union of concerned scientists...

8/7/2019 Union of Concerned Scientists --No-Sure-Fix-Prospects-for-Reducing-Nitrogen-Fertilizer-Pollution-Thr_9Y9

1/46


2/46


3/46

Doug Gurian-Sherman

Noel Gurwick

Union of Concerned Scientists

December 2009

Prospects for Reducing Nitrogen Fertilizer Pollution

through Genetic Engineering


4/46

iiUnion of Concerned Scientists

2009 Union of Concerned ScientistsAll rights reserved

Doug Gurian-Sherman and Noel Gurwickare senior scientists in the Unionof Concerned Scientists (UCS) Food and Environment Program.

The Union of Concerned Scientists (UCS) is the leading science-basednonprofit working for a healthy environment and a safer world. UCScombines independent scientific research and citizen action to developinnovative, practical solutions and to secure responsible changes ingovernment policy, corporate practices, and consumer choices.

The goal of the UCS Food and Environment Program is a food systemthat encourages innovative and environmentally sustainable ways to producehigh-quality, safe, and affordable food, while ensuring that citizens have avoice in how their food is grown.

More information is available on the UCS website atwww.ucsusa.org/food_and_agriculture.

This report is available on the UCS website (in PDF format) atwww.ucsusa.org/publicationsor may be obtained from:

UCS Publications2 Brattle SquareCambridge, MA 02238-9105

Or, email [email protected] call (617) 547-5552.

Design: Catalano DesignCover image: Todd Andraski/University of Wisconsin-Extension


5/46

No Sure Fix ii i

Contents

Text Boxes, Figures, and Tables ivAcknowledgments v

Executive Summary 1

Chapter 1: Introduction: Genetic Engineering and Nitrogen

in Agriculture 5

Key Terms Used in This Report 5Report Organization 6The Impact of Nitrogen Fertilizer Use in Agriculture 6The Role of Reactive Nitrogen 8

Chapter 2: Nitrogen Use Efficiency in GE Plants and Crops 10

How We Evaluated GEs Prospects for Improving NUE 11Studies of GE NUE Crops 12Approved Field Trials of GE NUE Crops 15Possible Risks Related to GE NUE Genes 16Commercialization of GE NUE Crops 17

Chapter 3: Improving NUE through Traditional and Enhanced

Breeding Methods 19

NUE Improvements in Commercial Varieties 19The Impact of Higher Yield on NUE 19Genetic Variability of NUE-Related Traits in Major Crops 20Strengths and Limitations of Breeding Compared with GE 22

Chapter 4: The Ecosystem Approach to NUE 24

A Big-Picture Perspective 24

The Time Is Ripe for a New Approach 24

Chapter 5: Other Means of Improving NUE 26

Cover Crops 27Precision Farming 28

Chapter 6: Conclusions and Recommendations 30

The Promise and Pitfalls of Non-GE Approaches 30What the United States Should Do 31

References 33


6/46

ivUnion of Concerned Scientists

Text Boxes

1. How Engineered Genes Contribute to Plant Traits 10

2. Methods Used to Study Crop NUE 11

Figures

1. The Nitrogen Cycle 7

2. Rise in Reactive Nitrogen Production 8

3. USDA-Approved Field Trials of GE Crops 15

Tables

1. Genes Used to Improve NUE through Genetic Engineering 13

2. Improvements in Nitrogen Use Efficiency 20

Text Boxes, Figures, and Tables


7/46

vNo Sure Fix

This report was made possible in part through the generousfinancial support of C.S. Fund, Clif Bar Family Foundation,CornerStone Campaign, Deer Creek Foundation, The EducationalFoundation of America, The David B. Gold Foundation, The JohnMerck Fund, Newmans Own Foundation, Next Door Fund of theBoston Foundation, The David and Lucile Packard Foundation,and UCS members.

The authors would like to thank Walter Goldstein of the Michael Fields

Agricultural Research Institute, Linda Pollack of the U.S. Departmentof Agricultures Agricultural Research Service, and Christina Tonittoof Cornell University. The time they spent in reviewing this report isgreatly appreciated and significantly enhanced the final product.

Here at UCS, the invaluable insights provided by Mardi Mellon andKaren Perry Stillerman helped clarify and strengthen the report as well.Brenda Ekwurzel contributed valuable suggestions regarding climate-change-related aspects of the report. The authors also thank HeatherSisan for research assistance that made everything go more smoothly.

Finally, the report was made more readable by the expert copyediting ofBryan Wadsworth.

The opinions expressed in this report do not necessarily reflect theopinions of the foundations that support the work, or the individualswho reviewed and commented on it. Both the opinions and theinformation contained herein are the sole responsibility of the authors.

Acknowledgments


8/46


9/46

1No Sure Fix

Nitrogen is essential for life. It is the mostcommon element in Earths atmosphereand a primary component of crucial bio-logical molecules, including proteins and nucleicacids such as DNA and RNAthe building blocksof life.

Crops need large amounts of nitrogen in orderto thrive and grow, but only certain chemicalforms collectively referred to as reactive nitrogen

can be readily used by most organisms, includingcrops. And because soils frequently do not containenough reactive nitrogen (especially ammonia andnitrate) to attain maximum productivity, manyfarmers add substantial quantities to their soils,often in the form of chemical fertilizer.

Unfortunately, this added nitrogen is a majorsource of global pollution. Current agriculturalpractices aimed at producing high crop yields oftenresult in excess reactive nitrogen because of the dif-

ficulty in matching fertilizer application rates andtiming to the needs of a given crop. The excessreactive nitrogen, which is mobile in air and water,can escape from the farm and enter the globalnitrogen cyclea complex web in which nitrogenis exchanged between organisms and the physicalenvironmentbecoming one of the worlds majorsources of water and air pollution.

The challenge facing farmers and farm policymakers is therefore to attain a level of crop produc-

tivity high enough to feed a growing world popula-tion while reducing the enormous impact ofnitrogen pollution. Crop genetic engineering hasbeen proposed as a means of reducing the loss ofreactive nitrogen from agriculture. This reportrepresents a first step in evaluating the prospectsof genetic engineering to achieve this goal whileincreasing crop productivity, in comparison with

other methods such as traditional crop breeding,precision farming, and the use of cover crops thatsupply reactive nitrogen to the soil naturally.

The Importance of Nitrogen Use

Efficiency (NUE)

Crops vary in their ability to absorb nitrogen, butnone absorb all of the nitrogen supplied to them.The degree to which crops utilize nitrogen is callednitrogen use efficiency (NUE), which can be mea-sured in the form of crop yield per unit of addednitrogen. NUE is affected by how much nitrogenis added as fertilizer, since excess added nitrogenresults in lower NUE. Some agricultural practicesare aimed at optimizing the nitrogen applied tomatch the needs of the crop; other practices, suchas planting cover crops, can actually remove excessreactive nitrogen from the soil.

In the United States, where large volumes

of chemical fertilizers are used, NUE is typicallybelow 50 percent for corn and other major cropsin other words, more than half of all added reactivenitrogen is lost from farms. This lost nitrogen isthe largest contributor to the dead zone in theGulf of Mexicoan area the size of Connecticutand Delaware combined, in which excess nutrientshave caused microbial populations to boom, rob-bing the water of oxygen needed by fish and shell-fish. Furthermore, nitrogen in the form of nitrate

seeps into drinking water, where it can become ahealth risk (especially to pregnant women andchildren), and nitrogen entering the air as ammo-nia contributes to smog and respiratory disease aswell as to acid rain that damages forests and otherhabitats. Agriculture is also the largest human-caused domestic source of nitrous oxide, anotherreactive form of nitrogen that contributes to global

Executive Summary


10/46

2Union of Concerned Scientists

warming and reduces the stratospheric ozone thatprotects us from ultraviolet radiation.

Nitrogen is therefore a key threat to our globalenvironment. A recent scientific assessment of nineglobal environmental challenges that may make theearth unfavorable for continued human develop-ment identified nitrogen pollution as one of onlythreealong with climate change and loss of bio-diversitythat have already crossed a boundarythat could result in disastrous consequences if notcorrected. One important strategy for avoiding thisoutcome is to improve crop NUE, thereby reduc-ing pollution from reactive nitrogen.

Can Genetic Engineering Increase NUE?

Genetic engineering (GE) is the laboratory-basedinsertion of genes into the genetic material oforganisms that may be unrelated to the sourceof the genes. Several genes involved in nitrogenmetabolism in plants are currently being used inGE crops in an attempt to improve NUE. Ourstudy of these efforts found that:

Approval has been given for approximately125 field trials of NUE GE crops in the UnitedStates (primarily corn, soybeans, and canola),

mostly in the last 10 years. This compares withseveral thousand field trials each for insect resis-tance and herbicide tolerance.

