The Nitrogen Cascade Author(s): James N. Galloway, John D. Aber, Jan Willem Erisman, Sybil P. Seitzinger, Robert W. Howarth, Ellis B. Cowling, B. Jack Cosby Source: BioScience, Vol. 53, No. 4 (Apr., 2003), pp. 341-356 Published by: American Institute of Biological Sciences Stable URL: http://www.jstor.org/stable/1314367 Accessed: 12/04/2010 10:58 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=aibs. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. American Institute of Biological Sciences is collaborating with JSTOR to digitize, preserve and extend access to BioScience. http://www.jstor.org oArticles The Nitrogen Cascade JAMESN. GALLOWAY, JOHND. ABER,JAN WILLEMERISMAN,SYBILP SEITZINGER, ROBERTW. HOWARTH, ELLISB. COWLING, AND B. JACKCOSBY Humanproductionoffood and energyis the dominantcontinentalprocessthat breaksthe triplebondin molecularnitrogen(N2) and createsreac- Nr in Earth's andbiosphere hasa widevarietyofconsequences, tivenitrogen (Nr)species.Circulation ofanthropogenic atmosphere, hydrosphere, Nr can cause whicharemagnified withtimeasNr movesalongitsbiogeochemical The same atom multipleeffectsin theatmosphere, pathway. of Asthecascascade. in terrestrial infreshwater andmarinesystems, andonhumanhealth.Wecallthissequence ecosystems, ofeffectsthenitrogen the rate all environmental not at same cadeprogresses, theoriginofNr becomes Reactive does cascade systems; through unimportant. nitrogen andresult Theselagsslowthecascade somesystemshavetheabilitytoaccumulate Nr,whichleadstolagtimesin thecontinuation of thecascade. NraccumuTheonlywaytoeliminate whichin turncanenhancetheeffectsofNr on thatenvironment. inNr accumulation in certainreservoirs, Nr backto nonreactive lationandstopthecascadeis toconvert N2. ozone,denitrification cascade, Keywords: fertilizer, forestdieback, eutrophication, nitrogen T he chemicalelementsnitrogen(N), carbon(C), phosphorus(P), oxygen (0), and sulfur(S) areall necessaryforlife.Withone exception,theyaregenerallyavailable in globalreservoirsto sustainlife formsrangingfromsinglecell organismsto vertebrates.Of these elements,N has the greatesttotalabundancein Earth'satmosphere,hydrosphere, and biosphere;it is ironic that N is the elementleast readily availableto sustain life. The total amount of N in the at4 x 1021 mosphere,soils,andwatersof Earthis approximately of these other elof all four than the mass grams(g)-more than more ements combined (Mackenzie1998). However, 99%of this N is not availableto more than 99%of livingorganisms.The reason for this seeming contradictionis that while thereis an abundanceof N in nature,it is almost entirelyin the formof molecularnitrogen,a chemicalformthat is not usableby most organisms.Breakingthe triple bond holding the two N atoms together requires a significant amount of energy-energy that can be mustered only in high-temperatureprocessesor by a small number of specializedN-fixing microbes. We divide the N compounds in natureinto two groups: nonreactiveand reactive.NonreactiveN is N2; reactiveN (Nr) includes all biologically,photochemically,and radiatively activeN compounds in Earth'satmosphereand biosphere. Thus, Nr includes inorganic reduced forms of N (e.g., ammonia [NH3] and ammonium [NH4+]), inorganic oxidizedforms(e.g.,nitrogenoxide[NOx],nitricacid[HNO3], nitrousoxide [N20 ], and nitrate[NO3-]), and organiccompounds (e.g., urea,amines,proteins,and nucleicacids). In the prehumanworld,creationof Nr from N2 occurred primarilythrough two processes,lightning and biological nitrogenfixation(BNF).ReactiveN did not accumulatein environmental reservoirsbecause microbial N fixation and denitrificationprocesseswere approximatelyequal (Ayres et al. 1994). This is no longer the case.ReactiveN is now accumulating in the environmenton all spatialscales-local, regional, andglobal(Gallowayet al. 1995).Duringthe lastfewdecades, productionof Nr by humanshas been greaterthan production from all naturalterrestrialsystems.The globalincrease in Nr productionhas threemain causes:(1) widespreadcultivationof legumes,rice,and other cropsthat promoteconversionof N2 to organicN throughBNF;(2) combustionof fossilfuels,which convertsboth atmosphericN2 and fossilN to reactiveNOx; and (3) the Haber-Boschprocess,which convertsnonreactiveN2 to reactiveNH3to sustainfood production and some industrialactivities. The globalrateof increasein Nr creationby humanswas relativelyslowfrom1860to 1960.Since1960,however,the rate of increasehas acceleratedsharply(figure la). CultivationinducedNr creationincreasedfrom approximately15 teragrams(Tg) N peryearin 1860to approximately33 Tg N per year in 2000. ReactiveN creationthrough fossil fuel combustion increasedfrom less than 1 Tg N per yearin 1860to approximately25 Tg N per yearin 2000.ReactiveN creation from the Haber-Boschprocesswent from 0 before 1910 to andB.JackCosbyareprofessors JamesN. Galloway(e-mail:[email protected]) in the Departmentof EnvironmentalSciencesat the Universityof Virginia, VA22904.JohnD. Aberis a professorat the Institutefor the Charlottesville, StudyofEarth,Oceans,and Space,UniversityofNew Hampshire,Durham, NH 03824. Jan WillemErismanis head of the Departmentof Integrated Assessmentat the EnergyResearchCenterof the Netherlands,PO Box 1, 1755ZGPetten,TheNetherlands.SybilP Seitzingeris a professorat theInstituteof Marineand CoastalSciences,RutgersUniversity,71 DudleyRoad, New Brunswick,NJ 08901.Robert Howarthis a professorin theDepartW. ment of Ecologyand EvolutionaryBiology,CornellUniversity,Ithaca,NY North 14853.EllisB. Cowlingis a professorin theCollegeofNaturalResources, CarolinaState University,1509 VarsityDrive,Raleigh,NC 27606. @2003 AmericanInstituteofBiologicalSciences. April2003 / Vol.53 No. 4 * BioScience 341 Articles 4 200 6 a 4 100 2/ .150 0.. - .. ... 1910 ' 0 1850 1870 - Population 1890 . Haber-Bosch .... ; . 1930 1950 1990 1970 - Fossil fuel C-BNF i 0 2010 Total Nr 40 0.3 b 30 20 0.1 10 0.0 1850 1870 - Population 1890 1910 - Fertilizer 1930 C-BNF 1950 - 1990 1970 Fossilfuel - Total Nr 0 2010 morethan 100TgN peryearin 2000, with about85%used in the production of fertilizers. Thus,between1860 and 2000,the anthropogenicNr creation rate increasedfrom approximately 15 Tg N per yearto approximately165Tg N peryear,with about fivetimesmoreNr comingfromfood production than from energy production (Gallowayet al. 2002). As in the global system, the Nr creationratein the UnitedStateshas increasedoverthelastfewdecades.In 1961, the United States createdNr at a rateof approximately8 Tg N per year(figure1ib).By 1997,the Nr creationratewasapproximately 25 Tg N per year.Companionpapersin this issue (Aberet al. 2003, Driscollet al. 2003, Fenn et al. 2003a, 2003b) discuss Nr biogeochemistry in the United States,with a focus on the Northeastand the West. The remarkablechangesin the N cycle haveresultedin a wide variety of changes,both beneficialanddetrimental,to the healthand welfareof people and ecosystems.A largeportion of the humanpopulationof the worldis sustainedtodaybecauseNr is providedas syntheticfertilizersand cultivation-induced BNF (Smil 2000). But there are also some significant worrisome consequences. First,Nr is widely dispersedby hydrologicand atmospherictransport processes.Second,Nr is accumulating in the environmentbecauseNr creationratesaregreaterthanratesof Figure1. (a) Globalpopulation trendsfrom 1860 to 2000 (billions, left axis) and reactivenitrogen (Nr) creation (teragrams nitrogen[Tg N] per year, right axis). "Haber-Bosch" representsNr creation throughthe Haber-Boschprocess,including ammonia For 1920, 1930, and 1940, we assumedthat global total Nr production productionof for nonfertilizerpurposes. the Haber-Bosch was to through process equivalent global anthropogenicfertilizerproduction (Smil2001). For 1950 onward, data on Nr creation throughthe Haber-Boschprocesswere obtainedfrom USGSMinerals (Kramer1999). "C-BNF" (cultivation-inducedbiologicalnitrogenfixation) representsNr creationfrom cultivation of legumes,rice,and sugarcane. The C-BNFratefor 1900 is estimated to be approximately15 TgN per year (VaclavSmil, Universityof Manitoba, Winnipeg, Canada,personal communication,January2002). The C-BNFratesfor 1860, 1870, 1880, and 1890 were estimatedfrom population, using the 1900 data on population and Nr creation.For 1961-1999, Nr creation rateswere calculatedfrom crop-spe2000) and fixation rates (Smil 1999). Decadal datafrom 1910 to 1950 were interpocific data on harvestedareas (FAOSTAT lated between 1900 and 1961. "Fossilfuel" representsNr createdfrom fossilfuel combustion.The data from 1860 to 1990 are from a compilationfrom ElisabethHolland, basedon Miiler (1992), Keeling(1993), and Holland and Lamarque(1997). Thesedata agree well with those recentlypublishedby van Aardenneand colleagues(2001)for decadal time stepsfrom 1890 to 1990. The data for 1991 to 2000 were estimatedby scaling emissionsof nitrogenoxides to increasesin fossilfuel combustion over the same period. "TotalNr" representsthe sum createdby these threeprocesses.(b) USpopulation trendsfrom 1961 to 1997 (billions,left axis; FAO2000) and Nr creation (Tg N per year, right axis; Howarth et al. 2002a). 342 BioScience * April2003 / Vol.53 No. 4 Articles Nrremovalthroughdenitrification to nonreactive N2.Third, Nrcreationandaccumulation is projected to continueto increasein thefutureashumanpopulations andpercapitaresourceuseincrease. contributes to Fourth,Nraccumulation environmental For manycontemporary problems. example: * Increases in Nrleadto production of tropospheric ozoneandaerosols thatinduceseriousrespiratory andcardiac diseasein humans(Popeet illness,cancer, al.1995,FollettandFollett2001,WolfeandPatz2002). * Forestandgrassland increase andthen productivity decrease wherever Nrdeposition increases atmospheric andcriticalthresholds areexceeded; Nr significantly additions also decrease in probably biodiversity many natural habitats (Aberet al.1995). * Reactive N is responsible withS)foracidifi(together cationandlossof biodiversity in lakesandstreams in et al.1997). manyregionsof theworld(Vitousek * Reactive N is responsible foreutrophication, hypoxia, lossof biodiversity, andhabitatdegradation in coastal Itis nowconsidered thebiggestpollution ecosystems. in coastal waters Howarth et al.2000, problem (e.g., NRC2000,Rabalais 2002). * Reactive N contributes to globalclimatechangeand ozonedepletion, bothof whichhave stratospheric health(e.g.,Cowling impactson humanandecosystem etal. 1998). Thisarticlefocuseson the multiple linkagesamong the ecological and human health effectsof Nr moleculesas they move from (oS . I oneenvironmental systemto another.Thisphenomenonis called the N cascade (Galloway1998), which we define as the sequential transferof Nr through environmental systems and which resultsin