that these nutrient management concepts and principles could potentially be used to reduce N transport in the environment. We propose that the best approach to reduce off-site N transport is to work at a field level, starting with a good conservation and nutrient management plan that reduces excessive N inputs. We believe that the application of an N trading concept could help increase the implementation of best management practices for N at the field level and expand management considerations to include the entire N cycle. Applying an N trading concept could also increase the development of precision conservation at a watershed level that could include strategic placement and management of nutrient farming devices such as denitrification traps and better management of irrigation and drainage ditches and wetlands that reduce off-site N transport.

A New GIS Nitrogen Trading Tool Concept for Conservation

121

Improvement of N management, including the use of precision conservation practices across agricultural systems worldwide, will be critical to the sustainability of agriculture, maximization of yields, and the conservation of our biosphere during the twenty-first century. These practices will become even more important in this century if we are to reduce the continual increase in nitrous oxide (N2O) emissions, which may contribute to global warming, and atmospheric reactive N deposition, which impacts the ecological balance in natural systems. The increasing demand for biofuels presents another reason for conservation-focused N management. Since most of the agriculturally viable land in the world is already being used to produce food for the current population (Baligar et al., 2001), the world population continues to increase, and biofuel cropping can compete for land area and water resources that are already being used for food production, and the sustainability and productivity of agricultural land are of utmost importance. Additionally, removal of crop residue may increase nitrate (NO3-N) leaching and N2O-N emissions (Delgado et al., 2010) and erosion (Lal, 1999). Considering the continued reports about the possibility of global warming, climate change, extreme weather events (droughts and floods), depletion of important aquifers in some of the most productive regions in the world, desertification, a rise in sea level, and other ecological events that may impact food production, sustainable practices and maximum production in all agroecosystems will be necessary to ensure future worldwide food security (Eggleston et al., 2006; Hatfield and Prueger, 2004; Houghton et al., 1992; Hu et al., 2005; Lal, 1995, 2000; Nearing et al., 2004). Increasing the sustainability and yield per unit area will also relieve the pressure to cultivate marginal lands and forested areas, pressure that is otherwise likely to increase with population. Several authors have reported on how excessive N applications that increase the potential for N cycle leaks can impact the quality of air, surface water, and groundwater (Follett and Walker, 1989; Follett et al., 1991). For example, excessive N applications increase the potential for NO3-N leaching losses, which can impact groundwater quality (Hallberg, 1989; JuergensGschwind, 1989). Anthropogenic N sources have been tied to losses of nitrogen that contribute to the Gulf of Mexico Hypoxia (Antweiler et al., 1995; Goolsby et al., 2001). Increased N inputs are also tied to increased emissions of trace gases such as N2O, which increase the potential for global warming (Eggleston et al., 2006; Houghton et al., 1992; Mosier et al., 1991). Other pathways that contribute to N losses include off-site surface transport (Bjorneberg, et al., 2002) and ammonia (NH3-N) volatilization (Peoples et al., 1995), which impact water and air quality, respectively. It has been shown that N management can improve the synchronization of N sources and sinks, but knowledge about how weather, the hydrologic cycle, irrigation, off-site factors, and cropping systems interact with the soil


122

Crops

Fertilizer

Manure Irrigation

Management

Soi

ls

ate

Clim

Off-site

Figure 2 Essential components of NO3-N Leaching Index (NLI) (from Shaffer and Delgado, 2002).

Reduced

Oxidized

Nitrification

NO

NO NO

NH+4

NH2OH

NO−2

[HNO]

NO−3

[X] N2 O Denitrification

N2

Figure 3 2002).

N 2O

NO

NO−2

NO−3

Diagram of nitrification and denitrification processes (from Mosier et al.,

N pools, N dynamics, and N fate and transport in a given landscape is invaluable (Delgado and Shaffer, 2008; Shaffer and Delgado, 2002) (Fig. 2). Basic management principles are often all that are necessary to minimize both NO3-N leaching (Meisinger and Delgado, 2002) and N2O emissions (Mosier et al., 2002) (Figs. 1 and 3). Delgado and Lemunyon (2006) reported that it is important for nutrient managers to continually seek education regarding nutrient management to stay current in the newest advances in technique and technology. Knowledge of these advances is critical to ensure good, effective management decisions (Delgado and Lemunyon, 2006). There are new tools that can be used to supply and integrate some of this information, which can help nutrient managers understand the potential for N loss savings resulting from the implementation of precision conservation management. New advances and practices such as controlled-release fertilizers, management zones, remote sensing, growing season in situ testing, cover crops, and limited irrigation are proven improvements to N


123

management that result in significantly reduced N losses. Trading systems may become significant considerations during the management implementation process in the future. In this chapter, we present the concept of using a GIS Nitrogen Trading Tool (NTT) approach to assess N management and conservation practices to reduce reactive N losses to the environment. We propose that a GIS NTT based on computer models can help identify where the higher N losses are occurring across a field and how much savings in N may be achievable in a given field to be traded in water and air quality markets.

2. Understanding the Nitrogen Cycle with Respect to Nitrogen Management and Trading Nitrogen management principles that can be used to increase nitrogen use efficiencies should be considered when evaluating the potential for increasing nitrogen trading. The NTT concept defined by Delgado et al. (2008c) assessed the differences in N losses between a new management scenario and a given baseline management practice. The NTT can conduct quick analysis about N management for the new scenario and baseline scenario using nitrogen and water budgets. Since implementation of a new N management practice for 1 year can impact soil nitrogen pools and increase the release of nitrogen long after its initial application, the differences to the baseline are evaluated over a long time (24 years). This long-term evaluation integrates any changes to nitrogen pools or N sequestration (Al-Sheikh et al., 2005) that could affect N dynamics. This could help ensure that the implementation of today’s practices and the potential for trading will not create negative effects 5 or 10 years later due to changes in nitrogen dynamics. Because nitrogen management will affect the N pools and dynamics, it is important to use a mass balance for N and water budgets to track inputs and outputs over the long term. It is important that the long-term evaluations take into account the interaction of management practices and field characteristics that consider the soil–crop–hydrologic cycle, which is site specific. To take advantage of best management practices that reduce N losses to the environment, we need to understand how the principles for nitrogen management can be used to reduce N losses and N transport to water bodies (Meisinger and Delgado, 2002; Randall et al., 2008) and/or to the atmosphere (Mosier et al., 2002). Additionally, an NTT that uses a mass balance analysis for nitrogen and water also helps to avoid simultaneously accounting for reductions in N inputs and losses. Nitrogen use efficiencies have been reported to be around 50% in general and as low as 33% for cereals (Baligar et al., 2001; Raun and Johnson, 1999). Baligar et al. (2001) discussed several different definitions

124


of nutrient use efficiencies, including the nutrient use efficiency ratio (Gerloff and Gabelman, 1983), physiological efficiency, agronomic efficiency, agrophysiological efficiency, and apparent recovery efficiency. Delgado (1998) and Delgado et al. (2001a) assessed the effects of best management practice implementation on system N use efficiency with a modeling approach. This modeling approach considered an N mass balance and how best management practices for N increased the N use efficiency, reduced NO3-N leaching losses, and mined NO3-N from underground waters. Evaluations of multiple cropping systems showed that the deeperrooted crops acted like vertical filter strips, recovering NO3-N from groundwater, as well as reducing NO3-N leaching (Delgado, 1998, 2001; Delgado et al., 2001a). Delgado et al. (2008c) proposed that a similar mass balance approach should be used to quantify the potential for savings in nitrogen that can be traded in water and air quality markets assuming the implementation of a determined set of management practices. This new nutrient trading concept may provide an additional factor for consideration by managers deciding what practices to implement to increase N use efficiencies. Several other researchers have reported on the potential to use environmental quality market credits to account for reductions of agricultural N losses and prevention of their transport into water bodies (Glebe, 2006; Greenhalch and Sauer, 2003; Hey, 2002; Hey et al., 2005; Ribaudo et al., 2005). However, we need to be realistic and consider that the dynamics of the N cycle make the quantification of these reductions in N losses difficult, especially when one considers interactions with the temporally and spatially variable hydrologic cycle, weather, soils, management, crop rotations, and other uncontrollable and isolated factors (such as thunderstorms), which may increase leaching and/or denitrification (Delgado, 2002). Delgado et al. (2008c) and Gross et al. (2008) described the potential use of quick, new NTTs to help quantify the effect of conservation practices and N management on reactive N losses to the environment. The new concept of the NTT was defined within the context of the N cycle and considers an N mass balance approach for the cropping systems (Delgado et al., 2008c). The NTT difference in reactive N losses (NTTDNLreac) draws comparisons between a baseline and new management scenarios. A positive NTT-DNLreac means that a new N management practice increases the savings in reactive N, while a negative number means that there are no savings in reactive N. In other words, a positive number means that there is potential to trade these savings, while a negative number means that there is no potential for trade. The NTT-DNLreac can be thought of as a bank account balance. A positive number means that there is N in the bank for trade and a negative number means that there is no N in the bank to trade. The new GIS concept that we are presenting in this chapter can be applied across the field and considers spatial and temporal variability.


125

The new N trading concept, a stand-alone NTT, and a new Internet prototype of an NTT were developed by the Natural Resources Conservation Service (NRCS), in cooperation with the Agriculture Research Service Soil Plant Nutrient Research Unit (ARS-SPNR) (Delgado et al., 2008c; Gross et al., 2008) (Figs. 4 and 5). Both the web-based version and stand-alone prototype allow users of this new technology to quickly determine how many potential N credits their farming operations can generate. The new Internet and stand-alone NTT are the only tools with the level of rigor to allow producers to calculate potential N credits for air and water quality markets as a function of conservation measure implementation. Environmental aggregators, brokers, and water quality traders may also use these tools (Delgado et al., 2008c; EPA-WQTN, 2007). The development of the N trading concept and the NTT is part of the national agreement between the USDA-NRCS and the EPA Office of Water to participate in potential water-quality trading programs (EPA-WQTN, 2007). We suggest that such a tool could be used for air quality markets and for direct and indirect carbon sequestration equivalent markets. Further, we propose in this chapter that the new NTT-GIS can be used to quickly

Figure 4 Nitrogen Trading Tool: the web-based prototype (from Gross et al., 2008).

126


Figure 5 A stand-alone version of the NTT prototype (from Delgado et al., 2008c).

identify the scenario that shows the greatest potential to maximize field-level savings in reactive N for environmental conservation and to earn N credits for trade. Delgado and Follett (2002) reported that carbon management should also be a fundamental part of any nutrient management plan, integrating N and carbon input management with data about existing soil contents of these elements (Fig. 6). They reported that nutrient managers who manage in a way that increases the soil carbon content need to adjust for the greater N cycling and soil N mineralization potential by reducing N inputs (Fig. 6). In other words, we need to account for management practices that will increase soil organic matter (SOM) and N cycling by adjusting future N recommendations according to higher N mineralization rates (Delgado and Follett, 2002) (Fig. 6). Delgado et al. (2008c) considered the complexity of the N cycle when they proposed a conceptual framework for nitrogen trading in which the effect of management practices is assessed using a computer model simulating a long period of time with an N mass balance approach. To avoid double accounting for N inputs, the N mass balance approach proposed by

127


Effects of organic carbon on nutrient cycling and productivity

Practices that contribute to loss of organic carbon

Higher offsite transport

Higher fertilizer inputs

Lower flow of organic C contributes to lower nutrient cycling

Higher nutrient leaching losses

Productivity aeration aggregates drainage WHC CEC MB Reduce Improve Lower offsite transport

Practices that contribute to reduced loss of organic carbon

Lower fertilizer inputs

Higher flow of organic C contributes to higher nutrient cycling

Lower nutrient leaching losses

Figure 6 Potential organic C contribution to nitrogen cycling (þ organic C; þ N cycling; N fertilizer); use efficiency (þ organic C; þ N use efficiency); nitrogen leaching (þ organic C; N leaching); and nitrogen losses, under best management practices (from Delgado and Follett, 2002).

Delgado et al. (2008c) accounts for all N inputs and N outputs, as well as N transformations (e.g., N releases, sequestration, mineralization, etc.).

2.1. Understanding the relationships between the soil–crop–hydrologic cycle and nitrogen trading 2.1.1. Soil–crop–hydrologic cycle One key principle for maximizing nitrogen trading is to understand the relationship between N management practices and the soil–crop–hydrologic cycle for a given region. This understanding could inform management decisions, and therefore help to avoid excess N applications while maintaining high crop yields, and help to increase the synchronization of applied N with crop N uptake sinks. Site-specific soil textures, hydrological properties, and crop water use all affect the soil water content and aeration and alter N dynamics (e.g., mineralization rates) and pathways for N losses (e.g., denitrification and nitrate leaching) (Meisinger and Delgado, 2002). One example of the relationships between the hydrologic and N cycles is that high precipitation or irrigations can create water-logged conditions that favor potential losses of N due to denitrification (Meisinger and Randall, 1991). Other events that may contribute to water-logged conditions include the seasonal increase in water table. Meisinger and Randall (1991) summarized these relationships and reported that the denitrification potential for a well-drained soil with 1% SOM content on a semidry system will be about

128


3%. For the same case scenario under a humid or irrigated system, the denitrification potential increases by three times to about 9%. If the percentage of SOM is higher or if manure is applied, the denitrification potential will be higher (Meisinger and Randall, 1991). These denitrification losses are also driven by the lack of oxygen in the soil (Fig. 3). Williams and Kissel (1991) reported on the interaction of soil hydrology and nitrogen losses. They reported that the threshold precipitation is less than 406 mm for dryland systems where NO3-N leaching is zero or minimal. Evans et al. (1994) and Westfall et al. (1996) reported a similar relationship between the hydrologic cycle and NO3-N leaching. However, when there are high precipitation events, NO3-N leaching is much larger in soils with sandier and coarser texture that have hydrological properties that are conducive to a faster movement of water out of the root zone (Delgado et al., 2001a; Follett and Walker, 1989). Williams and Kissel (1991) developed an index that incorporated these relationships between soil hydrological properties, weather, and water leaching. Another pathway that contributes to N losses and is closely related to the hydrologic cycle is surface runoff. Surface transport caused by irrigation and/or precipitation is one means by which soil particles, SOM, organic N, and other N that may be bound to clay particles or dissolve in water can be transferred off-site and lost from the system. By understanding this property of the soil–crop–hydrologic cycle, nutrient managers could anticipate when periods of higher denitrification, leaching and/or erosion potential may occur for a given landscape crop combination, and implement conservation management practices that would reduce N losses. Results from Williams and Kissel (1991) were adapted and presented in Figs. 7–9 using the same soil nitrate N concentrations across Ames, Iowa, Brookings, North Dakota, and Caldwell County, Kentucky. The adapted data from Williams and Kissel (1991) for the four major hydrologic groups in a high precipitation site such as Ames, Iowa, show that the potential for leaching is much higher in Ames than in a dryland region site such as Brookings, North Dakota (Figs. 7 and 8). These adapted data from Williams and Kissel (1991) are in agreement with the NTT results from Delgado et al. (2008c). They show that practices that reduce nitrate leaching will be advantageous for trading nitrogen on coarse texture sandier systems that more readily leach higher quantities of nitrate, especially under areas with higher precipitation or irrigation. Seasonal timing of precipitation is also important to consider when managing nitrogen. For example, early or winter precipitation will help lead to increased leaching potential of available NO3-N if there are no crops growing that can use water or uptake nitrogen, assuming there is nitrate available to leach in the soil profile. Figure 9 shows that the NO3-N leaching losses were minimal when the precipitation was mainly occurring during the crop growing season (15 kg NO3-N ha1 y1). The same amount of precipitation, with a higher proportion occurring before planting,

129


125 A

Leacheate (mm)

100 75 50 25 0 B

Run-off (mm)

100 75 50 25

Nitrate leaching (kg NO3-N/ha)

0 C

A = 72 kg NO3-N ha−1 y−1

4 B = 32 kg NO3-N ha−1 y−1

3 C = 16 kg NO3-N ha−1 y−1

2 D = 4 kg NO3-N ha−1 y−1

1 0 Jan Feb

Mar

Apr

May

Group A

Jun

Jul

Group B

Aug

Sep

Group C

Oct

Nov

Dec

Group D

Figure 7 Leaching (A) and run-off (B) water volume (A and B adapted from Williams and Kissel, 1991). Estimated NO3-N leaching (C) assuming same NO3-N l1 leachate concentrations for an irrigated corn grown in the main hydrologic soil groups (A, B, C, and D) under a rain-fed humid climate (Ames, Iowa) (from Delgado, 2004).


130

125

A

Leacheate (mm)

100 75 50 25 0 B

Run-off (mm)

100 75 50 25


0 C A = 4kg NO3-N ha−1 y−1 4 B = 0.0kg NO3-N ha−1 y−1 3 C = 0.0kg NO3-N ha−1 y−1 2 D = 0.0kg NO3-N ha−1 y−1 1 0

Jan

Feb

Mar

Apr

Group A

May Jun Group B

Jul

Aug Sep Group C

Oct

Nov Dec

Group D

Figure 8 Leaching (A) and run-off (B) water volume (A and B adapted from Williams and Kissel, 1991). Estimated NO3-N leaching (C) assuming same mg NO3-N l1 leachate concentrations for an irrigated corn grown in the main hydrologic soil groups (A, B, C, and D) under a rain-fed dry climate (Brookings, North Dakota) (from Delgado, 2004).

131


Growing season 4/15–9/15

A

Precipitation or leaching (mm)

85 kg NO3-N ha−1 y−1 150

30

100

20

50

10

0

0


40

200

B

150

30

100

20

50

10

0


Precipitation or leaching (mm)

15 kg NO3-N ha−1 y−1

0 Jan Feb Mar Apr May Jun

Jul Aug Sep Oct Nov Dec

Precipitation

kg NO3-N/ha

Leaching

Figure 9 Effect of high (A) and low (B) in-season precipitation during corn growing season at Caldwell County, Kentucky (adapted from Williams and Kissel, 1991).

significantly increased the leaching potential by about six times if there was nitrate available to leach (85 kg NO3-N ha1 y1) (Fig. 9). Reducing the available NO3-N to leach can reduce soil susceptibility to leaching during the winter months (Meisinger and Delgado, 2002). 2.1.2. Limited irrigation Limited irrigation can improve water use efficiency while maintaining yields and a viable cropping system (Hu et al., 2005). Delgado et al. (2007) reported that cover crops with limited irrigation can save water, reduce

132


nitrate leaching, and even increase yields of subsequent potato crops. Incorporating a viable summer cover crop that can be grown with half of the water requirement of a traditional potato or winter wheat crop can save a significant amount of water. Figure 10 and Table 1 show the positive effects of using cover crops for hay and/or green manure across a region of south central Colorado. Limited irrigation has potential as a management tool to maintain viable cropping systems while increasing the potential for trading N and carbon sequestration equivalents (Delgado et al., 2008b) (Fig. 11, Table 1).

2.2. Inputs The average N use efficiencies are reported to be about 50% and as low as 33% for grains (Baligar et al., 2001; Raun and Johnson, 1999); however, these N use efficiencies can be around 30% for irrigated shallow-rooted crops grown on sandy coarse soils (Delgado, 2001; Delgado et al., 2001a,b) and lower than 30% when excessive N (750 to 1900 kg N ha 1) is applied (Zhu and Chen, 2002; Zhang et al., 1996). Nonetheless, for most agricultural systems N inputs are needed to maintain agricultural production, maximize yields, and quality and to supply the N that is removed with crop harvesting. Organic, inorganic, and biological (N fixation) sources can be used as N inputs for agricultural systems and may be applied using many different techniques. 2.2.1. Amount of N inputs Management of N inputs can be done through a mass balance approach, in which all N sinks are considered along with crop uptake. Any N applied in excess of the crop uptake will increase the N available for leaching and the overall potential for N losses (Fig. 12). In order to increase the N use efficiency for the applied N, all N sources already present and available for uptake should be subtracted from the needed N. Examples of N sources to subtract include: residual soil NO3-N that is available within the root zone or at least for the top surface foot, N that will be released from mineralization of SOM during the growing season, background NO3-N applied with irrigation water while plant N uptake is active, and N that will be mineralized from the previous crop residue. Factoring the residue N from the previous crop is particularly important if the crop residue is from a leguminous crop that was incorporated into the soil or from a cover crop or vegetable crop with low carbon to nitrogen ratios. A nutrient manager can calculate the needed N using an efficiency factor that accounts for management practices. Some states and regions have developed N uptake formulae that include efficiency factors to supplement data about N sink from crops. For example, in Colorado, the calculation of the appropriate amount of N fertilizer application to corn is based

kg N year−1

20

40

60 Legend

N

Rivers Vegetables Potato Meadow Grain Alfalfa SLV_soils Lakes Highways Cities

80 0

N 600 1200

2400 Unknown units

W

E S

Farmer with 5 center irrigated pivots (13,400kg N year−1 reactive N) Aggregator with 94 center irrigated pivots (317,000kg N year−1 reactive N) 0 8,750 17,500

35,000

52,500

70,000 Meters

Figure 10 A stand-alone NTT-GIS prototype can be used to quickly evaluate the effects of management practices on total reactive N losses and the resultant potential to trade across regions (hypothetical example; adapted from Delgado et al., 2008b).


134

Table 1 Assessment of potential reductions in nitrous oxide emissions (N2O-N), in reactive N losses, and carbon sequestration equivalents

State

Virginia Virginia Ohio

BMP

a

Add legume Improved NM Improved MM-Inc Ohio Improved MM-Spring Colorado Add SCC-LI Colorado Improved NM-CC

Carbon sequestration equivalents

N2O creditsb (kg N)

Total N creditsb (kg N)

Directc (kg C)

Indirectd Total (kg C) (kg C)

240 70 350

700 900 7600

32,000 9200 46,900

500 800 8800

32,500 10,000 55,700

430

6200

56,500

6800

63,300

100 80

5300 5000

13,700 10,200

5100 5000

18,800 15,200

a

NM, nutrient management; MM-Inc, manure management incorporated only spring application; MM-Spring, manure management only spring application; SCC-LI, summer cover crop with limited irrigation; CC, cover crop. b The baseline for Ohio was a manure application in spring before corn planting (249 kg N ha 1) and fall after soybean harvesting (also 249 kg N ha 1). The baseline for Virginia was a continued conventional corn–corn rotation at 224 kg N ha 1. The baseline for Colorado was a continued potato–potato rotation at 269 kg N ha 1. The soils type across the 100 ha were loam, loamy fine sand and loam for Ohio, Virginia and Colorado, respectively. c Direct carbon sequestration equivalents ðDDCO2 CseN2 O Þ were calculated for 100 ha by using the equation: DDCO2 CseN2 O ¼ DN2 O N 310 0:2727 1:571. d Indirect carbon sequestration equivalents ðDICO2 CseN2 O Þ were calculated for 100 ha by using the equation: DICO2 CseN2 O ¼ ½ððDNO3 N þ DNst N þ DNer Þ 0:0075 310 1:571Þ þðDNH3 N 0:01 310 1:571Þ 0:2727. The estimated potential for direct, indirect, and total carbon sequestration equivalents calculated using the nitrogen trading tool are also presented here for different best management practices (BMP).

on the Mortvedt et al. (1996) algorithm for N fertilizer applications: [N rate ¼ 35 þ (1.2 EY) (8 soil ppm NO3-N) (0.14 EY OM) (other N credits)], where EY is expected yield and OM is organic matter. The basic principle is to apply the correct amount of N, so as to avoid the excessive application of N. On average, using an N budget approach, whether by considering the N sinks and sources or using already calibrated formulas such as those developed by Mortvedt et al. (1996), will help increase the N use efficiencies by prescribing N inputs more closely aligned with the N needs of the system. 2.2.2. Types of N inputs Several sources of inorganic N fertilizer are available. Among the most important are ammonia (NH3), nitrogen solutions (combinations of ammonium nitrate (NH4NO3), urea, and water), ammonium nitrate, urea,

Direct carbon sequestration potential (Kg C ha−1) 44–69 >69–93 >93–125 >125–275

Legend

N

Rivers Vegetables Potato Meadow Grain Alfalfa SLV_soils Lakes Highways Cities

0 8,750 17,500 35,000

52,500

70,000 Meters

Figure 11 When released, the stand-alone NTT-GIS will be able to quickly evaluate the effects of management practices on direct carbon sequestration equivalents (kg C ha 1).