About half a dozen genes (or variants of thesegenes) appear to be of primary interest. The exactnumber of NUE genes is impossible to deter-mine because the genes under consideration bycompanies are often not revealed to the public.

No GE NUE crop has been approved byregulatory agencies in any country or com-mercialized, although at least one gene (andprobably more) has been in field trials for abouteight years.

Improvements in NUE for experimental GEcrops, mostly in controlled environments,have typically ranged from about 10 to 50 per-cent for grain crops, with some higher values.

There have been few reports of values from thefield, which may differ considerably from lab-based performance.

By comparison, improvement of corn NUE

through currently available methods has beenestimated at roughly 36 percent over the pastfew decades in the United States. Japan hasimproved rice NUE by an estimated 32 percentand the United Kingdom has improved cerealgrain NUE by 23 percent.

Similarly, estimates for wheat from France showan NUE increase from traditional breeding ofabout 29 percent over 35 years, and Mexico hasimproved wheat NUE by about 42 percent over

35 years.Available information about the crops and

genes in development for improved NUE suggeststhat these genes interact with plant genes in com-plex ways, such that a single engineered NUE genemay affect the function of many other genes. Forexample:

In one of the most advanced GE NUE crops,the function of several unrelated genes thathelp protect the plant against disease has beenreduced.

Another NUE gene unexpectedly altered theoutput of tobacco genes that could change theplants toxicological properties.

Many unexpected changes in the function ofplant genes will not prove harmful, but some maymake it difficult for the crops to gain regulatoryapproval due to potential harm to the environmentor human health, or may present agricultural draw-

backs even if they improve NUE. For the mostadvanced of the genes in the research pipeline,commercialization will probably not occur until atleast 2012, and it will likely take longer for most ofthese genes to achieve commercializationif theyprove effective at improving NUE. At this point,the prospects for GE contributing substantially toimproved NUE are uncertain.


11/46

3No Sure Fix

Other Methods for Reducing Nitrogen Pollution

Traditional or enhanced breeding techniquescanuse many of the same or similar genes that arebeing used in GE, and these methods are likely to

be as quick, or quicker, than GE in many cases.Traditional breeding may have advantages in com-bining several NUE genes at once.

Precision farmingthe careful matching ofnitrogen supply to crop needs over the courseof the growing seasonhas shown the ability toincrease NUE in experimental trials. Some of thesepractices are already improving NUE, but adop-tion of some of the more technologically sophisti-cated and precise methods has been slow.

Cover cropsare planted to cover and protectthe soil during those months when a cash cropsuch as corn is not growing, often as a componentof an organic or similar farming system. Some cansupply nitrogen to crops in lieu of synthetic fertil-izers, and can remove excess nitrogen from the soil;in several studies, cover crops reduced nitrogenlosses into groundwater by about 40 to 70 percent.

Cover crops and other low-external-inputmethods (i.e., those that limit use of syntheticfertilizers and pesticides) may also offer otherbenefits such as improving soil water retention(and drought tolerance) and increasing soil organicmatter. An increase in organic matter that containsnitrogen can reduce the need for externally sup-plied nitrogen over time.

With the help of increased public investment,these methods should be developed and evaluatedfully, using an ecosystem approach that is bestsuited to determine how reactive nitrogen is lostfrom the farm and how NUE can be improved in

a comprehensive way. Crop breeding or GE aloneis not sufficient because they do not fully addressthe nitrogen cycle on real farms, where nitrogen lossvaries over time and space, such as those times whencropsconventional or GEare not growing.

Conclusions

GE crops now being developed for NUE mayeventually enter the marketplace, but such crops

are not uniquely beneficial or easy to produce.There is already sufficient genetic variety for NUEtraits in crops, and probably in close relatives ofimportant crops, for traditional breeding to buildon its successful track record and develop moreefficient varieties.

Other methods such as the use of cover cropsand precision farming can also improve NUE andreduce nitrogen pollution substantially.

Recommendations

The challenge of optimizing nitrogen use in a hun-gry world is far too important to rely on any oneapproach or technology as a solution. We thereforerecommend that research on improving crop NUE

continue. For traditional breeding to succeed,public research support is essential and should beincreased in proportion to this methods substantialpotential.

We also recommend that system-basedapproaches to increasing NUEcover crops, preci-sion application of fertilizer, and organic or similarfarming methodsshould be vigorously pursuedand supported. These approaches are complemen-tary to crop improvement because each addresses a

different aspect of nitrogen use. For example, whilebreeding for NUE reduces the amount of nitrogenneeded by crops, precision farming reduces theamount of nitrogen applied. Cover crops removeexcess nitrogen and may supply nitrogen to cashcrops in a more manageable form.

Along with adequate public funding, incentivesthat lead farmers to adopt these practices are alsoneeded. Although the private sector does exploretraditional breeding along with its heavy invest-

ment in the development of GE crops, it is notlikely to provide adequate support for the develop-ment of non-GE varieties, crops that can better usenitrogen from organic sources, or improved covercrops that remove excess nitrogen from soil. Wemust ensure that broad societal goals are addressedand important options are pursued nevertheless.

In short, there are considerable opportunitiesto address the problems caused by our current


12/46


overuse of synthetic nitrogen in agriculture if weare willing to make the necessary investments. Theglobal impact of excess reactive nitrogen will wors-en as our need to produce more food increases, sostrong actionsincluding significant investmentsin technologies and methods now largely ignoredby industrial agriculturewill be required to lessenthe impact.


13/46

5No Sure Fix

The need to raise global food productionperhaps as much as 100 percent by themiddle of the century poses one of themajor challenges currently facing the worldasdoes reducing the pollution caused by many cur-

rent agricultural practices. Because plant growthis often constrained by the amount of nitrogen inthe soil that plants can access, adding more nitro-gen to agricultural fields will almost certainly playa role in meeting the challenge of increased cropproductivity. Unfortunately, some of the nitrogensources readily available to farmers across muchof the globe are already chief contributors tonitrogen pollution.

Dobermann and Cassman (2005) project a

need to increase grain production 38 percent by2025, and assert that this may be done with anitrogen crop yield response increase of 20 percentusing current technologies, with a net increase innitrogen of 30 percent if current losses of agricul-tural land do not continue. Other estimates,however, note that a 45 percent reduction in nitro-gen pollution in the Gulf of Mexico is likely neededto have a substantial impact on the dead zone there(EPA 2009b). Pouring on even more fertilizer toincrease food production would aggravate thisand other problems and carry potentially highcosts. What we need are ways to increase foodproduction on existing farmland while reducingnitrogen pollution.

Strategies for reducing nitrogen loss from farmswithout reducing productivity include vegetationbuffer strips planted along waterways adjacent tocrop fields; such buffers have captured significant

amounts of nitrogen that would otherwise reachstreams and rivers. Also, better timing of nitrogenfertilizer applicationto be performed only when itis actually needed by a given crop during the grow-ing seasonreduces the amount of nitrogen applied.

Key Terms Used in This Report

Improving the nitrogen use efficiency (NUE)ofcrops is another strategy for reducing nitrogen lossfrom farmsand consequent downstream nitrogenpollutionin this case by increasing the amountof plant growth that occurs for each pound ofnitrogen added to the soil. Improved NUE reducesthe need for nitrogen fertilizer. This can poten-tially be done in two ways: through traditional or

enhanced methods of crop breeding, or throughgenetic engineering.NUE can also be improved in order to reduce

nitrogen loss from farm fields rather than toincrease crop yield. The use of cover crops andbetter-timed fertilizer applications often serve thispurpose. It should be noted that because differentmethods for measuring NUE can arrive at differentvalues, it may be difficult to make direct compari-sons between NUE values found in this report andelsewhere.

Traditional breedinginvolves controlled matingbetween plant parents selected for their desirabletraits. This powerful technology, responsible formost genetic improvement in crops over the last100 years, can now be enhanced with new genomictechnologies that assist scientists in identifyingprospective traits. Using information about plantgenetics to inform breeding does not constitute

Chapter1

Introduction: Genetic Engineering and Nitrogenin Agriculture


14/46


genetic engineering, and the promise offered bythese two approaches may differ dramatically.

Genetic engineering (GE)refers specificallyto the isolation and removal of genesspecifi-cally, genes that determine traits of scientific oreconomic interestfrom one organism and theirinsertion into another, where they become part ofthe inherited genetic material. In relation to crops,GE can add genes to plants from virtually anysource and achieve gene combinations not pos-sible in nature. For example, most commercializedGE crops contain genes from bacteria that makethe crops immune to certain herbicides or protectthem against insect pests.

GE and traditional breeding have different

advantages and limitations as techniques for devel-oping new crop varieties. GE enables us to com-bine genes from organisms that cannot reproducewith each other, but its success depends on howspecific genes (and specific combinations of genes)influence plant growth. Very few plant traits arecontrolled by a single gene, and our understandingof how multi-gene systems influence plant growthis limited, especially when considering the variedenvironmental conditions under which plants grow

and the changes in gene function and metabolismthat occur over the life of the plant.