environmentalchanges as Nr movesthroughor is temporarilystoredwithineachsystem. We use figure2 as a frame of referencefor two illustrativescenarios of the Nr cascade.In the firstexample,energyproduction by fossilfuelcombustionresultsin the conversionof atmosphericN2 (or fossil Nr) into NOx. In sequence,an atom of N mobilized as NOx in the atmosphere can first increase ozone concentrations,then decreaseatmospheric visibilityand increaseconcentrations of small particles, and finallyincreaseprecipitationacidity. Following deposition to the terrestrialecosystem,the sameN atom can increasesoil acidity (if a base cation is lost from the system),decreasebiodiversity,and eitherincreaseor decreaseecosystemproductivity. If dischargedto the aquaticecosystem,the N atom can increasesurfacewateracidityand lead to coastaleutrophication. If the N atom is convertedto N20 and emitted back into the atmosphere,it can firstincreasegreenhousewarming potentialand then decreasestratosphericozone. The second exampleillustratesa similarcascadeof effects of Nr from food production.In this case,atmosphericN2 is convertedto NH3 in the Haber-Boschprocess.The NH3 is used primarilyto producefertilizer.Abouthalf the Nr fertilizerappliedto globalagroecosystems is incorporated into crops that areharvestedfrom fieldsand used for human food and livestockfeed (Smil 1999,2001). The otherhalf is transferred to the atmosphere as NH3, NO, N20, or N2 or is lost to aquaticecosystems,primarilyas NO3-.Once transferredto these downstreamor downwindsystems,the N atom is part of the cascade.Dependingon its chemicalform,Nr will enter the cascadeat differentplaces.An importantcharacteristic of the cascadeis thatonceit starts,thesourceof theNr (e.g., fossil fuel combustionor fertilizerproduction)becomes irrelevant.Nr speciescan be rapidlyinterconvertedfrom one Nr formto another.Thus,the criticalstep is the formationof Nr in the firstplace. The simplifiedconceptualoverviewin figure2 does not cover some importantaspectsof the N cascade.It ignores the internal cycling of Nr within each component of the tratospheric Effects I I NOF(1 Ene Visibility Effects Ec !" sts nFc rgyA Production I I N2AT N iFood Production NH3 Effects Agroecosystem NH Forests .& . Grassland Animal Crop - Peoplesoi (Food; Fiber) -N--1 G Groundwater Human Activities Effects Efects water Ioe.. TheNitrogen Surfiace ETects Efects Cascade 1 Ecosyste Aquatic s Figure2. Illustrationof the nitrogen (N) cascadeshowing the sequential effectsthat a single atom of N can have in various reservoirsafter it has been convertedfrom a nonreactive to a reactiveform. Abbreviations:GH,greenhouseeffect;NHY ammonia;NO,, nitrate; NO., nitrogenoxide;N20, nitrous oxide;PM, particulate matter. April2003 / Vol.53 No. 4 * BioScience 343 Articles 4 ecosystem;it does not accountfor long-termNr storage;and it does not accountfor all the pathwaysthat Nr followsas it flowsfrom one "box"or effectto another.The followingsections describethese characteristicsin greaterdetail,first for the atmosphere(troposphereand stratosphere),next forterrestrialecosystems(agroecosystems,forests,andgrasslands), and finallyfor aquaticecosystems(groundwater,wetlands, streams,lakes,rivers,and marinecoastalregions).The coverage for these systems is not meant to be exhaustivebut ratherto give an overviewof how each systembehavesand interactswithothersystemswithinthe cascade.Wedo not considerthe N cycleof the open ocean,becauselittle of the Nr createdby human activitieson land reachesthe open ocean (Nixon et al. 1996, Seitzingerand Giblin 1996, Chen and Wang 1999). The cascadeof Nr fromone systemto anotheris enhanced if thereis limitedpotentialforNr accumulationor loss of N2 through denitrificationwithin a given system and thus increasedpotentialfortransferto the next system.Thereis a lag in the cascadeif there is a largepotentialfor accumulation within a system.The cascadedecreasesif thereis a largepotential for denitrificationto N2 within a system (table 1). Thus,at each stageof the N cascade,we evaluatethe potential for * accumulationand cyclingof Nr withinthe system; * loss of Nr throughconversionto N2 by denitrification; * transferof Nr to othersystems;and * effectsof Nr withinthe system. Atmosphere The atmospherereceivesNr mainlyas airemissionsof NOx, NH3,and N20 from aquaticand terrestrialecosystemsand and fromcombustionof biomassor fossilfuels. of NOx NOx NHx (NH3 andNH4+) canaccumulatein the troposphereon a regionalscale.ButbecauseNOxand NHxhavea shortresidencetime in the atmosphereand lackpotentialto form N2 by denitrification,almost all Nr emitted as NOxand NHx is transferredback to Earth'ssurface within hours to days. Thereis some internalcyclingof Nr within the atmosphere. Togetherwith volatileorganicC compounds,increasedconcentrationsof can lead to increasedconcentrationsof NOx ozone and other photochemicaloxidantsin the atmosphere. of the NOxis convertedto HNO3,which is much Ultimately eitherconvertedto an aerosol(e.g.,ammoniumnitrate)or deposited on land,surfacewaters,or othersurfaces.NH3 emitted to the atmosphereis either deposited or transformed into an ammonium aerosol (e.g., ammonium bisulfateor ammoniumsulfate).Beforedeposition,ammoniumaerosols contributeto fine particulatematterand regionalhaze concentrationsin the atmosphere. Sixmajoratmosphericeffectsareassociatedwith increased NOx and NH3 emissions: (1) Fine particulatematter decreasesatmosphericvisibility;(2) elevatedozone concentrationsenhancethe greenhousepotentialof the atmosphere;(3) ozone and fine particulatematterhave serious impacts on human health (Pope et al. 1995);(4) ammoniaplaysan importantrole in the directand indirecteffectsof aerosolson radiativeforcingand thus on globalclimatechange(Seinfeld and Pandis1998,Penneret al. 2001;RussellDickerson,Universityof Maryland,CollegePark,MD 20742,personalcommunication,March2003);(5) ozone depositioncan decrease productivityof crops,forests,andnaturalecosystems;and (6) atmospheric deposition of NH , NO (all oxidized forms of nitrogenother than N20), and organicforms of Nr can contribute to ecosystem acidification, fertilization, and eutrophication. N20 and NOxareproducedduringboth nitrificationand denitrification.Addition of Nr to agroecosystemsleads to increased N20 and NOx emissions. Nitrous oxide has a Table1. Characteristicsof differentsystemsrelevantfor the nitrogen cascade. System Accumulation potential Transfer potential N2production potential Linksto systems down the cascade Atmosphere Low Veryhigh None Allbut groundwater Humanand ecosystem health, climate change Agroecosystems Lowto moderate Veryhigh Lowto moderate All Humanand ecosystem health, climate change Forests High Moderate, high in places Low All Biodiversity,net primary mortality,groundwater productivity, Grasslands High Moderate, high in places Low All Biodiversity,net primary productivity, groundwater Groundwater Moderate Moderate Moderate Surface water, atmosphere Humanand ecosystem health, climate change Wetlands,streams, lakes, rivers Low Veryhigh Moderateto high Atmosphere,marine coastal systems Biodiversity,ecological structure,fish Marinecoastal regions Lowto moderate Moderate High Atmosphere Biodiversity,ecological structure,fish, harmfulalgal blooms 344 BioScience o April2003 / Vol.53 No. 4 Effects potential Articles troposphericresidencetime of approximately100yearsand is increasingin the troposphereat a rate of approximately 0.25%peryear(Pratheret al.2001).Nitrousoxideis a greenhouse gas in the troposphereand, when transferredto the stratosphere,decreasesthe concentrationof stratospheric ozone.As discussedabove,NOxcan contributeto increased troposphericozone and decreasedatmosphericvisibility. In summary,the residencetime for most Nr speciesin the atmosphereis short.Thereis an internalcascadeof effects: NOx increases the potential first for ozone and then for aerosolformation.Exceptfor N20, thereis verylimitedpotential for long-term storage of Nr (and thus limited lag time),buttherearesignificanteffectsfromNr whileit remains in the atmosphere.Thereis no potentialfor denitrification andthereis a largepotential backto N2 withinthetroposphere, for Nr transferto the next receptor-terrestrialand aquatic ecosystems(table 1). Terrestrial ecosystems This section discusses the nitrogen cascade in three types of terrestrial ecosystems-agroecosystems, forests, and grasslands. Agroecosystems.Intensivelymanagedagroecosystemsand, evenmore,concentratedanimalfeedingoperations(CAFOs, alsocalled"factoryfarms")areonlythe latesttechnicalmeans by whichhumanneedsand dietarypreferencesaremajordrivers of changein the N cycle.About 75%of the Nr created to sustainfood globallyby humansis addedto agroecosystems production.About 70%of thatNr is from the Haber-Bosch processandabout30%fromculBNF(figurela). tivation-induced Smil(2001)estimatesthatabout 40%of thepeoplealivetodayowe theirlife to the productionand wideuseof fertilizers produced by theHaber-Bosch process. The agroecosystemsof the world consist of two general types: (1) crop production systems,composedof primary and (2) animalproproducers; ductionsystems,composedof producers. Cropagrosecondary producecerealgrains, ecosystems and commerfruits,vegetables, cial fibersthroughthe uptake of mostlyinorganic Nr,othernutrients,andwater,usingsunlight as the energy source. Crop productionis sustainedby indigenoussourcesof Nr in the soil;by wet and dry deposition of Nr fromthe atmosphere; by biologicalN fixation;by recycling of crop residues,animal manure,and human waste;and by applicationof synthetic Nr fertilizers.The residencetime of Nr within crop agroecosystemscan be yearsto decadesbecauseof the largepool of organicmatterin the soil.Cassmanand colleagues(2002) reportsubstantialindigenoussoil Nr reserves;for example, a typicalirrigatedrice soil in Asiacontainsabout 2800 kilograms (kg) N per hectare(ha) in the upper 20 centimeters (cm). Fertileprairiesoil in the CornBeltof the UnitedStates often containsabout4000 kg N per ha in the upper20 cm of the soil profile.However,usuallyless than 5%of this indigenous supply is availablefor crop uptakein a given season. Thus, the annualcrop yield is determinedprimarilyby the amountof Nr added.Nearlyall of this addedNr is lost from most agroecosystemsover the courseof a year.On a global basis,about 120Tg N fromnew Nr (fertilizerand cultivationinducedBNF)and about50 Tg N frompreviouslycreatedNr (e.g., crop residue,animalmanure,and atmosphericdeposition) is addedannuallyto cropagroecosystems(Smil2001, 2002). Of this amount,about 33 Tg N per yearis consumed by animalsand about 16 Tg N per yearis consumedby humans (figure3). The remainder(about 121Tg N peryear)is lost to the atmosphere(as NOx,NH3, N20, and N2) or to waters (as dissolvedand particulateNr) or is partof