136

Sum residual and leached NO3-N (kg NO3-N/ha)

A 350

y = 1.1295x + 59.505

300

R2 = 0.8322

250 200 150 100 50 0 0

50

100

B

150 200 Excess N fertilizer (kg N/ha)

250

300

350


350 y = 0.9546x + 42.778

300

R2 = 0.8999

250 200 150 100 50 0 0


C

50

100


250

300

350

350 y = 1.0189x + 88.374

300

R2 = 0.9975

250 200 150 100 50 0 0

50

100


250

300

350

Figure 12 Correlation between excess N fertilizer and sum of residual soil NO3-N and NO3-N leached for low (A), medium (B), and high (C) productivity management zones. Residual and leached fertilizer was simulated with NLEAP (data adapted from Delgado and Bausch, 2005; Delgado et al., 2005). Excess N use fertilizer was defined as: N fertilizer (N uptake by crop N uptake by control or zero fertilizer).

ammonium sulfate, and several other sources, such as ammonium phosphates and calcium nitrate (Boswell et al., 1985). The source of N is important to consider when managing N, and careful choice of source can be used to increase the efficiency of a given system. For example, NO3-N sources should not be applied to systems that will be submerged in water, such as rice fields, due to the high potential for N losses via denitrification. In a submerged system an NH4-N source should be used


137

instead. The NO3-N sources can also increase the leaching potential on sandy irrigated soils; therefore, application of urea or NH4-N sources, rather than NO3-N application, can likely increase the nitrogen use efficiency and reduce the NO3-N leaching potential for these systems. Other N sources, such as controlled-release fertilizers, can be used to correlate the timing of N release with times of greater N uptake (Shoji and Gandeza, 1992). Controlled-release fertilizers contain the N source inside a capsule and release the N slowly to correspond better with periods of crop N uptake, thereby reducing the time that the N is susceptible to losses (Amans and Slangen, 1994; Mikkelsen et al., 1994; Rauch and Murakami, 1994; Shoji and Gandeza, 1992; Shoji and Kanno, 1994; Wang and Alva, 1996). Several field studies have shown that, when using controlled-release fertilizer, nutrient managers can apply 50% of the traditional amount of fertilizer and still produce the same yields as with traditional fertilizer practices (Shoji and Gandeza, 1992; Shoji et al., 2001). In other words, the fertilizer use efficiency of the controlled release fertilizer is much higher than that achieved using traditional fertilizer practices, helping to reduce agricultural N2O emissions (Delgado and Mosier, 1996; Shoji and Gandeza, 1992, Shoji et al., 2001). Nitrification inhibitors (NI) can help increase N use efficiencies by slowing down the nitrification of NH4-N to NO3-N (Freney et al., 1992; Yadvinder-Singh et al., 1994). The NH4-N is less susceptible to leaching, binds more to the clay particles, and is not affected by denitrification. Nitrification inhibitors also reduce the emissions of N2O (Bronson and Mosier, 1993; Delgado and Mosier, 1996) and have been reported to reduce NO3-N leaching (Owens, 1987; Timmons, 1984). Organic N sources such as manure can also be used to provide N to agricultural crops. Significant amounts of manure N can be cycled to the subsequent crops (Eghball et al., 2002). Eghball et al. (2002) reported that composted manure can cycle 18% of its N content during the first year, while cattle feedlot manure can cycle 30% of the N content. They reported that the total N available from feedlot manure is double the total N available from composted manure (Davis et al., 2002; Eghball et al., 2002). Kirchmann and Bergstrom (2001) reported that N management is more important than N source in terms of controlling NO3-N leaching losses when organic farming practices are compared to traditional farming practices. In either case, overapplication of N will contribute to increased NO3N leaching problems. They concluded that reduction in NO3-N leaching was not as much a question of organic versus conventional farming as it was a question of adequate management practices. It is very important to practice effective N management with manure applications to avoid environmental degradation that can result from excessive application. However, Delgado et al. (2010) reanalyzed unique 15N crop residue exchange studies that used the Delgado et al. (2004) method and reported that N losses from organic crop residue are much lower (about 13%) than N

138


losses from inorganic N fertilizer (about 31); these results conflict with the Kirchmann and Bergstrom (2001) study. Delgado et al. (2010) also conducted DAYCENT simulation analysis to evaluate the N losses from inorganic N fertilizer versus crop residue, and they found that the NO3N leaching losses and N2O emissions were much lower from crop residues than from inorganic N fertilizer inputs. 2.2.3. Method and time of N inputs The method by which N is applied, whether or not the N is applied in split (multiple) applications, the equipment used for application, and the location of application are important management factors that can be manipulated to increase N use efficiencies. It is important that we closely match the N inputs with N sinks (Meisinger and Delgado, 2002). The time of N application can be adjusted in order to reduce the time that the N is susceptible to losses if the periods of N availability are synchronized with the periods of more active rooting. For example, 15N isotopic studies show that spring N applications are used more efficiently than fall applications (Delgado et al., 1996). The spring 15N isotopic fertilization recoveries in plants and soil were 60 and 71% for urea and NH4NO3, respectively, in contrast to the 42 and 57% recoveries from fall applications (Delgado et al., 1996). Several scientists have reported on the benefits of splitting N applications into preplant, side-dress, and fertigations in order to match greatest N availability with the periods of greatest N sinks (Gunasena and Harris, 1968; Oberle and Keeney, 1990; Russelle et al., 1981; Sowers et al., 1994; Stanford and Legg, 1984; Westermann and Kleinkopf, 1985). Split N applications that reduce the amount of total N applied and increase the number of N applications will improve N use efficiency and crop yield while reducing the potential for N losses (Alva and Paramasivam, 1998). Good water management practices are important to increase N use efficiencies and reduce NO3-N leaching losses to the environment (Meisinger and Delgado, 2002). There are best management practices that can help minimize NO3-N leaching losses (Alva and Paramasivam, 1998; Hergert, 1986; Smika et al., 1977; Thompson and Doerge, 1996a,b; Westermann et al., 1988). Management systems under sprinkler irrigation that use fertigations can contribute to higher N use efficiencies, especially for shallower-rooted cropping systems and vegetables that are grown in sandier coarse textured soils with a lower capacity to hold water (Westermann et al., 1988). For coarser soils, a high number of fertigations (5–8) help increase N use efficiencies (Doerge et al., 1991). The application of N below the surface can increase N use efficiencies compared to broadcast methods, especially when NH3-N volatilization is reduced (Meisinger and Randall, 1991; Peoples et al., 1995).


139

2.2.4. Advanced technologies New technologies such as precision farming techniques have the potential to improve N use efficiencies (Gotway et al., 1996; Hergert et al., 1996; Redulla et al., 1996). Variable rate maps and/or management zones can be used to improve the accuracy of N fertilizer applications (Delgado and Bausch, 2005; Delgado et al., 2005; Ferguson et al., 1996; Khosla et al., 2002). New technologies can help improve N management by providing information to nutrient managers about the potential N uptake and N status throughout the growing season. Nutrient managers can use this information to develop N management plans that better synchronize N inputs with crop N sinks from preplanting through harvest. Some of these technologies can provide spatial and temporal information during the growing season and help nutrient managers identify areas that are deficient or overfertilized with nitrogen. Soil samples can be collected using a Global Position System (GPS), then analyzed in a laboratory to provide information about the spatial variability of residual soil inorganic N, SOM, and mineralization potential. There are other technologies such as remote sensing, that can be used to instantly provide information about the N status of large field areas. Site-specific management zones (SSMZ) can be used to manage N based on yield history, soil color from aerial photographs, topography, and the producer’s past management experiences (Fleming et al., 1999). SSMZ can be used to develop an N management plan that considers the variability in N sinks using realistic yields across the field. Additionally, management zones integrate the potential N from SOM, residual NO3-N and other sources that are representative of each zone instead of using a yield average. Recent research has shown that these new technologies can help increase N use efficiencies (Khosla et al., 2002) and reduce NO3-N leaching (Delgado et al., 2005). The lower yield, sandier, coarser areas, which received greater N applications, had greater leaching losses because the N sink was much lower than areas with higher yields. Applications of N according to management zones or spatial variability of N sinks can increase agronomic N use efficiencies and reduce losses of N to the environment by minimizing NO3-N leaching (Delgado et al., 2005; Khosla et al., 2002). We evaluated the data presented by Delgado et al. (2005) and Delgado and Bausch (2005). We estimated excessive N applications with the following formula: ENFA ¼ ½NFA ðCU CUWFÞ ð1Þ where ENFA is excessive N fertilizer application, NFA is N fertilizer applied, CU is aboveground crop uptake at the given fertilizer rate, and CUWF is aboveground crop N uptake by plant without fertilizer. This definition of excessive N fertilizer, based on the net N uptake from the added N fertilizer, was correlated with the sum of simulated residual soil


140

NO3-N and NO3-N leaching by zones (P < 0.01). The areas of the fields with higher sand content (low-productivity zones) had lower residual soil NO3-N content (Fig. 13). It is clear that any N applications greater than the crop N uptake will increase NO3-N leaching and NO3-N available to leach across all of the zones (Fig. 12). Figure 12 is in agreement with Andraski et al. (2000), who defined excessive N fertilizer application as the applied N fertilizer rate minus the economically optimum N fertilizer rates, correlated with soil water NO3-N concentrations. Our definition of excessive N application calculates the N that is available for loss to the environment, accounting for a site-specific N uptake of zero N fertilizer, and assessing all sources of N except N inputs from fertilizer or manure. Figure 12 is also in agreement with Pratt (1979), who reported that we cannot completely eliminate NO3-N leaching losses. Delgado et al. (2006, 2008a) N index ranks NO3-N leaching losses from the system as very low (28 kg N ha 1 y 1), low (> 28 56 kg N ha 1 y 1), medium (> 56 112 kg N ha 1 y 1), high (> 112 168 kg N ha 1 y 1), and very high (>168 kg N ha 1 y 1). Delgado et al. (2005) and Delgado and Bausch (2005) showed that using SSMZ and remote sensing in conjunction with reducing excessive N fertilizer applications can significantly reduce NO3-N leaching losses. Delgado et al. (2005) concluded that spatially variable N management based on productivity zones produces less NO3-N leaching than uniform strategies while maintaining maximum yield. They estimated that we can cut NO3-N leaching losses by 25% during the first year by using an SSMZbased nutrient management plan. 250 r2 = 0.55 Soil NO3-N kg N/ha

200

150

100

50 0 60

70

80 Sand (%)

90

100

Figure 13 Correlation between the residual soil NO3-N in the top 1.5 m of soil with the respective sand content at each site during the 2000 growing season (from Delgado and Bausch, 2005).


141

Remote-sensing techniques can be used to monitor spectral reflectance to determine crop N status, including deficiency levels that may reduce yields (Al-Abbas et al., 1974; Stanhill et al., 1972; Thomas and Gausman, 1977). Remote sensing has allowed the development of reflectance indices used to monitor N status during the growing season, such as the N Reflectance Index (NRI) by Bausch and Duke (1996), and the Normalized Difference Vegetation Index (NDVI) by Tucker (1979) and Wood et al. (1999). These techniques and indices can quickly provide in situ information to help determine the need for N applications (Raun and Schepers, 2008). These N indices and remote-sensing techniques have allowed us to determine spatially variable N status across fields (Bausch et al., 1996; Blackmer et al., 1996; Franzen et al., 1999; McMurtrey et al., 1994; Raun and Schepers, 2008; Scharf et al., 2002). For example, crop N information gathered with remote sensing was used to cut N applications to 50% of traditional application rates (Bausch and Delgado, 2003), reducing NO3-N leaching losses by 47% (Delgado and Bausch, 2005). Other relatively new tools include chlorophyll meters and portable electrodes that can help monitor N levels during the growing season to further increase the N use efficiency through split N applications (Follett et al., 1992; Schepers et al., 1992a,b; Turner and Jund, 1991). Chlorophyll readings can be compared with N application rates to identify areas requiring additional N applications (Schepers et al., 1992a,b). Delgado et al. (2001b) reported a correlation between the leaf chlorophyll readings and potato tuber yield and quality. These tools have the potential to be used to determine N status and the need for N fertilizer applications, especially under irrigated systems. Another relatively new method is the use of field test strips to assess N status by determining sap NO3-N concentration for vegetables (Prasad and Spiers, 1984; Scaife and Stevens, 1983; Williams and Maier, 1990) and small grains (Papastylianou, 1989). Portable NO3-N ion-selective instruments can also be used to measure sap NO3-N concentrations for vegetables (Errebhi et al., 1998; Hartz et al., 1994; Kubota et al., 1996, 1997; Westcott et al., 1993) and winter cover crops (Delgado and Follett, 1998). Collecting plant samples for laboratory testing is a more traditional method for determining N status. This approach may require additional time compared to remote-sensing techniques, chlorophyll meters, portable electrodes, and field strips, because of the time needed to run the samples in the laboratory to get a recommendation about N status. Laboratory results can also be combined with SSMZs and precision farming techniques if the samples are collected using Global Position Systems. An example of laboratory-based tissue analysis is the potato petiole NO3-N test (King et al., 1999). The presidedress soil NO3-N test (PSNT) can also be used to monitor crop N status (Bundy and Meisinger, 1994). The PSNT is commonly used in the Northern Corn Belt and the

142


northeastern United States to assess the available soil NO3-N pool to identify if N levels are sufficient and/or to provide a basis for sidedress fertilizer N recommendations (Bundy and Meisinger, 1994). The PSNT can help increase N use efficiencies and lower NO3-N leaching potential (Durieux et al., 1995; Guillard et al., 1999). 2.2.5. Models and index An N budget based on an estimation of the percentage of applied N taken by the crop could be used to conduct a quick assessment of the potential for N losses (Bock and Hergert, 1991). However, with the Bock and Hergert (1991) N use efficiency index, there is no information about what may happen to the N that is not absorbed by the crop. Shaffer and Delgado (2002) discussed advantages and disadvantages of several Nitrogen Indexes that can be used to assess N management. A Nitrogen Index that considers N losses to the environment could potentially be used to conduct an assessment of how N management practices are affecting N losses (Delgado et al., 2006, 2008a). This new qualitative/quantitative N index can be joined to GIS to discern practices that have very low, low, and medium potential for N losses from practices that have high and very high potential risk for these losses (De Paz et al., 2008). Although N indexes could be used to conduct quick assessments of N losses, an NTT requires a more robust approach such as the use of an N model that can integrate detailed layers of information about soil–crop– hydrologic systems to assess losses of nitrogen from the nitrogen cycle (Delgado et al., 2008c). There are several national and international models that can be used to assess N losses to the environment. Examples of these models include the Nitrate Leaching and Economic Analysis Package (NLEAP) (Delgado et al., 1998; Shaffer et al., 1991), the Crop Estimation through Resource and Environmental Synthesis (CERES) (Ritchie et al., 1985), Erosion Productivity Impact Calculator (EPIC) (Williams et al., 1983), Nitrogen Tillage Residue Management Model (NTRM) (Shaffer and Larson, 1987), Root Zone Water Quality Model, RZWQM (Shaffer et al., 2000), LEACHM (Wagenet and Hutson, 1989), and the Great Plains Framework for Agricultural Resource Management (GPFARM) (Ascough et al., 2001). For additional information on other national and international models that simulate N dynamics and transport, see Shaffer et al. (2001). The initial prototype of the NTT used NLEAP (Delgado et al., 2008c; Gross et al., 2008), but if nitrogen trading markets are more widely implemented throughout the world, and/or nitrogen trading is integrated with the trading of potential carbon sequestration credits, it is possible that we could see a series of other NTTs developed in the near future for national and international users.


143

2.2.6. Identifying and managing spatial and temporal variability There are new advances in software that can be used to identify spatial variability (Berry, 2003a,b, 2007a,b). Recent advances in current geospatial research have been refocusing on data structure and analysis (Berry, 2007a). Delgado and Berry (2008) reported on how to identify spatial patterns and to manage spatial variability with precision conservation to reduce environmental impacts. Watershed models such as the Agricultural Non-Point Source Pollution (AGNPS) model (Young et al., 1987) and the Soil and Water Assessment Tool (SWAT) model (Arnold et al., 1993) can be used to assess erosion losses. These models are also being used to assess nutrient losses. The assessment of chemical movement, runoff, and erosion was also conducted using the Agricultural Management Systems (CREAMS) (Smith and Williams, 1980). Renschler and Lee (2005) used three models and GIS to evaluate the effects of best management practices. The models used were the Water Erosion Prediction Project (WEPP), the Geospatial interface for WEPP (GeoWEPP), and SWAT. Bonilla et al. (2007) used the Precision Agricultural-Landscape Modeling System (PALMS) and reported that PALMS can evaluate the effects of local soil properties and microtopography on changes in soil detachment and deposition across short distances and has the capability to quantify a series of spatial and temporal parameters. Modeling can be used to assess spatial erosion and N losses across the environment. There is also potential to use the NLEAP GIS 4.2 prototype to assess spatial N losses at the field level and/or nitrogen trading at the field level (Delgado et al., 2008b,c). For additional details about precision conservation and identifying and managing spatial and temporal variability, see Delgado and Berry (2008). 2.2.7. Rotation of crops Nitrogen management can be improved with crop rotations and more efficient crop varieties. Deeper-rooted crops can be rotated into shallower-rooted systems to increase the N use efficiency of the system. The deeper-rooted crops recover NO3-N from groundwater, minimizing the net NO3-N leaching from the system and contributing to water conservation (Delgado, 1998, 2001). Deeper root depth was correlated with less NO3-N leaching, greater NO3-N mining, and higher N use efficiencies (Delgado, 1998, 2001; Delgado et al., 2006). Some researchers have reported on the potential of winter cover crops to reduce NO3-N leaching (Delgado, 1998; Meisinger et al., 1991; Shipley et al., 1992). The inclusion of winter cover crops in rotations can increase system N use efficiencies, not only by recovering N from the previous crop but also by reducing N losses from the next crop (Delgado, 1998, 2001).

144


Scavenger cover crops can increase N cycling by increasing the N sink during the fallow period (Delgado, 1998; Meisinger et al., 1991; Shipley et al., 1992). Delgado et al. (2007) reported that summer cover crops with limited irrigation can increase the N sink during the fallow season, significantly increase N use efficiency, and improve yield and quality of the following crop. Multiple crops per year can also involve grasses, which may be harvested multiple times and which may help increase the potential for N trading. Adding a legume to the crop rotation can further increase the N use efficiency of the systems and reduce N losses (Meisinger and Delgado, 2002; Randall et al., 2008). Because leguminous crops can fix N from the atmosphere they require lower or zero N inputs, which, combined with the residue N cycling to the following crop, reduces NO3-N leaching potential even more (Kanwar et al., 1997; Randall et al., 1997). These studies show the potential for NO3-N leaching reduction in tile systems when the leguminous crops are included in grain rotations. In Virginia, this practice increased the potential for N trading over the baseline, with the potential to trade N (Delgado et al., 2008c). The savings in N2O were up to 4 kg N ha 1, which generated the potential to trade 500 kg C ha 1 as carbon sequestration equivalents (Delgado et al., 2008c; Lal et al., 2009). It is clear that quantification of N losses to the environment is difficult; however, we can use isotopic 15N techniques to assess N losses. Delgado et al. (2004) developed a crop residue exchange method to assess N cycling, fate and losses from crop residues on a large plot scale that was used for cover crop residue exchange studies in Colorado and the Pacific Northwest (Collins et al., 2007). The results from these studies show that the 72 and 58% N recovered (soil and plant) from fertilizer in Colorado and the Pacific Northwest, respectively, were much lower than the 85 and 95% recovered from crop residue. These cover crop studies are important in that they point out that N losses from fertilizer are two times greater than N losses from crop residues (Delgado et al., 2004, 2007). In other words, cover crops not only increase the system N use efficiency and NO3-N mining from underground water, but also increase N cycling to the subsequent crops, leading to higher N use efficiencies and fewer N losses to the environment than fertilizer inputs. Delgado et al. (2010) reported that these unique 15N crop residue exchange studies and simulations conducted with the DAYCENT model showed that the N losses from crop residue are much lower than those from inorganic N fertilizer, including lower emissions of N2O and NO3-N leaching. Delgado et al. (2010) recommended the use of lower coefficients for N2O emissions from crop residue, especially if they have a high C/N ratio (>30). Delgado et al. (2007) reported that summer cover crops with limited irrigation are being used by farmers in Colorado. If farmers were to implement a summer cover crop with limited irrigation program more widely


145

than is done currently, the potential savings in reactive N to the environment across this region could be as much as 300,000 kg N y 1 for approximately every 94 center irrigated pivots (about 60 kg N ha 1 y 1). The savings in N that would then be available to trade would also generate 860,000 kg C sequestration equivalents due to direct and indirect reductions in emissions of N2O (Figs. 10 and 11, Table 1). Results from these studies are in agreement with the Al-Sheikh et al. (2005) report that increasing rotation of deep-rooted crops and incorporation of crop residue increase the N sequestration in this region. In addition to the advantages just described, farmers would also benefit from tremendous savings in irrigation water. 2.2.8. Summary of N inputs Crop rotations, lower N inputs, split N applications, leguminous crops, cover crops, and modified methods of applications (such as incorporation of manures), and other practices all can be used to reduce N losses the environment and increase N savings (Delgado et al., 2008c). These N savings may be even more substantial depending on the practice(s) used and soil combinations present (Delgado et al., 2008c). An Internet-based or stand-alone NTT can be used to assess potential N savings and the potential to trade N in conservation markets (Delgado et al., 2008c; Gross et al., 2008).

2.3. Transformations and pathways for reactive and total nitrogen losses Several scientists have reported that it may be possible to use denitrification as a method to reduce the losses of reactive N to the environment (Hey, 2002; Hey et al., 2005; Hunter, 2001; Mosier et al., 2002). This can be achieved by adding a carbon source to the system (Mosier et al., 2002), strategically placing denitrification traps (Hunter, 2001), strategically managing water levels of drainage systems (Strock et al., 2007), and strategically locating wetlands to increase denitrification and removal of NO3-N from surface water (Hey, 2002; Hey et al., 2005). This strategic use of denitrification-based management practices is another example of how precision conservation that considers spatial and temporal variability can be used to reduce N transport in the environment and increase N trading potential. Since some scientists recommend denitrification as a positive pathway for removing NO3-N from surface and groundwater flows, we defined the NTT as the quantification of the mathematical difference between a base scenario and a new N management scenario by adding individual pathways of the N cycle. Since denitrification (N2-N) loss has been reportedly beneficial in some cases by reducing the effects of reactive N on the environment (Hey, 2002; Hey et al., 2005; Hunter, 2001; Mosier et al., 2002), we calculated the NTT-DNLreac using Eqs. (2)–(7).