Traditional breeding, which is sometimesinformed by a detailed understanding of the parentplants genetics, also rearranges the genetic mate-rial of the crop. But in this case, because all of theparents genes are involved, some undesired genesmay end up in the resulting crop along with thegenes of interest. And unlike GE it uses only thosegenes already found in the crop or closely related

plant species. The ability of traditional breeding tobring many genes from sexually compatible plantstogether can be advantageous for improving themany traits controlled by multiple genes. Whileknowledge of genetics can inform traditionalbreeding, this method can also achieve the desiredtraits even when the genetic basis is not thoroughlyunderstood.

Report Organization

This report describes the status of GE as a toolfor producing crops with improved NUE, and isdivided along the following lines:

The next section of Chapter 1 describes the roleof the nitrogen cycle.

Chapter 2 provides definitions for NUE relevantto this report and discusses the implications ofusing different conceptual frameworks to mea-sure NUE. We then evaluate GEs prospects forproviding food and feed crops with enhancedNUE, based on an examination of the scientificliterature and government databases.

Chapter 3 evaluates traditional breedings pros-

pects for providing food and feed crops withenhanced NUE. Covered technologies includemarker-assisted breeding and other advances ingenomics, and the identification of crop genesinvolved in nitrogen metabolism. Importantdifferences between traditional breeding andGE are considered.

To provide appropriate context, Chapter 4discusses the value of an ecosystem approachto evaluating nitrogen pollution and solutions,and Chapter 5 reviews two other approaches forreducing fertilizer use and nitrogen pollution:precision farming and cover cropping.

Finally, Chapter 6 offers several recommenda-tions for public policies that can help reducenitrogen pollution.

The Impact of Nitrogen Fertilizer Use in

Agriculture

The addition of nitrogen fertilizers, along withother changes in agriculture, has greatly increasedcrop productivity in many parts of the world,allowing global food production to remain aheadof rapid population growth in the second half ofthe twentieth century (Vitousek et al. 2009). Butareas where soils are exceptionally deficient innitrogen, such as much of Africa (Sanchez 2002),


15/46

7No Sure Fix

have not kept pace in producing enough food, andimprovements in soil fertility are urgently needed.

While essential to food production, nitrogencompounds added to agricultural ecosystems arealso some of the most important sources of pol-lution nationally and globally. Consequences ofnitrogen pollution include toxic algal blooms,oxygen-depleted dead zones in coastal waters, andthe exacerbation of global climate change, acidrain, and biodiversity loss (Krupa 2003; McCubbinet al. 2002; Vitousek et al. 1997). Reactive nitro-gen entering the Mississippi River from cropfields comprises about 42 percent of the nitrogencausing the dead zone in the Gulf of Mexicoat16,500 sq. km in recent years (EPA 2008), an areathe size of Delaware and Connecticut combined.

Fertilizer-intensive agriculture practices arealso the United States major anthropogenic (i.e.,

human-caused) source of nitrous oxide (N2O), apotent heat-trapping gas that also contributes tothe destruction of stratospheric ozone. Agriculturalsoils are responsible for about two-thirds of theanthropogenic nitrous oxide produced in theUnited States (EPA 2009a). In addition, gaseousammonia released from nitrogen fertilizer contrib-utes to fine particulate matter that causes respira-tory disease and acid rain (Anderson, Strader, andDavidson 2003; Krupa 2003; McCubbin et al.2002; Vitousek et al. 1997). Nitrate concentrationsabove 10 parts per million in drinking water havebeen implicated as a cause of methemoglobinemia,or blue baby syndrome (Fan and Steinberg 1996).

Recently, it has been suggested that disruptionof the global nitrogen cyclethe complex web inwhich nitrogen is exchanged between organismsand the physical environment (Figure 1)caused

Nitrogenfertilizers

N 0, N2 2

N 0, NO, N2 2

NH3

Crop residue

Soil organic matter

Manure, urine

Fixation Ammonia volatilization

Decomposition mineralization Plant uptakeconsumption

Oceans, lakes

N2 Nitrogen gas

NO Nitric oxide

NO2 Nitrite

NO3 Nitrate

N2O Nitrous oxide

NH3 Ammonia

NH4 Ammonium

Symbols

Rivers and streams

Groundwater

Denitrification

Denitrification N O2

N2

NH4

Leaching

NO3

Nitrogen inputs

N O2

NONitrifica

tion

NO2

The nitrogen cycle is a highly complex, global cycle that continuously transforms nitrogen into various chemical forms.Industrial agriculturewith its inefficient use of synthetic fertilizersalters this cycle by adding excessive amounts ofreactive nitrogen to the local and global environments.

Source: Adapted from Government of South Australia, Primary Industries and Resources SA.

Figure 1. The Nitrogen Cycle


16/46


by added nitrogen now exceeds the planets capac-ity to maintain a desirable state for human survivaland development (Rockstrm et al. 2009). Of thenine significant planetary processes or conditionsdescribed in that report, only climate change andloss of biodiversity have also passed such a point.This assessment underscores the enormous impactthat excess nitrogen is having on the environment.

The Role of Reactive Nitrogen

The dramatic consequences of nitrogen fertilizeruse, both positive and negative, are understandablewhen we appreciate the extent to which humanactivity altered the nitrogen cycle in the twentiethcentury, especially following the green revolutionof the 1960s (Figure 2). Overall, production of reac-tive nitrogen increased by a factor of 11, from about

15 teragrams (Tg), or trillion grams, of nitrogenper year in 1860 to about 165 Tg per year in 2000.About 80 percent of this nitrogen has been used incrop production (Galloway et al. 2002).

Those forms of nitrogen called reactive nitro-gen are critically important in the context ofcrop production and its environmental impact.Although nitrogen exists in many forms in theenvironment and is abundant in the atmosphereas nitrogen gas (N2), this report focuses on twoof the many reactive nitrogen compounds mostreadily used by crops: ammonia and nitrate. Thesecompounds are readily used by both plants andmicrobes, hence are commonly referred to as reac-tive nitrogen. By contrast, N2 cannot be used bymost organisms. Reactive nitrogen enters agricul-tural systems from several sources:

The amount of human-caused reactive nitrogen in the global environment has increased 11-fold since the nineteenthcentury and about eight-fold since the 1960s, which marked the beginning of the green revolution in agriculture.Agriculture is responsible for about 80 percent of the reactive nitrogen produced worldwide.

Source: Adapted from Galloway et al. 2003. 2003, American Institute of Biological Sciences. Used by permission. All rights reserved.

6

4

2

0

1850 1870 1890 1910 1930 1950 1970 1990 2010

World

Population

Total Reactive

Nitrogen

Industrially

Produced Reactive

Nitrogen

Biologically

Produced Reactive

Nitrogen

200

150

100

50

0

Population(billions)

ReactiveNitrogenCreat

ed

(teragramsperyear)

Figure 2. Rise in Reactive Nitrogen Production


17/46

9No Sure Fix

Industrial production of synthetic fertilizer,which combines natural gas and N2 to produceammonia

Microbe-driven decomposition of organic matter

Bacterial nitrogen fixation, the process in whichmicrobes, often associated with legumes such assoybeans and alfalfa, break the N2 bond

Lightning, which can split the N2 bond

Agriculture is often the most important sourceof several reactive nitrogen compounds in theenvironment. Nitrate, for example, is one of themajor forms of reactive nitrogen in fertilizer, anda major source of water pollution. Much of theother major forms of reactive nitrogen in fertil-

izer, ammonia and urea, are rapidly converted tonitrate. Nitrate is a particular problem because it isespecially mobile in the soil, and therefore readilylost through leaching.

The mobility of several forms of reactivenitrogen means that nitrogen can pollute the

environment at local, regional, and global lev-els. In addition, microbes in soils often convertless mobile forms of reactive nitrogen into moremobile forms such as ammonia and nitrous oxide,which are mobile in the air, further contributing tothe spread of nitrogen pollution from farms.

We thus face the dilemma of expandingour food supply to meet the needs of a growingglobal populationfor which we currently rely onincreased nitrogen usewhile reducing pollutionfrom nitrogen. Whether supplied as synthetic fer-tilizer or via the addition of biological componentslike legumes, nitrogen is an expensive input intoan agricultural system, so farmers already want touse it as efficiently as possible. But this objective

has gained new urgency as we witness the impactof nitrogen overuse on global ecosystems. It is nowimperative that we develop new ways of using nitro-gen efficiently if we are to avoid even greater harmto the environment in our quest for more food.