the material reintroducedto crop systemsduringthe next cropping cycle.Smil (1999) estimatesthatonly a smallamount (about 4 Tg N per year) of the 170 Tg N per yearintroducedaccumulatesin cropagroecosystems.Generally,the moreNr that is addedto cropagroecosystems,the more is lost throughair and waterpathways.Once lost, the releasedNr can cascade throughnaturalecosystems,whereit alterstheirdynamicsand THE AGROECOSYSTEM Nr inputs 50 " Cropland 49 S'Io121 , AFO 1 Humans 4 Fromfish. Sgrazrs ,8r an w25 Lossesto soil, airandwater Figure3. Majorreactivenitrogen (Nr)flows in cropproductionand animal production componentsof the global agroecosystem.Croplandscreatevegetableprotein through primaryproduction;animal productionutilizes secondaryproductionto createanimal protein. Reactivenitrogeninputs representnew Nr, createdthroughthe Haber-Bosch processand throughcultivation-inducedbiologicalnitrogenfixation, and existingNr that is reintroducedin theform of crop residues,manure,atmosphericdeposition,irrigation water,and seeds.Portionsof the Nr lossesto soil, air, and water are reintroducedinto the croplandcomponentof the agroecosystem(Smil2001, 2002). Numbersrepresent teragramsof nitrogenper year;AFO,animal feeding operations. April2003/ Vol.53 No.4 * BioScience345 Articles in manycasesdecreasestheirabilityto provideecosystemservices (table 1). Animal agroecosystemsproduce dietary protein (milk, eggs, and edible meat) from the consumption of proteins produced by crop agroecosystems.Just as in crop agroecosystems, most of the Nr that enters the animal agroecosystemis lost to the environmentoverthe courseof a year. AnimalsreceiveNr primarilythrough the consumption of aminoacidsin grainsandothervegetation.Theefficiencywith which food animals use N from forage and grain varies greatly.Typicalrates of N-use efficiencyfor production of human-digestibleprotein from feed grains and forageson farmsareabout 50%to 60%for fish, about40%to 50%for poultry,about35%to 40%for dairy,and about 15%to 30% forbeef.On a globalbasis,33 Tg N peryearof grainproduced by the cropagroecosystemis fed to animals(figure3). Of this amount,about 15%is consumedby humans.The remaining Nr is lost as manure and waste (Smil 2001, 2002). The manure can be reused as fertilizer,but the Nr is often lost through air emissions (e.g., NH3, N20, NO, and N2) and throughleachingof NO3-to groundor surfacewaters. Humansconsumeabout25 TgN peryear,mostlyas dietary protein.Of this amount,64%comes from grain,20%from CAFOs,andabout16%fromfishandpasturegrazinganimals (figure3). A portionof the Nr addedto agroecosystemsis denitrified back to nonreactiveN2 in systemsthat have high NO3- or NH4+concentrations,high organicmatter,and low 02 concentrations,such as agriculturalfieldswith high watercontent (e.g., wetlandrice), those that receivelargeamounts of precipitationovera shorttime period(e.g.,springrains),and anaerobicmanurestoragesystems.Unfortunately, the amount in denitrification agroecosystemsis poorly ofN2 lost through known. Oenemaand colleagues(2001) reviewedNOx,N20, andN2 emissionsfrompastures;fromanimalhousingsystems; frommanureslurryin tanks,silos,andlagoons;frommanure heaps;and fromslurryand manureappliedto the soil. They concludedthatour knowledgeof gaseouslossesfrom animal manure is greatestfor NH3 and successivelyless for N20, NO, and N2.Smil (1999) estimatesthat on a globalbasis6% is denitrifiedto N2. to 12%of the Nr addedto agroecosystems In an analysisof agroecosystemsin the GreatPlainsregionof the United States,Del Grosso and colleagues (2001) used modelto showthatN2 fluxesweresmallcomthe DAYCENT pared with NO and N20 fluxes, primarilybecause of the of the region.Theyestimatethatless semiaridcharacteristics than 5%of the appliedNr is lost as N2. Forlargewatersheds of the easternUnited States,van Breemenand colleagues (2002) estimateby differencethat up to 49%of the Nr input to agroecosystemsis denitrified.In the Netherlands,with high levelsof Nr production,30%to 40%of the Nr applied is eitherstoredor denitrified(Erismanet al.2001).Theselarge differencesreflectin partregionalvariabilityin the conditions that promote denitrification (e.g., soil moisture) and the generaluncertaintyof the magnitudeof denitrification. 346 BioScience s April2003 / Vol.53 No. 4 In summary,global crop agroecosystemsreceive about 75% of the Nr createdby human activity (Gallowayand to othersystems Cowling2002).Mostof thisNr is transferred the N a much smaller cascade; along portion globally is denitrifiedto N2 (table1). Improvingthe efficiencyof N use in majorgrainandanimalproductionsystemswillrequirecollaboration among ecologists, agronomists, soil scientists, agriculturaleconomists,and politicians.A greatneed exists for accuratemeasurementsof actual fertilizerN-use efficiency,N losses,andloss pathwaysin majorcropand animal systems.Onlyin thiswaycanwe (a) identifyopportunitiesfor increasedefficiencyof N use through improved crop and soil management;(b) quantify N-loss pathwaysin major food crops,includingCAFOs;and (c) improvehuman understandingof local, regional,and globalN balancesand N losses from majorcrop and animalsystems. Forests. Forestscan be a major reservoirand a short- to long-termsinkwithin the N cascade.TotalNr stocksin soils and biomass can rangeas high as 500 g N per squaremeter (m2). Inputsof Nr in unpollutedregionsare0.1 to 0.2 g N per m2per yearbut can reach5 or,very rarely,10 g N per m2per year(Dise andWright1995).Outputsas dissolvedorganicN in steamwaterare low and relativelyconstant across sites (e.g.,Goodaleet al.2000),whiledissolvedinorganicN (DIN) loss is quite variableand can be as low as zero or equal to deposition,dependingon foresthistoryand condition (Dise et al. 1998,Gundersenet al. 1998,Fennet al. 2003a).In general,this meansthatNr residencetime withina forestis measuredin hundredsto thousandsof years,and thus thereis a substantialopportunityfor lags in the N cascade.However, human activitycan shortenthis residencetime eitherby increasinginputsor by removingNr throughharvests,fires,or conversionto agriculture.Largereservoirsize and relatively low turnoverrates mean that human-induced changes in the Nr statusof forestscan havelong-termeffects,and previouslandusehistorycanplaya largerolein conditioningforest responseto Nr additions(Goodaleet al. 2000,Aberet al. 2003). Our understandingof the responsesof historicallyNrlimited foreststo increasesin Nr depositionhavebeen presentedin summarydiagramsof the processknown as N saturation(figure4;Aberet al. 1998). Fourgeneralstagesalong this continuum have been discussed.In highly Nr-limited (stage0) systems,low Nr availabilityis reflectedin low foliar N contentand low net primaryproductivity.Plantsand root and symbionts compete effectivelyfor mineralized while N20 net nitrificationand NO3-leachingareminimal,NH4, effluxis low to undetectable.In stage 1, increasedNr depositiongraduallyrelievesNr limitationson biologicalfunctions. Eitherthrough plant uptake or through direct incorporation into soil organic matter,added Nr enters into the N cycle of the forest and increasesmineralizationand cycling rates.FoliarNr increases,as does productivity. Therearetwo criticalthresholdsin the N saturationprocess. The firstis induction of net nitrification(stage2). Thus,the Articles presenceof net nitrificationand extractablesoil NO3- arekey indicatorsof ecosystemNr status. Field experimentshave shown that there can be a significant delay between the initiationof Nr additionsandthe inductionof detectablerates of net nitrification(Magillet al. 2000). Both NO3- leaching and N20 effluxesarelow but detectableduringthis period. The second criticalthresholdoccurswhen Nr becomes a nonlimitingelementfor plantgrowth(stage3). At thispoint, biologicalretentionprocessesbecomelesseffective,andNO3leaching losses increasesubstantially.Effluxesof NO and N20 also may increasebecauseof nitrificationor denitrificationprocesses(Davidsonet al. 2000), which may become significantif soilsareimperfectlyor variablydrainedsoils.In stage 3, experimental additions of Nr may decrease tree growth(e.g.,Magillet al. 2000), while decreasesin Nr depositionmayincreasetreegrowth(Beieret al. 1998,Boxmanet al. 1998).ExcessNr may damageforestsby causingnutrient imbalancesandby increasingsensitivityto factorssuchasfrost and attacksby fungi (Erismanand de Vries2000). Thereis substantialfielddocumentationthatN saturation does occur and that it is relatedto increasesin atmospheric Aber inputs. In a companionarticlein thisissueof BioScience, and colleagues(2003) analyzefoliar,soil, and surfacewater chemistryfroma largenumberof sitesto testforpatternsthat coincide with strong Nr deposition gradients across the northeasternUnited States.Surfacewatersyield the clearest patternsbecausethey integrateover largeareas;the surface water data show increasingNr leachingin responseto increasingNr deposition (Aberet al. 2003). These resultsare verysimilarto patternsobservedin Europeanforests.In an analysis of 139 forests,Dise and colleagues (1998) found that inorganicNr leachingincreasedwith Nr depositionas well as with changesin humus compositionand soil acidity (figure5). How quickly Nr status or degree of N saturation can changeis poorly known.VeryNr-poor sites receivingconcentrateddosesof DIN can retainover 100g N perm2 before nitrificationand NO3- leaching are induced (Magill et al. 2000). Richersitesreceivingmore dilutesolutionscan show an increasein NO3- loss immediately(Kahlet al. 1993).Less N appearsto be requiredto move needle-leavedevergreens from stage 1 to stage3 than is requiredfor broad-leaveddeciduousforests(Aberet al. 1995, 1998).Nitrogensaturation may occur quicklynear largepoint sourcesof Nr (Erisman and de Vries2000). It is clearthat most Nr retainedin forestsis held in soils (Nadelhofferet al. 1999, Magill et al. 2000). Recent work suggeststhateitherabioticreactionsbetweenDIN andsoil organicmatteror assimilationand conversionby mycorrhizae maybe important(Aberet al. 1998,Johnsonet al.2000,Dail et al.2001).Understandingthe kineticsand capacityof these or other retentionmechanismsis key to predictingNr retention ratesand the role of forestsin the N cascade. To summarizethe roleof forestsin the N cascade:(a) The residencetimes and lag times of Nr in forestscan be yearsto centuriesdependingon foresthistory,type,andNr inputs;(b) the effects of Nr accumulation in forests are numerous, mostlyrelatingto changesin forestandmicrobialproductivity and function; (c) there is significantpotentialfor Nr to be transferredto the atmosphereas NO andN20, and especially to surfacewatersas NO3-, once Nr additionsor availand (d) relativeto inabilityexceedbioticrequirements; with Nr in areas puts high deposition,thereis a limited to be removed from the cascadeby for Nr