146

The following equations are used to calculate reactive N losses, which include nitrate leaching (DNO3-N, Eq. (2)), nitrous oxide losses (DN2O-N, Eq. (3)), ammonia volatilization (DNH3-N, Eq. (4)), surface N transport not connected to soil erosion (DNst, Eq. (5)), surface N transport caused by soil erosion (DNer, Eq. (6)), and NTT-DNLreac (Eq. (7)): DNO3 N ¼ NO3 Nbms NO3 Nnms

ð2Þ

DN2 O N ¼ N2 O Nbms N2 O Nnms

ð3Þ

DNH3 N ¼ NH3 Nbms NH3 Nnms

ð4Þ

DNst N ¼ Nst Nbms Nst Nnms

ð5Þ

DNer ¼ Ner Nbms Ner Nnms

ð6Þ

NTT DNLreac ¼ DNO3 N þ DN2 O N þ DNH3 N þ DNst þ DNer

ð7Þ

If the nutrient managers are also interested in N use efficiencies in the cropping system, they will want to know the effects of nonreactive N losses due to denitrification. To calculate total N losses, Eq. (8) is used to calculate N2-N denitrification (DN2-N) and Eq. (9) is used to calculate the NTT difference in total N losses (NTT-DNLtot). For Eqs. (2)–(9), bms refers to the base management scenario, and nms refers to the new management scenario: DN2 N ¼ N2 Nbms N2 Nnms

ð8Þ

NTT DNLtot ¼ NTT DNLreac þ DN2 N

ð9Þ

Some users will be interested in trading N in air quality markets as carbon sequestration equivalents (Delgado et al., 2008c; Lal et al., 2009). The carbon sequestration unit equivalents earned through the reduction of N2O-N losses to the atmosphere can be estimated with Eq. (10) (DN2ON132.8). The International Panel on Climate Change (IPCC) methodology also accounts for indirect N2O emissions from reactive N losses to the environment. The IPCC’s methodology assumes that 30% of fertilizer N input is leached and/or lost as runoff and that 0.75% of these losses are emitted as N2O-N (Eggleston et al., 2006; Houghton et al., 1992). Additionally, the IPCC methodology assumes that 10% of the N fertilizer (20% of the manure N) is lost through NH3-N/NOx-N volatilization and that 1.0% of these losses are also emitted indirectly as N2O-N (Eggleston et al., 2006). The indirect savings in carbon sequestration equivalents due to the reduction in direct N2O losses are estimated with Eq. (11). The total savings


147

in carbon sequestration equivalents due to the reduction in direct and indirect N2O losses are estimated with Eq. (12): DDCO2 CseN2 O ¼ DN2 O N 310 0:2727 1:571 DICO2 CseN2 O ¼½ððDNO3 N þ DNst N þ DNer Þ 0:0075 310 1:571Þ

ð10Þ ð11Þ

þðDNH3 N 0:01 310 1:571Þ 0:2727 DTCO2 CseN2 O ¼ DDCO2 CseN2 O þ DICO2 CseN2 O

ð12Þ

2.3.1. Gaseous pathways There are several gaseous pathways by which N gases may be emitted from soils (Fig. 1). Researchers have conducted in situ field and laboratory studies to measure the effects of management practices on emissions of N gases and how management of gaseous losses affects N use efficiencies. One of the most important pathways for N loss is denitrification (Figs. 1 and 3). The acetylene technique is based on the discovery by Federova et al. (1973) that the reduction from N2O to N2 in the denitrification process can be inhibited with acetylene. Isotopic 15N labeled N has been used to trace the effects of management on denitrification. The process of denitrification has been studied very closely by Firestone and Davidson (1989), Hutchinson (1995), and Mosier and Klemedtsson (1994), among others. Biogeochemical reactions of nitrification and denitrification drive emissions of N2O/NO/N2 (Fig. 3). Although emissions of N2O are minimal and reported to be an average 1% of the applied N fertilizer (Eggleston et al., 2006), the losses of N2 due to denitrification could be significant (Meisinger and Randall, 1991; Peoples et al., 1995). Denitrification potential has been correlated with surface texture and drainage characteristics by several scientists. Peoples et al. (1995) reported that potential denitrification for poorly drained clay soils was 35%, seven times higher than the 5.5% for the well-drained sandy soils. Similarly, Meisinger and Randall (1991) reported that potential denitrification was 25–55% for poorly drained soils with over 5% SOM, while the potential denitrification was about 6–20% for the well-drained soils. Mosier et al. (2002) reported that we can manage denitrification with water and nitrogen management practices and carbon inputs. Nitrification inhibitors (Bronson and Mosier, 1993; Freney et al., 1992) and controlledrelease fertilizers (Delgado and Mosier, 1996; Shoji and Gandeza, 1992; Shoji and Kanno, 1994; Shoji et al., 2001) can be used to further reduce N2O emissions. Mosier et al. (2002) recommended that the best practice for reducing N2O emissions is to develop a management plan that increases N use efficiencies and reduces N inputs.

148


The Global Warming Potential over a 100-year time frame for N2O is about 310 (USEPA, 2007; http://www.epa.gov/OMS/climate/420f05002. htm#global). In other words, a management practice that reduces N2O emissions by 1.0 kg N2O-N is equivalent to the sequestration equivalents of 132.8 kg CO2-C. An NTT can be used to evaluate the effectiveness of these techniques for reducing N2O emissions and the resultant ability to trade these reductions as carbon sequestration equivalents in air quality markets (Delgado et al., 2008c). Finer soils with greater denitrification potential and greater N2O emission potential offer an advantage for trading carbon sequestration equivalents, particularly under irrigated systems, because greater reductions of N2O emissions can be achieved than with coarser soils (Delgado et al., 2008c). NTT results show that practices that match the N application with N uptake or reduce excessive N applications mitigate denitrification, N2O and NO3-N leaching losses and increase the potential for N trading and trading of carbon sequestration equivalents (Eqs. (10)– (12)). Mosier et al. (2002) reported that management practices that increase N use efficiencies, such as using N budgets to avoid overapplication, using the right N source with respect to water management, splitting N into multiple applications, improving water management, using source types to reduce denitrification, and other N management methods can lead to reduced denitrification losses. Management of soil denitrification will also be correlated with management of soil oxygen concentrations and water-filled pore space (e.g., soil water content) (Freney et al., 1992; Gilliam and Boswell, 1984; Hey et al., 2005; Linn and Doran, 1984; Meisinger and Randall, 1991; Mosier et al., 2002; Peoples et al., 1995; Steenvoorden, 1985). Additionally, management of soil denitrification will also be correlated with management of carbon inputs (Firestone and Davidson, 1989; Hunter, 2001; Meisinger and Randall, 1991; Mosier et al., 2002; Peoples et al., 1995; Weier et al., 1993, 1994). Management of denitrification can be used as a mitigation alternative to reduce the off-site transport of N across the environment. We could use management of oxygen levels in soils by managing water levels to increase denitrification rates for the removal of NO3-N, thereby reducing its transport in the environment (Gilliam and Boswell, 1984; Hey et al., 2005; Hunter, 2001; Mosier et al., 2002; Steenvoorden, 1985). Alternatively, we could add carbon sources to increase dentrification rates of nitrate that has been leached out of the system (Hunter, 2001). Delgado et al. (2008c) recommended that denitrification should not be accounted for when evaluating the potential reduction of reactive N losses to the environment, and that any methods that reduce the NO3-N transport at a farm or field level should be counted as a practice that reduces the transport of reactive N losses to the environment. Thus, if a management practice increases denitrification losses and reduces the transport of reactive


149

NO3-N, the new practice will be basically credited with savings as far as reducing potential N losses of reactive N over the baseline scenario. However, a full analysis should consider N2O emissions, since this management practice may also increase N2O emissions under a higher denitrification potential, depending on the oxygen levels (Mosier et al., 2002). An NTT could provide the advantage of conducting a mass balance analysis of both pathways simultaneously to determine if the reduction of NO3-N transport due to denitrification may increase N2O emissions. At a watershed level, the concept of nutrient farming proposed by Hey et al. (2005) is a very valuable one and can serve as a key precision conservation practice (Delgado and Berry, 2008). However, it remains to be sorted out how nitrogen trading systems will credit farmers for reducing the transport of NO3-N out of their fields at an upstream watershed while simultaneously crediting a nutrient harvesting farm downstream without double accounting. We suggest that farmers who reduce the NO3-N transport may get a credit at a farm level, while the implementation of a wetland area or riparian forest downstream may be credited with the balance between NO3-N transport into the system and NO3-N coming out, since these systems will serve as potential filters for NO3-N. However, the effect of the denitrification on potential N2O emissions, the emissions of other gases such as methane, and even on carbon sequestration may also have to be sorted out. Other critical factors such as distance to water bodies (like streams and rivers) would need to be considered as well, but will not be covered in this chapter. A full analysis for the nutrient harvesting wetland may be needed to determine the balance between carbon and nitrogen pools. Another important form in which N is lost to the atmosphere is NOx, a pathway that does not result in as many losses as N2, but generally presents N losses much greater than N2O. For example, it has been reported that the 1.3 kg NOx ha 1 y 1 lost from a Colorado short grass steppe was about 10 times greater than the N2O emissions (Martin et al., 1998) and was driven mainly by N mineralization. These results were in agreement with Hutchinson (1995), who reported that NO is formed in the denitrification process, but is not considered a major product of denitrification because of the combined effect of high water content restricting NO diffusion into the atmosphere and the further reduction of NO into N2O and N2. Another significant pathway for gaseous losses of N is NH3-N volatilization from fertilizers and animal wastes that contain urea and NH4-N (Peoples et al., 1995). Peoples et al. (1995) reported that losses due to NH3-N volatilization can be significant in every part of the world, especially in sensitive systems such as flooded rice in Australia, China, India, and the Philippines (45–78%) and sugarcane fields in Australia (47–61%). However, these losses can be significantly reduced through proper management. Studies have indicated that higher levels of NH3-N volatilization correlate with higher pH. For example, the volatilization of urea in flooded rice was

150


reported to be about 9%, much lower than the 30% observed when the site was in a flooded calcareous soil (Peoples et al., 1995). Peoples reported that small grain systems such as barley, sorghum, and wheat usually receive broadcast applications and incorporation, with reported decreases in losses via NH3-N volatilization (90%) are autumn sown and harvested in early summer. Only a small fraction of the Australian crop area is devoted to summer crops such as sorghum, sugarcane, rice, sunflower, and maize. With total crop sowings exceeding 20 million ha annually and ranging over 40º of longitude and 25º of latitude, there is a considerable potential for the growth and development of field crops to vary with local seasonal conditions. Just over half of the Australian cropping area receives winter dominant rainfall, 40% has rainfall distributed more evenly throughout the year, and about 5% of the area is sown under a summer-dominate rainfall climate (Unkovich et al., 2009). Across this cropping region, soils range from light sands to heavy clays, are variable in depth, can be uniform or exhibit strong textural contrasts, and may exhibit severe physical and chemical constraints within their subsoils (Adcock et al., 2007; Dang et al., 2006). Water holding capacity in the crop root zone varies from only 30 mm to more than 400 mm (McKenzie et al., 2004), and potential seasonal cereal crop yields may thus vary from near 0 to more than 5 t/ha. Within this scenario, the application of single grain yield values and harvest indices for calculation of net C balances is unlikely to be reliable.

177

Variation in Crop Harvest Index

2. Determinants of Harvest Index 2.1. Energy investment in seeds, fruits, and storage organs Comparisons among crop species reveal that the energy or carbon cost of producing seeds, the portion of the crop most often harvested, varies enormously. The quantity of seed that a given crop species can produce from a given amount of photosynthate (energy) is constrained by differences in plant biochemistry. In Table 1, we assemble available data on seed composition and approximate energy (glucose) costs of synthesis (following Penning de Vries et al., 1974; Sinclair and De Wit, 1975). These calculations include the following assumptions:

The principal energy costs are in synthesis of carbohydrates (CHO), proteins, and lipids; the equivalent glucose costs of which are protein from NH3—1.62 g glucose/g protein, protein from NO3—2.48 g glucose/g protein, carbohydrate—1.21 g/g, lipids—3.03 g/g, lignin— 2.12 g/g, and organic acid—0.91 g/g (from Penning de Vries et al., 1983). N for protein is acquired as nitrate for nonlegume crops (2.48 g glucose required/g N), and for legumes, the cost of symbiotic N fixation is the same as for ammonium assimilation (1.62 g glucose required/g N) (Pate and Layzell, 1990) N requirement approximates mg protein N in seed/total glucose consumed (mg/g), and protein is 15% N Carbon (energy) requirement, biomass productivity (PV, g seed/g glucose consumed), and N requirement can then be calculated according to Eqs. (2)–(4), respectively. X mass glucose required ðgÞ ¼ ð%CHO 1:21Þ þ ð%protein 2:48Þ þ ð%lipid 3:03Þ P %CHO þ protein þ lipid þ ash PV ¼ mass glucose required N requirement ðmg=gÞ ¼

protein 0:15 1000 mass glucose required

ð2Þ ð3Þ ð4Þ

Our estimated productivity and N requirement data (Table 1) vary slightly from those of Sinclair and De Wit (1975), perhaps because we have used the conversion values of Penning de Vries et al. (1983) without rounding. The data should only be considered gross approximations for reproductive investment, as they do not include the cost of pods and other

Murray Unkovich et al.

178

Table 1 Carbohydrate (CHO), protein, lipid and ash content, biomass productivity (PV), and nitrogen demand of seeds of crop species

Crop

Barley (Hordeum vulgare) Rice (Oryza sativa) Rye (Secale cereale) Wheat (Triticum esculentum) Oat (Avena sativa) Mean for C3 cereals Sorghum (Sorghum vulgare) Popcorn (Zea mays praecox) Corn (Zea mays) Mean for C4 cereals Bean, lima (Phaseolus vulgaris) Chickpea (Cicer arietinum)—kabuli Cowpea (Vigna unguiculata) Faba bean (Vicia faba) Field pea (Pisum sativum) Pigeon pea (Cajanus cajan) Chickpea (Cicer arietinum)—desi Bean, mung (Phaseolus aureus) Lentil (Lens culinaris) Lupin (Lupinus angustifolius) Lupin (Lupinus cosentinii) Vetch (Vicia sativa) Lupin (Lupinus albus) Lupin (Lupinus luteus) Soybean (Glycine max)

N PV requirement (g/g) (mg/g) Refs.

CHO Protein Lipid (%) (%) (%)

Ash (%)

80

9

1

4

0.75 11

1

88 82 82

8 14 14

2 2 2

2 2 2

0.75 9 0.72 15 0.71 15

1 1 1

77 82 82

13 12 12

5 2.4 4

5 3.0 2

0.70 14 0.73 13 0.70 13

1

80

13

5

2

0.69 14

1

84 82 70

10 12 24

5 4.7 1.8

4 2.7 3.9

0.64 11 0.68 13 0.67 24

1

74

20

3.8

2.6

0.67 20

1, 3

69

26

2

3

0.66 25

1

72 70

24 25

1.2 1.6

2.7 2.8

0.66 24 0.66 24

3 1, 3

69

25

2

4

0.66 25

1

71

22

5

2.7

0.65 21

1, 3

66

26

4.5

3.8

0.64 25

1, 3

66 59

26 33

5 5.5

3.1 2.9

0.63 25 0.59 29

1, 3 2, 3

58

35

4

3

0.59 31

2

61 51 49

28 37 42

9 9.2 5

2.5 3.2 4

0.59 25 0.56 31 0.56 35

3 2, 3 2

38

38

20

4

0.50 28

1

1

1, 3

179


Table 1

(continued)

Crop

Peanut (Arachis hypogaea) Mean for legumes Cotton (Gossypium hirsutum) Safflower (Carthamus tinctorius) Sunflower (Helianthus annus) Flax (Linum usitatissum) Hemp (Cannabis sativa) Canola (Brassica napus) Sesame (Sesamum indicum)

N requirement PV Refs. (g/g) (mg/g)

CHO Protein Lipid (%) (%) (%)

Ash (%)

25

27

45

3

0.43 17

1

60 47

29 25

7.8 25

32 3

0.61 26 0.52 19

1

50

14

33

3

0.52 11

1

48

20

29

3

0.51 15

1

32

26

38

4

0.46 18

1

27

29

41

3

0.44 19

1

25

23

48

4

0.43 15

1

19

20

54

7

0.42 13

1

CHO data for Petterson and Mackintosh, 1994 (Ref. 2 data) is approximated as (100 (protein þ lipid þ ash)). We have recalculated the PV and N requirement of Sinclair and De Wit (1975) (Ref. 1) and Perry et al. (1986) (Ref. 2) and added data from Petterson and Mackintosh (1994) (Ref. 3).

reproductive structures, but only seeds. Some protein content conversions also suffer from estimation, and some seed components are left out (lignin, organic acids), although these come with very little C cost (Penning de Vries et al., 1974). Nevertheless the rankings among crop species are valid. On average, C3 cereal crops turn out to be the most efficient in their conversion of photosynthate to grain with a biomass productivity (PV) averaging 0.73, due to the low lipid and protein contents of their seed. Barley is highlighted as being the most efficient. Flood and Martin (2001) noted a negative relationship between grain protein content and HI in wheat. Legumes tend to have a higher carbon cost for seed production, averaging 0.61 g seed dry matter/g glucose invested. Peanut stands out as having a very high cost of seed production (0.43 g/g) due to high seed oil content (45%). This is similar to a range of other high oil species such as cotton, canola, and sunflower. On a biochemical basis one would thus expect cereals to have higher harvest indices than grain legumes, which in turn would be higher than the other species with an oil-rich seed (Table 1). Compared to noncereal crops,

180


the plasticity in grain protein concentration of wheat (Flood and Martin, 2001), and other cereals, may reduce variations in harvest index. In contrast, Lupin (Lupinus angustifolius) seed, for example, tends to have a much more stable protein content (Duthion and Pigeaire, 1993) and grain N concentration may dictate grain yield more so than for cereal crops, although N supply such in N2 fixing legumes should be nonlimiting. Manufacture of seeds and other reproductive structures comes at a significant energy cost, both in terms of storage and biosynthesis, especially where seeds contain lipids and proteins, and thus seed biology impinges directly on the amount of seed that can be produced for a given amount of photosynthate.

2.2. Breeding The HI of crop plants has increased over time due to breeding for higher yield (see Calderini et al., 1999; Perry and D’Antuono, 1989; Whitehead et al., 2000), and more recently, specifically for HI (Sinclair, 1998). Based on the data presented in (Fig. 2) and Turner (1997), we calculate that the rate of HI increase due to cereal breeding in Australia is about 0.015 per decade, compared to about 0.02 achieved in the United Kingdom (Turner, 1997). Shorter-statured, modern crop cultivars have higher harvest indices than their taller forebears, although total dry matter production is most often very similar (Evans, 1993). A recurring theme in HI gain has also been a shortening of the vegetative phase of crop plants, providing a proportionately

Figure 2 Changes in harvest index of wheat for Australian wheat varieties 1884–1983, all grown between 1979 and 1981 at a range of sites in Western Australia (redrawn from data of Perry and D’Antuono, 1989).


181

longer grain-filling period (Lopez Pereira et al., 2000; Siddique et al., 1989). A similar effect can often be achieved by delayed sowing (e.g., Batten and Khan, 1987b; Gomez-Macpherson and Richards, 1995). These increases in HI of crops over time may be an unrecognized source of error where C accounting is conducted over extended historical time periods (e.g., Adger and Subak, 1996) in which significant shifts in HI have occurred and crop residue returns are calculated from measured grain yields using a constant harvest index.

2.3. Determinacy Most cereal crop cultivars are determinate, flowering within a very small window of time and then entering a grain-filling period such that vegetative and reproductive growth are separated in time. Determinate crops generally have higher harvest indices as most crop resources are diverted to grain production once flowering has commenced. In contrast, for indeterminate crops, such as most legumes and oilseeds, concurrent vegetative and reproductive growth is maintained, resulting in competition for resources between vegetative and reproductive sinks (Constable and Gleeson, 1977; Pate and Armstrong, 1996), and flowering occurs over an extended period during which flowers and pods may be subjected to a range of climatic stresses. Harvest index could thus be more variable in indeterminate than determinate crops. In cases where both determinate and indeterminate cultivars are available (e.g., faba bean, soybean), the determinate varieties tend to have higher harvest indices (De Costa et al., 1997; Erskine et al., 1988). While determinate crops are also thought to have less variable harvest indices, we cannot find sufficient data to reliably assess this. While flowering over an extended period in indeterminate crops means that developing flowers and pods are subject to a broader range of climatic conditions, some of which may cause abortion of reproductive parts, it also means that aborted flowers or seeds may be compensated for by later formed flowers.

2.4. Effects of pests and diseases Crop harvest indices can be reduced by pests and diseases through impacts on flowers or seed and pod development, or indirectly through destruction of grain. Diseases and pests can reduce kernel number or size, for example, Hoffman and Kolb (1998) showed that HI in wheat was correlated with the incidence of bean yellow mosaic virus, and McKirdy et al. (2002) showed that this virus caused shriveled grain in wheat, but less so in oat. In Australia pea weevil (Bruchus pisorum) and other insect pests can inflict significant yield losses on grain legume crops (Hardie et al., 1995), significantly reducing economic yield and harvest index.

182


2.5. Temperature extremes during flowering and grain fill For most plants reproductive development is sensitive to a wide range of environmental impacts, with many grain crops having a preferred temperature range for flower and seed development. Large variations in temperature outside of this range may cause problems with flower or seed growth. Damage can be caused by freezing injury or due to temperatures just above freezing point. A definition of absolute values of these is problematic at the field level as there will be some differences among crop species, cultivars and age, and interactions with local landscape factors. For chickpea, Croser et al. (2003) defined freezing temperatures as rainfed > rain excluded. Thus having more water available to canola plants postanthesis increased harvest index. This relationship was shown quite eloquently for wheat in the glasshouse experiments of Passioura (1977), as shown in Fig. 3. In both these experiments HI was found to be independent of total water use. Richards and Townley-Smith (1987) found a similar relationship in later glasshouse experiments. The curvilinear relationship they observed (Fig. 4) was attributed to a variable contribution of stored C from preanthesis photosynthesis to grain yield. It is not clear how applicable these essentially glasshouse-derived relationships between pre/postanthesis water use and HI are to actual field conditions. Field data to assess this are scarce. For rainfed field crops it is difficult to manage water supply to match current crop demand. The balance between pre- and postanthesis growth will be a function of available

Figure 3 Harvest index of pot cultured wheat as a function of the percentage of total water supply used after anthesis (replotted from Passioura, 1977 with permission). Different symbols are for different cultivars.