18/46

10 Union of Concerned Scientists

The variety of strategies available for increas-ing NUE (and thereby reducing nitrogenpollution) reflects the different spatialand time scales at which NUE can be analyzed.At the scale of the individual plant, NUE can beincreased by enhancing the capacity of that cropspecies to acquire nitrogen from the soil or bet-

ter use nitrogen within the plant. For example, aplant with a mature root system that continuesto acquire nitrogen even when concentrations inthe soil are lowor that acquires nitrogen morerapidly even when concentrations in the soil arehighwill use more of the available nitrogen inthe soil than a comparable plant with lower NUE.

Similarly, plants that transfer more nitrogen to thegrain or increase grain yields will also use nitrogenmore efficiently.

Plant characteristics that influence NUEinclude the amount of energy allocated to rootsystems (more extensive root systems can enablegreater utilization of soil nitrogen) and the specific

characteristics of enzyme systems used to acquirenitrogen and allocate acquired nitrogen to differentparts of the plant, such as the seed of grain plants.Because the main advantage of GE is its ability totarget specific plant traits (Box 1), we here reviewthe status of GE technology for improving NUE,primarily at the scale of the individual plant.

Chapter2

Nitrogen Use Efficiency in GE Plants and Crops

Genes can be thought of as consisting of two parts: the

part that carries information needed to produce proteins

that underlie traits (the structural gene), and the part

that directs when and how much of the protein is

produced, especially the part called the promoter.

Gene expression refers to the timing and amount of

protein production, which strongly influences plant

function and development. Typically, the most important

regulator of gene expression is the promoter. Genetic

engineers typically alter the timing or amount of protein

production by adding a new promoter to the gene that

causes high expression.

The promoter and the structural gene may each

originate from different genes and different organisms,

and can be brought together in new combinations. For

example, a promoter from a rice gene can be attached to

a structural gene from a bacterium.

Some genes directly control the expression of several

genes. The proteins produced by such genes are called

transcription factors. Transcription factors sometimes

have advantages for the engineering of genetically com-

plex traits (such as NUE) that are controlled by several

genes. But they can also affect the expression of genes

that control traits other than the intended onea result

that may have undesirable consequences. Such a result

can also occur if the expression of single genes that are

not transcription factors is altered.

Altering gene expression has so far proved to be as

important for improving NUE through GE as have struc-

tural genes. Most experimental increases in NUE have

come from increasing the expression of existing structural

genes (or similar genes from other organisms) rather than

using genes that are fundamentally different from those

already found in the crop.

Box 1. How Engineered Genes Contribute to Plant Traits


19/46

11No Sure Fix

How We Evaluated GEs Prospects for

Improving NUE

Ideally, to evaluate the efficacy of a new cropdesigned to increase NUE, we would study the

plants as they are grown on a variety of workingfarmsin the field with varying soil conditions,plant densities, rainfall patterns (over a periodof years), and other factors that influence plantgrowth. Such studies provide realistic estimates ofcommercial promise and reveal unintended conse-quences on and off the farm.

Because on-farm studies are costly, a series ofpreliminary, controlled, and more easily interpretedexperiments are usually performed first. For exam-

ple, new GE plants are typically evaluated first bygrowing them individually in pots in a greenhouse.

Laboratory and greenhouse studies have greatvalue because they show how genetic manipula-tions manifest themselves in plants, rather than ina bacterium in a Petri dish. They do not, however,enable us to evaluate how a crop will contribute toa farming system that may retain or lose nutrientsto the surrounding landscape, air, and water (seeBox 2 for a discussion of different testing environ-ments for GE plants).

The publicly available information on GEcrops with NUE genes comes primarily from con-trolled studies conducted in growth chambers orgreenhouses, and U.S. Department of Agriculture(USDA) records indicate that no such crops have

yet been approved or commercialized. On-farmexperiments, therefore, have not been conducted.

The performance of new NUE crops may be assessed

by growing them within structures or outdoors. The

different methods have their own strengths and weak-

nesses: growth chambers provide the greatest control

over growing conditions and the most precise compari-

sons, while commercial-scale studies provide the most

realistic environment.

Greenhouse and growth chamber studies involve

growing the experimental crop under highly controlled

settings. Though greenhouses typically use ambient light

and may not fully control temperature, they still represent

an artificial environment compared with the exposed

conditions of a crop field. Growth chambers are enclosed

structures that typically control all aspects of crop growth

including temperature, light, and humidity. Plants are often

grown in pots rather than in groups or rows as on a farm.

Field trials test crops outdoors, but under conditions

that can be monitored and treated in a controlled manner.

Although field trials approach commercial crop produc-

tion in terms of exposure to environmental conditions,

they are much more limited in size (plots are often less

than an acre), duration (often for only a few years), and

geographic distribution.

Commercial-scale studies typically involve monitor-

ing crop growth on commercial farm fields that are much

larger than field trials, and may continue (continuously or

intermittently) for many years. Commercial-scale stud-

ies may sometimes be performed like field trials, but at a

much larger scale and for a longer duration.

Growth chambers and greenhouses cannot repli-

cate the complex interactions between a plant and the

environment that occur outdoors, including conditions

that may lead to undesirable side effects. Field trials can

begin to assess environmental effects, but sporadic phe-

nomena such as pests and severe weather may not be

present during the limited duration of a field trial.

Therefore, commercial-scale studies over a long

period of time are needed to reliably detect the effects of

sporadic, but important, environmental phenomena, as

well as processes that take a long time to develop (such

as the accumulation of organic nitrogen in the soil). Such

studies may also provide considerable information about

how plants affect each others growth and about NUE,

including nutrient loss from agricultural systems.

Box 2. Methods Used to Study Crop NUE


20/46


A relatively small number of field trials (whichrepresent an intermediate step between growthchamber and on-farm studies) have been con-ducted, but the results of those trialsconsideredconfidential business informationhave not beenreleased. Without comprehensive field studies, wecannot evaluate the promise of GE NUE cropsunder commercial conditions, or whether seriousdrawbacks such as impaired responses to droughtor pathogens may emerge in the field.

Nonetheless, the available data provide a use-ful assessment of the state of development of GENUE crops. Although many such crops appear tobe in relatively early stages of development, andface several possible hurdles, there are a number of

examples in the scientific literature (beginning inthe 1990s, but primarily since 2000) of genes thathave shown promise for improving NUE. Progressin this area mirrors our increased understandingof nitrogen metabolism by the genes involved inNUE, gained with the use of traditional geneticmethods as well as tools from physiology andmolecular biology (Hirel et al. 2007).

Studies of GE NUE Crops

Researchers have focused much of their efforts todevelop GE NUE crops on seven genes, primar-ily in major grain crops (rice, corn, and wheat)and the oilseed crop canola. Soybeans have been acommon subject of USDA field trials for improvedNUE, but the genes used in these trials are notknown to the public. Most of the research in thepublic literature has centered on plant-derivedgenes important to nitrogen metabolism in plants,though some genes have come from bacteria

(which resemble plants in some aspects of nitrogenmetabolism). Many of these genes have been iso-lated and analyzed in experimental plants such asArabidopsisas well as crops.

Genes that have been evaluated in the litera-ture include:

genes that code for nitrate and ammonium trans-porters that assimilate nitrogen from the soil;

genes such as nitrate and nitrite reductases,which alter the form of nitrogen in the crop so

it may be incorporated into organic (carbon-containing) molecules;

genes that synthesize nitrogen compounds suchas glutamine synthetase, which produces theamino acid glutamine (used to transport nitro-gen through the plant); and

genes responsible for remobilizing nitrogen fromthe vegetative parts of plants into the seed.1

The following discussion of studies describedin the scientific literature focuses on those genes

that have attracted the most attention and haveshown the greatest promise for improving NUE.

In most cases, the GE strategy for nitrogenmetabolism genes has been to boost their expres-sion with gene promoters that cause the gene tobe turned on at high levels in many plant tissuesmost of the time (Box 1) (Good, Shrawat, andMuench 2004). Boosting gene expression through-out a plant means that the protein product of geneexpression will occur in plant tissues where it is not

normally found, or in atypical amounts. This wide-spread change may increase the chance of undesir-able side effects (or pleiotropy, discussed below).

Concern about the likelihood of unintendedconsequences stems in part from our understand-ing that most aspects of plant molecular biology(including nitrogen metabolism) are highly regu-lated and respond to changes in plant biochem-istry. Therefore, atypical expression of nitrogenmetabolism genes will likely cause some reactions

by the plant. Whether these reactions will manifestthemselves in plant growth and cause agricultural,environmental, or human safety problems is usual-ly not entirely predictable given our current knowl-edge of plant biochemistry and metabolic networks(Sweetlove, Fell, and Fernie 2008).

1 A more detailed list and discussion about these genes can be found in Good, Shrawat, and Muench (2004).


21/46

13No Sure Fix

Using promoters that express nitrogen metabo-lism genes at high levels in many parts of theplant, in most cases, has resulted in increased NUEin experimental crops. Below and in Table 1 is alist of the gene-crop combinations of potentialinterest to genetic engineers.