potential meansofN2 formation. Succession, N Deposition Harvest,Fire,Agriculture FoliarN 200 N Mineralization NPP) S100 Soil C:NRatio Methane Consumption > Nitrification Leaching, N20 Efflux 0 0 2 1 3 Stage Figure4. Changesin severalecosystemfunctions with increasingNr availability or degreeofN saturation (modifiedfrom Aberet al. 1998).Abbreviations:C, carbon;N, nitrogen;N20, nitrous oxide; NPP, net primaryproductivity. Grasslands.Unmanaged grasslandsreceive most of their Nr from BNF and atmosphericdeposition;the lattersource is much more importantwhere deposition ratesare large.As with forests,temperategrasslands have the potential to be a major storagereservoir and a short- to long-term sink within the N cascade.At the Konza Prairie(tallgrass)site, for example,Nr stocksin soils and in biomassareon the order of 625 g per m2 and 6 to 25 g per m2, respectively (Blairet al. 1998). Inputsof Nr in unpollutedregions are less than 1 g N per m2 per yearbut can reach 5 g N per m2 per year (Gallowayand Cowling2002). The residencetimes of Nr in grasslandscan be decadesto centuries, because most biomass is subsurfaceand decompositionratesareslow (Blairet al. 1998,Epstein et al. 2001). Giventhe N-limitednatureof most grasslandsand the long residencetimes,thereis a largepotential for significantinternal cycling (including redistribution by grazing animals) and for Nr accumulation. o April2003 / Vol.53 No. 4 BioScience 347 Articles 4 Aquaticecosystems 60 a m 50 - Nin< 65% NH4+-N Nin> 65% NH4 -N This section discussesthe N cascadein threetypes of aquatic ecosystems:(1) groundwater;(2) wetlands, streams,lakes,and rivers;and (3) coastalsystems. TheprimaryanthropogenicNr sourcefor Groundwater. groundwateron a globalbasis is leachingfrom agroS * ecosystems,althoughin some regionshumanwastedisposal can also be important(Puckettet al. 1999,Refsgaardet al. 1999,Hudak2000,Nolan and Stoner2000, Nolan2001,vanEgmondet al.2002).Nitrateis themost e 20 common Nr species(Burkartand Stoner2001). As in othersystemsconsideredin this article,therearethree AAOA fatesof Nr in groundwater:accumulation,conversion 10 20 30 40 50 60 70 80 to N2, and distributionto other systemsthroughhyS or atmosphericpathdrologicpathways(e.g.,as NO3-) as or ways (e.g., N20 NO). Althoughpollutioncan reDINflux(kg N perha peryear) sult in high levels of NO3- in some groundwater Throughfall aquifers,the accumulationof N in groundwateris not a majorsink for the N mobilizedby human activityat Figure5. Dissolved inorganicnitrogen(DIN, in kilogramsper hectare either globalor regionalscales.In regionsof intenseagriper year) lost in runoffand seepage.Siteswith throughfallDIN domicultural activityin Europeandthe UnitedStates,the avnated by nitrogenfrom ammonium (NH4+-N)are shown in open trirate of accumulationof NO3- in groundwater erage angles (Dise et al. 1998). Ni, representsN throughfallinputs. amountsto at most a few percentof the N inputsfrom fertilizerand other sources(Howarthet al. 1996). Nr losses from grasslandscan occur through hydrologic Nitrogen is lost from groundwater both through denitrification to N2 and through losses of Nr to surface losses and atmospheric emission. Because precipitation watersandthe atmosphere,but the relativemix of thesefates ratesandthus runoffratesin grasslandsarelow,atmospheric is site dependent,as is the residencetime of Nr in groundemissions are important, especiallyif fire is frequent;firearea waterreservoirs.Forexample,in an intensiveagricultural induced Nr losses can be approximatelyequal to input by in west-centralMinnesota,about40%of the Nr thatentered atmosphericdeposition.Forexample,in ungrazedregionsof the groundwaterwas denitrified,and the rest accumulated the KonzaPrairie,Nr lossesfromfirecan approximateor exin the aquiferwith little dischargeto surfacewaters(Puckceed atmosphericNr deposition inputs (Blairet al. 1998). ett et al. 1999, Puckett and Cowdery2002). In the southHowever,some of theselossesarein the form of N2 (Crutzen eastern United States, some systems have high losses of and Andreae 1990). For biomass burning in general, Kuhlbuschand colleagues (1991) estimate that N2 constiNO3- and concomitant low NO3- concentrations, while other systemshavelow losses of NO3- and high NO3- contutes the largest flux of N gases emitted. Thus, if fire is centrations (Nolan 1999). A review of several studies of Nr from some atmospheric deposition resulting frequent, denitrificationin groundwaterby Groffmanand colleagues human action maybe returnedto the atmosphere,partlyas Nr and partlyas N2. (1998) found similardegreesof variability. In additionto denitrificationthroughfire,Nr can alsobe Althoughthe accumulationof Nr in groundwateris not a convertedto N2 and lost from the grasslandthroughmicroregionallyor globallyimportantsink relativeto the amount do pose of Nr created,theeffectsof elevatedNr in groundwater soils are because most bialdenitrification. However, grassland a significanthuman healthrisk,becausedrinkingwatercan well aerated, conversion of Nr to N2 through microbial become contaminated.In the human body, NO3- is condenitrificationis probablynot an important process (Del to nitrite,whichcan causemethemoglobinemiaby inverted in the accumulated If et retained al. Grosso 2001). grassland, of loss in and Nr canleadto increases productivity terferingwith the abilityof hemoglobinto takeup 02. Most biodiversity cases of methemoglobinemiaoccur afterconsumingwater and Howarth 2000). (Tartowski suswith are animal for Grasslandsmanaged high concentrationsof NO3; infantsareparticularly production(e.g.,cattle) treatment who receive are as Nr. The adloss of added with to much more"leaky" kidneydialysis people ceptible, respect dition of fertilizeror grazinganimalsincreasesthe amount (Follettand Follett2001). Forthis reason,the WorldHealth of Nr that is availablefor loss, especiallythroughthe atmosOrganization recommends that NO3- concentrations in and and drinkingwatershouldbe lessthan 10 milligrams(mg) N per Hutchings 1997] phere (e.g., NH3 [Sommer N20 exceedthislevel liter.In the UnitedStates,NO3-concentrations the effective in [Fowleret al. 1997]).Thus, managedecosystems 4 of the 33 from than of in more 15% unless than for to be the have residencetimes groundwatersamples potential as sourcesof used most commonly majorregionalaquifers managedsystems. 40- 40 348 BioScience * April2003 / Vol.53 No. 4 Articles drinkingwater(Nolan and Stoner2000). Othereffectsassociated with elevated concentrations of NO3- in drinking waterincluderespiratoryinfection,alterationof thyroidmetabolism,and cancersinducedby conversionof NO3-to Nnitroso compoundsin the body (Follettand Follett2001). In summary,groundwatersystemsareaccumulatingNr at low ratescomparedwith ratesof human mobilizationof Nr in the landscape,but the effects of contaminatedgroundwatercan nonethelessbe significant.Nr is lost from groundwaterthroughdenitrification,throughadvectionof NO3- to surfacewaters,and throughconversionto gaseousNr forms that diffuseor flow to the atmosphere.Justas thereis a lag in Nr releasein forests because of N saturationphenomena (figure4), there is also a lag for Nr releasefrom groundwaterto surfacewater,althoughthis is highlyvariableamong sites (Groffmanet al. 1998,Puckettand Cowdery2002). At sometime,the Nr introducedinto groundwaterwill be transferredto surfacewaterif the Nr is not permanentlystoredin groundwateror denitrified to N2. If the inputs of Nr to groundwater systems decrease, then the decrease of the groundwaterNO3- burden can be delayedfor a significant amount of time becauseof the significantresidencetime of Nr in groundwaterreservoirs. Wetlands,streams,lakes, and rivers.Surfacefreshwater ecosystems consist of wetlands (e.g., bogs, fens, marshes, swamps,and prairiepotholes),streams,lakes(includingartificialreservoirs),and rivers.Surfacefreshwaterecosystems receivemostof theirNr fromtheirassociatedwatersheds, from BNF within the and from atmosphericdeposition, system.Niin fixation is more important eutrophic generally trogen lakes and in some wetlands, where the process has been found to contributebetween5%and 80%of totalNr inputs (Howarthet al. 1988). Thereis limited potentialfor Nr accumulationwithin surfacewaterecosystemsbecausethe residencetime of Nr within surfacewaters,likethe wateritself, is verybrief.Residencetimes may be relativelylongerin the sedimentsassociatedwithwetlandsand somelargerlakes,but they arestill shortwhen comparedwith the residencetimes in terrestrialecosystems. Despitea shortresidencetime, Nr cyclingin surfacefresh watercan be very complex.In additionto dissolvedN2, inorganicoxidizedand reducedforms of Nr can occur,as can organicformsof Nr.In mostheadwatersystems,the inorganic formsof Nr (NO3-and NH4+)arepresentin low concentrationsunlessthe watersareassociatedwithN-saturatedforests, grasslands,agroecosystems,or suburbanlandscapes.This is primarilybecausethe intensecyclingof Nr in the terrestrial portions of the watershedresults in little runoff of Nr to streams.The little inorganicNr that does reachthe streams in pristineregionsis usuallydenitrifiedor quicklyincorporatedintobiomassby aquaticplantsandthenrecycledthrough the hierarchyof consumersand decomposersunlessNr concentrationsarehigh (Petersonet al. 2001). However,many headwaterstreamsand lakes are now in highly disturbed landscapesand thus havehigh NO3- concentrations,which can leadto eutrophicationproblemslocallyor fartherdownstream.In addition,forheadwaterstreamsandlakesdraining poorlybufferedsoils,increasedNO3- concentrationscan result in streamacidification,with resultantimpactson biota. In undisturbedareasin the temperatezone, the major terrestrialflux of Nr into surfacewatersmay be organicNr in theformof detritusor dissolvedorganicmatterwashedinto lakes and streams (Lewis 2002, Perakisand Hedin 2002). Undisturbedtropicalregions can have significantlosses of NO3- (Lewis et al. 1999). In the presenceof increasedatmosphericdepositionof inorganicNr or significantland-use perturbations(e.g., urbanizationor agriculture)in the terrestrialwatershed,the amountof Nr deliveredto streamsand lakesmay easilyexceedthe retentioncapacityof the aquatic system.Therewill then be little delayin the transportof N down the cascadeunlessthe Nr is denitrifiedbackto N2.As discussedbelow,the potentialfor denitrificationin wetlands, streams,lakes,and riversis large. Wetlandscangenerallybe consideredaggradingecosystems wherethe additionalNr can come from adjacentwatersand, in some cases,from BNF and atmosphericdeposition.Wetlandsareso efficientat removingNr throughdenitrification (seebelow)thattheyarefrequentlyconstructedto removeNr from effluent waters from a number of human activities. However,humans arealso acceleratingwetlandremoval.In many areasof the