184

0.5

Harvest index

0.4 1 3 2 3

0.3

1 2

0.2

0.1

10

20 30 40 50 % water used after anthesis

60

Figure 4 Relation between HI and the percentage of total water used after anthesis for winter (○) or spring-summer (●) sown wheat in the glasshouse (reproduced from Richards and Townley-Smith, 1987 with permission).

soil water (rainfall soil type soil evaporation) and the rate it is used by the crop. Preanthesis transpiration can be increased by promoting more rapid development of crop leaf area. This can be done with the application of fertilizers, especially N, and with crop rotations which reduce root diseases and promote faster root growth. Together these could lead to faster leaf area development and earlier exhaustion of soil water. Field experiments in NSW employing irrigation and rain out shelters (Kirkegaard et al., 2007) examined seasonal water use and yield of wheat. In one experiment additional postanthesis water use of ca 10 mm increased grain yield but had no impact on harvest index, while in a second experiment where a severe water stress was imposed, extraction of an additional 30 mm of water increased harvest index from 0.31 to 0.38. The only substantial field datasets on pre- and postanthesis water use and crop HI that we have been able to find for given sites in Australia are assembled in Fig. 5. Here it can be seen that there is little evidence for a relationship between pre:postanthesis water use and HI of these field-grown wheat crops. Averaging across site year may mask variations in fractional postanthesis water use and HI due to crop sequence (e.g., Lenssen et al., 2007; O’Connell et al., 2002); however, we found that if data were plotted without averaging site year, the correlation was even weaker than that seen in Fig. 5. Wahbi and Gregory (1989) presented a negative relationship between the percentage of total water use (20–35%) after anthesis and HI of barley grown in Syria and concluded that water use during grain filling was a poor indicator of harvest index.


185

Figure 5 Scatter plot of percentage water used after anthesis and harvest index of wheat at Dooen (Cantero-Martinez et al., 1995) and Walpeup (from O’Connell et al., 2002; O’Leary and Connor, 1997) in Victoria. The triangle represents an outlier in the Walpeup dataset. Excluding this outlier there was a weak negative correlation between the fraction of water used after anthesis and wheat harvest index (R2 ¼ 0.39, HI ¼ 0.1905 fraction water use after anthesis þ 0.3217).

Reasons for a weaker relationship between pre:postanthesis water use and harvest index of wheat in field versus glasshouse studies could be that the range in % of water used after anthesis for field-grown crops at a given site may be less than that engineered in glasshouse experiments, or in field experiments employing irrigation (e.g., Goyne et al., 1993; Palta et al., 2003; Wright et al., 1992). For example, for wheat over a 10-year period (1981– 1991) at Dooen in Victoria, postanthesis water use of wheat varied from 18% to 31% of total crop water use, but with 70% of values being between 25% and 30%. At Walpeup, a drier environment in north-west Victoria, postanthesis water use of wheat similarly ranged from 18% to 46% of total water use. This is probably the least sensitive of the range in fractional water use in the data of Richards and Townley-Smith (1987) shown in Fig. 4. The plasticity of the wheat plant may also contribute to reduced variation in harvest index at the field level, with the number of ears m 2, the number of spikelets per ear, number of grains per spikelet, and grain size all being variable and allowing for considerable compensation among yield components, especially since these components are formed sequentially. Other factors, such as shoot:root development (Shepherd et al., 1987) and mobilization of preanthesis assimilates to grain, can operate to stabilize harvest index at the field level (Palta and Fillery, 1995; van Herwaarden et al., 1998a). Relationships between pre and postanthesis water use and harvest index have also been reported for indeterminate plants. In glasshouse-grown lentil (Shrestha et al., 2006) harvest index ranged from 0.36 to 0.45 when plants were water stressed during the reproductive phase, compared to 0.53–0.59

186


for well-watered lentil. Reductions in seed yield due to water stress during reproductive development resulted from reduced flower number and from seed abortion, but not seed growth rate or seed size. In this case, postanthesis water use was proportionately lower in the droughted treatment. Water stress during chickpea pod development resulted in pod drop and lower harvest indices (Thomas and Fukai, 1995). For lupin crops in Western Australia, pod fill does not commence until the majority of soil water has been used, restricting pod fill and harvest index (Dracup et al., 1998a). Efforts to reduce competition for assimilates between concurrent vegetative and seed pod growth in lupin, by breeding lines with reduced levels of branching, have contributed to increased yield (Hamblin et al., 1986), most probably through earlier flowering, but with only a slight increase in harvest index (Dracup et al., 1998a). In glasshouse-grown lupin, transient water deficits reduced dry matter production by 28% and seed yield by 33%, but harvest index was only reduced by 9% (Palta and Plaut, 1999). Identifying such small differences in HI due to fractional water use at the field level is problematic. Significant positive HI responses to late season watering have been demonstrated in irrigation experiments for crops such as lupin (Dracup et al., 1998a), chickpea (Leport et al., 1999), and canola (Wright et al., 1992) in Australia, but substantial field datasets for rainfed crops to clearly demonstrate direct effects of differences in fractional water use postanthesis are not readily apparent. Across most of the Australian cropping zone crop water deficits during pod and grain fill are the norm (Nix, 1975), while alleviating this can potentially increase grain yields and crop harvest index, increased parallel vegetative growth may act to moderate significant positive impacts of this in indeterminate broadleafed crops. The ability of broadleafed crops to translocate C from vegetative parts to pods or seeds varies among species and among cultivars within species. Leport et al. (1999) highlighted that this may be a way in which HI of chickpea might be improved. Lupin on the other hand appears to have a limited capacity to mobilize earlier fixed C and is more reliant on current photosynthesis to fill seeds (Pate et al., 1980). The variable carbon cost of seed production for different species and varieties may mask the relationship between seasonal water use and HI. Although Sadras and Connor (1991) were able to take this into account and replicate the relationship seen in Fig. 4 in response to fractional water use for the data of Richards and Townley-Smith (1987), and for container-grown sunflower in the field, recalculating the harvest index production value (HIpv) for the data of Fig. 5 did not strengthen the relationship between the fraction of water used after anthesis and HI. Under field conditions where water supply is not controlled, other factors may counterbalance any effects of fractional seasonal water use on HI as observed in glasshouse or controlled environment studies, and field evidence for fractional seasonal crop water use impacting on crop HI is weak. A similar conclusion was reached by Turner (1997).

187


2.7. Water use, soil mineral nitrogen, and harvest index of wheat The field studies of van Herwaarden et al. (1998a) provide some evidence for there being a relationship between fractional water use after anthesis and harvest index; however, this dataset is confounded by interactions with N fertilizer applications. Indeed N fertilizer rates may have an influence on HI that does not operate through seasonal water use patterns. The work of van Herwaarden et al. (1998a,b) indicates that high soil N fertility might result in a decrease in the deposition water-soluble carbohydrates in wheat stems, tillers, and leaves and an increase in structural carbon, leading to less mobilization of vegetative C to grain and a lowering of harvest index. Examining wheat crops in SE Australia, Angus and van Herwaarden (2001) concluded that vigorous preanthesis growth can increase plant transpiration, at the expense of soil evaporation, and lead to a higher C fixation. However, structural carbohydrate laid down earlier in the growing season is not readily mobilized to the grain. In irrigation experiments with wheat receiving 44 kg N/ha (Kirkegaard et al., 2007), a greater fraction of stem water-soluble carbohydrates were mobilized to grain in more water-stressed plants, but this did not result in increased harvest index. In a second experiment, there were no significant differences in watersoluble carbohydrates between irrigation treatments. Replotting the HI dataset of Cantero-Martinez et al. (1995) on the basis of available N (Fig. 6) shows a weak negative relationship between available N at sowing and the harvest index of wheat, whereas no relationship was apparent for fractional water use after anthesis and crop harvest index in this same dataset. A similar relationship is indicated in the data of Adcock (2006) from the Eyre Peninsula SA. Here harvest index of wheat in 2 years was 0.50 0.45

Harvest index

0.40 0.35 0.30 0.25 0.20 0.15 y = −0.0004x + 0.4331 R2 = 0.2067, P = 0.002

0.10 0.05 0 50

100

150

200

Available N (kg/ha)

Figure 6 Correlation between soil nitrate-N at sowing (0–1 m) and harvest index of wheat at Dooen Victoria (1981–1990). Data from Cantero-Martinez et al. (1995).


188

lower after legumes (vetch or medic) than after nonlegumes (canola, barley, or wheat), although there were minimal differences in the fraction of water used after anthesis between rotations within a given year. The timing of seasonal water use was governed by absolute water availability (rainfall) and not through deferral of water use within the crop-growing season. In this study, soil mineral N was higher when the preceding crop was a legume (133–225 kg N/ha) than cereal or canola (81–149 kg N/ha), and this may have contributed to the differences in HI rather than seasonal water use patterns by the crops. The importance of nitrogen availability in Western Australia is yet to be determined, but the generally lower soil fertility, lower N fertilizer applications resulting from a greater uncertainty of responses to N fertilizers due to climate variability, and cultivars with greater concentrations of water-soluble carbohydrates (van Herwaarden and Richards, 2002), may reduce the likelihood of poor mobilization of stored carbohydrates to grain in the western wheatbelt. In a pot study on barley, Fathi et al. (1997) found that reductions in grain yield and increases in grain protein due to postanthesis water stress were greater when plants were grown under high N supply than when the supply of N was limited. This result is in concordance with the aforementioned studies on wheat. A theoretical framework relating seasonal water use and remobilization of stored C was presented by Fischer (1979).

2.8. Summary of factors influencing HI In Australia, environment rather than genotype is seen as the major determinant of harvest index for most field crops for a given site year genotype (Turner et al., 1999). Frost and cold temperatures at flowering, or high temperatures during flowering or seed development can contribute to substantial variation in HI. The balance between pre- and postanthesis water use is not yet clearly implicated in variation in HI of field crops, but in some regions vigorous growth early due to high soil N status, perhaps in conjunction with available water, reduces HI of wheat, and possibly barley. Field data on seasonal water use and HI for other crops are scant.

3. A Database of Crop Dry Matter, Grain Yields, and Harvest Indices We have collated data on grain yield, crop dry matter, or harvest index for a range of crops grown in Australia. Data were from readily available literature and from generous contributions from colleagues. We restricted our data search to 1980 as changes in crop growth and yield over time due to agronomy and breeding (yield) may otherwise confound some


189

interpretations. Data were included from small plot experiments, from farmer-managed paddocks, and in the case of HI only, also from glasshouse experiments. At the time of writing >365 data sources have been used to accumulate 20,100 cereal crop entries, 4500 legume crop entries, and 5500 entries for canola crops. The assembled data are available as accessory material on the publisher’s Web site (URL). It should be interpreted with some caution since it may not be a representative sample of grain crops across Australia, but is a sample of primarily research data from an assemblage of experiments and sites which were not distributed equally across the grain-growing regions in time or space. Differences between experimental yields and farmer best practice yields are not known. In cases where commercial crops have had the same inputs as experimental plots, yields have been equivalent (Evans, 1993). However, there is a tendency for experimental plots to be located on better land within a field, and for inputs to be higher than for commercial crops, thus perhaps skewing research data to achievable but more economically risky yields.

3.1. Problems with interpretation of agronomic harvest indices Grain/produce yield data are usually reported on an ‘‘as harvested’’ basis. Typically this means that they are not reported on a dry weight basis, but at some nominal moisture content, conforming to the receival quality assurance standards dictated by wholesalers or processors. For some crops at physiological maturity the moisture content may be as high as 30% (e.g., maize, Table 2), but this may be left to dry out, or managed to obtain a lower moisture content before and/or after harvest. Many reports of grain yield do not indicate moisture content, even in the research literature. In our database we have not specifically corrected for moisture content and this would introduce some error. In addition to moisture content, variations in calculated HI values can occur depending on the timing of sampling or the extent of leaf drop and crop attrition during senescence. For crops which drop leaves prior to maturity, HI should be calculated from peak biomass dry matter, including fallen leaves, not just from standing dry matter at maturity. For crops which shed leaves, HI calculated in this latter way is referred to as ‘‘apparent harvest index’’ (Schapaugh and Wilcox, 1980), and these are usually higher than actual HI (e.g., apparent harvest index of 0.75 for soybean Mayer et al., 1991). Some of the harvest indices in our database might be apparent rather than actual HI per se, and comparisons among crop species should be carefully considered. We tried to exclude any data which did not include fallen leaves although this was rarely explicitly stated within the original data sources used.


190

Table 2 Typical moisture content for harvested material, and harvest losses due to shattering/harvest inefficiencies

a

Crop

Typical moisture content of harvested product (%)

Wheat

11

Barley Canola Lupin

11 7 > 12

Source

Typical harvest losses (%)

Grain sorghum 13.5 Oats for grain 11 Field pea 14

Receival standarda Receival standard Receival standard Dracup et al. 20 (1998a,b) Receival standard Receival standard Receival standard 5–24

Triticale Chick pea

11 14

Receival standard Receival standard 10–20

Faba bean Rice Lentil Mung bean

14 21 14 9

Sunflower Maize

9 13.5

Receival standard Receival standard Receival standard Petterson and Mackintosh (1994) Receival standard Spenceley (2005) 22

Soybean

15

Rose et al. (2005)

Source

Snowball (1986)

Cassells and Armstrong (1998) O’Mara (undated)

Robertson et al. (2003)

Receival standards are maximum values for Australian grain delivered to silo, average values are likely to be 1–2% lower.

The influence that harvesting method (hand or machine harvested) and inclusion of fallen leaves can have on calculated harvest index was indicated for lupin in Western Australia (Table 3). The data showed that HI varied from 0.16 to 0.29 depending on the rigor of sampling, and crop shoot dry matter attrition between peak dry matter (DM) and harvesting, leading to an apparent increase in harvest index. Similarly, Dracup (1994) estimated harvest index of a lupin crop to be 0.19 at the time of harvest, but 0.15 if fallen leaves were also taken into account. For pot-grown chickpea, Hooda et al. (1986) calculated harvest indices of 0.25 if fallen leaves were excluded

191


Table 3 Harvest index for a lupin crop calculated from various sampling strategies (from Unkovich, 1991) HI formula used

HI

Hand-harvested seed/standing biomass Mechanically harvested seed/standing biomass Hand-harvested seed/(standing biomass þ fallen leaves) Mechanically harvested seed/(standing biomass þ fallen leaves) Hand-harvested seed/peak standing biomass

0.29 0.21 0.22 0.16 0.20

Table 4 Harvest index of chickpea (Cicer arietinum), calculated for C, N, and dry matter, and including different plant fractions (from Hooda et al., 1986) Carbon

Including above-ground biomass only Excluding fallen leaves 0.26 Including fallen leaves 0.20 Including roots Excluding fallen leaves 0.20 Including fallen leaves 0.16

Nitrogen

Dry matter

0.50 0.38

0.25 0.18

0.37 0.31

0.16 0.13

and 0.18 if they were included (Table 4). Saxena et al. (1983) indicated that excluding fallen leaves (pinnae) from HI calculations reduced HI in chickpea by 10%. In experiments with canola (Hocking et al., 1997) HI was estimated to average 0.28, or 0.25 if shed leaves were included. Even where fallen leaves were carefully collected with sequential samplings, total dry matter of lupin crops declined between a peak around midgrain fill and maturity (Greenwood et al., 1975; Unkovich et al., 1994). Prince et al. (2001) used the relationship between apparent and actual HI in soybean of Schapaugh and Wilcox (1980) for estimating crop DM from regional yield data (Actual HI ¼ 9.76 þ 0.96 apparent HI). Whether such relationships have general applicability is not known and we could not find sufficient data to derive such relationships for field-grown crops in Australia. Reported HI values may also be influenced by the proportion of yield produced that is harvestable. Lodging due to weak stems, storm events, insect damage, or other processes can reduce the proportion of grain produced that is harvested and thus induce error in HI estimates or estimates of crop residues where average HI values are used. McDonald (2006) reported hand-harvested wheat yields in a commercial crop to be double that of the machine-harvested yields. On the upper Eyre Peninsula SA, harvest losses as high as 17.5% for field pea and 58% for lupin have been reported (Egan et al., 1994). For short-statured crops such as lentil and

192


chickpea, many of the lower pods may remain unharvested, especially on uneven ground. Schwenke et al. (1998) compared farmer yields with measured (quadrat) yields in commercial crops over 2 years in northern NSW. On average, farmers reaped only 48% of the hand-harvested chickpea yields and 65% of faba bean yield. Problems with low harvestability and high shattering losses are probably a lesser issue in research plots which are more intensively managed than commercial crops. Nevertheless it illustrates that in some instances, plot yields, harvested either by hand (quadrat) or machine are potentially higher than commercial grower-harvested yields. For the purposes of carbon-accounting systems, HI data are required to define the amount of carbon retained in the form of crop residues at a site after harvest. It is clear that HI data used for this purpose must be reflective of typical agricultural harvesting practices and account or compensate for any leaf drop prior to harvest. Problems are likely to be more significant for noncereal crops for which leaf drop is more pronounced.

4. Dry Matter, Grain Yield, and Harvest Index Values for Crops Lower limits for crop dry matter, grain yield, and harvest index would all be near zero for rainfed crops, due to occasional drought or other catastrophic crop failure (e.g., pests, disease, weeds, frost). Unfortunately very low and zero crop yields are rarely reported in the scientific literature, partly because under situations of crop failure, or poor crop growth, many research plots are not harvested or are unharvestable. The high reliance on data reported in research publications within our database may skew the dataset toward greater average yields. Nevertheless, for some crops zero and very low grain yields have been reported (Table 5), and although zero yield crops would by default have zero harvest indices, this was not specifically reported for these particular crops. Across the dataset mean harvest index for wheat and barley (0.37) was the same as that averaged for all grain legumes (0.37, lupin, field pea, faba bean, lentil, chickpea, and vetch), while oat harvest index was lowest (0.21) followed by canola (0.28). The low HI of oat may reflect its primary role as a forage crop and a low intensity of breeding for grain yield, and for canola the heavy energy (oil) investment in seeds (see Table 1). In Fig. 7 we plot the cost of seed production against mean HI for crops in Australia where it can be seen that high energy cost of seed production is correlated with lower crop harvest index. Again the position of oat likely reflects a lack of breeding effort for high grain yield compared to that achieved with wheat and barley. The apparently high HI value for faba bean may partly be due to non collection of fallen leaves in HI estimations.

Table 5 Gross summary of grain crop dry matter, grain yield, and harvest index data reported for dryland crops in Australia (from Unkovich et al., 2006) Shoot dry matter (kg/ha)

Grain yield (kg/ha)

Crop

Species

Mean

n

Min

Max

Mean

n

Wheat Barley Lupin

Triticum aestivum Hordeum vulgare Lupinus angustifolius Brassica napus Avena sativa Sorghum bicolor Triticum durum Secale cereale Pisum sativum Cicer arietinum Vicia faba Lens culinaris Zea mays Vicia sativa Arachis hypogea

6699 7777 5148

1015 96 165

140 12 714

22,250 19,300 13,200

2449 2594 1575

6829 9669 6855 9568

129 163 78 6

51 1050 460 6065

17,793 28,102 20,763 11,230

4706 3783 5130 3817 20,733 3701 3730

167 180 49 42 12 24 16

1140 80 1330 1440 15,500 222 940

11,580 10,245 10,740 9310 28,000 9890 12,378

Canola Oat Sorghum Triticale Field pea Chickpea Faba bean Lentil Maize Vetch Peanut

Harvest index Max

Mean

n

Min

Max

14,535 19 2992 5 919 17

8600 8180 4482

0.37 0.38 0.30

1266 117 229

0.08 0.09 0.04

0.56 0.57 0.5

1569 2535 3232 3109

5449 872 109 409

20 3 80 52

5160 6840 10,000 7264

0.28 0.21 0.46 0.34

117 21 78 10

0.04 0.11 0.14 0.28

0.41 0.48 0.7 0.46

1486 1128 1727 1307 5659 1070 1553

708 830 720 261 51 130 50

150 0 163 0 1445 0 51

4535 4033 5752 4134 14,400 2830 6000

0.36 0.37 0.44 0.33 0.52 0.38 0.33

185 188 48 44 7 23 32

0.06 0.06 0.11 0.06 0.41 0.16 0.02

0.58 0.55 0.58 0.51 0.62 0.47 0.57

Min

Data are from an exhaustive summary of the literature, not a rigorous strategic sampling across all regions and years. Crops are listed in decreasing order of sowing area.


194

0.50

Mean harvest index

0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.35

0.45

0.55

0.65 PV(g/g)

0.75

0.85

Figure 7 Scatterplot of the cost of seed production (PV, g seed/g carbohydrate) and mean harvest index for field crops in Australia. Excluding oat, the PV accounted for 33% of the variation in HI across crops (HI ¼ 0.3133 PV þ 0.1725).

In Table 6 we have extracted the maximum records of dry matter, yield, and harvest index and their origins as a matter of interest. Such values are often difficult to find and may prove useful for those interested in upper physiological or agronomic limits, or modeling potential C fluxes. Reported crop dry matter accumulation maxima in Australia (Table 6) range from 30, that is, for wheat, barley, lupin, chickpea, field pea, faba bean, and canola. Data were first averaged by site year and then organized into 0.05 HI interval increments. The form of the fitted distribution was BetaGeneral (a1, a2, min, and max) with the minimum and maximum values set to 0 and 0.65, respectively.

kurtosis have a distinct peak near the mean, decline rapidly, and have heavy tails. Datasets with low kurtosis have a flat top near the mean rather than a sharp peak. The values obtained for each of these statistics for the measured HI frequency distributions will be discussed in the subsequent sections where the HI data for individual crops are examined. The fitted distributions were obtained with @Risk (version 5, Palisade Corporation, Ithaca, New York)


199

using a BetaGeneral distribution with four fitting parameters: a1, a2, min, and max. The minimum HI (min) was set to 0 and the maximum HI (max) to 0.65, and then the values for a1 and a2 were fitted to the measured data. For all seven crops, the fitted frequency distributions appear to provide a reasonable estimate of the measured frequency distributions. We follow with a brief summary of what is known of HI for each crop, in decreasing order of importance, in terms of the area sown in Australia (Unkovich et al., 2009). We conclude each crop with summary of our HI data averaged on a site year basis which we believe provides the safest interpretation given the nature and purpose of the dataset. This site year averaged dataset can be made available to readers on request. Please contact the corresponding author.

4.1. Wheat For wheat more than 14,500 yield estimates were obtained and >1200 for harvest index, many more than for any other crop (Table 5). Mean shoot dry matter yield for rainfed wheat in the crops reported was 6.7 t/ha with a maximum of 22.5 t/ha. Harvest index values ranged from 0.04 to 0.59. Variation in HI at a given site over time can be examined in the rotation trial at Tarlee, South Australia. Here, HI for wheat ranged from 0.23 to 0.50 (1980–1987, calculated from reports, see Schultz, 1995). Estimates of HI at Merredin (WA) sites in our database cover 11 years from various locations, ranging from 0.25 to 0.54. Over 17 years at Wagga NSW, HI of wheat has ranged from 0.08 to 0.42, and at Walpeup (Vic) HI for wheat has ranged from 0.09 to 0.50 across 13 years of data. Maximum growth, yield, and harvest index of wheat were examined in a series of experiments on irrigated wheat in southern NSW during 1983– 1985 (Stapper and Fischer, 1990a–c), and combinations of N, sowing rate/ spacing, and sowing date were employed. Within the treatments and years HI varied from 0.28 to 0.51. Maximum potential grain yield in this environment with extant material was considered to be 9–10 t/ha at 12% moisture. At Condobolin and Cowra in NSW, HI of wheat increased with later sowing, ranging from 0.28 to 0.31 for April-sown plots to 0.38–0.42 for June-sown plots (Batten et al., 1999. The range for all treatments/sites was 0.21–0.46). Batten and Khan (1987a) grew wheat at Wagga over 3 years with an N and P fertilizer factorial. They related HI to N and P nutrition. The crop HI ranged from 0.31 to 0.44, but there was no attempt to relate HI to seasonal rainfall. In NSW, increasing the sowing rate and applying N fertilizer increased DM production at the expense of HI (Duggan et al., 2005). In the work of Mason and Fischer (1986), at one site in southern NSW, HI in wheat varied from 0.25 to 0.42 across 3 years and six tillage treatments. Averaging across data source site year provided 194 estimates of HI for rainfed wheat, a mean of 0.36, a median of 0.36, and a mode of


200

approximately 0.33. More than 50% of site year HI estimates fell within the 0.3–0.4 range (Fig. 8A). The distribution of data was negatively skewed and more peaked (kurtosis >3) than a normal distribution indicating a greater frequency of values near the mean and a longer left tail (low HI values, Table 8).