Perhaps the most widely explored genes forimproved NUE are those that control productionof glutamine synthetase (GS). Several versions ofthese genes, called a gene family, appear to becentral to nitrogen metabolism because glutamineis the primary compound involved in the move-ment of nitrogen throughout the plant, includinginto the growing seed. Versions of GS genes arefound in the root and in the green parts of the

plant. GS has been engineered into several crops.Glutamine synthetase in wheat. GE wheat

was developed using a bean GS gene and a strongpromoter from a rice gene (Habash et al. 2001).

Plants were grown under controlled light and tem-perature in a growth chamber using a soil pottingmix. The over-expression of this gene, comparedwith the normal wheat GS gene, in the green tis-sues of the plant resulted in an increased grainyield of about 10 percent, and increased grainnitrogen by a somewhat larger amount, under nor-mal nitrogen fertilization. This occurs by increas-ing the reallocation of nitrogen in the plant fromthe leaves to the seed.

The root system of the GE GS wheat plantswas also enhanced compared with non-GE wheatplants. While this may be a beneficial result, pos-sibly enhancing nitrogen assimilation, it illustratesthe side effects that often occur with the altered

expression of engineered genes.Glutamine synthetase in maize. A maize GS

gene, normally expressed in leaves, was over-expressed using a promoter taken from a plant

Table 1. Genes Used to Improve NUE through Genetic Engineering 1

GeneGene Source

(Gene/promoter)Engineered Plant

NUE Improvement 2

(Percent)Grown in the Field?

3

Glutamine synthetase (GS) Bean/rice Wheat 10 No

Glutamine synthetase (GS) Corn/plant virus Corn 30 No

Glutamate synthase (GOGAT) Rice/rice Rice 80 No

Asparagine synthetase (AS) Arabidopsis/plant virus Arabidopsis 21 No

Glutamate dehydrogenase E. coli/plant virus Tobacco 10 Yes

Dof1 Corn/plant virus ArabidopsisNitrogen content: 30;

growth: ~65No

Alanine aminotransferase(ALA)

Barley/canola Canola 40 Yes

Alanine aminotransferase(ALA)

Barley/rice Rice 3154 Yes4

Notes:

1 As reported in the public literature; other genes may be under private study by companies and universities.

2 Values for NUE are measured in different ways in different experiments. Therefore the values presented here are not directly comparable.

3 It is possible that field trials for these genes have been conducted but not disclosed to the public.

4 USDA field trials have been approved for this gene, but the results have not been reported to the public.


22/46


virus that produces GS in most plant tissues.Plants were grown in a greenhouse in pots, andproduced about 30 percent more grain under low-level nitrogen fertilization (Martin et al. 2006).

Glutamate synthase in rice. Glutamate synthase(GOGAT) genes represent another gene familyimportant in plant nitrogen metabolism, and havebeen used in experiments to improve NUE in rice.Genetically engineered indica ricethe primarysubspecies grown in India and several other parts ofAsiawas developed using an indica GOGAT geneunder the control of a GOGAT promoter from adifferent rice subspecies, japonica rice (Yamayaet al. 2002).2 Grain yields for GE indica plantsgrown in pots in controlled conditions were 80 per-

cent higher than for the non-GE indica plants.Asparagine synthase inArabidopsis. As with

the GS gene, the asparagine synthase (AS) genecontrols the synthesis of an amino acid that canbe important for transporting nitrogen through aplant. AS was over-expressed in the experimentalplant, Arabidopsis, using a strong promoter from aplant virus that produces high levels of AS in mostplant tissues (Lam et al. 2003). The GE plantswere grown in pots under controlled light and

temperature and normal levels of nitrogen. Seedprotein content increased by about 21 percent.

Glutamate dehydrogenase in tobacco. Underfield conditions in Illinois, a bacterial glutamatedehydrogenase gene (from E. coli) expressed athigh levels in tobacco using a promoter froma plant virus produced up to about 10 percentmore plant biomass than the non-GE plants overa period of three years (Ameziane, Bernhard, andLightfoot 2000). Increased crop yield appeared to

occur only at normal nitrogen fertilization levels.Dof1 transcription factor inArabidopsis. The

maize Dof1 gene is a transcription factor (Box 1)that controls the expression of several genesinvolved in carbon metabolism (Yanagisawa et al.2004). Carbon and nitrogen metabolism are linkedin plants, and because many plant molecules

contain significant amounts of both carbon andnitrogen, increased expression of a gene for carboncompounds may also boost nitrogen in the plant.The GE Arabidopsisplants containing Dof1 at highlevels accumulated more nitrogen than normalplantsin some cases more than twice as muchwhen grown in the laboratory on an artificial agar-based medium containing low amounts of nitrogen.The GE plants also showed greater growth thantheir non-GE counterparts, although the amount ofgrowth difference was not quantified.

Alanine aminotransferase in canola. The ala-nine aminotransferase (ALA) gene is one of the fewnitrogen metabolism genes that has been expressedfrom a promoter restricted to specific plant tissues

and environmental conditions, and grown in thefield rather than only in greenhouses or growthchambers. Investigators combined a barley ALAgene with a promoter that functions in the roots ofcanola plants and used the resulting combination togenetically alter canola plants (Good et al. 2007).

In field trials over a two-year period, and withnitrogen fertilizer application rates 40 percentbelow normal, they observed canola seed yieldsequivalent to those achieved at typical soil nitrogen

levels. At more typical application rates, the GEcanola exhibited a yield increase of approximately33 percent. At high application rates (280 kg/hect-are), no yield advantage was reported.

Alanine aminotransferase in rice. A barleyALA gene was expressed by a root-tissue-specificpromoter in GE rice (Shrawat et al. 2008). Undercontrolled conditions, grain yield increasedbetween 31 and 54 percent compared with thenon-GE rice. Root and fine root biomass also

increased considerably, as did nitrogen uptake. TheUSDA has also approved field trials of ALA rice,but the results have not been released to the public.

Summary. Our review of the literature revealedseveral genes important to plant nitrogen metabo-lism that have drawn the interest of genetic engi-neers. Of these, GS genes have probably attracted

2 There are several distinct types of Asian riceindica, japonica, and javanicaall of the species Oryza sativa, and all generally inter-fertile. Indica rice varieties are the most widelygrown.


23/46

15No Sure Fix

the widest interest. Promising results have alsobeen observed with GOGAT and ALA. Work onthe latter appears to be the most advanced, withfield trials lasting several years (see below).

The studies described above, mostly conductedin controlled environments, demonstrate thatNUE genes can increase both seed yield (at low,normal, or high nitrogen fertilizer levels) and plantnitrogen content. Grain yield increases in green-house tests have ranged from approximately 10percent to 80 percent (Table 1). However, tests incontrolled environments may not identify undesir-able genetic side effects that manifest themselvesunder certain environmental conditions, and maynot detect other limitations imposed by commer-

cial-scale crop production.

Approved Field Trials of GE NUE Crops

Field trials test experimental GE crops under con-ditions that may approach those on farms, andafford the opportunity to assess a variety of pos-sible environmental impacts as well as NUE atthe scale of a crop field (rather than an individualplant). However, secrecy about genes and fieldtrial results greatly limits our ability to evaluate the

prospects of these genes. Field data are critical toassess the success of efforts to produce high-NUE

crops because, for example, an individual plantmay have high NUE when grown in a pot butlower NUE in the field if fertilizer is applied beforeroot systems have developed sufficiently to colo-nize most of the fields soil. Nutrient losses oftendepend on the timing of not only fertilizer applica-tion but also irrigation and/or rainfall.

U.S. field trials of GE crops must receiveUSDA approval, and are listed in the USDAs pub-licly available GE field trial database. This databasetherefore provides the number of all approvedNUE field trials in this country, and offers a gener-al sense of how advanced this research is comparedwith other GE traits.

Between 1987, when the USDA initiated its

field trial program, and 2000, only 26 field tri-als for nitrogen metabolism were approved, but99 have been approved since then (Animal PlantHealth Inspection Service 2009). This substantialincrease over the past decade suggests growinginterest in, and identification of, possible NUEgenes. Nevertheless, the total number representsonly a fraction of the field trials approved forinsect-resistant and herbicide-tolerant GE crops:there have been 4,623 field trials approved for

herbicide tolerance and 3,630 for insect resistancethrough 2008 (Gurian-Sherman 2009) (Figure 3).

5,000

4,000

3,000

2,000

1,000

0Insect

Resistance

Herbicide

Tolerance

NUE

3,360

4,623

125

* Field trials for herbicide tolerance and insect resistance approved through February 2009. Field trials for NUE approved throughAugust 2009. Source: USDA, APHIS Biotechnology Regulatory Services, online at www.isb.vt.edu/cfdocs/fieldtests1.cfm.