world,wetlandsarebeing drained;nearly half the wetlandsin the United Stateshave been destroyed since 1780 (Gleick1993).Whilewetlandscan be efficientNr sinks,their arealextentis limited (and becomingmore limited) comparedwith that of terrestrialand oceanic sinks. In summary,the potentialforNr accumulationin streams, lakes,rivers,and associatedwetlandsis small (table1).While changesin Nr inputs to surfacewatersmay significantlyalter the internalcycles in these systems,from a global perspectivesuchchangesaremerelyextremelyefficientavenues for propagatingthe effectsof the N cascadefrom higherto lower components. Even though wetlands may delay or preventthis transferlocally,and denitrificationmay shortcircuitsome of the Nr transportalongthe way,surfacefreshwaterecosystemsessentiallymoveNr fromthe mountainsto the sea, ensuring that perturbations at one point in the cascadequicklylead to changeselsewhere. Coastalsystems. Coastalecosystems(e.g., estuaries)receive most of their Nr from riverine and groundwaterinputs; direct atmospheric deposition is an important source in some systems,and inputs from the ocean are importantin others.These inputs have increasedseveral-foldas a consequence of human activities(Howarthet al. 1996,Seitzinger and Kroeze1998,NRC2000,Howarthet al.2002b).Because of the dynamicnatureof coastalecosystems,thereis limited potentialfor Nr accumulation.In addition,althoughthe potentialforNr transferto continentalshelfregionsis large,there is limited transportto the shelf becauseof the high ratesof * April2003 / Vol.53 No. 4 BioScience 349 Articles( denitrification(mostlyas N2), and the Nr that is transferred is mostly convertedto N2 before its transportto the open ocean.With coastalsystemsactingas Nr sinks,atmospheric depositionbecomesa potentiallyimportantsourceof Nr for the open ocean, especiallyin oligotrophicmidocean gyres (figure2). AlthoughNr has a short residencetime in coastalecosystems comparedwith terrestrialecosystems,the time it does spendtherecanhavea profoundimpacton the coastalecosystem. Primaryproductionin most coastalrivers,bays,andseas of the temperatezone is limitedby Nr supplies(Vitousekand Howarth 1991, Nixon et al. 1996, NRC 2000). As a consequence,greaterNr inputslead to increasedgrowthof algae. In moderation,this canbe viewedas beneficial,as it canlead to increasedproductionof harvestablefish (Nixon 1988,Jorgensenand Richardson1996).However,high Nr inputs can also lead to excessivealgalgrowth,or eutrophication.In the United States,the increasedNr flux is now viewed as the most seriouspollution problemin coastalwaters(Howarth et al. 2000, Rabalaiset al. 2002). One-thirdof the nation's coastalriversandbaysareseverelydegraded,andanotherthird havebeenmoderatelydegradedfromnutrientoverenrichment (Brickeret al. 1999).The situationis probablyequallysevere in otherregionsof the globewherehumanactivityis leading to highNr inputsto the coast(e.g.,theNorthSeaandtheBaltic Sea). In the tropics,P ratherthan N often limits relativelypristine coastalecosystems.However,increasednutrientloading can shiftthesesystemstowardNr limitation(McGlatheryet al. 1994,Howarthet al. 1995) and,as in temperate-zonesystems, Nr is a majorcontributorto coastaleutrophicationin tropical coastal systems (Corredoret al. 1999, NRC 2000, Rabalais2002). One of the most obviousconsequencesof increasingNr inputs to coastalwatersoverthe past few decadeshas been an increasein the sizeof watermassesthatareanoxic(completely devoid of 02) or hypoxic (with concentrationsof 02 less than 2 to 3 mg per liter).Theseso-calleddeadzones now occur in the Gulf of Mexico,the ChesapeakeBay,LongIsland Sound,FloridaBay,the BalticSea,the AdriaticSea,and many othercoastalareas(DiazandRosenberg1995,NRC2000).The hypoxic water mass in the northern Gulf of Mexico has grownin sizein recentsummersto about20,000squarekiloattributableto Nr pollution running meters and is dclearly down the MississippiRiver(Goolsbyet al. 1999,Rabalaiset al. 1999,2002, NRC 2000). Othermajoreffectsof increasingNr in coastalregionsinclude loss of seagrassbeds, macroalgalbeds,and changesin coralreefs(Lapointeand O'Connell1989,Valielaet al. 1997, Howarthet al.2000,NRC2000).ReactiveN additionscan increasethe incidenceand durationof harmfulalgalblooms (NRC 2000). Anoxic or hypoxic events and harmful algal bloomscanleadto fishkills.ReactiveN pollutioncanalsolead to more subtleeffects,with alterationsof marinefood webs that leadto decreasedfish production(NRC2000). Reactive 350 BioScience * April2003 / Vol.53 No. 4 N pollution is a leading cause of loss of biotic diversityin marineecosystems(NRC 1996,2000). In summary,Nr inputs to coastal ecosystems have increasedsignificantlyoverthe lastfew decades.Althoughmost Nr is eventuallydenitrifiedto N2 within the coastalecosystems and associatedshelf,Nr pollution has significantand widespreadimpactson variousecosystemcomponentsand on human health (table 1). Denitrificationpotentialin the wetland-stream-riverestuary-shelfcontinuum.Alongthe entireaquaticcontinuum, fromwetlandsto headwaterstreamsto the continentalshelf and eventuallyto the open ocean, not only is Nr rapidly cycledamong the variousforms (e.g., NH4, NO3-, and particulateand dissolvedorganicN), thereis greatpotentialfor loss of Nr from the biosphere through the conversion of NO3-to N2 (denitrification).In this sectionwe aim to identifyplacesalongthe N cascadewhereNr canbe convertedback into N2 or N20. Conditionsrequiredforthisconversioninclude(a) thepresence of NO3-or nitrite(referredto collectivelyin this section as nitrate),(b) the presenceof labileorganicmatter,and (c) the absenceor low concentrationof dissolved02. The most activesites for denitrificationin aquaticsystemsarebenthic sediments,which are often anoxic below the first few millimeters,eventhoughthe overlyingwateris well oxygenated. Evenwhen nitrateconcentrationsarelow in sediments,denitrificationrates could be high if other conditions are favorable,becauseof the close spatialand temporalcoupling alsocanocof nitrificationanddenitrification. Denitrification curin anoxiczoneswithinthewatercolumn,withthe primary sourceof nitratecoming from outsidethe zone of anoxia. Increasedinputsof Nr fromanthropogenicsourcescanincreaseratesof denitrificationalong the entire aquaticcontinuum.As Nr inputs increase,they can increasethe nitrate concentrationin the watercolumn and thus increasethe diffusivesupplyof nitrateto the sediments.Increasedinputsof Nr can also enhanceprimaryproduction,particularlyin Nrlimited estuarineand continentalshelfwaters,thus increasing organicmatterdepositionto the sedimentsandsubsequent sediment nitrificationand denitrification.However,if Nr inputs result in the water column becoming anoxic, sediment nitrification and consequently denitrification can markedlydecrease.Forexample,in the ChesapeakeBay,denitrificationin sedimentsunderlyinganoxicwaterswas low comparedwith periods of well-oxygenatedbottom waters (Kemp et al. 1990). However,the net effect of persistent anoxicwaterin estuarineandcontinentalshelfsystems,where nitrateconcentrationsand ratesof wateradvectionarelow, on total denitrificationrates (in sedimentplus water)is not well documented.On the other hand, in riverssuch as the Seineand Scheldt,with high inputsof nitratefromupstream agriculturalsources,moreNr was removedin the riverby denitrificationwhen the water was severelydepleted in 02, comparedwithtimesafterimprovementsin sewagetreatment Articles whenthe 02 concentrations in the riverincreased(Billen estuariesor discharged bylargeriversdirectlyontothecontinentalshelf.Inestuaries, waterresidence timeagainisanim1990). Theproportionof Nr inputsremovedthroughdenitrififactor the portant controlling proportionof Nr inputsthat cationhas a similarrangefor lakes,rivers,and estuaries areremovedby denitrification. A similarfunctionalrelafromlessthan10%to morethan80%).Istherean the to ones demonstrated forlakesandriversapplies (generally tionship overallecosystempropertythatcan accountfor the wide to estuaries. Inestuaries withawaterresidence timeranging intheproportion oftheNrinputsremoved from 0.1 month to over a the total Nr variability through year, inputsremoved denitrification? Theresidence timeof wateris animportant denitrification from less than to approxi10% by ranged factorcontrollingremovalof Nr throughdenitrification in 75% et al. (Nixon 1996). mately Theeffectof waterresiNr thathasnotbeenremovedbydenitrification in rivers streams,lakes,rivers,andestuaries. dencetimeis probablyrelatedto thetimethatNr hasto reorestuaries issubjectto removal onthecontinental shelf(e.g., actin theecosystembeforeit is transported to the adjacent DevolandChristensen 1993,LaursenandSeitzinger 2001). downstream betweentheproportion In fact,Nr removedby denitrification in shelfsediments system.Therelationship of Nrinputsthataredenitrified andthewaterresidence time exceedsNr exportedto coastalareasbyrivers.Deprobably in anaquaticenvironment wasoriginally shownforwell-oxytailedNrbudgetsforcontinental shelfregionsthroughout the North Atlanticbasin suggest that denitrificationin contigenatedandwell-mixedshallowlakes(Kellyet al. 1987). Thisrelationship wasextended to riverreachesandreservoirs nentalshelfsedimentsis equalto or greaterthaninputsof Nr and indicatedthata smallerproportionof the Nr inputs fromland-basedsources(SeitzingerandGiblin1996).Theadwereretainedin a riverreach(often5%to 20%)compared ditionalNr requiredto supportthe estimateddenitrification withmanylakesandreservoirs ratesin shelf sedimentsprobablycomes from the transport (fromabout10%to morethan this relationship for of Nr fromoceanicregionsacrossthe slope-shelfboundary. 