4.2. Barley Average dry matter (8.7 t/ha) and grain yield (2.6 t/ha) of barley were higher than those obtained for wheat (Table 5). Harvest index across the whole database (n ¼ 127) averaged 0.38. Harvest index in barley tends to be higher than for wheat, perhaps a consequence of its lower grain N content (see Table 1), with 70% of records being between 0.35 and 0.45 (0.275–0.475, Fig. 8B). While it would thus appear to be less variable than HI in wheat (Fig. 8A), the dataset is smaller. Values range from 0.09 in 2001 at St George in Qld, to 0.57 at Northfield SA in 1990 (Fathi et al., 1997). In Queensland, the HI of two barley varieties ranged from 0.31 to 0.41, depending on seasonal water supplied (Thomas and Fukai, 1995). There are few specific studies of HI in barley in Australia. Averaging our data for barley by source site year gave 35 estimates of HI for rainfed barley, with a mean of 0.38, a median of 0.39, and mode of Table 8 Statistics for measured and fitted frequency distribution for crop harvest index from the database where the number of site year observations >30, that is, for wheat, barley, lupin, chickpea, field pea, faba bean, and canola Min

Max

Mean Mode Median s.d.

Skewness Kurtosis n

Wheat 0.08 Fitted 0.00 Barley 0.10 Fitted 0.00 Lupin 0.07 Fitted 0.00 Chickpea 0.07 Fitted 0.00 Field pea 0.06 Fitted 0.00 Faba bean 0.11 Fitted 0.00 Canola 0.10 Fitted 0.00

0.52 0.65 0.53 0.65 0.44 0.65 0.55 0.65 0.56 0.65 0.58 0.65 0.38 0.65

0.36 0.36 0.38 0.38 0.28 0.28 0.36 0.36 0.36 0.36 0.45 0.44 0.27 0.27

0.74 0.08 0.90 0.19 0.03 0.15 0.62 0.13 0.48 0.13 1.20 0.47 0.75 0.14

0.33 0.36 0.47 0.39 0.20 0.27 0.37 0.37 0.42 0.37 0.52 0.48 0.30 0.26

0.36 0.36 0.39 0.39 0.29 0.27 0.38 0.36 0.37 0.36 0.46 0.45 0.28 0.26

0.07 0.07 0.09 0.09 0.08 0.08 0.11 0.11 0.12 0.12 0.11 0.10 0.06 0.06

3.89 2.73 3.91 2.63 2.21 2.67 2.86 2.45 2.60 2.37 4.38 2.73 3.32 2.82

194 35 49 52 39 37 44

Data were first averaged by site year and then organized into 0.05 HI interval increments. The form of the fitted distribution was BetaGeneral (a1, a2, min, and max) with the minimum and maximum values set to 0 and 0.65, respectively.


201

approximately 0.47 (Table 8). As noted for wheat, the distribution of HI data was negatively skewed and more peaked than a normal distribution.

4.3. Lupin Database entries for lupin total 165 for DM yield, 919 for GY, and 229 for harvest index. Mean values across the database were 5.2 t/ha for DM yield, 1.6 t/ha for GY, and 0.30 for HI (Table 5). Maximum DM yield of 12.6 t/ha for irrigated lupin (Dracup et al., 1998a) and 11.1 t/ha for rainfed lupin at Eradu, WA, have been reported. The grain yield maxima of 4.4 t/ha in the database was recorded at Cressy, TAS (Bishop and Mendham, 1996). Across the database HI of lupin ranged from 0.04 to 0.50, with an average of 0.30 (Table 5). While the upper limit of 0.5 used in the modeling of Farre et al. (2004) seems sensible in the light of the aforementioned data, they included roots in their HI calculation. Large variation in lupin HI has been suggested (Dracup et al., 1998a), but we can only find 49 site year estimates of HI, not enough to provide substantial evidence to demonstrate this. Even the article of Delane et al. (1988), which is often cited as evidence of variable HI (e.g., Dracup and Kirby, 1996; Palta and Ludwig, 1998), concluded that ‘‘overall, results suggest that very poor pod setting on the main stem and low HI are not as widespread a problem as previously thought.’’ In the studies of Delane et al., across 38 sites, crop dry matter explained 67–83% of yield variation of three genotypes, indicating that factors influencing pod set probably account for no more than 20% of yield variation. Using supplementary irrigation, Dracup et al. (1998a) showed that terminal drought can have a substantial influence on yield and HI of lupin. HI increased from 0.14 in unirrigated plants to 0.22 with the addition of ca 190 mm water. Additional water during pod fill was seen as key to increased HI in these experiments. Irrigation increased pod survival which led to the increased yield and harvest index. In lupin, mild transient water stress may also increase HI by redirecting plant resources from vegetative growth to seed fill (French and Turner, 1991). In lupin, up to 94% of flowers and pods may be aborted (Atkins and Pigeaire, 1993); however, keeping flowers on such plants does not necessarily lead to increased seed yield (Palta and Ludwig, 1996), and seed filling rather than increased pod number would appear to be more likely to increase lupin HI and yield (Palta et al., 2003). Increasing C supply during pod fill (Palta and Ludwig, 2000) increased seed yield but not HI as concurrent vegetative growth maintained its share of the C sink. Averaging across site year provided 49 estimates of HI for lupin (Fig. 8C) with a mean of 0.28, a median of 0.29, and mode of 0.20 (Table 8). Approximately two-thirds of the HI values fell between the 0.20 and 0.30 class intervals. Little evidence for skewness existed and the distribution was flatter than a normal distribution.

202


4.4. Sorghum Seventy five percent of the crop entries in the database were for rainfed production (n ¼ 113). Mean DM yield for rainfed sorghum in the database was 6.8 t/ha (n ¼ 78, Table 5) and 11.4 t/ha for irrigated crops (n ¼ 18). Grain yield varied from 0.8 to 10.0 t/ha for rainfed crops and 2.7–12.4 t/ha for irrigated crops. Harvest index for rainfed crops ranged from 0.14 to 0.70 (mean ¼ 0.46), and for irrigated crops 0.26–0.52 (mean ¼ 0.43). There were insufficient data (n ¼ 19) to provide meaningful insights from site year averaging.

4.5. Chickpea Dry matter production of chickpea in the database averages 3.8 t/ha and grain yield 1.1 t/ha. The maximum DM yield recorded was 10.7 t/ha for an irrigated crop at Merredin, WA (Leport et al., 1999), with an accompanying grain yield of 4.2 t/ha. For rainfed crops the maxima were very similar at 10.2 t DM/ha (Warwick Qld Berger et al., 2004 and personal communication) and 4 t grain/ha at Tamworth, NSW (Haigh, 2006). More than half (141) of the 212 chickpea HI entries were from 13 varieties over five sites and 2 years provided courtesy of Berger et al. (2004) and personal communication. There were 9 years of data from Merredin (WA) with HI there varying from 0.07 to 0.54. In Queensland (Thomas and Fukai, 1995) HI of chickpea ranged from 0.19 to 0.52, depending on seasonal water supply under controlled conditions. Across sites and years (n ¼ 52) chickpea HI averaged 0.36 with a median of 0.38 and mode of 0.37 (Table 8). The distribution (Fig. 8D) was negatively skewed and approximated the normal distribution in terms of its peakedness.

4.6. Field pea Mean dry matter production of field pea in our database is 4.7 t/ha and grain yield 1.5 t/ha. Maxima were 11.6 t/ha for dry matter (Mt Barker WA, Armstrong and Pate, 1994a) and 4.5 t/ha for grain yield (Bool Lagoon SA, McMurray, 2005). These are well below yields attained for the crop in Europe (e.g., Jensen et al., 2004). For HI, absolute minima of 0.06 (Condobolin, NSW, 1986) and maxima of 0.58 (see French and Ewing, 1989) were recorded, although the latter may not have included fallen leaves. The mean across the database is 0.36 (Table 5). The data for field pea appear to be more variable than for most other crops (Table 7). Armstrong and Pate (1994b) showed that HI of semidwarf types were higher than taller, conventional cultivars, as has been shown for other crop species. Harvest indices of up to 0.6 were reported for their glasshouse-grown plants.


203

For our 39 site year averaged data (Fig. 8E), HI ranged from 0.06 to 0.56, with 70% of values in the 0.35–0.50 class intervals. The distribution of HI values was negatively skewed (Table 8) and flatter than a normal distribution.

4.7. Faba Bean Dry matter of faba bean in the database averages 5 t/ha and grain yield 1.7 t/ha, with absolute maxima of 10.7 t/ha for DM and 5.7 t/ha for grain (Dongara, WA see Loss and Siddique, 1997). As for field pea, HI in faba bean is quite variable (Table 7). The relatively high mean of 0.46 may be more apparent than real, with possible noncollection of fallen leaves skewing the values to the high end (see De Costa et al., 1997). For fababean grown at the Waite Institute (Adisarwanto and Knight, 1997) HI decreased with increasing plant density across a range of sowing dates. This resulted from fewer pods/plant at higher sowing densities. Harvest index ranged from 0.54 (later sowing, low density) to 0.4 (early sowing, high density). L’opez-Bellido et al. (2005) reviewed HI in faba bean in relation to plant density and presented values ranging from 0.25 to 0.52. In a series of experiments with determinate and indeterminate fababean and irrigation in the United Kingdom (De Costa et al., 1997), harvest index ranged from 0.34 to 0.68, for both irrigated, rainfed, and rainexcluded treatments. Across genotypes and water stress there was no trend in HI, excepting that where water was provided continuously through trickle irrigation, HI was lower than all other treatments. Higher availability of water increased assimilate partitioning to vegetative matter. Averaging our database across site year (n ¼ 37) for faba bean gives a mean HI of 0.45 (Table 8), a negatively skewed distribution and a more peaked distribution (Fig. 8F).

4.8. Canola Dry matter production of canola in our database averages 6.8 t/ha and grain yield 1.6 t/ha. The maximum rainfed grain yield of 5.16 t/ha (Zhang and Evans, 2004) was very similar to the irrigated yield of 5.2 t/ha recorded at Tamworth NSW (Wright et al., 1992). Maximum dry matter production (15 t/ha) was recorded in a rainfed crop at Katanning WA (Ward et al., 2002). Across our database HI for canola averaged 0.28 (n ¼ 117), ranging from 0.04 (Thomas, 2005) to 0.41 (Hirth et al., 2001). In field experiments examining canola response to N (Hocking et al., 1997), HI varied from 0.27 to 0.30 at one site in NSW, and from 0.29 to 0.34 at another. Nitrogen nutrition would thus not appear to exert a large influence on Canola HI. Robertson et al. (2004) examined HI in canola in response to sowing date. They found that, aside from very early or very late

204


(after mid-June) sowings, HI was relatively stable and yield principally a function of dry matter accumulation, with yield ¼ 0.335 DM 27.5 (R2 ¼ 0.93), very similar to earlier work in WA (slope 0.35 and intercept 96 g/m2). In our database 66% of values fall into the 0.25 and 0.30 class intervals (Fig. 8G), indicating that HI in canola is fairly constrained. Only 44 site year records for HI of canola were found. The low mean value of 0.27 is to a large extent a consequence of the large investment in oil and protein in seed (PV ¼ 0.43 g/g, see Table 1). Median and mode HI values of 0.28 and 0.30 (Table 8) were obtained and the frequency distribution (Fig. 8G) was negatively skewed and more peaked than a normal distribution. Variation in HI of canola was less than that of most other field crops examined (Table 7).

4.9. Rice Rice crop grain yields in Australia are among the highest in the world, with the maximum reported here (14 t/ha) being close to the potential for this crop (ca 15 t/ha Sheehy et al., 2004). Crop dry matter yield peaked at 31.6 t/ha at Walkamin, Qld (Ockerby and Fukai, 2001), or in a more traditional rice-growing region 27.5 t/ha (Yanco, NSW Horie et al., 1997). Across our database HI ranges from 0.01 (Williams and Angus, 1994) to 0.71 (Dunn and Beecher, 1994). The low protein and lipid content of rice may serve to increase its harvest index. All rice grown in Australia is irrigated, and it has been shown that water deficits can be managed to increase harvest index, especially for high N crops (Zhang and Yang, 2004). If senescence of the rice crop is delayed a long, slow grain-filling period ensues, reducing HI, this is typical under high N fertility but can be circumvented by a controlled soil drying. Low temperatures during reproductive development can also have a substantial impact on HI, but this is controlled to some extent by judicious floodwater management which can effect increases in canopy temperatures (Williams and Angus, 1994). Treatments within years can have a substantial effect on HI in rice, especially N application, for example, HI for rice varied from 0.01 to 0.63 in a single experiment in 1 year (Williams and Angus, 1994) as a function of sowing date (26 Sept. or 31 Oct.), N rate (0–250 kg N), and floodwater depth (5 or 20 cm). There was a strong negative interaction between floodwater depth and N rate. Of course such a range in environmental conditions would not normally be experienced for commercial crops which would be expected to be sown near a single optimum time and with optimized fertilizer rates. The experimental data showed that a combination of temperature and plant N content (kg/ha) at microspore development could be used to model harvest index quite effectively with an R2 of 0.54–0.77, depending on cultivar. HI values >0.6 have been reported in the literature, for example, 0.47–0.71 (Dunn and Beecher, 1994) and


205

0.5–0.6 (Williams and Angus, 1994), but these were more likely to be extreme examples of what can be induced experimentally rather than what is likely in commercial crops. The maximum HI of 0.40–0.53 (depending on cultivar) in the model of Williams and Angus (1994) seems realistic. There were only seven site year combinations for rice in our database, insufficient to present frequency distribution data. Mean site year HI values ranged from 0.28 at Deniliquin NSW to 0.64 in Yanco NSW, averaging 0.40.

4.10. Sunflower The database includes only 28 records for sunflower for which the mean DM yield is 7.8 t/ha, GY 1.4 t/ha, and HI 0.4. Variation in HI is only minor with all values falling between 0.32 and 0.54 (n ¼ 25). Records of maximum DM yield and GY for rainfed crops were 15.6 and 5.7 t/ha, respectively. For an irrigated crop at Gatton Qld, sown at optimum time and provided with luxuriant nutrients, total dry matter production was 17.1 t/ha at anthesis and 24.2 t/ha at maturity. Grain yield (0% moisture) was 6.02 t/ ha and harvest index 0.4. Average yields for irrigated crops in the region were 1.8–3.0 t/ha (Bange et al., 1977). Connor et al. (1985) studied variation in harvest index as a function of the timing and quantity of irrigation. Harvest index increased where water was predominantly applied postanthesis, with postanthesis irrigated crops having an HI of 0.53, cf the unirrigated crop HI of 0.38. Across the 2 years of experimentation and 12 irrigation treatments, HI for sunflower varied between 0.32 and 0.53, averaging 0.44. In India Nanja Reddy et al. (2003) reported HI in sunflower to vary between 0.14 and 0.44 across genotypes.

4.11. Lentil The lentil database consists of 42 entries for DM yield, 261 grain yield estimates and 44 harvest index entries. Dry matter yield averaged 3.8 t/ha and grain yield 1.3 t/ha. Maximum DM yield of 9.3 t/ha was reported at Cunderdin, WA (Siddique et al., 1998), and grain yield 4.1 t/ha at Laura, SA. (NVT, 2008). Studying a wide range of germplasms from across the globe, from wild types, through old and modern cultivars Whitehead et al. (2000) found HI in Lentil to reach 0.52, including fallen leaves, etc. Maximum dry matter production was 18.3 t/ha and crop residues of 14.1 t/ha were recorded across all the cultivars studied, and seed yield >4.2 t/ha. In a glasshouse study (Shrestha et al., 2006) harvest index ranged from 0.36 to 0.45 for waterstressed lentil, to 0.53–0.59 for well-watered lentil. Our data across site year provided only 11 HI estimates for lentil, ranging from 0.06 to 0.50, insufficient to present frequency distribution

206


data. The mean, median, and mode of these data were 0.33, 0.35, and 0.35, respectively.

4.12. Maize The database includes only 33 records for harvest index in maize, with 40 DM yield estimates and 117 grain yield estimates. The mean values across the database were 19.3 t/ha (DM yield), 5.5 t/ha (GY), and 0.49 (HI), including both rainfed and irrigated crops. The mean HI was very close to that observed in the United States (Prince et al., 2001). Maximum maize DM yield in the database was 28.0 t/ha and GY 20.5 t/ha (Birch et al., 2006). Maize grown in temperate environments tends to have a higher HI than maize grown in tropical environments (Fischer and Palmer, 1983) due to a shorter grain-filling period in the latter, caused by water limitation.

4.13. Peanut Across our database, there were 113 records of pod yield for peanut (60 rainfed and 53 irrigated), 56 DM yield records (34 rainfed), and 62 HI estimates (40 rainfed). Shoot dry matter for peanut in the database ranged from 0.9 t/ha for a rainfed crop to 13.3 t/ha for an irrigated crop (Bell et al., 1991). Pod yields ranged from 0.51 (rainfed) to 6 t/ha for irrigated peanut. Mean values across the database for irrigated peanut were DM yield, 7.0 t/ha, pod yield 2.2 t/ha and HI 0.30. For rainfed peanut, the values were 5.5 t/ha (DM yield), 2.0 t/ha (GY), and 0.36 for HI. In one study under irrigated conditions DM yield ranged from 9 to 14.2 t/ha and HI from 0.33 to 0.56, depending on cultivar (Bell et al., 1991). Averaging our data across site year produced only five HI data points, insufficient to warrant more detailed presentation.

4.14. Oilseed poppies Little data were available for oilseed poppies (Papaver somniferum), an important crop in Tasmania (10–15,000 ha) grown for the production of the opiate alkaloids, morphine, codeine, and thebaine, and which accounts for >30% of all broadacre crop sowings in Tasmania (Unkovich et al., 2009). Detailed data are kept as commercial ‘‘in confidence,’’ and the ABS AgStats (Australian Bureau of Statistics, 2002) does not contain yield information. Straw harvest data available from the International Narcotics Control Board (INCB: http://www.incb.org/) include only the seed capsule (not seed) and a small amount of attached stalk. This only represents about 50% of the dry matter harvested, or about 25% of total crop dry matter (P. Jolly, Tasmanian Alkaloids Pty Ltd, personal communication). Some of the harvested seed may enter the culinary market (morphine), or be used as furnace fuel (thebaine) in opiate production. Poppy seed makes up about 25% of

207


total crop dry matter. After opiate extraction the waste capsule/straw biomass may be returned to the field for disposal. We have been unable to find direct measurement data to substantiate these approximations.

5. Crop Harvest Indices and C Accounting Fixed harvest indices have been used to estimate net primary productivity or C inputs of croplands from statistical data on crop yields in the United Kingdom (Adger and Subak, 1996), across Europe (Vleeshouwers and Verhagen, 2002), the United States ( Johnson et al., 2006; Prince et al., 2001), and Canada (Bolinder et al., 2007). Some of these harvest indices used are given in Table 9, along with the average (site year) values from the present Australian database for rainfed crops for comparison. Interestingly there were very significant differences in the HI values for some crops in the two United States studies highlighted in Table 9, for example, sunflower values were 0.27 ( Prince et al., 2001) and 0.40 ( Johnson et al., 2006) and oat was given as 0.52 (Prince et al., 2001) or 0.44 ( Johnson et al., 2006). Smil (1999) estimated global crop residue resources using single HI values of 0.40 for cereals and 0.49 for legumes. With the exception of sunflower and grain sorghum, the values in our current dataset tend to be lower than those used by Prince et al. (2001), Johnson et al. (2006), and Bolinder et al. (2007) (Table 9). Our HI review indicates that under Australian conditions a higher quantity of residues would be returned per unit production than noted in other regions of the world. In addition, none of the abovementioned studies considered variation in HI or included such an exhaustive assessment of crop HI. The approach of using a single value for HI belies the substantial variation that exists in HI of Table 9 Crop harvest indices used in the studies of Prince et al. (2001) and Johnson et al. (2006) in the United States and Bolinder et al. (2007) in Canada, along with mean values from the Australian database (averaged site year)

a

Crop

From United States (Prince et al., 2001)

From United States (Johnson et al., 2006)

From Canada (Bolinder et al., 2007)

Australian means

Corn grain Soybeana Oats Barley Wheat Sunflower Sorghum

0.53 0.42 0.52 0.50 0.50 0.27 0.44

0.53 0.46 0.44 0.50 0.45 0.40 0.47

0.50 0.40 0.53 0.53 0.40 – 0.25

0.52 0.24 0.30 0.38 0.36 0.40 0.45

Australian data are for n ¼ 3 only, from Herridge and Holland (1992).


208

field crops in Australia which would create significant uncertainty in C balances. Developing an understanding of the factors operating to cause variation in HI of field-grown crops may enable HI to be varied according to local and seasonal factors and increase the accuracy of estimated crop C balances. For some crops, such as canola, which appear to have less variable HI, using a single value for C-accounting purposes may be adequate, but for many other crops, including wheat, the affect of seasonal and regional variation in HI on C accounting needs to be assessed as a potential source of error in Australia. As harvest index is known to vary as a function of environment (Prihar and Stewart, 1990) and some crops are grown across a large range of environments in Australia, it is not surprising that such variation in HI is observed.

6. Summary and Conclusions The efficiency with which crops produce grain from dry matter is the first determinant of harvest index. In this context, cereals, which have the highest carbohydrate content, tend to have the highest harvest indices. The high energy cost of lipid production makes species with a high oil content (canola, soybean, peanut) inefficient at producing grain from assimilate, with high protein seeded species (pea, lentil, chickpea) being intermediate. For most grain crops in Australia, a harvest index rarely exceeds 0.5, excepting where fallen leaves are not taken into account in which case apparent HI may exceed 0.6. Certainly values above 0.5 should be carefully scrutinized in Australia, especially for grain legumes and oilseeds. In more favorable environments where the grain-filling period is cooler (e.g., some parts of Europe and New Zealand), harvest indices above 0.5 may be more readily achieved than in Australia. The crop yields and dry matter data summarized here may be skewed toward higher values compared with commercial crops due to probable higher average inputs on experimental sites than commercial crops, higher grain losses during harvest of commercial crops, and the range of sites for which detailed data are available not being equally distributed across the crops sowing regions. However, with HI being a ratio and not an absolute amount our dataset may more closely represent HI of crops under commercial growing conditions. Delayed sowing reduces the vegetative phase of crop growth and increases the harvest index for most crops, as the reproductive phase then becomes proportionately longer. Predicting or modeling crop yield or harvest index responses to temperature extremes during the reproductive phase is problematic due to differences among genotypes, and an inability to


209

accurately model both actual ambient temperatures and to reliably quantify grain damage from temperature extremes. Clear relationships between fractional water use after anthesis and crop harvest index have been observed for glasshouse plants; however, while a theoretical framework has been provided for fractional postanthesis water use impacting on crop harvest index, unequivocal field data illustrating a pivotal role for profligate early water use reducing crop harvest index are lacking. Relations between HI and the fraction of water transpired after anthesis observed in controlled experiments may thus not be easily seen in field crops since other mitigating factors are not controlled, and furthermore, a complex interactions of cultivar, site, season, and management may dampen variation in crops HI. It is clear that crop nitrogen supply is implicated in variable harvest index of wheat and barley. Gross changes in HI might be able to be predicted for determinate crops from an assessment of matching climatic and soil fertility data. For indeterminate crops this challenge will be greater due to concurrent vegetative and reproductive growth and development. Crop harvest index is a critical term in C-accounting systems for agriculture and we present here the first thorough review of variation in crop harvest indices. It must be stressed that the data presented in the present analysis were a sample from the research literature, not from all crops in all seasons. As such they represent a biased sample based on where and when research plots are located and not a random sample of all Australian crops, or even a strategic sampling or survey of Australian commercial crops. We will continue to build our database and are conducting an analysis of the variation in crop HI in our dataset as a function of seasonal climatic conditions, with a view to finding relationships between local climate and HI which may be used to vary HI temporally and spatially and improve C-accounting practices for Australian agriculture.