Figure 3. USDA-Approved Field Trials of GE Crops*


24/46


The relatively small number of field trials forNUE shows that advances in this research are morerecent than that on other traits, and less advanced.It is also consistent with the small number of genesthe public literature suggests have attracted themost interest. For example, if the number of fieldtrials for the ALA gene is representative of otherNUE genes, then dividing the total number ofNUE trials (125) by ALA field trials (17) suggestsabout seven NUE genes being studied in field tri-als. On the other hand, it is also possible that sometrials may involve several NUE genes.

All of the field trials conducted through2004as well as several conducted afterwardusethe general term nitrogen metabolism altered

to describe the GE trait. It is unclear how manyof these 60 approved field trials were attempt-ing to increase NUE specifically, but because theterms nitrogen metabolism altered and NUE areused to describe the same gene at different times,we have included these trials in our total underthe assumption that at least some had the goal ofimproving NUE.

The USDA database also provides a windowon which institutions are investing in enhanced

NUE via GE, and which crops have receivedattention. The large majority of field trials, forexample, have been conduced by either Monsantoor Pioneer Hybrid. Several have also been con-ducted by Arcadia, which is using the ALA gene.This company appears to be collaborating withMonsanto, as revealed by a paper discussing GEALA in canola that was co-authored by scientistsemployed by both companies (Good et al. 2007).Most of the NUE field trials involve corn, with

many involving soybeans, canola, and rice as well.A few have been conducted using other crops, buthave not been carried forward to recent years; oneinvolving the potential biofuel crop switchgrasswas approved in 2009.

Because of current limits on the public avail-ability of field trial data, we must rely on infer-ences about the genetic and physiological effects

of GE NUE genes on the plant to evaluate theirprospects for success.

Possible Risks Related to GE NUE Genes

Limited testing has already revealed several pos-sible undesirable or harmful unintended changes inthe expression of plant genes due to the engineer-ing of NUE genes. Even when GE NUE cropsshow promise in greenhouse tests, the possibilityof undesirable or harmful side effects (pleiotropiceffects) when those crops are grown in the fieldmay reduce the value of the gene. Field trials con-ducted for several years are more likely to detectundesirable side effects, but some may only beobserved in response to occasional occurrences,such as extreme heat or cold or an outbreak ofpathogens, that may not occur during field trials.

One particularly worrisome side effect of GENUE genes is that they may indirectly increasethe production of harmful substances in the edibleparts of crops. Most crops have genes that pro-duce harmful substances, but these genes are notexpressed, or are expressed at low levels, in theedible parts of crops. Engineered genes, however(or genes manipulated through traditional breed-

ing), may have the opposite effect due to complexinteractions between the engineered gene and cropgenes (National Research Council 2000).

Consider the E. coliglutamate dehydrogenasegene, which was studied as a possible NUE gene(Ameziane, Bernhard, and Lightfoot 2000). Whenexpressed in tobacco it altered the production ofmany plant compounds (some were increased andsome were decreased), most notably the amountsof nine known carcinogens and 14 potential drugs

(Munger et al. 2005). Although tobacco is notedible, this example illustrates the possibility ofunpredictable and potentially harmful changes infood crops.

Because we know that the nitrogen statusof plants affects various aspects of their physiol-ogy, including defense against pests (Craine et al.2003; Vitousek et al. 2002), it is reasonable to ask


25/46

17No Sure Fix

whether altering nitrogen metabolism with NUEtransgenes could influence the amounts and typesof important plant components.

Recent tests have found that overexpressionof ALA in rice causes a significant change in theexpression of 91 other genes in the roots andshoots of rice plants grown hydroponically (Beattyet al. 2009). Seventeen of these genes had alteredexpression in two independently created ALA riceplants. The identified rice genes are involved invarious aspects of plant function: several have beenassociated with defense against pathogens, one ofwhich (called the osmotin-like, thaumatin-likegene) was expressed at a two- to three-fold lowerlevel than in normal rice plants. Two genes of the

PR10 type (a pathogenesis-related protein impli-cated in the defense of plants against disease) werealso found to have significantly reduced expression.Reduced expression of these genes raises a questionabout the possible increased susceptibility of ALAGE rice to disease.

In summary, pleiotropic effects are a distinctpossibility for GE NUE crops, but have yet to beexplored in the public literature. Because they arelargely unpredictable and may only occur under

specific environmental conditions, these side effectsmay not be revealed by the types of experimentsthus far performed (mostly under greenhouse con-ditions). Even when such crops are grown in thefield, some changes in gene expression may onlybe detected through sophisticated testing of plantgenes or compounds, as was done for tobaccocontaining a glutamate dehydrogenase gene andrice containing an overexpressed ALA gene; suchtesting is not required under current U.S. regula-

tions. Many side effects may be harmless or incon-sequential for crop production, but the possibilitythat some could be undesirable should be carefullyevaluated.

Commercialization of GE NUE Crops

There is not enough detailed information aboutthe performance of GE NUE crops at this time to

clearly understand their prospects for commercial-ization. Commercial potential is therefore generallyinferred from available information about a) theefficacy of NUE genes and b) possible hurdles thatmay be faced as these crops are tested under morerealistic conditions and as they proceed throughthe regulatory process.

The NUE values obtained for GE crops inrecent tests, most of which were conducted incontrolled environments and with limited dura-tions, are unlikely to be maintained on commercialfarms under real-world conditions. In addition, theapparently limited number of comparisons withexisting crop varieties that may differ in NUE alsosuggests that NUE values for GE crops may be

lower than reported (see Chapter 3).Only the actual performance of GE NUE

crops will determine whether these varieties areeconomically viable and attractive compared withother technologies for improving NUE. The NUEvalues of GE crops need to be high enough tojustify the costs of development, production, andmarketing, as well as the extra costs farmers mustpay for GE seed.

Undesirable side effects, where they exist, may

reduce the efficacy of these crops, force farmersto pay additional costs, and affect how widely thecrops are adopted if approved. For example, ifplant diseases are exacerbated in some instances(see above), higher costs for disease control couldreduce the adoption rate of the crop and, in turn,the practical impact on NUE. When side effectsare harmful to the environment, they may alsoprevent regulatory approval.

The ALA gene shows the most promise for

commercialization based on publicly availableinformation. It is the only gene identified inUSDA field trials, 17 of whichalmost 14 percentof all NUE field trialshave been approved since2002. This long record suggests that the ALA genemay be approaching the late stages of testing.On the other hand, the lack of large-scale fieldtrialsnone of more than five acresthat are


26/46


usually conducted within several years of regula-tory approval may suggest that commercializationis at least several years away.

This gene also reduced expression of severalgenes in rice that help the plant defend itselfagainst disease. It is unclear at this time what prac-tical effects this may have.

No petition for deregulation of any NUEcropa prerequisite for commercializationhasyet been announced by the USDA in its publicdatabase (Animal and Plant Health InspectionService 2009b). Examination of the petitiondatabase shows that deregulation decisions gener-ally require at least two years. It seems unlikely

therefore that any NUE crop will be commercial-ized in the United States before 2012.

Most of the GE NUE crops reported in thescientific literature appear to be at relatively earlystages of development, with the possible exceptionof the barley ALA gene in canola and rice. If othergenes are in more advanced stages of development,the work is occurring behind closed doors. Basedon the information available to us, the prospectsfor commercialization of GE NUE crops must beconsidered largely uncertain at this time. Theseprospects should be compared with those of othermethods and technologies for addressing nitrogenpollution, which are addressed in the next chapters.


27/46


28/46


crop breeding (Duvick 2005). If that relationshiphas continued during the past dozen years, thentraditional breeding may account for roughly halfof the improvement in NUE over this period basedon yield per unit of added nitrogen, or about halfof the increased yield value.

This provides a very rough estimate of tradi-tional breedings contribution to improved NUEin this country in recent years. Overall, consider-able improvement has occurred over the past sev-eral decades, as a result of both breeding and othermeans (Table 2).

Genetic Variability of NUE-Related Traits in

Major Crops

As past studies suggest, there is considerablepotential for improving NUE through traditionalbreeding methods, but this potential depends onthe variability of NUE-related traits, and theircorresponding genes, within a crop or its wild rela-tives. Much of the genetic potential of major cropsremains untapped for many traits (Hoisingtonet al. 1999), which likely include NUE.

Hoisington et al. found that only a small por-tion of the genetic variation in corn and wheathas been utilized in current crop varieties. Thisunder-utilization is especially true for wild relativessuch as Tripsacumand Teosintespecies for maizeimprovement and Aegilops, Agropyron, and non-wheat Triticumspecies for wheat improvement.Only about 1 percent of the U.S. maize germ-plasm base, and only about 5 percent of the glob-ally available germplasm base, has utilized theseresources (Hoisington et al. 1999). For wheat, onlyan estimated 10 percent of varieties as of 1986 mayhave used the genetic resources from exotic wheatvarieties (called landraces) to improve existingwheat varieties.