90%;Howarthet al. 1996).Recently, streamandriverreacheswasincorporated intoa streamand Confirmingthis observationin another region, Chen and rivernetworkwatershed modelthatwasappliedto 15river ratein the East Wang(1999)foundthatthe net denitrification networksin thenortheastern UnitedStates(Seitzinger et al. ChinaSeashelfis greaterthanthe totalriverinesupplyto the 2002).ThemodelresultsindicatethattheNr removaleffiregion.Additionalmeasurementsof denitrificationand Nr ciencyvariesthroughouta networkof riversandstreams. inputs for other continental shelf regions throughout the insmallerstreams Reaches removea greater of the world'soceans are needed to betterunderstandthe Nr balproportion Nrinputsto thosereaches thanreachesin largerriversperse. anceandfinalfateof land-based Nr inputs. thetotalamountof Nrremovedbydenitrification In summary, However, nearlyallof theNr thatis injectedintosurin downstream isgreater becausethetotalNrinputs, facewatersis denitrified reaches, alongthestream-river-estuary-shelf fromtheupstream arelarger. Atthescale continuum(figure6).Whilemostof thisNr is converted to watershed, primarily of awholewetland-stream-river thecumulative efa is fraction converted to and NO. Model estimates network, N2, N20 fectof continuedNrremovalalongtheentire flowpathfromwetlandsto smallstreamsto Denitrificationreduces the downstream transport of N riversdownstream canresultindenitrifilarger cationof asmuchas30%to 70%of thetotal externalNr inputsto the river,althoughthe of Nrinputsremovedbydenitrifiproportion cationin a particular reachis generallyquite 1 small(often1%to 20%). alterations to rivers,suchaschanPhysical candecreasewaterresidencetime nelization, 30-70%entering andthusdecrease denitrification withina river. G-N. Channelization of riversalsodestroys riparian whichhavea considerable wetlands, capacity for Nr removal(Billenand Garnier2000). Wetlandrestoration andthe construction of estuaries denitrified newwetlands havebeenconsidered asoptions fordecreasing Nrloadingatlocalandregional > 80%entering scales.Whileindividualwetlandsareclearly shelves denitrified activesitesfordenitrification, theoverallcontributionof wetlandsto Nr removalat the whole-watershed scalegenerallyis not well ] Ocean documented andwarrants furtheranalysis. Reactive N not removed within the Figure6. Thetransportof reactivenitrogenfromterrestrialto oceanicsystems at eachstepalongtheriver-estuarine-continental wetland-rivernetworkis transportedto decreases shelfsystem. - 351 April2003/ Vol.53No.4 o BioScience Articles i of globalN20 emissionssuggestthatrivers,estuaries,andcontinentalshelvesaccountfor approximately30%of the total globalanthropogenicN20 emissions(Seitzingeret al.2000). As injections of Nr to rivers increase,so will the rates of creationof NO and N20 (Seitzingerand Kroeze1998). Thenitrogencascade:Linkageswithotherelements ElevatedatmosphericNr can increaseozone, increaseatmospheric fine particulateloadings, and enhance the impacts of aerosol sulfate on the atmosphericradiationbalance. These are not the only important linkages.Because many naturalecosystemsareN limited (e.g., temperateand borealforests,grasslands,and temperatecoastalwaters),the increasedabundanceof Nr leadsto increasesin productivity, whichin turnleadto increasesin uptakeof allotherelements tied to productivity(e.g., calcium [Ca],C, P,potassium,and magnesium).Eventually,anotherelement may replaceN as the limitingelement.Forexample,when temperateforested ecosystemsreachstage3 of N saturation,N becomesa nonlimiting element for plant growth (figure4). At that point, something else becomes limiting, and the basic biogeochemicalnatureof the forestchanges.Otherpossiblelimiting materialsarebasecations,P,light,andwater(e.g.,Schulze 1989,DeHayeset al. 1999). Matsonand colleagues(2002) discusshow the globalization of Nr deposition raises questions concerning consequencesof anthropogenicNr for ecosystemsin tropicalregions,whereP or Camost likelylimitsproductionin humid tropicalforestandsavannaecosystemswith highlyweathered soils.In manyof these systems,Nr mayalreadybe a nutrient in excess,so manytropicalhumid forestsmaybe naturallyN saturated(Hall and Matson 1999). Thenitrogencascade:Possibilitiesfor intervention Nr accumulationin environmentalreservoirsenhancesthe N cascadeand its consequenceson people and ecosystems. However,just as anthropogenicactivitieshave substantially increasedthe rateof Nr formation,it is possibleto intervene at criticalpoints alongthe N cascadeand makeNr less abundant.Therearetwo waysto decreasetotalNr:(1) decreasethe rateof Nr creationduringenergyandfood productionor (2) convertNr backto N2 followingNr creationand use. Decreasingthe rateof reactivenitrogencreation.Thereis no benefit to the Nr createdduringfossil fuel combustion.Nitrogen oxides are formed inadvertentlyduringcombustion eitherthrough oxidation of fossil-organicNr in the fuel or throughoxidationof atmosphericN2. In both cases,the options for significantlydecreasingNO emissionsarenumerous, eitherby using an alternativemethod to provideenergy fuelcells)or by eliminatingNOxand (e.g.,hydrocarbon-based other Nr species from the combustion products (Bradley and Jones2002, Moomaw 2002). It is now technicallyfeasible to decreaseNr creationfrom fossil fuel combustionto a pointwhereit becomesjusta minordisturbanceto the global 352 BioScience * April2003 / Vol.53 No. 4 cycle(Cowlinget al. 2002). If that occurs,NO depositionto global ecosystemswill decreaseby about half, and the remainingmajorNOxsourceswill be emissionsfrombiomass burning and agriculturalsoils. Nr deposition to regional systemswill also decrease.In the northeasternUnitedStates, if fossilfuel combustionwereno longera sourceof NOx,Nr deposition to the researchsites in the study by Aber and colleagues(2003)woulddecreaseby morethan50%(Ollinger et al. 1993). Kroezeand colleagues(2001) examinedthe effectof institutingthe maximumtechnicalpotentialto decrease emissionsin Europeon DIN exportby Europeanrivers NOx in the year2050.This scenarioassumedthat all countriesapply maximum emission control in electricity generation, transport,and industry.This resultedin a decreaseof approximately80%in NOydepositionto Europeanwatersheds relativeto a business-as-usualscenario. In food production,thereis a benefitto Nr creationthrough the Haber-Boschprocessand through cultivation-induced BNE However,thereis also inefficiency.Of about 170 Tg N of Nr added to global crop agroecosystemsin 1995, only about 12% entered human mouths (figure 3; Smil 1999, 2002). Most of the restwas distributedto the environment without servingthe purposefor which it was created. Thereis ampleopportunityto increasethe efficiencyof Nr use in food productionandthusdecreasethe Nr creationrate. Therearemanywaysof achievingthis goal:(a) increasethe efficiencyof N use in cropand animalagriculture(Cassman et al. 2002), (b) increaseNr recyclingwithin agroecosystems (i.e.,if Nr is not incorporatedinto food the firsttime around, send it aroundagain;Smil 2002), (c) increaseuse of cultivation-inducedBNF (Royet al.2002), (d) provideincentivesto reduceoverfertilization(Howarthet al. 2002b), and (e) redistributeNr from areaswith high Nr productionto areas wherethereis a need for Nr for food production(Erismanet al. 2001). CAFOsofferlargeopportunitiesforimprovementof N-use efficiency.In recentdecades,livestockand meat processing industrieshavebeentransformedin manycountries.Bothverticaland horizontalintegrationhavegreatpotentialto maximizeN-use efficiencywithintegratedeconomicandadvisoryservicelinkagesamongfarmers,feedsuppliers,animal-rearing advisers,and food-processing companies. These changes often lead to largelyunforeseenNr-inducedenvironmental problems,mainly on local and regional scales.Thus, emphasison economicefficiencywithoutattentionto Nr-induced rather healthand environmentalrisksleadsto externalization than internalizationof these realcosts. If societies decreasedtheir consumption of meat, there wouldbe lessdemandforNr createdto producefood (Bleken 1997,Kroezeet al.2001,Smil2002).Howarthand colleagues (2002a)projectthatif the US populationadopteda Mediterraneandiet of approximately6.3 kg meatper capitaperyear (about one-seventhof the supplyin the UnitedStates),total inorganicN fertilizerconsumptionwould decreaseto about 6.9 Tg N per year by 2030-a 65% decrease.Kroezeand Articles colleagues(2001)reportsimilarfindingsforthe UnitedStates and Europe. Theprojectionsfor futureNr creationthroughthe HaberBosch processdepend to a largeextenton the degreeof intervention and on assumptionsand preferencesregarding human diets.A studyof futurefertilizerrequirementsup to the year2030 projectsthat annualincreasesin fertilizeruse will range from 0.7% to 1.3%,dependingon assumptions aboutN-use efficiency(FAO2000, Fixenand West2002). In absoluteterms,this means that N fertilizeruse in 2030 will rangefromabout96 Tg N peryearto 118Tg N peryear,compared with about 78 Tg N per year for the base period 1995-1997.Tilmanand colleagues(2001)projectthatN fertilizeruse will increasefrom 87 Tg N peryearin 2000 to 135 Tg N per yearin 2020.Whilethe actualproductionof Nr in the futurefor fertilizeruse is uncertain,it is clearthat N demands for food productionwill increase. theconversion ofreactive to N2.Inmost nitrogen Increasing environmentalreservoirs,N2 formationis verylimited (e.g., atmosphere,grasslands,andforestedecosystems)or counterproductiveto the purposeof the system(e.g.,agroecosystems). The wetland-stream-river-estuary-shelfcontinuum is the only systemthat providesthis servicenaturally,at ratesthat arelargerelativeto Nr inputs.However,beforeNr is converted to N2 in the continuum,a numberof detrimentaleffectsoccur in forests,grasslands,and surfacewaterecosystems.For this reason,the next opportunityfor interventionis immediatelyafterNr is used as a resourcebut beforeit is distributed to the environment.While there are probablyother suitablepoints for intervention,we choosetwo: (1) Nr losses from animal and human waste and (2) Nr transferfrom agroecosystemsto surfacewaters. Globally,animalsand humans excreteabout 75 Tg N per year and about 23 Tg N per year,respectively(Smil 1999). While this is about 80%of the Nr createdfor food production, some of it is reused,and not all of the remainderis easily collectible.Foranimals,of about 75 Tg N per year,about 28 Tg N per yearareproducedin CAFOs,and of that about 18 Tg N per year are recycledto agroecosystems,leaving about 10 Tg N per yearas easilycollectiblewasteNr.Forhumans,in 2000,47%of the world'spopulationlivedin an urban environment;of this 47%,84%had theirwastecollected by a sewagesystem(6 March2003;www.worldbank.org/data/ So, of about 23 Tg N per yearcontainedin dataquery.html). humanwastethatis generatedannually,municipalsewageoperationscollectabout9 Tg N peryear.Thus,of about 100Tg N per yearproducedas Nr-containingwasteby animalsand humans,about21 TgN peryearis not reusedandis easilycolinlectible,whichincreasesthe potentialfor a "denitrification tervention"unlessothervalue-addedproductscan be made using this N-rich resource (Cassmanet al. 2002, Oenema and Pietrzak2002). The secondinterventionpoint is betweenagroecosystems and streamsor rivers.Acceptingthat Nr leakageswill occur from most agroecosystemsand some forests, there is an opportunityto takeadvantageof the factthat denitrification does occur in wetlandand riparianareas(Hill 1996,Hill et al.2000,Steinhartet al.2000,Mitschet al.2001), and specific managementactivitiescan be used to createan environment that will increase this rate (e.g., Schipper and VojvodicVukovic 2001). However,it is not enough to increasedetheproductionof N2 mustalsobe optimizedrelnitrification; ativeto N20 formation. Summary The atmospheredirectlyreceivesabout15%of the Nr created by human activitiesas a consequenceof energyproduction; all of it is deposited as Nr. Agroecosystemsreceive 75%, most