ACKNOWLEDGMENTS This work was supported by funding from NCAS (the National Carbon Accounting System) within the Department of Climate Change. We would also like to express our sincere gratitude to the many colleagues who so kindly contributed their data to this exercise. Dr John Angus from CSIRO provided valuable advice on a draft of this manuscript.

REFERENCES Adcock, D. (2006). Soil water and nitrogen dynamics of farming systems on the upper Eyre Peninsula, South Australia. PhD, University of Adelaide, Adelaide. Adcock, D., McNeill, A. M., McDonald, G. K., and Armstrong, R. D. (2007). Subsoil constraints to crop production on neutral and alkaline soils in south-eastern Australia: A review of current knowledge and management strategies. Aust. J. Exp. Agric. 47, 1245–1261.

210


Adger, W. N., and Subak, S. (1996). Estimating above-ground carbon fluxes from UK agricultural land. Geogr. J. 162, 191–204. Adisarwanto, T., and Knight, R. (1997). Effect of sowing date and plant density on yield and yield components in the faba bean. Aust. J. Agric. Res. 48, 1161–1168. Allen, H. M., Pumpa, J. K., and Batten, G. D. (2001). Effect of frost on the quality of samples of Janz wheat. Aust. J. Exp. Agric. 41, 641–647. Angus, J., and Fischer, R. A. (1991). Grain and protein responses to nitrogen applied to wheat growing on a red earth. Aust. J. Agric. Res. 42, 735–746. Angus, J., and van Herwaarden, A. (1989). Comparative productivity of oilseeds and cereals. In ‘‘5th Australian Agronomy Conference’’(M. W. Perry, Ed.), p. 525. Aust. Soc. Agron., Perth, Western Australia. Angus, J., and van Herwaarden, A. (2001). Increasing water use and water use efficiency in dryland wheat. Agron. J. 93, 290–298. Armstrong, E. L., and Pate, J. S. (1994a). The field pea crop in S.W. Australia I. Patterns of biomass production and photosynthetic performance in genotypes of contrasting morphology. Aust. J. Agric. Res. 45, 1347–1362. Armstrong, E. L., and Pate, J. S. (1994b). The field pea crop in S.W. Australia II. Effects of contrasting morphology and environment on reproductive performance. Aust. J. Agric. Res. 45, 1363–1378. Armstrong, E. A., Heenan, D. P., Pate, J. S., and Unkovich, M. J. (1997). Nitrogen benefits of lupins, field pea and chickpea to wheat production in south-eastern Australia. Aust. J. Agric. Res. 48, 39–47. Atkins, C., and Pigeaire, A. (1993). Application of cytokinins to flowers to increase pod set in Lupinus angustifolius L. Aust. J. Agric. Res. 44, 1799–1819. Austin, R., Bingham, J., Blackwell, R., Evans, L. T., Ford, M., Morgan, C., and Taylor, M. (1980). Genetic improvement in winter wheat yields since 1900 and associated physiological changes. J. Agric. Sci. (Camb.) 94, 675–689. Australian Bureau of Statistics. (2002). AgStats on the GSP Australian Bureau of Statistics, Canberra. Database on compact disc. Bakker, D. (2008). Crop establishment and productivity with improved root zone drainage. In ‘‘GRDC Farming Systems Crop Updates’’ (D. Abrecht, Ed.), pp. 89–93. Agriculture WA, Perth. Bange, M., Hammer, G., and Rickert, K. (1977). Environmental control of potential yield of sunflower in the subtropics. Aust. J. Agric. Res. 48, 231–240. Batten, G., and Khan, M. (1987a). Uptake and utilisation of phosphorus and nitrogen by bread wheats grown under natural rainfall. Aust. J. Exp. Agric. 27, 405–410. Batten, G., and Khan, M. A. (1987b). Effect of time of sowing on grain yield, and nutrient uptake of wheats with contrasting phenology. Aust. J. Exp. Agric. 27, 881–887. Batten, G., Fettell, N., Mead, J., and Khan, M. (1999). Effect of sowing date on the uptake and utilisation of phosphorus by wheat (cv. Osprey) grown in central New South Wales. Aust. J. Exp. Agric. 39, 161–170. Beecher, H. G., Thompson, J. A., Bacon, P. E., and Heenan, D. P. (1994). Effect of cropping sequences on soil nitrogen levels, rice growth, and grain yields. Aust. J. Exp. Agric. 34, 977–986. Bell, M., Harch, G., and Wright, G. (1991). Plant population studies on peanut (Arachis hypogaea L.) in subtropical Australia 1. Growth under fully irrigated conditions. Aust. J. Exp. Agric. 31, 535–543. Berger, J., Turner, N., Siddique, K. H. M., Knights, E., Brinsmead, R., Mock, I., Edmonson, C., and Khan, T. (2004). Genotype by environment studies across Australia reveal the importance of phenology for chickpea (Cicer arietinum L.) improvement. Aust. J. Agric. Res. 55, 1071–1084. Birch, C., McLean, G., and Sawers, A. (2006). Analysis of high yielding maize production— A study based on a commercial crop. In ‘‘‘‘Water to Gold’’ 6th Triennial Conference’’


211

(E. Humphreys, K. O’Keefe, N. Hutchins, and R. Gill, Eds.), pp. 55–65. Maize Association of Australia, Griffith. Bishop, A. C., and Mendham, N. J. (1996). Quantifying the yield-density relationship for narrow-leafed lupin (Lupinus angustifolius) in Tasmania. In ‘‘8th Australian Agronomy Conference’’ (M. Asghar, Ed.), pp. 100–103. Australian Society of Agronomy, Toowoomba. Boer, R., Campbell, L. C., and Fletcher, D. J. (1993). Characteristics of frost in a major wheat-growing region of Australia. Aust. J. Agric. Res. 44, 1731–1743. Bolinder, M. A., Angers, D. A., and Dubuc, J. P. (1997). Estimating shoot to root ratios and annual carbon inputs in soils for cereal crops. Agric. Ecosyst. Environ. 61–66, 61. Bolinder, M. A., Janzen, H. H., Gregorich, E. G., Angers, D. A., and VandenBygaart, A. J. (2007). An approach for estimating net primary productivity and annual carbon inputs to soil for common agricultural crops in Canada. Agric. Ecosyst. Environ. 118, 29–42. Bondeau, A., Smith, P. C., Zaehle, S., Schaphoff, S., Lucht, W., Cramer, W., Gerten, D., Lotze-Campen, H., Muller, C., Reichstein, M., and Smith, B. (2007). Modelling the role of agriculture for the 20th century global terrestrial carbon balance. Glob. Change Biol. 13, 679–706. Boonchoo, S., Fukai, S., and Hetherington, S. E. (1998). Barley yield and grain protein concentration as affected by assimilate and nitrogen availability. Aust. J. Agric. Res. 49, 695–706. Calderini, D. F., Reynolds, M. P., and Slafer, G. A. (1999). Genetic gains in wheat yield and associated physiological changes during the twentieth century. In ‘‘Wheat ecophysiology and yield determination’’ (E. Satorre and G. Slafer, Eds.), pp. 351–377. Food Products Press, New York. Cantero-Martinez, C., O’Leary, G., and Connor, D. (1995). Stubble retention and nitrogen fertilisation in a fallow-wheat rainfed cropping system. 1. Soil water and nitrogen conservation, crop growth and yield. Soil Till. Res. 34, 79–94. Cassells, J., and Armstrong, E. L. (1998). Timing all important in boosting quality. Farming Ahead 81, 63–64. Connor, D., Jones, T., and Palta, J. (1985). Response of sunflower to strategies of irrigation I. Growth, yield and the efficiency of water use. Field Crops Res. 10, 15–36. Constable, G. A., and Gleeson, A. C. (1977). Growth and distribution of dry matter in cotton (Gossypium hirsutum L.). Aust. J. Agric. Res. 28, 249–256. Cornish, P., and Murray, G. (1989). Low rainfall rarely limits wheat yields in southern New South Wales. Aust. J. Exp. Agric. 29, 77–83. Croser, J. S., Clarke, H. J., Siddique, K. H. M., and Khan, T. N. (2003). Low-temperature stress: Implications for Chickpea (Cicer arietinum L.) improvement. Crit. Rev. Plant Sci. 22, 185. Dang, Y. P., Dalal, R. C., Routley, R., Schwenke, G. D., and Daniells, I. (2006). Subsoil constraints to grain production in the cropping soils of the north-eastern region of Australia: An overview. Aust. J. Exp. Agric. 46, 19–35. De Costa, W. A. J. M., Dennett, M. D., Ratnaweera, U., and Nyalemegbe, K. (1997). Effects of different water regimes on field-grown determinate and indeterminate faba bean (Vicia faba L.). II. Yield, yield components and harvest index. Field Crops Res. 52, 169–178. Dear, B. S., Kaiser, A., and Piltz, J. (2005). Yield and digestibility of legume and oat forages, Department of Primary Industries (NSW), Primefact 52. Delane, R. J., Matthews, J. C., and Hamblin, J. (1988). Effect of genotype and environment on harvest index of L. angustifolius. In ‘‘Proceedings 5th International Lupin Conference, Poznan, Poland’’, p. 527. Donald, C. M., and Hamblin, J. (1976). The biological yield and harvest index of cereals as agronomic and plant breeding criteria. Adv. Agron. 28, 361–405.

212


Dracup, M. (1994). Physiological aspects to lupin improvement for water-limited Mediterranean environments. In ‘‘First Australian Lupin Technical Symposium’’ (M. Dracup and J. Palta, Eds.), pp. 140–151. Agriculture Western Australia, Perth. Dracup, M., and Kirby, E. J. M. (1996). Pod and seed growth and development of narrowleafed lupin in a water limited Mediterranean-type environment. Field Crops Res. 48, 209–222. Dracup, M., Reader, M., and Palta, J. (1998a). Variation in yield of narrow-leafed lupin caused by terminal drought. Aust. J. Agric. Res. 49, 799–810. Dracup, M., Turner, N. C., Tang, C., Reader, M., and Palta, J. (1998b). Responses to abiotic stresses. In ‘‘Lupins as Crop Plants: Biology, Production and Utilisation’’ ( J. S. Gladstones, C. A. Atkins, and J. Hamblin, Eds.), CABI International, Wallingford. Duggan, B., Richards, R., van Hewaarden, A., and Fettell, N. (2005). Agronomic evaluation of a tiller inhibition gene (tin) in wheat. I. Effect on yield, yield components, and grain protein. Aust. J. Agric. Res. 56, 169–178. Dunn, B. W., and Beecher, H. G. (1994). Green manuring legume pasture for aerial-sown rice. Aust. J. Exp. Agric. 34, 967–975. Duthion, C., and Pigeaire, A. (1993). Stability of the nitrogen concentration of inoculated white lupins during pod development. Ann. Bot. 72, 55–61. Egan, J., and Crouch, J. (2004). Faba bean—2004 variety trials. In ‘‘South Australian Field Crop Evaluation Program 2004/2005’’ (R. Wheeler and L. McMurray, Eds.), p. 9. Primary Industries and Resources, South Australia, Adelaide. Egan, J., McCormack, P., Bennie, M., and McEvoy, K. (1994). Grain legumes for low rainfall areas. ‘‘Harvest and Livestock Report for Upper Eyre Peninsula 1993/94’’, pp. 35–37. Primary Industries South Australia, Port Lincoln. Erskine, W., Nassib, A., and Telaye, A. (1988). Breeding for morphological traits. In ‘‘World crops: Cool season food legumes’’ (R. J. Summerfield, Ed.), pp. 117–127. Kluwer Academic, Dordrecht. Evans, L. T. (1993). Crop evolution, adaptation and yield. Cambridge University Press, Cambridge. Farre, I., Robertson, M., Asseng, S., French, R., and Dracup, M. (2004). Simulating lupin development, growth, and yield in a Mediterranean environment. Aust. J. Agric. Res. 55, 863–877. Fathi, G., McDonald, G., and Lance, R. (1997). Effects of post-anthesis water stress on the yield and grain protein concentration of barley grown at two levels of nitrogen. Aust. J. Agric. Res. 48, 67–80. Ferraris, R., and Charles-Edwards, D. (1986). A comparative analysis of the growth of sweet and forage sorghum crops. I Dry matter production, phenology and morphology. Aust. J. Agric. Res. 37, 495–512. Fischer, R. A. (1979). Growth and water limitation to dryland wheat yield in Australia: A physiological framework. J. Aust. Inst. Agric. Sci. 45, 83–94. Fischer, K. S., and Palmer, A. (1983). Maize. In ‘‘Potential productivity of field crops under different environments’’ (Anon, Ed.), pp. 155–180. International Rice Research Institute, Los Banos, Philipines. Flood, R., and Martin, P. (2001). Nitrogen accumulation and distribution at anthesis and maturity in ten wheats grown at three sites in north-western Victoria. Aust. J. Exp. Agric. 41, 533–540. French, R. J., and Ewing, M. A. (1989). Soil type influences the relative yields of different cereals and crop legumes in the Western Australian wheatbelt. Aust. J. Exp. Agric. 29, 829–835. French, R. J., and Turner, N. C. (1991). Water deficits change dry matter partitioning and seed yield in narrow-leafed lupins (Lupinus angustifolius L.). Aust. J. Agric. Res. 42, 471–484.


213

Gomez-Macpherson, H., and Richards, R. A. (1995). Effect of sowing time on yield and agronomic characteristics of wheat in south-eastern Australia. Aust. J. Agric. Res. 46, 1381–1399. Goyne, P., Milroy, S., Lilley, J., and Hare, J. (1993). Radiation interception, radiation use efficiency and growth of barley cultivars. Aust. J. Agric. Res. 44, 1351–1366. Greenwood, E. A., Farrington, P., and Beresford, J. D. (1975). Characteristics of the canopy, root system and grain yield of a crop of Lupin angustifolius cv. Unicrop. Aust. J. Agric. Res. 26, 497–510. Gregory, P. J., Palta, J. A., and Batts, G. R. (1995). Root systems and root:mass ratio-carbon allocation under current and projected atmospheric conditions in arable crops. Plant Soil 187, 221–228. Haigh, B. (2006). Personal communication. NSW Agriculture, Tamworth. Hall, A. J., Connor, D. J., and Whitfield, D. M. (1989). Contribution of pre-anthesis assimilates to grain-filling in irrigated and water-stressed sunflower crops I. Estimates using labelled carbon. Field Crops Res. 20, 95–112. Hamblin, J., Delane, R., Bishop, A., and Gadstones, J. Y. (1986). Yield potential of reduced branching lupins (Lupinus angustifolius) on sandy soils in a short season environment. Aust. J. Agric. Res. 37, 611–620. Hardie, D. C., Baker, G. J., and Marshall, D. R. (1995). Field screening of Pisum accessions to evaluate their susceptibility to the pea weevil (Coleoptera: Bruchidae). Euphytica 84, 155–161. Herridge, D. F., and Holland, J. F. (1992). Production of summer crops in northern New South Wales. I. Effects of tillage and double cropping on growth, grain and N yields of six crops. Aust. J. Agric. Res. 43, 105–122. Hirth, J., Haines, P., Ridley, A. M., and Wilson, K. (2001). Lucerne in crop rotations on the Riverine Plains. 2. Biomass and grain yields, water use efficiency, soil nitrogen, and profitability. Aust. J. Agric. Res. 52, 279–293. Hocking, P., Randall, P., and DeMarco, D. (1997). The response of dryland canola to nitrogen fertilizer: Partitioning and mobilization of dry matter and nitrogen, and nitrogen effects on yield components. Field Crops Res. 54, 201–220. Hodgson, A. S. (1978). Rapeseed adaptation in northern New South Wales. II Predicting plant development of Brassica campestris L. and Brassica napus L. and its implications for planting time, designed to avoid water deficit and frost. Aust. J. Agric. Res. 29, 711–726. Hoffman, T. K., and Kolb, F. L. (1998). Effects of barley yellow dwarf virus on yield and yield components of drilled winter wheat. Plant Dis. 82, 620–624. Hooda, R. S., Rao, A. S., Luthra, Y. P., Sheoran, I. S., and Singh, R. (1986). Partitioning and utilization of carbon and nitrogen for dry matter and protein production in chickpea (Cicer arietinum L.). J. Exp. Bot. 37, 1492–1502. Holland, J. F., and Felton, W. L. (1989). Effect of tillage practice on grain sorghum production in northern New South Wales. Aust. J. Exp. Agric. 29, 843–848. Horie, T., Ohnishi, M., Angus, J. F., Lewin, L. G., Tsukaguchi, T., and Matano, T. (1997). Physiological characteristics of high-yielding rice inferred from cross-location experiments. Field Crops Res. 52, 55–67. Huhn, M. (1990a). Comments on the calculation of mean harvest indices. J. Agron. Crop Sci. 165, 86–93. Huhn, M. (1990b). On the variability of harvest indices. J. Agron. Crop Sci. 164, 271–281. Humphreys, E., Smith, D. J., White, R. J., Xevi, E., and Godwin, D. C. (2003). Maizeman—A decision support system for water and nitrogen management. In ‘‘5th Australian Maize Conference’’ (C. J. Birch and S. R. Wilson, Eds.), pp. 131–134. Maize Association of Australia, Toowoomba. Institute of Sustainable Irrigated Agriculture (Vic). (1995). Biennial Research Report (1994–1995). Tatura.

214


Jackson, R. B., Canadell, J. J., Mooney, H. O., and Schulze, E. (1996). A global analysis of root distributions for terrestrial biomes. Oecologia 108, 389–411. Jensen, C. R., Joernsgaard, B., Andersen, M. N., Christiansen, J. L., Mogensen, V. O., Friis, P., and Petersen, C. T. (2004). The effect of lupins as compared with peas and oats on the yield of the subsequent winter barley crop. Eur. J. Agron. 20, 405–418. Johnson, J. M. F., Allmaras, R. R., and Reicosky, D. C. (2006). Estimating source carbon from crop residues, roots and rhizodeposits using the National grain-yield database. Agron. J. 98, 622–636. Kirkegaard, J., Lilley, J., Howe, G., and Graham, N. (2007). Impact of subsoil water use on wheat yield. Aust. J. Agric. Res. 58, 303–315. Lenssen, A. W., Johnson, G. D., and Carlson, G. R. (2007). Cropping sequence and tillage system influences annual crop production and water use in semiarid Montana, USA. Field Crops Res. 100, 32–43. Leport, L., Turner, N. C., French, R. J., Barr, M. D., Duda, R., Davies, S. L., Tennant, D., and Siddique, K. H. M. (1999). Physiological responses of chickpea genotypes to terminal drought in a Mediterranean-type environment. Eur. J. Agron. 11, 279–291. Lobell, D. B., Hicke, J. A., Asner, G. P., Field, C. B., Tucker, C. J., and Los, S. O. (2002). Satellite estimates of productivity and light use efficiency in United States agriculture, 1982–98. Glob. Change Biol. 8, 722–735. L’opez-Bellido, F., L’opez-Bellido, L., and L’opez-Bellido, R. J. (2005). Competition, growth and yield of faba bean (Vicia faba L.). Eur. J. Agron. 23, 359–378. Lopez Pereira, M., Trapani, N., and Sadras, V. O. (2000). Genetic improvement of sunflower in Argentina between 1930 and 1995: Part III. Dry matter partitioning and grain composition. Field Crops Res. 67, 215–221. Loss, S. P., and Siddique, K. (1997). Adaptation of faba bean (Vicia faba L.) to dryland Mediterranean-type environments I. Seed yield and yield components. Field Crops Res. 52, 17–28. Loss, S., Siddique, K., Jettner, R., and Martin, L. (1998). Responses of faba bean (Vicia faba L.) to sowing rate in south-western Australia I. Seed yield and economic optimum plant density. Aust. J. Agric. Res. 49, 989–997. Mason, I., and Fischer, R. A. (1986). Tillage practices and the growth and yield of wheat in southern New South Wales: Lockhart, in a 450 mm rainfall region. Aust. J. Exp. Agric. 26, 457–468. Mayer, J. D., Lawn, R. J., and Byth, D. E. (1991). Adaptation of soybean [Glycine max (L.) Merrill] to the dry season of the tropics. II. Effects of genotype and environment on biomass and seed yield. Aust. J. Agric. Res. 42, 517–530. McDonald, G. (2006). Effects of soil properties on variations in growth, grain yield and nutrient concentration of wheat and barley. Aust. J. Exp. Agric. 46, 93–105. McKenzie, N., Jacquier, D., Isbell, R., and Brown, K., (Eds.) (2004). In ‘‘Australian Soils and Landscapes: An Illustrated Compendium’’. CSIRO Publishing, Collingwood. McKirdy, S. J., Jones, R. A. C., and Nutter, F. W. (2002). Quantification of yield losses caused by barley yellow dwarf virus in wheat and oats. Plant Dis. 86, 769–773. McMurray, L. (2005). Filed pea trials. ‘‘Crop Variety Evaluation 2004’’, South Australian Research and Development Institute, Adelaide. McMurray, L., and Matic, R. (2003). Vetch trials. ‘‘Crop Variety Evaluation 2003’’, South Australian Research and Development Institute, Adelaide. Mitchell, J. H., Fukai, S., and Cooper, M. (1996). Influence of phenology on grain yield variation among barley cultivars grown under terminal drought. Aust. J. Agric. Res. 47, 757–774. Nanja Reddy, Y. A., Uma Shaanker, R., Prasad, T. G., and Udaya Kumar, M. (2003). Physiological approaches to improving harvest index and productivity in sunflower. Helia 26, 81–90.