Despite this minimal use of the availablegenetic diversity, tremendous contributions havealready been made to maize and wheat improve-ment, including numerous genes for disease resis-tance, insect resistance, stress tolerance (such as fordrought), quality traits, and yield (Hoisingtonet al. 1999)suggesting there is also potential herefor improving NUE. Another possible resource is

Table 2. Improvements in Nitrogen Use Efficiency

CropTime Frame*

(Years)Country

Source of

NUE Gain

NUE Gain

(Percent)Reference

Wheat 35 Mexico Breeding 42 (59 kg/ha/year)Ortiz-Monesterio

et al. 1997

Wheat 35 France Breeding 29Brancourt-Hulmel

et al. 2003

Rice ~15 Japan Unknown 32Dobermann and

Cassman 2005

Maize ~20 United States Unknown 36Dobermann andCassman 2005

Cereal crops 1520 United Kingdom Unknown 23Dobermann andCassman 2005

* All studies were conducted in the second half of the twentieth century.


29/46

21No Sure Fix

the existence of sexually compatible wild relativesfor virtually all other major food crops, such as therice relative Oryza rufipogonand the soybean rela-tive Glycines soja(Ellstrand 2003). NUE has notbeen widely investigated in these genetic resources.

Some research with nitrogen metabolism genesshows some of the genetic variability for traits orgenes associated with NUE within a crop species.Perhaps most striking is the genetic variabilityfound in rice for a GOGAT gene (Yamaya et al.2002). The gene used for this experiment (or moreproperly, the gene promoter) originated in onetype of rice (japonica) and was inserted into anoth-er type of sexually compatible rice (indica) afterbeing attached to the indica GOGAT gene, result-

ing in an 80 percent increase in yield. Because thegenes were from sexually compatible varieties ofrice, a similar yield improvement may be accom-plished using traditional breeding techniques.3

Related research shows a similarly high level ofvariation in the amount of another gene and pro-tein widely studied and used to develop GE NUEcrops. GS protein in rice leaves ranged from 2.55to 16.18 microgramsmore than a six-fold dif-ferencein the offspring of two varieties, one a

japonica and the other an indica type (Obara et al.2001). To the extent that levels of GS expression areimportant for improving NUE, as has been seen inother experiments, this variation suggests substantialpotential for improving NUE through breeding.

Obara et al. identified this variability by com-paring only two varieties of rice, albeit varietiesthat are genetically divergent. These two variet-ies do not contain all of the genetic variabilitycontained in all rice varieties or their wild rela-

tives, and therefore do not reveal how much addi-tional variabilitywhich may be used to increaseNUEcould be found if more of the rice genepool was examined.

Research in corn has revealed numerous chro-mosomal locations associated with variation indifferent aspects of NUE (Coque and Galais 2006;Galais and Hirel 2004). Several of these regions,called quantitative trait loci (QTL), are also associ-ated with genes that have been used experimentallyin GE to enhance NUE, including several GSgenes and glutamate dehydrogenase. Genetic mark-ers that are linked to QTL and function as geneticfingerprints can be used to track the QTL duringbreeding in a process called marker-assisted breed-ing, which can greatly accelerate breeding.

QTL and significant genetic variability forNUE have also been found in barley (Mickelsonet al. 2003). Preliminary work in wheat has iden-

tified several QTL associated with NUE, whichinclude one or more GS genes in the flag leaf (theleaf nearest the wheat seed head), which is knownto be important for grain yield (Habash et al.2007). Development of improved crop varieties thatuse QTL can be difficult to accomplish in practicalbreeding programs because they may perform wellin one environment or in one variety of the crop,but not as well in others (Bernardo 2008; Dekkersand Hospital 2002; but see Heffner, Sorrels, and

Jannick 2009 for a more optimistic view).Higher values for NUE-associated traits have

been observed in progeny than in either parent(Mickelson et al. 2003).4 This demonstrates that,at least for the varieties tested, improvement ofNUE-associated traits found in parent varieties ispossible. Traditional breedings ability to improveNUE can thus be enhanced by new methods basedon the identification of specific genes or regions ofcrop genomes.

Even if there is considerable genetic diversitywithin a crop, however, it is possible that currentcommercial varieties may already contain verygood genes for NUE, which could reduce the

3 Traditional breeding would not provide exactly the same result, because the promoter from the japonica rice gene was combined with a structural gene from indica rice using GE,which would not typically occur with traditional breeding. We are assuming that overexpression (rather than some other activity) of the GOGAT enzyme is primarily responsible forthe results, but it is possible that overexpression is not responsible for the entire difference in NUE.

4 This is a characteristic called transgressive hybridization, which may be associated with strongly adaptive traits. Many traits in the progeny of varietal crosses, however, show valuesbetween those of the two parents.


30/46


potential for further improvement through tradi-tional breeding. Other possible barriers includeundesirable pleiotropic effects similar to those thatmay occur through GE.

The potential for improving NUE throughtraditional breeding is likely to be considerable (butnot unlimited). Recent research on crop geneticvariability, and variability for NUE traits specifi-cally, suggests considerable variation exists for traitsassociated with NUE. The extended crop gene poolthat includes sexually compatible wild relatives doesnot seem to have been explored for the purposeof improving NUE, but may provide additionalopportunity to improve NUE through breeding.

Strengths and Limitations of BreedingCompared with GE

Given the need to allocate public research moneyjudiciously, it would be wise to compare the rela-tive prospects of GE and traditional breeding forimproving NUE. In general, both methods havethe capacity to generate improved crop varieties,but only traditional breeding has thus far succeed-ed in bringing varieties with improved NUE to themarketplace.

For both GE and breeding, reported valuesof improvement in NUE should be viewed aspreliminary prior to extensive field testing thatincludes comparisons with the best current variet-ies of the crop. Many of these values were derivedfrom studies of plants grown in pots, inside growthchambers or greenhouses where light, water, andtemperature were controlled. Growing conditionsin the field can be expected to introduce stressesand other environmental effects that may negative-

ly affect NUE values. In addition, the genetic andphysiological complexity of nitrogen metabolismin plants presents a considerable challenge for GEapproaches relying on single genes (Lawlor 2002),which may lead to problems in the field.

Equally important, the values reported in theliterature for GE NUE crops are typically deter-mined by comparing the GE variety only with

its non-GE progenitor. Such comparisons do notreflect the variation in NUE that exists in commer-cial or other available varieties of the crop. Thus,because some other varieties of the crop may deliv-er better NUE than the one used for comparisonwith the GE variety, the relative NUE advantage ofthe engineered gene would be less than the valuereported in the literature.

In theory, GE should be capable of develop-ing new crop varieties more quickly than breed-ing (Long et al. 2006) because it involves addingonly one or a few genes, while breeding combinesthe entire genomes of the two parents. Removingundesirable genes to arrive at the desired combina-tion of genes typically requires years of breeding.

But, this presumed advantage of GE appearsto be minimal or absent in practice. First, theGE process itself introduces mutations and otherchanges in the plant that may be undesirable(National Research Council 2004). Althoughplants are initially screened for obvious unintendedalterations, many potential changes can involveplant metabolism or occur only under certainenvironmental conditionsfactors that would notbe detected during the initial screening. Many of

these mutations can be eliminated by the samekind of iterative process used to improve plantsthrough traditional breeding, but this requires con-siderable time.

Time is also added to the GE process becausethe effects of new engineered genes under differ-ent growing conditions are not predictable. NewGE varieties must therefore be grown in field trialsfor several years (as must new varieties developedthrough traditional breeding).

Meanwhile, the breeding process has beenimproved by our increasing understanding of plantgenetics, physiology, and biochemistry. This hasled to selection methods that can accelerate thebreeding process substantially, further reducing anyadvantages of GE.

Certain studies have found that GE cropsrequire more than a decade to be developed and


31/46

23No Sure Fix

deployedsimilar to the amount of time neededfor traditionally bred crops (Goodman 2004;Gepts 2002).

Finally, as noted above, the supposed advantageof GE over breeding in providing expanded accessto genetic resources has yet to result in improvedNUE. The available research papers that have pro-vided preliminary quantification of NUE improve-ment through GE or breeding have not revealed adistinct advantage for either approach.

To reiterate, although GE and traditionalbreeding both have the potential to produce newcrop varieties with higher NUE, only traditionalbreeding has succeeded in bringing such varietiesto market (GEs attempts are limited to the past

10 to 12 years). Whatever advantages GE is pre-sumed to have in generating new varieties, they arenot apparent in this arena or any other involvingcomplex traits.

So, while there is no reason to abandon ongo-ing GE efforts, there is no reason to expect morefrom them than traditional breeding, and theyshould not be favored in the allocation of scarceresources. Evidence shows that public resourcesfor traditional breeding have declined globallyin recent decades (Kloppenburg 2005) despiteits success with complex traits such as NUE. Wemust ensure that public sector traditional breedingreceives a level of support commensurate with itsdemonstrated potential.