of which is eithertransferreddirectlyinto the atmosphereor hydrosphereor lost to the environmentduringthe processof food productionand consumption.The remaining 10%of the Nr is usedin industrialprocesses.Theprimary beneficial effect of Nr introduced into agroecosystemsis human nutrition;other effectsoccur as the Nr from agroecosystemscascadesthroughother environmentalsystems. In the atmosphere,increasedNr concentrationshave directandindirectimpactson humanand ecosystemhealthon a regionaland globalbasis.Most of the Nr emittedto the atmosphereis depositedbackto Earth'ssurfaceafterhours to days.IncreasedNr depositionto grasslandsand forestshas a residencetime of yearsto centuries,introducingthe potentialfor a largelagtime in the cascade.Becausethe production of N2 is smallrelativeto Nr inputs,N accumulatesin a reactive form, resultingin initially increasedproductivityand some loss of biodiversity,and ultimatelysome loss of productivity.Severalregionsof the world havealreadyreached the point whereaccumulationhas slowed and loss of Nr to the atmosphereand hydrospherehas increased. Transferof Nr from the atmosphere, agroecosystems, forests, and grasslands into the wetland-stream-riverestuary hydrospherecontinuum is increasingand has resultedin numerouseffects,includingacidification,eutrophication,and human healthproblems.However,throughout the continuumthereis a largepotentialfor conversionof Nr to N2, especiallyin wetlands,largerivers,estuaries,andthe continentalshelf.Thus,while the N cascadebegins at the point of Nr creation,and while Nr will accumulatein and cycle among,most terrestrialsystems,the cascadereachesan end at the continentalmargins,where its primarycontinuation is N20 productionduringnitrification. Acknowledgments A topic of this scalerequiresexpertisefrom a diversecollection of disciplines.In that regard,we aregratefulfor the advice andcounselof BethBoyer,EricDavidson,LindaDeegan, SteveDel Grosso,RussellDickerson,NancyDise,HowardEpstein, Christy Goodale, Peter Groffman, Bill Keene, Tom Nolan, and JeffStoner.We also appreciatethe constructive commentsof TonBresser,PaulFixen,RickHaeuber,Kathleen Lohse, JerryMelillo, Bill Moomaw, Oene Oenema, Nancy Rabalais,RabindraRoy,VaclavSmil,Rob Socolow,Klaasvan * April2003 / Vol.53 No. 4 BioScience 353 Articles Egmond,Dan Walters,FordWest,and XunhuaZheng,who providedcommentsaboutthe N cascadeduringthe Plenary SpeakerWorkshopheld in April2001 beforethe SecondInternationalNitrogenConferenceheld the followingOctober. We thank MaryAnn Seifertfor putting the paper into the correct format and Sue Donovan for assistancein figure preparation.JamesN. Gallowayis gratefulto the Ecosystems Centerof the MarineBiologicalLaboratoryand the Woods Hole OceanographicInstitutionfor providinga sabbatical home to writethis paperand to the Universityof Virginiafor the SesquicentennialFellowship. cited References AberJD,MagillA, McNultySG,Boone RD,NadelhofferKJ,Downs M, Hallett R. 1995.Forestbiogeochemistryand primaryproductionalteredby nitrogensaturation.Water,Air and Soil Pollution85: 1665-1670. AberJD,McDowellWH, NadelhofferKJ,MagillA, BerntsonG, KamakeaM, McNultySG, CurrieW,RustadL,FernandezI. 1998.Nitrogensaturation in temperateforest ecosystems:Hypotheses revisited.BioScience48: 921-934. AberJD,GoodaleCL,OllingerSV,SmithM-L,MagillAH, MartinME,Hallett RA,StoddardJL.2003. Is nitrogendepositionalteringthe nitrogen statusof northeasternforests?BioScience53: 375-389. AyresRU,SchlesingerWH, SocolowRH. 1994.Humanimpactson the carbon andnitrogencycles.Pages121-155in SocolowRH,AndrewsC,BerkhoutR,ThomasV,eds.IndustrialEcologyandGlobalChange.Cambridge (MA):CambridgeUniversityPress. Beier C, BlanckK, BredemeierM, LamersdorfN, RasmussenL, Xu Y-J. 1998. Field-scale"clean"rain treatmentsto two Norwaysprucestands withinthe EXMANproject--effectson soil solutionchemistry,foliarnutritionand treegrowth.ForestEcologyand Management101:111-124. BillenG. 1990.N-budgetof the majorriversdischarginginto the continental coastalzone of the North Sea:The nitrogenparadox.Pages153-172 in LancelotC, BillenG, BarthH, eds. Eutrophicationand AlgalBloom in North SeaZones, the Balticand AdjacentAreas:Predictionand Assessment of PreventionActions. Brussels(Belgium):Commission of the EuropeanCommunities.WaterPollutionResearchReport12. Billen G, GarnierJ. 2000. Nitrogen transfersthrough the Seine drainage network:A budgetbasedon the applicationof the"Riverstrahler" model. Hydrobiologia410: 139-150. BlairJM,NellisMD,TurnerCL,HenebryGM,Su Haiping.1998.Terrestrial nutrientcyclingin tallgrassprairie.Pages222-243 in KnappAK,Briggs JM,HartnettDC, CollinsSL,eds.GrasslandDynamics:Long-TermEcologicalResearchin TallgrassPrairie.NewYork:OxfordUniversityPress. BlekenMA. 1997.Food consumptionand nitrogenlosses fromagriculture. Pages19-31 in LagJ,ed. Some GeomedicalConsequencesof Nitrogen CirculationProcesses.Oslo (Norway):NorwegianAcademyof Science and Letters. BoxmanAW,BlanckK, BrandrudT-E,EmmettBA,GundersenP,Hogervorst RF,KjonaasOJ,PerssonH, TimmermannV. 1998.Vegetationand soil biotaresponseto experimentally-changed nitrogeninputsin coniferous forestecosystemsof the NITREXproject.ForestEcologyand Management 101:65-80. BradleyMJ,JonesBM.2002.ReducingglobalNOxemissions:Promotingthe development of advanced energy and transportation technologies. Ambio 31: 141-149. BrickerSB,ClementCG,PirhallaDE,OrlandSP,FarrowDGG. 1999.National EstuarineEutrophicationAssessment:A Summaryof Conditions,HistoricalTrends,andFutureOutlook.SilverSpring(MD):NationalOceanic and AtmosphericAdministration,NationalOceanService. BurkartMR, Stoner JD. 2001. Nitrogen in groundwaterassociatedwith systems.Pages123-145in FollettRF,HatfieldJL,eds.Nitrogen agricultural in the Environment:Sources,Problemsand Management.Amsterdam (Netherlands):ElsevierScience. 354 BioScience - April2003 / Vol.53 No. 4 CassmanKG,DobermannAD,WaltersD. 2002.Agroecosystems,nitrogenuse efficiency,and nitrogenmanagement.Ambio 31: 132-140. ChenAC-T,WangS-L.1999.Carbon,alkalinityand nutrientbudgetson the EastChinaSea continentalshelf.Journalof GeophysicalResearch104: 20,675-20,686. CorredorJE,HowarthRW,TwilleyRR,MorellJM. 1999.Nitrogencycling and anthropogenicimpact in the tropicalinteramericanseas.Biogeochemistry46:163-178. CowlingE, ErismanJW,SmeuldersSM, Holman SC,Nicholson BM. 1998. Optimizingairqualitymanagementin EuropeandNorth America:Justification for integratedmanagement of both oxidized and reduced forms of nitrogen.EnvironmentalPollution 102:599-608. CowlingEB,et al. 2002. Optimizingnitrogenmanagementin food and energy production and environmentalprotection:Summarystatement from the Second InternationalNitrogen Conference.In GallowayJN, Cowling EB, Erisman JW,Wisniewski J, JordanC, eds. Optimizing Nitrogen Managementin Food and EnergyProductionand Environmental Protection:ContributedPapersfrom the SecondInternational Nitrogen Conference,14-18 October2001, Potomac,Maryland.Lisse (Netherlands):A. A. Balkema. CrutzenPJ,AndreaeMO. 1990.Biomassburningin the tropics:Impacton atmospheric chemistry and biogeochemical cycles. Science 250: 1669-1678. DailDB,DavidsonEA,ChoroverJ.2001.Rapidabiotictransformationof nitratein an acid forestsoil. Biogeochemistry54: 131-146. Davidson EA,KellerM, EricksonHE,VerchotLV,VeldkempE. 2000. Testing a conceptualmodelof soil emissionsof nitrousandnitricoxides.BioScience50: 667-680. DeHayesDH, SchabergPG,HawleyGJ,StrimbeckGR. 1999.Acidrainimpact on calciumnutritionand foresthealth.BioScience49: 789-800. Del GrossoSJ,Parton,WJ,MosierAR,HartmanMD, KeoughCA,Peterson GA,OjimaDS, SchimelDS. 2001. Simulatedeffectsof landuse, soil texture and precipitation on N gas emissions using DAYCENT.Pages 413-432 in FollettRF,HatfieldJL,eds. Nitrogen in the Environment: Sources, Problems and Management. Amsterdam (Netherlands): ElsevierScience. Devol AH, ChristensenJP.1993.Benthicfluxesand nitrogencyclingin sediments of the continentalmarginof the easternNorth Pacific.Journal of MarineResearch51: 345-372. Diaz R, RosenbergR. 1995.Marinebenthichypoxia:A reviewof its ecologicaleffectsandthe behavioralresponsesof benthicmacrofauna. Oceanographyand MarineBiologyAnnualReview33: 245-303. Dise NB,WrightRF.1995.Nitrogenleachingin Europeanforestsin relation to nitrogendeposition.ForestEcologyand Management71: 153-162. Dise NB,MatznerE,GundersenP 1998.Synthesisof nitrogenpoolsandfluxes from Europeanforest ecosystems.Water,Air and Soil Pollution 105: 143-154. DriscollC, et al.2003.Nitrogenpollutionin the northeasternUnitedStates: Sources,effects,and managementoptions. BioScience53: 357-374. EpsteinHE,BurkeIC,MosierAR.2001. Planteffectson nitrogenretention in shortgrasssteppe2 yearsafter'"Naddition.Oecologia128:422-430. Erisman W, de VriesW.2000.Nitrogendepositionand effectsin European forests.EnvironmentalReviews8: 65-93. ErismanJW,de VriesW, KrosH, Oenema O, van der EerdenL, Smeulders S. 2001. An outlook for a nationalintegratednitrogenpolicy.EnvironmentalScienceand Policy4: 87-95. [FAO]Food and AgricultureOrganizationof the United Nations. 2000. FertilizerRequirementsin 2015 and 2030. Rome:FAO. 2000.(4 March2003;http://apps.fao.org) FAOstatisticaldatabases. [FAOSTAT] FennME,et al. 2003a.Ecologicaleffectsof nitrogendepositionin the western United States.BioScience53: 404-420. Fenn ME,et al. 2003b.Nitrogen emissions,deposition,and monitoringin the westernUnited States.BioScience53: 391-403. FixenPE,WestFB.2002.Nitrogenfertilizers:Meetingthe challenge.Ambio 31: 169-176. Follett JR,Follett RF. 2001. Utilization and metabolism of nitrogen by humans. Pages 65-92 in Follett R, Hatfield JL,eds. Nitrogen in the Articles Environment: Sources,ProblemsandManagement.Amsterdam(Netherlands):ElsevierScience. FowlerD, SkibaU,Hargreaves KJ.1997.Emissionsof nitrousoxidefromgrasslands.Pages147-164 in JarvisSC,PainBF,eds. GaseousNitrogenEmissions from Grasslands.North Wyke (United Kingdom):Institute of Grasslandand EnvironmentalResearch. GallowayJN. 1998.The globalnitrogencycle:Changesand consequences. EnvironmentalPollution 102 (Sl): 15-24. GallowayJN,CowlingEB.2002.Reactivenitrogenand the world:Twohundredyearsof change.Ambio 31: 64-71. GallowayJN,SchlesingerWH, LevyH II,MichaelsA, SchnoorJL.1995.Nitrogenfixation:Anthropogenicenhancement-environmentalresponse. GlobalBiogeochemicalCycles9: 235-252. GallowayJN,CowlingEB,SeitzingerSJ,SocolowR. 2002.Reactivenitrogen: Too much of a good thing?Ambio 31: 60-63. GleickPH, ed. 1993.Waterin Crisis:A Guideto the World'sFreshWaterResources.New York:OxfordUniversityPress. GoodaleCL,AberJD,McDowellWH. 2000.Long-termeffectsof disturbance on organicand inorganicnitrogenexportin theWhiteMountains,New Hampshire.Ecosystems3: 433-450. GoolsbyDA, BattaglinWA,LawrenceGB,Artz RS,AulenbachBT,Hooper RP,KeeneyDR, StenslandGJ.1999.Fluxand sourcesof nutrientsin the RiverBasin:Topic3 Reportfor the Integrated Mississippi-Atchafalaya Assessmenton Hypoxiain the Gulfof Mexico.SilverSpring(MD):NationalOceanicandAtmosphericAdministration.NOAACoastalOcean ProgramDecisionAnalysisSeriesno. 