215

Nicholls, N., Drosdowsky, W., and Lavery, B. (1997). Australian rainfall variability and change. Weather 52, 66–71. Nix, H. A. (1975). The Australian climate and its effects on grain yield and quality. In ‘‘Australian field crops: Volume 1. Wheat and other temperate cereals’’ (A. Lazenby and E. Matheson, Eds.), pp. 183–226. Angus and Robertson, North Ryde. NVT. (2008). National Variety Trials Australian Crop Accreditation System (ACAS), Horsham (Online. Vol. 2008). Ockerby, S. E., and Fukai, S. (2001). The management of rice grown on raised beds with continuous furrow irrigation. Field Crops Res. 69, 215–226. O’Connell, M., Connor, D., and O’Leary, G. (2002). Crop growth, yield and water use in long fallow and continuous cropping sequences in the Victorian mallee. Aust. J. Exp. Agric. 42, 971–983. O’Leary, G. J., and Connor, D. (1997). Stubble retention and tillage in a semi-arid environment: 3. Response of wheat. Field Crops Res. 54, 39–50. O’Mara, B. (undated). Chickpea Harvest and Seed Storage. Pulse Australia, Edgecliffe, NSW. Palta, J., and Fillery, I. (1995). N application increases pre-anthesis contribution of dry matter to grain yield in wheat grown on a duplex soil. Aust. J. Agric. Res. 46, 507–518. Palta, J., and Ludwig, C. (1996). Pod set and seed yield as affected by cytokinin application and terminal drought in narrow-leafed lupin. Aust. J. Agric. Res. 48, 81–90. Palta, J., and Ludwig, C. (1998). Yield response of narrow-leafed lupin plants to variations in pod number. Aust. J. Agric. Res. 49, 63–68. Palta, J., and Ludwig, C. (2000). Elevated CO2 during pod filling increased seed yield but not harvest index in indeterminate narrow-leafed lupin. Aust. J. Agric. Res. 51, 279–286. Palta, J., and Plaut, Z. (1999). Yield and components of seed yield of indeterminate narrowleafed lupin (Lupinus angustifolius L.) subjected to transient water deficit. Aust. J. Agric. Res. 50, 1225–1232. Palta, J., Turner, N. C., French, R. J., and Buirchell, B. (2003). Selection for high lupin yield under terminal drought. ‘‘Agribusiness Crop Updates 2003’’, Agriculture WA, Perth. Passioura, J. (1977). Grain yield, harvest index, and water use of wheat. J. Aust. Inst. Agric. Sci. 43, 117–120. Passioura, J. (2006). Increasing crop productivity when water is scarce—From breeding to field management. Agric. Water Manag. 80, 176–196. Pate, J. S., and Armstrong, E. L. (1996). Pea. In ‘‘Photoassimilate distribution in plants and crops’’ (E. Zamski and A. Schaffer, Eds.), pp. 625–642. Marcel Dekker Inc., New York. Pate, J. S., and Layzell, D. B. (1990). Energetics and costs of nitrogen assimilation. In ‘‘The Biochemistry of Plants’’ (B. Mifflin and P. Lea, Eds.), Vol. 16, pp. 1–42. Academic Press. Pate, J. S., Atkins, C. A., and Perry, M. W. (1980). Significance of photosynthate produced at different stages of growth as carbon source for fruit filling and seed reserve accumulation in Lupinus angustifolius L. Aust. J. Plant Physiol. 7, 283–297. Penning de Vries, F. W. T., Brunsting, A. H. M., and Van Laar, H. H. (1974). Products, requirements and efficiency of biosynthesis a quantitative approach. J. Theor. Biol. 45, 339–377. Penning de Vries, F. W., Van Laar, H., and Chardon, M. (1983). Bioenergetics of growth of seeds, fruits and storage organs. In ‘‘Potential Productivity of Field Crops Under Different Environments’’ (Anon, Ed.), pp. 37–59. International Rice Research Institute, Los Banos, Philipines. Perry, M. W., and D’Antuono, M. F. (1989). Yield improvement and associated characteristics of some Australian spring wheat cultivars introduced between 1860 and 1982. Aust. J. Agric. Res. 40, 457–472. Perry, M. W., Delane, R. J., Tennant, D., and Hamblin, A. P. (1986). The growth and grain yield of lupins in relation to the soil water balance. ‘‘IV International Lupin Conference’’,

216


pp. 112–128. Western Australian Department of Agriculture: International Lupin Association, Geraldton. Petterson, D. S., and Mackintosh, J. B. (1994). The chemical composition and nutritive value of Australian grain legumes.Grains Research and Development Corporation, Canberra. Prihar, S. S., and Stewart, B. A. (1990). Using upper-bound slope through origin to estimate genetic harvest index. Agron. J. 82, 1160–1165. Prince, S., Haskett, J., Steininger, M., Strand, H., and Wright, R. (2001). Net primary production of the US mid-West croplands from agricultural harvest yield data. Ecol. Appl. 11, 1194–1205. Pritchard, K., and Small, S. (1992). Soil and water management for irrigated crops on a redbrown earth soil. In ‘‘6th Australian Agronomy Conference’’ (D. Hutchinson and P. Vickery, Eds.), p. 251. Australian Society of Agronomy, Armidale. Reader, M. A., Dracup, M., and Atkins, C. A. (1997). Transient high temperatures during seed growth in narrow-leafed lupin (Lupinus angustifolius L.) II. Injuriously high pod temperatures are likely in Western Australia. Aust. J. Agric. Res. 48, 1179–1186. Richards, R., and Townley-Smith, T. (1987). Variation in leaf area development and its effect on water use, yield and harvest index of droughted wheat. Aust. J. Agric. Res. 38, 983–992. Robertson, M. J., Holland, J. F., Cawley, S., Bambach, R., Cocks, B., and Watkinson, A. (2001). Phenology of canola cultivars in the northern region and implications for frost risk. ‘‘Science and technology: Delivering results for agriculture. Proc. 10th Australian Agronomy Conference’’, Australian Society of Agronomy, Hobart. Robertson, M., Cawthray, S., Birch, C., Crawford, M., Dalgleish, N., and Hammer, G. (2003). Managing the risks of growing dryland maize in the northern region. In ‘‘5th Australian Maize Conference’’ (C. Birch and S. R. Wilson, Eds.), pp. 112–119. Maize Association of Australia, Toowoomba. Robertson, M. J., Holland, J. F., and Bambach, R. (2004). Response of canola and Indian mustard to sowing date in the grain belt of north-eastern Australia. Aust. J. Exp. Agric. 44, 43–52. Robinson, D. (2004). Scaling the depths: Below-ground allocation in plants, forests and biomes. Funct. Ecol. 18, 290–295. Rose, I., McCaffery, D., Lowien, J., and McGufficke, R. (2005). Soybeans: Inland northern NSW planting guide 2005–2006. Department of Primary Industries NSWRep. No. 103. Sadras, V. O., and Connor, D. (1991). Physiological basis of the response of harvest index to the fraction of water transpired after anthesis: A simple model to estimate harvest index for determinate species. Field Crops Res. 26, 227–239. Saxena, N. P., Natarajan, M., and Reddy, M. S. (1983). Chickpea, pigeon pea and groundnut. In ‘‘Potential productivity of field crops under different environments’’ (Anon, Ed.), pp. 281–305. International Rice Research Institute, Los Banos, Philipines. Schapaugh, W., and Wilcox, J. (1980). Relationships between harvest indices and other plant characteristics in soybeans. Crop Sci. 20, 529–533. Schultz, J. E. (1995). Crop production in a rotation trial at Tarlee, South Australia. Aust. J. Exp. Agric. 35, 865–876. Schwenke, G., Peoples, G., Turner, G., and Herridge, D (1998). Does nitrogen fixation of commercial dryland chickpea and faba bean crops in north-west New South Wales maintain or enhance soil nitrogen? Aust. J. Exp. Agric. 38, 61–70. Seymour, M. (1996). Response of narrow-leafed lupin (Lupinus angustifolius L.) to foliar application of growth regulators in Western Australia. Aust. J. Exp. Agric. 36, 473–478. Sheehy, J. E., Peng, S., Dobermann, A., Mitchell, P. L., Ferrer, A., Yang, J., Zou, Y., Zhong, X., and Huang, J. (2004). Fantastic yields in the system of rice intensification: Fact or fallacy? Field Crops Res. 88, 1–8.


217

Shepherd, K. D., Cooper, P. J., Allan, A., Drennan, D., and Keatinge, D. (1987). Growth, water use and yield of barley in Mediterranean-type environments. J. Agric. Sci. (Camb.) 108, 365–378. Shrestha, R., Turner, N. C., Siddique, K. H. M., Turner, D., and Speijers, J. (2006). A water deficit during pod development in lentils reduces flower and pod numbers but not seed size. Aust. J. Agric. Res. 57, 427–438. Si, P., and Walton, G. (2004). Determinants of oil concentration and seed yield in canola and Indian mustard in the lower rainfall areas of Western Australia. Aust. J. Agric. Res. 55, 367–377. Siddique, K. H. M., Belford, R. K., Perry, M. W., and Tennant, D. (1989). Growth, development and light interception of old and modern wheat cultivars in a Mediterranean-type environment. Aust. J. Agric. Res. 40, 473–487. Siddique, K., Loss, S. P., Regan, K. L., and Pritchard, D. L. (1998). Adaptation of lentil (Lens culinaris Medik) to short season Mediterranean-type environments: Response to sowing rates. Aust. J. Agric. Res. 49, 1057–1066. Siddique, K. M. H., Regan, K. L., Tennant, D., and Thomson, B. D. (2001). Water use and water use efficiency of cool season grain legumes in low rainfall Mediterranean-type environments. Eur. J. Agron. 15, 267–280. Sinclair, T. (1993). Crop yield potential and fairy tales. In ‘‘International Crop Science 1’’ (D. Buxton, R. Shibles, R. Forsberg, B. Blad, K. Asay, G. Paulsen, and R. Wilson, Eds.), pp. 707–711. Crop Science Society America, Madison. Sinclair, T. (1994). Limits to crop yield? In ‘‘Physiology and determination of crop yield’’ (K. Boote, J. M. Bennett, T. Sinclair, and G. M. Paulsen, Eds.), pp. 509–532. ASA, CSSA, SSSA, Madison. Sinclair, T. (1998). Historical changes in harvest index and crop nitrogen accumulation. Crop Sci. 38, 638–643. Sinclair, T., and De Wit, C. T. (1975). Photosynthate and nitrogen requirements for seed production by various crops. Science 189, 565–567. Slafer, G., Araus, J., and Richards, R. (1999). Physiological traits that increase the yield potential of wheat. In ‘‘Wheat ecophysiology and yield determination’’ (E. Satorre and G. Slafer, Eds.), pp. 379–415. Food Products Press, New York. Smil, V. (1999). Crop Residues: Agriculture’s Largest Harvest. Biosci. 49, 299–308. Snowball, R. (1986). Swathing field crops in the south-west. J. Agric. West. Aust. 27, 84–88. Spenceley, J. (2005). Grain Sorghum. NSW Department Primary Industries, Sydney. Rep. No. P3.3.5. Stapper, M., and Fischer, R. A. (1990a). Genotype, sowing date and plant spacing influence on high-yielding irrigated wheat in southern New South Wales. I. Phasic development, canopy growth and spike production. Aust. J. Agric. Res. 41, 997–1019. Stapper, M., and Fischer, R. A. (1990b). Genotype, sowing date and plant spacing influence on high-yielding irrigated wheat in southern New South Wales. II. Growth, yield and nitrogen use. Aust. J. Agric. Res. 41, 1021–1041. Stapper, M., and Fischer, R. A. (1990c). Genotype, sowing date and plant spacing influence on high-yielding irrigated wheat in southern New South Wales. III. Potential yields and optimum flowering dates. Aust. J. Agric. Res. 41, 1041–1056. Stone, P., and Nicholas, M. (1995). A survey of the effects of high temperature during grain filling on yield and quality of 75 wheat cultivars. Aust. J. Agric. Res. 46, 475–492. ten Berge, H. F. M., Thiyagarajan, T. M., Shi, Q., Wopereis, M. C. S., Drenth, H., and Jansen, M. J. W. (1997). Numerical optimization of nitrogen application to rice. Part I. Description of MANAGE-N. Field Crops Res. 51, 29. Thomas, G. (2005). Grain yield and protein content for cropping systems at Dunkerry south trial 1996–2001.CSIRO, Toowoomba. Vol. 2006.

218


Thomas, Fukai, S. (1995). Growth and yield response of barley and chickpea to water stress under three environments in southeast Queensland. I. Light interception, crop growth and grain yield. Aust. J. Agric. Res. 46, 17–33. Thompson, J. A., and Heenan, D. P. (1989). The time of sowing of irrigated grain sorghum in the Murrumbidgee Valley. In ‘‘5th Australian Agronomy Conference’’ (M. W. Perry, Ed.), p. 509. Australian Society Agronomy, Perth. Thomas, Robertson, M. J., Fukai, S., and Peoples, M. B. (2004). The effect of timing and severity of water deficit on growth, development, yield accumulation and nitrogen fixation of mungbean. Field Crops Res. 86, 67. Turner, N. (1997). Further progress in crop water relations. Adv. Agron. 758, 293. Turner, N. C., Leport, L., French, R. J., Thomson, B. D., and Siddique, K. H. M. (1996). Adaptation of cool-season pulses to water-limited environments. In ‘‘8th Australian Agronomy Conference’’ (M. Asghar, Ed.), pp. 554–557. Australian Society of Agronomy, Toowoomba. Turner, N., Wright, G., and Siddique, K. H. M. (1999). Adaptation of grain legumes (pulses) to water-limited environments. Adv. Agron. 71, 193–229. Unkovich, M. J. (1991). The potential of the d15N natural abundance method for estimating the contribution of fixed nitrogen to the nitrogen economy of broadacre lupins (Lupinus angustifolius), M.Sc. Thesis, The University of Western Australia. Unkovich, M. J., Pate, J. S., and Hamblin, M. J. (1994). The nitrogen economy of broadacre lupin in southwest Australia. Aust. J. Agric. Res. 45, 149–164. Unkovich, M. J., Baldock, J., and Forbes, M. (2006). Australian Crop Yields and Harvest Indices (Microsoft Access Database).CSIRO Land and Water, Adelaide. Unkovich, M., Baldock, J., and Marvanek, S. (2009). Which broadacre crops and pastures should be included in a carbon accounting system for Australia’s cropping zone? Crop Pasture Sci. 60, 1–10. van Herwaarden, A., and Richards, R. (2002). Water-soluble carbohydrate accumulation in stems is related to breeding progress in Australian wheats. In ‘‘Plant Breeding for the 11th Millennium; 12th Australian Plant Breeding Conference’’ ( J. McComb, Ed.), pp. 878–882. Australasian Plant Breeding Conference, Perth, Vol. Compact Disc. van Herwaarden, A., Farquhar, G., Angus, J., Richards, R. A., and Howe, G. (1998a). ’Haying-off’, the negative grain yield response of dryland wheat to nitrogen fertiliser. I. Biomass, grain yield and water use. Aust. J. Agric. Res. 49, 1067–1081. van Herwaarden, A., Richards, R. A., Angus, J., and Farquhar, G. (1998b). Haying-off’, the negative grain yield response of dryland wheat to nitrogen fertiliser. II Carbohydrate and protein dynamics. Aust. J. Agric. Res. 49, 1083–1093. Vleeshouwers, L. M., and Verhagen, A. (2002). Carbon emission and sequestration by agricultural land use: A model study for Europe. Glob. Change Biol. 8, 519–530. Wahbi, A., and Gregory, P. J. (1989). Genotypic differences in root and shoot growth of barley (Hordeum vulgare) II. Field studies of growth and water use of crops grown in northern Syria. Aust. J. Exp. Agric. 25, 389–399. Ward, P. R., Dunin, F. X., and Micin, S. F. (2002). Water use and root growth by annual and perennial pastures and subsequent crops in a phase rotation. Agric. Water Manag. 53, 83–97. Wheeler, R., and McMurray, L., (Eds.) (2000). In ‘‘South Australian Field Crop Evaluation Program Annual Report 1999’’. South Australian Research and Development Institute, Adelaide. Whitehead, S. J., Summerfield, R. J., Muehlbauer, F. J., Coyneb, C. J., Ellis, R. H., and Wheeler, T. R. (2000). Crop improvement and the accumulation and partitioning of biomass and nitrogen in lentil. Crop Sci. 40, 110–120. Williams, R. L., and Angus, J. (1994). Deep floodwater protects high-nitrogen rice crops from low-temperature damage. Aust. J. Exp. Agric. 34, 927–932.


219

Wright, P., Morgan, J., and Cass, A. (1992). Effects of water stress on yield and the components of yield in canola and Indian mustard. ‘‘6th Australian Agronomy Conference’’, p. 617. Australian Society of Agronomy, Armidale. Zhang, X., and Evans, P. M. (2004). Grain yield production in relation to plant growth of wheat and canola following clover pastures in southern Victoria. Aust. J. Exp. Agric. 44, 1003–1012. Zhang, J., and Yang, J. (2004). Crop yield and water use efficiency: A case study in rice. In ‘‘Water Use Efficiency in Plant Biology’’ (M. A. Bacon, Ed.). Blackwell Publishing/ CRC Press, Oxford. Zubaidi, A., McDonald, G. K., and Hollamby, G. J. (1999). Shoot growth, root growth and grain yield of bread and durum wheat in South Australia. Aust. J. Exp. Agric. 39, 709–720. Zwer, P., and Hoppo, H. (2004). Oat trials. In ‘‘South Australian Field Crop Evaluation Program; Post Harvest Report 2003/2004’’ (R. Wheeler and L. McMurray, Eds.), South Australian Research and Development Institute, Adelaide.

C H A P T E R

S I X

The Role of Seed Ecology in Improving Weed Management Strategies in the Tropics Bhagirath S. Chauhan and David E. Johnson Contents 1. 2. 3. 4. 5.

Introduction Challenges for Weed Management in the Tropics Role of Seed Ecology in Determining Weed Populations Responses of Weed Seed Germination to Light Responses of Weed Seed Germination to Seed Scarification and Fire 5.1. Effect of seed scarification on germination 5.2. Effect of fire on germination 6. Responses of Weed Seed Germination to Seed Burial Depth, Tillage, and Surface Mulches 6.1. Effect of tillage systems on vertical seed distribution 6.2. Effect of seed burial depths on weed germination 6.3. Role of tillage in weed management in dry- and wet-land conditions 6.4. Role of surface mulches in rainfed and dry direct-seeded crops 7. Responses of Weed Seed Germination to Stresses 7.1. Effect of salt and moisture stress on weed germination 7.2. Role of flooding in lowland conditions 8. Harnessing Knowledge of Seed Ecology for Novel Improved Weed Management Strategies 9. Future Research Needs Acknowledgments References

222 223 224 226 229 229 230 232 232 232 236 239 242 242 244 247 249 252 252

Abstract Weed seed banks reflect past weed populations and management practices and are the source of weed infestations to come. The factors affecting weed seed germination, however, are often poorly understood. Depleting the soil seed Crop and Environmental Sciences Division, International Rice Research Institute, Metro Manila, Philippines Advances in Agronomy, Volume 105 ISSN 0065-2113, DOI: 10.1016/S0065-2113(10)05006-6

#


221

222

Bhagirath S. Chauhan and David E. Johnson

bank and influencing germination patterns are common goals of enduring cultural weed management practices. Greater understanding of the factors influencing the germination of weed seeds could facilitate the development of more effective cultural weed management practices through either suppressing germination or encouraging germination at times when seedlings can be readily controlled. Such cultural methods may contribute to overcoming problems such as feral crops (e.g., weedy rice), crop volunteers, and the evolution of herbicide resistance in weeds that have, in some systems, increased to a point where the lack of sustainable practices is a threat to productivity. Weed seed germination is commonly influenced by light exposure, soil moisture, burial depth through tillage, the use of mulches, fire for land clearance, and flooding of the soil. Harnessing these factors to influence germination can serve as major entry points for improved weed management. Diverse crop production systems provide a wide range of examples to illustrate how recent advances in the understanding of the responses of weed seed germination can be used to develop new and sustainable cultural management of weeds. Crop management practices, such as adopting no-till crops or delaying tillage, that increase weed seed exposure to predators (ants, beetles, etc.) could be incorporated into integrated weed management programs. Retention of crop residue on the soil surface under no-till systems can suppress weed seedling emergence, delay the time of emergence, and allow the crop to gain an advantage over weeds, and reduce the need for control. Rotation of tillage or crop establishment system could also be adopted to deflect the ‘‘trajectories’’ of likely weed population shifts. In rice, flooding after herbicide application or hand weeding can largely prevent the growth of weeds and reduce the need for further interventions.

1. Introduction Weeds are a major consideration for the effective management of all land and water resources, but their impact is greatest on agriculture. In most arable crop systems, management effort focuses on reducing weed density in the early stages of crop growth as later-emerging weeds often have little impact on crop yield (Zimdahl, 1988). Even weeds emerging late often produce some viable seeds and, once they enter the soil, contribute to the success of weeds (Cavers and Benoit, 1989; Gallandt, 2006). The weed seed bank, as a reservoir of weed seeds, largely determines the potential density and species composition of weeds that subsequently interfere with crops (Forcella, 1993). Seeds can be lost to the seed bank by germination, death, or predation, which in turn may be affected by the physical distribution of seeds in the soil, physiological changes in the seeds, and the flux of dormancies (Cousens and Mortimer, 1995). Seed bank dynamics regulate communities of many of the most important weed species; therefore, a better understanding of seed bank dynamics could contribute to the

Weed Seed Ecology

223

development of more efficient weed management systems. Instead of solely considering crop yield loss, management could also include strategies to deplete the weed seed bank and influence germination (Gallandt, 2006). The factors affecting weed seed germination, however, are often poorly understood. Greater understanding of the factors influencing the germination of weed seeds would facilitate the development of more effective cultural management practices through either suppressing germination or encouraging germination at times when seedlings can be readily controlled. Such cultural methods may be of increased importance as problems of feral crops, such as weedy rice (Oryza sativa L.), Avena fatua L. or other problem species, crop volunteers, and herbicide resistance have increased to a point where lack of sustainable control practices threatens production in some systems. Further, greater interest in cultural management practices is being generated by the demand to seek alternatives to herbicides for organic and low-input production systems. Weed seed germination is commonly influenced by soil moisture, seed burial depth due to tillage, the use of mulches, fire for land clearance, and flooding of the soil. Harnessing these factors to influence germination can provide a major entry point for improved weed management. Diverse crop production systems provide a wide range of examples to illustrate recent advances in the understanding of weed seed germination. These are then used to answer the question: ‘‘How can seed ecology contribute to improving weed management strategies in the tropics?’’

2. Challenges for Weed Management in the Tropics Weeds are a major biotic constraint to crop production in the tropics, and the pressing need to increase food production presents an immediate challenge for weed scientists. In Asia, actual losses in rice and wheat production due to weeds have been estimated as 10% and 13%, respectively (Oerke et al., 1994; Rao et al., 2007). Weed control in the tropics is commonly achieved by manual weeding, but this is becoming less common in many areas because of the increasing costs of labor resulting from the migration of rural labor to the cities. The declining availability of labor for agriculture has required farmers to seek alternatives to manual weeding, which has long provided farmers with the means to limit crop losses caused by weeds. Herbicide use in some countries has allowed a massive release of labor from agriculture (Nelson, 1996), and this trend is expected to continue. The use of herbicides, however, has been accompanied by concerns over the evolution of herbicide resistance in weeds, weed species population shifts, increased cost of chemical control measures, and concerns about

224


the environment (Buhler et al., 2002; Johnson and Mortimer, 2008; Primot et al., 2006). In India, for example, reliance on the herbicide isoproturon (substituted urea herbicide) led to a rapid increase in resistance in Phalaris minor Retz. in wheat (Triticum aestivum L.) crops (Malik and Singh, 1995). Weed surveys conducted during 2005–2006 revealed P. minor infestations in 82% of the wheat fields in Haryana, India (Singh, 2007). This weed has been reported to reduce wheat yield up to 80% (Singh et al., 1999). In addition to labor, concerns are increasing about supplies of irrigation water. Farmers in many areas, especially those growing rice (O. sativa L.), are likely to have only limited availability of irrigation water and, in the future, most of the dry-season areas in South and Southeast Asia will fall into an ‘‘economic water scarcity zone’’ (Bouman and Tuong, 2003). Water scarcity threatens the sustainability of production in these irrigated ecosystems since the crop may suffer from drought, and even limited shortages may make it unfeasible for farmers to flood rice fields to ensure adequate weed control. Changes in crop establishment methods have occurred in many countries. In the Indo-Gangetic Plains of India, for example, conventional tillage (CT) in wheat in the rice–wheat cropping system has been replaced by no-till (NT) in many areas (Singh, 2007). Similarly, different tillage systems for the establishment of corn (Zea mays L.) are also being explored that include reduced and minimum tillage (e.g., Nakamoto et al., 2006). Transplanting of rice in some countries has been replaced by direct seeding as farmers respond to increased costs or decreased availability of labor or water (Pandey and Velasco, 2005). The risk of crop yield loss due to competition from weeds in direct-seeded rice is higher than in transplanted rice because of the absence of the size differential between the crop and weeds and the suppressive effect of standing water on weed growth at crop establishment (Rao et al., 2007). Changes in crop management and establishment methods are likely to be associated with the size of weed infestations as well as a shift in the weed flora ( Johnson and Mortimer, 2005).