32/46


Ecosystem approaches consider the spatial,temporal, and species interactions that canaffect a crops NUEfactors not necessarilyconsidered during breeding for NUE, which oftenfocuses single-mindedly on crop yield per unit ofadded nitrogen. Viewed exclusively through thiscrop production lens, NUE may miss important

routes of nitrogen loss from the farm. For example,a crop with improved NUE may not reduce nitro-gen losses early in the growing season, prior tovigorous crop growth and root production.Ecosystem approaches thus represent a possibleroute to both higher crop yields and lower nitro-gen loss and pollution.

A Big-Picture Perspective

Ecosystem scientists view cover crops as part of a

holistic plant-soil system, and their approach tomeasurement reflects this view. Key data pointsoften include actual losses of reactive nitrogenfrom the farm, in the form of runoff, leachinginto groundwater, and gaseous emissions from thesoil (e.g., Tonitto, David, and Drinkwater 2006;Drinkwater, Waggoner, and Serrantonio 1998),and sometimes include a crops uptake of nitrogenas a percentage of the nitrogen applied. Throughthis lens, NUE could be defined as the amount

of plant matter or grain produced with the leastnitrogen pollution.

An ecosystem perspective also expands the timescale over which we consider NUE, drawing atten-tion to periods when plants are not actively grow-ing or when recently planted crops have immature

root systems that draw nitrogen from only a smallportion of the total soil volume. Reactive nitrogenthat goes unused when crops are not actively grow-ing can be a major source of nitrogen loss fromfarms (Tonitto, David, and Drinkwater 2006).Therefore, crop species that can be planted earlierin the seasonor that persist later in the grow-

ing seasoncan potentially reduce nitrogen lossby capturing more soil nitrogen than crops with ashorter growing season. This intersection betweenroot development, a plants nitrogen demand, andthe timing of fertilizer application also plays a rolein determining farm-level NUE.

Viewing the agricultural system at large spa-tial and temporal scales points to a variety ofapproaches (precision agriculture, use of covercrops) that can control the flow of nitrogen

between farm and adjacent systems (air, water).Cover crops, which are often used in organic orsimilar agricultural systems, and precision farming,which is used more often in traditional systems, arediscussed in Chapter 5.5

The Time Is Ripe for a New Approach

Any progress in nitrogen use we have made up tothis point has not led to the decrease in nitrogenpollution we need. For example, U.S. corn yields

have increased about 28 percent over the past13 years (Gurian-Sherman 2009), and productivityof other major crops such as soybeans and wheathave also increased. Nitrogen fertilizer use onmajor crops remained about the same during mostof that period (Wiebe and Gollehon 2006a),

Chapter4

The Ecosystem Approach to NUE

5 Means of reducing nitrogen pollution directly (e.g., planting vegetative buffer strips between crop fields and streams) are also important but not covered in this report.


33/46

25No Sure Fix

suggesting a substantial improvement in NUEbut several indicators suggest nitrogen pollutionhas not improved significantly.

For example, the so-called dead zone in the Gulfof Mexico, largely the result of agricultural nitrogenpollution, expanded during the 1990s, peaked in2002, and has remained at near-record size since.The U.S. Environmental Protection Agency (2008)has suggested that nitrogen pollution will need to bereduced about 45 percent to substantially shrink thedead zone. Other studies confirm that nitrogen

pollution remains a serious problem (Rockstrmet al. 2009; Vitousek et al. 2009).

Looking to the future, this analysis suggeststhat simply increasing the efficiency of crops (asdefined by yield per unit of nitrogen applied) isunlikely by itself to reduce pollution sufficiently.Any improvements in NUE must therefore beviewed from an ecosystem perspective that givesequal weight to preventing nitrogen loss andincreasing crop yields.


34/46


In addition to GE and traditional breeding,several other agricultural technologies or prac-tices show promise for improving NUE. Thischapter sets our evaluation of GE and breeding ina broader context by providing a brief overview ofprominent alternatives for improving NUE: preci-sion farming and organic or other low-external-

input farming systems 6 that use livestock manure,or green manure 7 from cover crops,8 as sourcesof crop nutrients. Both precision farming andcover crops can be incorporated into industrialagricultural systems; systems that use little or nopesticides or synthetic fertilizers require a morefundamental change from the predominant indus-trial farming system, but deliver a richer set ofenvironmental benefits.

Both precision farming and organic or simi-

lar systems attempt to improve NUE by manag-ing nitrogen input and the amount of nitrogenin the soil rather than altering the plant genome.Precision farming focuses on matching the nitro-gen supplied from synthetic fertilizers to the needsof the crop, avoiding the excesses that contributeto nitrogen pollution. Organic farming and similarsystems emphasize building soil quality and soilorganic matter, which provides multiple benefitsincluding reduced nitrogen loss from the farm.

In general, the negative environmental impactsof nitrogen, including air and water pollution andthe production of nitrous oxide, increase as theamount of inorganic nitrogen applied increases.

Industrial agriculture, which commonly applies alarge amount of synthetic, inorganic reactive nitro-gen at oncemore than crop roots can assimilateover a short period of timeis especially damag-ing. By contrast, methods that minimize the use ofsynthetic fertilizer, release nitrogen slowly over thegrowing season, or remove excess nitrogen from

the soil reduce the negative impacts of nitrogen.Both organic and precision farming take into

account the nitrogen sources already availablein soil (so-called indigenous nitrogen), which isprimarily organic (i.e., bound to carbon atoms)in form. Organic nitrogen breaks down into inor-ganic forms that are used by the crop but can causepollution if they find their way into water or air.9

It is generally desirable to increase the amountof indigenous nitrogen available as a source of

inorganic nitrogen for crop nutrition becauseit tends to contribute less to nitrogen pollu-tion (Cassman, Dobermann, and Walters 2002).Indigenous nitrogen generally releases inorganicnitrogen continuously, in amounts smaller thanindustrial agricultures typically large applicationsof synthetic fertilizer.

The amount of organic nitrogen in the soil andthe rate at which inorganic nitrogen is applied tothe soil or released from organic sources are impor-

tant considerations for both organic and precisionfarming. Specifically, the amount of synthetic inor-ganic nitrogen added to the soil should take intoaccount the amount released from the indigenous

Chapter5

Other Means of Improving NUE

6 Low-external-input systems emphasize the use of biological principles to achieve soil fertility and pest control, and include organic farming as well as methods that allow a minimaluse of synthetic fertilizers or pesticides.

7 Green manure refers to the use of plants as a means of supplying nutrients to other crops. Green manure crops are often grown during seasons when food crops are not grown; insteadof being harvested they are plowed into the soil, where they release their nutrients.

8 Cover crops are planted to protect soil that would otherwise lay bare (between cropping seasons, for example) and subject to erosion. Cover crops also take up inorganic nitrogen thatwould otherwise be lost from the field. Plowing cover crops into the soil prior to the planting of cash crops provides nutrients, improves soil quality, and increases soil carbon content.

9 Some organic compounds can be used by crops but are not as important as inorganic forms. Some can also move though soil into groundwater, but these are also generally unimportant.


35/46

27No Sure Fix

nitrogen supply. Because this can be challenging inpractice, it is not always done.

Cover Crops

Nitrogen can be supplied to crops by incorporat-ing livestock manure or leguminous plants used asgreen manure into the soil. Both kinds of manurecontain organic forms of nitrogen incorporatedinto large molecules such as proteins that are bro-ken down into the smaller inorganic forms usefulto crops.

Many of the major crops that are the targetof both GE and traditional breeding cannot pro-duce useable nitrogen, but otherslegumes, forexamplecan. Legumes include important food

and feed crops such as beans, peas, soybeans,peanuts, and alfalfa, as well as cover crops suchas vetches and clovers. These crops live in closeassociation with bacteria that can produce reac-tive nitrogen usable by the crop itself10 and bynon-legume crops planted in succeeding seasons.Because legume cover crops may supply most orall of the nitrogen needed for subsequent crops toproduce high yields, incorporation of legumes intoagricultural systems can reduce the need to sup-

ply synthetic nitrogen (thereby helping to reducenitrogen pollution).

Legumes supply nitrogen in the form oforganic molecules that are generally retained in thesoil for longer periods of time than synthetic nitro-genan additional advantage for reducing pol-lution. But because much of the organic nitrogenmay be converted into more reactive forms such asammonia or nitrate relatively quickly under certainconditions, the organic sources must be properly

managed to avoid causing nitrogen pollution.Manure and green manure also add carbonand other nutrients to soil, which may generallyimprove soil quality. For example, increasing soilorganic matter generally improves the soils water-holding properties and soil nitrogen levels, therebyimproving the ability of crops to survive drought

(Lotter, Seidel, and Liebhardt 2003). Use of covercrops on otherwise fallow soil

union of concerned scientists...

Documents