3. Role of Seed Ecology in Determining Weed Populations The seed bank in the soil is the primary source of annual weeds in most crop production systems (Buhler et al., 1997; Cavers, 1983). The seed bank commonly consists of both recent and older seeds shed and dispersed in an area (Dekker, 1999). In theory, depleting weed seed banks by preventing seed production and by management that provides a favorable environment for germination should be feasible (Buhler et al., 1997). In practice, however, managing seed banks is complex due to the difficulties of preventing

Weed Seed Ecology

225

the production and introduction of seeds, the persistence of a small percentage of the seed bank due to dormancy, and the high seed production potential of many weed species. It may be more realistic, rather than depleting the weed seed bank, to accept weed seed banks as a component of the agricultural environment and attempt to understand and predict their behavior, and devise management systems to minimize the impacts of the resultant weeds (Buhler et al., 1997). Most farmers would benefit from management practices that reduce weed seed inputs, increase seed losses, and reduce the probability that remaining seeds establish (Gallandt, 2006). Variable seed dormancy and germination are important survival strategies of plants. Seed dormancy is a classic feature of adaptation to environments with adverse conditions for growth and reproduction during some portion of the year; however, seeds of some species never experience dormancy (Forcella et al., 2000). In nondormant seeds, the primary factors governing seed germination are temperature and moisture. As weed seeds are distributed at various soil depths, they experience various soil temperatures and moisture levels, and germination is likely to be variable. Polymorphism is probably a further source of variability as considerable genetic variability exists within a given species (e.g., Danquah et al., 2002). The longevity of weed seeds in the seed bank will also depend upon the interactions of factors such as the inherent dormancy characteristics of the seed populations, the environmental conditions present in the soil that influence dormancy release (e.g., light, temperature, water, and gas environment), and biological (e.g., predation and allelopathy) interactions (Radosevich et al., 1996). Weed seeds persist in soil through resistance to predation and decay due to protective pods or impermeable seed coats, or variable seed dormancy (Chauhan et al., 2006d). Usually, seeds lose their viability through germination and mortality more rapidly when present near the soil surface than when buried deeper in the soil (Mohler, 1993). In temperate areas, for example, Froud-Williams (1983) reported that seed populations of Bromus sterilis L. declined by 85% during July and August in an uncultivated field and no viable seeds remained the following April. The rapid decline in viable seeds at the surface may be due to suitable germination conditions (Banting, 1966) and greater pathogen and predator incidence at the surface (Taylorson, 1970). In an Australian study, greater than 80% of weed seeds were reported to be removed by predators (Jacob Spafford et al., 2006). Similarly in the Philippines, 90% of seeds of Digitaria ciliaris (Retz.) Koel. on the soil surface were removed by seed predators (Chauhan et al., unpublished data). With adequate knowledge of the germination requirement of specific weed populations, situations can be identified where no or very low germination should occur, even if a high density of seeds is present in the soil. For example, weed seeds often fail to germinate deep in the soil profile. Such information could be used to design strategies involving crop establishment and tillage to suppress weed seedling emergence. The main factors affecting seed germination are discussed later in this chapter.

226


The germination patterns of species in natural and agronomic environments or the periodicity of germination often determines whether a plant competes successfully with its neighbors, is consumed by herbivores, is infected with disease, and whether it flowers, reproduces, and matures by the end of the growing season (Forcella et al., 2000). Germination patterns may also account for differences in weed species composition in different seasons (e.g., Johnson et al., 2004). Stoller and Wax (1973) concluded that weeds emerging early in the crop cycle are killed during land preparation for the sowing of corn or soybean (Glycine max L.). Knowledge of emergence patterns could therefore be used to time cultivation and postemergence application of herbicides to achieve maximum effect (Ogg and Dawson, 1984). In this way, delaying soybean sowing enabled improved weed control with cultivation (Buhler and Gunsolus, 1996). Light, soil moisture, soil salinity, fire, tillage, and surface residue (mulch) are important environmental factors affecting weed germination and emergence, and these can be manipulated more directly through management. Although these factors will be discussed individually later, interactions among them tend to be common. To illustrate the factors affecting weed seed germination, studies on some common weeds in rice, wheat, corn, and soybean (Table 1) are described. The persistence of these weed species is dependent on seeds with the exception of Cyperus esculentus L., Cyperus rotundus L., Imperata cylindrica (L.) Raeuschel, Panicum repens L., Paspalum distichum L., and Scirpus maritimus L., which can reproduce vegetatively. The discussion and examples given, however, are not limited to the species in Table 1.

4. Responses of Weed Seed Germination to Light Germination of various weed species has different responses to light and darkness. Exposure to light breaks dormancy and eventually increases germination in many species, especially small-seeded species (Cousens et al., 1993). Under normal daylight conditions, exposure of seeds to light flashes from several m s to 1 s is sufficient to stimulate germination (Hartmann and Nezadal, 1990; Scopel et al., 1994). Consequently, any soil tillage operation that brings seeds to the surface, even for an instant, probably provides sufficient light to meet the requirements for most seeds. Seed germination responses to light are species specific. Some species germinate equally in light and dark [e.g., A. fatua, Eleusine indica (L.) Gaertn., Melochia concatenata L., and Mimosa invisa Mart. ex Colla], whereas others [e.g., D. ciliaris, Echinochloa colona (L.) Link, and Portulaca oleracea L.] require light to stimulate germination (Table 2). Some species have an absolute light requirement for germination [e.g., Cyperus difformis L.,

227

Weed Seed Ecology

Table 1

Important weeds occurring in rice, wheat, corn, and soybean Crop

Weed species

Rice

Wheat

Corn

Soybean

Ageratum conyzoides L. Amaranthus spinosus L. Avena fatua L. Chenopodium album L. Cyperus iria L. Cyperus difformis L. Digitaria ciliaris (Retz.) Koel. Digitaria sanguinalis (L.) Scop. Echinochloa colona (L.) Link Echinochloa crus-galli (L.) P. Beauv. Eclipta prostrata (L.) L. Eleusine indica (L.) Gaertn. Fimbristylis miliacea (L.) Vahl Leptochloa chinensis (L.) Nees Ludwigia hyssopifolia (G. Don) Exell. Mimosa pudica L. Oryza sativa L. (weedy rice) Phalaris minor Retz. Portulaca oleracea L. Rottboellia cochinchinensis (Lour.) W.D. Clayton

þ þ þþ þþ þ þ þþ þþ þþ þ þþ þþ þþ

þ þþ þþ

þ þ þ þ þ þ þ þþ þþ þ þ þþ þ þ þ

þ þ þ þ þ þ þ þ þ þ þ

þ þþ þ þþa

þþ þ

þ þ þ

þ þþ

a Most important in upland/rainfed rice. Information from Holm et al. (1991, 1997) and Rao et al. (2007) was used to characterize the importance of weed species in these crops (þþ: most important; þ: important or reported to occur in the crop; –: not reported).

Digitaria longiflora (Retz.) Pers., and Eclipta prostrata (L.) L.], and these are described as positively photoblastic, a response thought to be controlled by phytochrome, a light-absorbing pigment within plants. In photoblastic seeds, light exposure may convert inactive-phytochrome ‘‘red’’ to activephytochrome ‘‘far-red’’ (Rollin, 1972). Species showing preference for light for germination have the potential to be more prevalent in continuous NT systems in which a large proportion of weed seeds remains on the surface and exposed to light (Chauhan and Johnson, 2008b; Chauhan et al., 2006c,d). Conversely, germination of seeds of species that require light is likely to be impeded by the implementation of NT. Seeds requiring light would remain dormant if buried in the soil and would eventually be lost to predation or decay. Even shallow burial can induce dormancy in light-requiring seeds (Pons, 1989; Wesson and Wareing, 1969) as less than 1%

228


Table 2 Effect of light conditions (light and dark) on seed germination of different weed species Germination (%) Weed species

Light

Dark

Avena fatua L. Borreria ocymoides (Burm. f.) DC. Celosia argentea L.

74 91 31 66 80 81 94 84 93 89 76 55 69 93 83 97 85 95 95 98 97 91 97 81 63 86 79

72 93 2 18 37 0 0 2 7 0 12 3 24 0 0 96 0 0 19 0 97 91 83 2 59 4 7

Chromolaena odorata (L.) R.M. King & H. Rob. Cyperus difformis L. Cyperus iria L. Digitaria ciliaris (Retz.) Koel. Digitaria longiflora (Retz.) Pers. Echinochloa colona (L.) Link. Echinochloa crus-galli (L.) P. Beauv. Eclipta prostrata (L.) L. Eleusine indica (L.) Gaertn. Fimbristylis miliacea (L.) Vahl Leptochloa chinensis (L.) Nees Ludwigia hyssopifolia (G. Don) Exell. Melochia concatenata L. Mimosa invisa Mart. ex Colla Phalaris minor Retz. Portulaca oleracea L. Sida rhombifolia L. Synedrella nodiflora (L.) Gaertn. Tridax procumbens L.

Sources: Altom and Murray (1996), Benvenuti et al. (2004), Boyd and Van Acker (2004), Chauhan and Johnson (2007, 2008a,c,d,e,f,g,i,j,k, 2009a,d,e,f,g), Chozin and Nakagawa (1988), Chun and Moody (1987), Okusanya (1980), Om et al. (2005).

of incident light penetrates beyond 2.2 mm into the soil (Egley, 1986; Woolley and Stoller, 1978). In this way, light acts as a ‘‘depth indicator’’ for seeds, preventing germination of deeply buried seeds and promoting germination of only those near the soil surface (Chauhan and Johnson, 2008c,e,g; Schu¨tz et al., 2002). The absence of a germination response to light in some species (Table 2) suggests that these seeds may germinate even when buried or after canopy closure in most crops grown in the humid tropics. Germination of deeply buried seeds will be fatal if emerging seedlings are unable to reach the surface,

229

Weed Seed Ecology

while seeds germinating after crop canopy closure are likely to experience greater competition for light, water, and nutrients from the crop and may result in less crop yield loss and weed seed production (O’Donovan et al., 1985; Uscanga-Mortera et al., 2007). Germination of photoblastic seeds is generally inhibited by low-red/far-red ratios usually present under a crop canopy (Taylorson, 1969) and could be important for seed germination after canopy closure. Information on these aspects is very limited, however, for weed species occurring in the tropics.

5. Responses of Weed Seed Germination to Seed Scarification and Fire 5.1. Effect of seed scarification on germination The seed coat itself as a physical barrier can impose dormancy (Turner et al., 2005). Hard seeds present a relatively absolute form of imposed dormancy due to the impermeability of the seed coat to water or gasses (Foley, 2001). Hard seeds generally require physical or chemical scarification, or weathering in the soil to enhance germination (Foley, 2001). Germination of hardcoated seeds is very low unless seeds are scarified (Table 3). Seeds with an impermeable seed coat may have a long life in the soil (Egley and Chandler, 1983). Seed coats and dormancy in species with hard seeds have been reviewed in detail by Egley (1989) and Kelly et al. (1992). In some species, such as Rottboellia cochinchinensis (Lour.) W.D. Clayton, dormancy is largely due to the ‘‘seed coat’’ surrounding the seeds (Mercado, 1978; Thomas and Table 3 Effect of seed scarification on germination of different weed species Germination (%) Weed species

Control

Scarified

Corchorus olitorius L. Malva parviflora L. Melochia concatenata L. Mimosa invisa Mart. ex Colla Mimosa pudica L. Sida rhombifolia L. Urena lobata L.

3 10 23 2 6 5 2

93 88 97 100 90 65 78

Seeds were scarified either by physical cut with a scalpel or chemically with sulfuric acid. Sources: Chauhan et al. (2006a), Chauhan and Johnson (2008a,i,j, 2009c), Wang et al. (2009).

230


Allison, 1975). Mechanisms that increase breakdown of the seed coat will increase germination and emergence of species with such seeds. Microbial and fungi attack and abrasion by soil particles may break dormancy in hard-seeded species, but little evidence supports this possibility (Baskin and Baskin, 2000; Kremer, 1993). Scarification during tillage will probably account for only a small proportion of seeds during any given tillage operation. Although these factors may scarify seeds, the role of tillage systems (intensity of soil disturbance) in scarification requires additional experimental investigation in order for the effects to be understood. Scarification may also occur through trampling by hoofed animals or by passage through an animal’s digestive tract; however, this will depend on farming systems and seed dispersal mechanisms. Other possible natural factors that may account for a dormancy break in hard-seeded species are extreme changes in temperature and moisture regimes, and fire (Baskin et al., 1998; Chauhan et al., 2006d). NT systems tend to leave most of the weed seeds on the soil surface after seed rain and even after crop planting (Chauhan et al., 2006b), where the seeds may experience fluctuating temperature and moisture regimes.

5.2. Effect of fire on germination In many countries, small farmers use fire to clear their land of straw or debris before crop planting (Roder et al., 1997). Further, in some areas farmers seldom practice soil tillage with upland rice but, instead, ‘‘dibble sow’’ the crop seed after burning the cut vegetation ( Johnson and Kent, 2002). Heat generated by burning affects populations of many species through its potential effects on the physical and biological properties of the soil (Uhl, 1982). Fire is one of the most important among several disturbances that may affect plant communities with persistent seed banks as it kills unburied seeds of susceptible species. In a study in India, wheat straw was removed or burned after combine harvesting, soil samples were taken and washed to separate the seeds of P. minor, and these seeds were tested for germination (Hari et al., 2003). The data showed that germination was 60% less in the field where wheat straw was burned after combine harvesting compared with its removal (87% germination). Similarly, 97% of ungerminated seeds of B. sterilis on the soil surface were destroyed and seedling numbers reduced by 94% with straw burning in Europe (Froud-Williams, 1983). Burning also destroys seeds of Alopecurus myosuroides Huds. (Moss, 1987) and A. fatua (Wilson and Cussans, 1975). The effect on germination of high temperatures, as seeds might experience if vegetation is cleared by burning, was evaluated in a range of weed species of upland crops (Table 4). Germination was completely inhibited when seeds, depending on weed species, were exposed at 160–200 C. Burning can increase the surface temperature to 550 C for 6 min (Cook, 1939), which is likely to destroy seeds on the soil

231

Weed Seed Ecology

Table 4 Temperature ( C) required for 50% inhibiting (T50) and completely inhibiting (T100) seed germination of different weed species Temperature ( C) Weed species

T50

T100

Digitaria ciliaris (Retz.) Koel. Digitaria longiflora (Retz.) Pers. Echinochloa colona (L.) Link Eclipta prostrata (L.) L. Eleusine indica (L.) Gaertn. Mimosa invisa Mart. ex Colla Mimosa pudica L. Synedrella nodiflora (L.) Gaertn.

148 116 139 167 133 96 and 153a 128 and 172a 152 and 165b

180 160 180 200 180 200 200 200

a

Germination was >50% between this range. Represents different kind of seeds (ray and disc). A known number of seeds were placed in an oven at different high temperatures for 5 min. The treated seeds were then tested for germination at optimum temperatures for 14 days. Sources: Chauhan and Johnson (2008d,e,g,i, 2009c,e,f ). b

surface. Soil temperatures, however, may decrease at a rate of 100 C cm 1 in the first 5 cm below the soil surface (Sanchez, 1976); this suggests that seeds (Table 4) buried below 4 cm may remain viable. Such seeds may then germinate if brought near the soil surface by cultivation or tillage operations. Fire has a strong influence in releasing physical dormancy in a range of hard-seeded species (Auld and O’Connel, 1991; Tieu et al., 2001; Whelan, 1995). The stimulant effect of fire may result from the physical effect of dry heat on seed structure, the physiological effect of dry heat on seed embryos, and/or the dormancy-breaking effects caused by volatile compounds, such as ethylene and ammonia (van Staden et al., 1995). Fire also creates open ground, which may result in a mass release of dormant seeds, especially if the fire coincides with hot and wet conditions (van Klinken et al., 2006). Scott (2006) demonstrated with the seeds of Parkinsonia aculeate L. that high temperatures can produce cracks on the seed coat, which, in turn, can increase imbibition and germination. Exposure to high temperature, as could be expected with vegetation burning, released the seeds from dormancy and stimulated the germination of nonscarified seeds of M. invisa, and raised germination from 4% pretreatment to 94% after exposure to 120 C for 5 min (Chauhan and Johnson, 2008i). Such high temperatures may scarify the seeds close to the soil surface but not affect seeds that lie beyond a few centimeters of depth. As species have differing responses to heat, the intensity and duration of fires may have a differential effect on the species in a community.

232


6. Responses of Weed Seed Germination to Seed Burial Depth, Tillage, and Surface Mulches 6.1. Effect of tillage systems on vertical seed distribution Tillage has a major influence on the vertical distribution of weed seeds in arable soils (Chauhan et al., 2006b; Cousens and Moss, 1990; Roberts, 1963; Staricka et al., 1990), and this pattern of seed distribution has a critical effect on seed germination and survival (Mohler, 1993). Different tillage systems have different effects on the distribution of weed seeds through the soil profile. An NT system was reported to retain 56% of the weed seeds in the top 1-cm soil layer, whereas the CT system buried 65% of the seeds to a depth of 1–5 cm and only 5% of the seeds remained in the top 1-cm soil layer (Chauhan et al., 2006b). Pareja et al. (1985) found 85% of all weed seeds in the upper 5 cm of soil in a reduced tillage system, but only 28% of seeds were found in this surface layer in a moldboard plow for CT. Likewise, 60% of the weed seeds were present in the top 1 cm of soil in an NT system compared with 30% in a chisel plow system (Yenish et al., 1992). A later study showed that more than 90% of weed seeds were found within 2 cm of the surface in an NT system (Yenish et al., 1996). In untilled soybean fields, D. ciliaris seeds were highly concentrated on or near the soil surface, whereas seeds were distributed uniformly throughout the soil profile in tilled soybean fields (Kobayashi and Oyanagi, 2005). Knowledge of the vertical distribution of weed seeds in the soil can enable prediction of the probability of a seed germinating and the likelihood of successful seedling emergence (Grundy and Mead, 1998). This information in the tropics, however, is limited in the literature. Nonetheless, tillage systems can be ranked in terms of the extent of soil disturbance as NT < chisel plow < moldboard plow. Low soil disturbance systems tend to leave most of the weed seeds in the top soil layer compared with high soil disturbance systems that bury seed more or less evenly throughout the cultivated depths. Different tillage systems therefore leave weed seeds at different depths, and this differential vertical distribution of the seeds in the soil has the potential to affect seedling emergence and weed population dynamics (Buhler, 1991; Harper, 1957).

6.2. Effect of seed burial depths on weed germination The distribution of seeds in the soil profile in arable soils is largely due to tillage implements. In NT systems, however, burial may also result from sowing implements, wheel traffic, animal traffic, soil cracking, and selfburial via structural characteristics, such as hydroscopic awn, of seeds

Weed Seed Ecology

233

(Forcella et al., 2000). Optimum burial depth is the depth from which a maximum number of sown seeds can produce established plants. In the field, this optimum depth is the shallowest depth at which germination is not obstructed by drought and where, at the same time, the seedlings can root easily and become sufficiently anchored in the soil. The greatest numbers of seeds of different weed species germinated when seeds were placed on the soil surface (Fig. 1; Table 5), but species differed in the response to seed burial depth. Species such as Cyperus iria L., C. difformis, Fimbristylis miliacea (L.) Vahl, and Leptochloa chinensis (L.) Nees were unable to emerge from a depth greater than 0.5 cm, whereas others, such as D. longiflora, Heliotropium indicum L., and P. oleracea, were completely inhibited by 2-cm burial. Burial by only 0.5-cm reduced emergence of Chromolaena odorata (L.) R.M. King & H. Rob., D. ciliaris, and E. colona by 70–88% compared with seeds on the soil surface. In contrast, some seedlings (11–21%) of E. indica, M. concatenata, and M. invisa emerged from a depth of 6 cm. A proportion of Echinochloa crus-galli (L.) P. Beauv. seedlings (

Advances in Agronomy, Volume 105

Advances in Agronomy, Volume 60

Advances in Agronomy, Volume 50

Advances in Agronomy, Volume 68

Advances in Agronomy, Volume 54

Advances in Agronomy, Volume 11

Advances in Agronomy, Volume 31

Advances in Agronomy, Volume 33

Advances in Agronomy, Volume 65

Advances in Agronomy, Volume 21

Advances in Agronomy, Volume 64

Advances in Agronomy, Volume 20

Advances in Agronomy, Volume 100

Advances in Agronomy, Volume 108

Advances in Agronomy, Volume 69

Advances in Agronomy, Volume 83

Advances in Agronomy, Volume 91

Advances in Agronomy, Volume 46

Advances in Agronomy, Volume 111

Advances in Agronomy, Volume 62

Advances in Agronomy, Volume 56

Advances in Agronomy, Volume 28

Advances in Agronomy, Volume 53

Advances in Agronomy, Volume 14

Advances in Agronomy, Volume 26

Advances in Agronomy, Volume 103

Advances in Agronomy, Volume 81

Advances in Agronomy, Volume 40

Advances in Agronomy, Volume 45

Advances in Agronomy, Volume 101

Advances in Agronomy, Volume 85

Advances in Agronomy, Volume 105

Advances in Agronomy, Volume 60

Advances in Agronomy, Volume 50

Advances in Agronomy, Volume 68

Advances in Agronomy, Volume 54

Advances in Agronomy, Volume 11

Advances in Agronomy, Volume 31

Advances in Agronomy, Volume 33

Advances in Agronomy, Volume 65

Advances in Agronomy, Volume 21

Advances in Agronomy, Volume 64

Advances in Agronomy, Volume 20

Advances in Agronomy, Volume 100

Advances in Agronomy, Volume 108

Advances in Agronomy, Volume 69

Advances in Agronomy, Volume 83

Advances in Agronomy, Volume 91

Advances in Agronomy, Volume 46

Advances in Agronomy, Volume 111

Advances in Agronomy, Volume 62

Advances in Agronomy, Volume 56

Advances in Agronomy, Volume 28

Advances in Agronomy, Volume 53

Advances in Agronomy, Volume 14

Advances in Agronomy, Volume 26

Advances in Agronomy, Volume 103

Advances in Agronomy, Volume 81

Advances in Agronomy, Volume 40

Advances in Agronomy, Volume 45

Advances in Agronomy, Volume 101

Advances in Agronomy, Volume 85

Recommend Documents