17. GroffmanPM, GoldAJ,JacinthePA. 1998.Nitrous oxide productionin riparianzones and groundwater.NutrientCyclingin Agroecosystems52: 179-186. GundersenP,EmmettBA,KjonaasOJ,KoopmansCJ,TietemaA. 1998.Impactof nitrogendepositionon nitrogencyclingin forests:A synthesisof NITREXdata.ForestEcologyand Management101:37-56. HallSJ,MatsonPA.1999.Nitrogenoxideemissionsafternitrogenadditions in tropicalforests.Nature400: 152-155. Hill AR. 1996.Nitrateremovalin streamriparianzones. Journalof EnvironmentalQuality25: 743-755. Hill AR,Devito KJ,CampagnoloS, SanmugadasK. 2000. Subsurfacedenitrificationin a forest riparianzone: Interactionsbetween hydrology 51:193-223. andsuppliesof nitrateandorganiccarbon.Biogeochemistry HollandEA,LamarqueJ-F.1997.Bio-atmosphericcouplingof the nitrogen cyclethroughNOxemissionsand NOYdeposition.NutrientCyclingin Agroecosystems48: 7-24. HowarthRW,MarinoR,LaneJ,ColeJJ.1988.Nitrogenfixationin freshwater, estuarine,and marineecosystems,1:Ratesand importance.Limnology and Oceanography33: 669-687. HowarthRW,JensenH, MarinoR, PostmaH. 1995.Transportto and processingof phosphorusin near-shoreand oceanicwaters.Pages323-345 in TiessenH, ed. Phosphorusin the GlobalEnvironment,SCOPE54. Chichester(United Kingdom):Wileyand Sons. HowarthRW,et al. 1996.Regionalnitrogenbudgetsand riverineN and P fluxesforthe drainagesto the NorthAtlanticOcean:Naturalandhuman influences.Biogeochemistry35: 75-139. - . 2000. Nutrientpollution of coastalrivers,bays,and seas.Issuesin Ecology7: 1-15. HowarthRW,BoyerE, PabichW, GallowayJN.2002a.Nitrogenuse in the United Statesfrom 1961-2000 and potentialfuturetrends.Ambio 31: 88-96. HowarthRW,SharpleyAW,WalkerD. 2002b.Sourcesof nutrientpollution to coastalwatersin the UnitedStates:Implicationsfor achievingcoastal waterqualitygoals.Estuaries25: 656-676. HudakPF 2000. Regionaltrendsin nitratecontent of Texasgroundwater. Journalof Hydrology228:37-47. JohnsonDW,ChengW, BurkeIC. 2000. Biotic and abioticnitrogenretention in a varietyof forestsoils. Soil ScienceSocietyof AmericaJournal 64: 1503-1514. JorgensenBB,RichardsonK. 1996.Eutrophicationin CoastalMarineSystems.Washington(DC):AmericanGeophysicalUnion. KahlJS,NortonSA,FernandezIJ,NadelhofferKJ,DriscollCT,AberJD.1993. Experimentalinducementof nitrogensaturationat the watershedscale. EnvironmentalScienceand Technology27: 565-568. KeelingCD. 1993.GlobalhistoricalCO2 emissions.Pages501-504 in Boden TA,KaiserDP,SepanskiRJ,Stoss FW,eds. Trends'93:A Compendium of Dataon GlobalChange.OakRidge(TN): CarbonDioxideInformation AnalysisCenter,Oak RidgeNationalLaboratory.ORNL/CDIAC65. KellyCA,RuddJWM,HessleinRH, SchindlerDW,Dillon PJ,DriscollCT, GheriniSA,HeckyRE.1987.Predictionof biologicalacidneutralization in acid-sensitivelakes.Biogeochemistry3: 129-140. KempWM,SampouPA,CaffreyJM,MayerM, HenriksenK,BoyntonWR. 1990.Ammoniumrecyclingversusdenitrification in ChesapeakeBaysediments.Limnologyand Oceanography35: 1545-1563. KramerDA. 1999.US GeologicalSurveyMineralsinformation:Nitrogen statistics and information. (4 March 2003; http://minerals.usgs.gov/ minerals/pubs/commodity/nitrogen) KroezeC, SeitzingerKP,DominguesR.2001.Futuretrendsin worldwideriver nitrogen transport and related nitrous oxide emissions: A scenario analysis.ScientificWorld1:328-335. KuhlbuschTA,LobertJM,CrutzenPJ,WarneckP.1991.Molecularnitrogen emissions from denitrificationduring biomass burning. Nature 351: 135-137. Lapointe BE, O'Connell JD. 1989. Nutrient-enhanced productivity of Cladophroaproliferain HarringtonSounds,Bermuda:Eutrophication of a confined,phosphorus-limitedmarineecosystem.Estuarine,Coastal and ShelfScience28: 347-360. LaursenAE,SeitzingerSP.2001. The role of denitrificationin nitrogenremovaland carbonmineralizationin mid-Atlanticbightsediments.ContinentalShelf Research22: 1397-1416. Lewis WM. 2002. Estimationof background nitrogen yields for North America by use of benchmarkwatersheds.Biogeochemistry57-58: 375-385. LewisWM, MelackJM,McDowellWH, McClaineM, RicheyJE.1999.Nitrogen yields from undisturbedwatershedsin the Americas.Biogeochemistry46: 149-162. MackenzieFT.1998.Our ChangingPlanet:An Introductionto EarthSystem Scienceand GlobalEnvironmentalChange.2nd ed. UpperSaddle River(NJ):Prentice-Hall. MagillA,AberJ,BerntsonG, McDowellW,NadelhofferK,MelilloJ,Steudler P.2000.Long-termnitrogenadditionsandnitrogensaturationin two temperateforests.Ecosystems3: 238-253. Matson P, Lohse K, Hall S. 2002. The globalizationof nitrogen:Consequencesfor terrestrialecosystems.Ambio 31: 113-119. McGlatheryKJ,MarinoR, HowarthRW.1994.Variableratesof phosphate uptakeby shallowmarine sediments:Mechanismsand ecologicalsignificance.Biogeochemistry25: 127-146. MitschWJ,DayJW,GilliamW, GroffmanPM,Hey DL,RandallGW,Wang N. 2001.Reducingnitrogenloadingto the Gulfof Mexicofromthe MississippiRiverBasin:Strategiesto countera persistentecologicalproblem. BioScience51: 373-388. MoomawWR.2002. Energy,industryand nitrogen:Strategiesfor reducing reactivenitrogenemissions.Ambio 31: 184-189. Miller J.1992.Geographicaldistributionand seasonalvariationof surface emissionsand depositionvelocitiesof atmospherictracegases.Journal of GeophysicalResearch97: 3787-3804. NadelhofferKJ,Downs MR, FryB. 1999.Sinksfor '5N-enrichedadditions to an oak forest and a red pine plantation.EcologicalApplications9: 72-86. Nixon SW.1988.Physicalenergyinputsand the comparativeecologyof lake and marineecosystems.Limnologyand Oceanography33: 1005-1025. Nixon SW,et al. 1996.The fate of nitrogenand phosphorusat the land-sea marginof the North AtlanticOcean.Biogeochemistry35: 141-180. Nolan BT. 1999. Nitrate behavior in ground waters of the Southeastern USA. Journalof EnvironmentalQuality28: 1518-1527. - . 2001. Relatingnitrogensourcesand aquifersusceptibilityto nitrate in shallowgroundwatersof the UnitedStates.GroundWater39:290-299. * April2003 / Vol.53 No. 4 BioScience 355 Articles Nolan BT,StonerJD.2000. Nutrientsin groundwatersof the conterminous United States,1992-1995. EnvironmentalScienceand Technology34: 1156-1165. [NRC]NationalResearchCouncil. 1996.UnderstandingMarineBiodiversity.Washington(DC): NationalAcademyPress. 2000. CleanCoastalWaters:Understandingand Reducingthe Ef--. fectsof NutrientPollution.Washington(DC):NationalAcademyPress. Oenema O, PietrzakS. 2002. Nutrient managementin food production: Achievingagronomicand environmentaltargets.Ambio 31: 159-168. OenemaO,BanninkA,SommerSG,VelthofGL.2001.Gaseousnitrogenemissions from livestockfarmingsystems.Pages255-290 in FollettRF,HatfieldJL,eds.Nitrogenin the Environment:Sources,Problemsand Management.Amsterdam(Netherlands):ElsevierScience. OllingerSV,AberJD,LovettGM, MillhamSE,LathropRG.1993.A spatial modelof atmosphericdepositionforthe northeasternU.S.EcologicalApplications3:459-472. Penner JE,et al. 2001. Aerosols, their direct and indirect effects. Pages 289-348 in Houghton JT,GriggsDJ,NoguerM, van der LindenPJ,Dai X, MaskellK,JohnsonCA,eds.ClimateChange2001:The ScientificBasis. Contributionsof WorkingGroupI to the ThirdAssessmentReport of the Intergovernmental Panelon ClimateChange.NewYork:Cambridge UniversityPress. PerakisSS,HedinLO.2002.Nitrogenloss fromunpollutedSouthAmerican forestsmainlyvia dissolvedorganiccompounds.Nature415:416-419. PetersonBJ,et al. 2001. Control of nitrogen export from watershedsof headwaterstreams.Science292:86-90. PopeCAIII,Thun MJ,NamboodiriMM,DockeryDW,EvansJS,SpeizerFE, HeathCW Jr.1995.Particulateairpollution as a predictorof mortality Critin a prospectivestudyof U.S.adults.American Journalof Respiratory ical CareMedicine 151:669-674. PratherM, et al. 2001.Atmosphericchemistryand greenhousegases.Pages 239-287 in HoughtonJT,GriggsDJ,NoguerM, van der LindenPJ,Dai X, MaskellK,JohnsonCA,eds.ClimateChange2001:The ScientificBasis. Contributionsof WorkingGroupI to the ThirdAssessmentReport of the Intergovernmental Panelon ClimateChange.NewYork:Cambridge UniversityPress. PuckettLJ,CowderyTK. 2002. Transportand fate of nitrate in a glacial outwashaquiferin relationto ground-waterage,land use practicesand redoxprocesses.Journalof EnvironmentalQuality31: 782-796. PuckettLJ,CowderyTK,LorenzDL,StonerJD.1999.Estimationof nitrate contaminationof an agro-ecosystemoutwashaquiferusing a nitrogen mass-balancebudget.Journalof EnvironmentalQuality28:2015-2025. RabalaisN. 2002.Nitrogenin aquaticecosystems.Ambio 31: 102-112. RabalaisNN, TurnerRE,JusticD, Dortch Q, WisemanWJ Jr.1999. Characterizationof Hypoxia:Topic 1. Reportfor the IntegratedAssessment of Hypoxiain the Gulfof Mexico.SilverSpring(MD):NationalOceanic and Atmospheric Administration. NOAA Coastal Ocean Program DecisionAnalysisSeriesReportno. 15. RabalaisNN, TurnerRE,ScaviaD. 2002.Beyondscienceinto policy:Gulfof Mexicohypoxiaand the MississippiRiver.BioScience52: 129-142. S, HansenS.1999.Largescale JC,ThorsenM, JensenJB,Kleeschulte Refsgaard modellingof groundwatercontaminationfrom nitrateleaching.Journal of Hydrology221: 117-140. Roy RN, MisraRV,MontanezA. 2002. Decreasingrelianceon mineralnitrogen,yet more food. Ambio 31: 177-183. SchipperLA,Vojvodic-VukovicM. 2001. Fiveyearsof nitrateremoval,dewall.WaterResearch nitrificationandcarbondynamicsin a denitrification 35: 3473-3477. 356 BioScience * April2003 / Vol.53 No. 4 SchulzeED. 1989.Airpollution and forestdeclinein a spruce (Piceaabies) forest.Science244: 776-783. SeinfeldJH,PandisSN. 1998.AtmosphericChemistryandPhysics:FromAir Pollutionto ClimateChange.NewYork:JohnWiley and Sons. SeitzingerSP,GiblinAE. 1996.Estimatingdenitrificationin NorthAtlantic continentalshelf sediments.Biogeochemistry35: 235-260. SeitzingerSP,KroezeC. 1998.Globaldistributionof nitrousoxide production and N inputs in freshwaterand coastalmarineecosystems.Global BiogeochemicalCycles12:93-113. SeitzingerSP,KroezeC, StylesRV.2000. Globaldistributionof N20 emissions fromaquaticsystems:Naturalemissionsandanthropogeniceffects. Chemosphere:GlobalChangeScience2:267-279. SeitzingerSP,StylesRV,BoyerE,AlexanderRB,BillenG,HowarthR,Mayer B, van BreemenN. 2002. Nitrogen retentionin rivers:Model development and applicationto watershedsin the easternU.S.Biogeochemistry 57-58: 199-237. Smil V. 1999. Nitrogen in crop production:An account of global flows. GlobalBiogeochemicalCycles13:647-662. . 2001. Enrichingthe Earth:FritzHaber,CarlBosch,and the Transformationof Food Production.Cambridge(MA):MITPress. . 2002. Nitrogen and food production:Proteinsfor human diets. Ambio 31: 126-131. SommerSG,HutchingsNJ. 1997. Componentsof ammoniavolatilization from cattleand sheepproduction.Pages79-93 in JarvisSC,PainBF,eds. Gaseous Nitrogen Emissions from Grasslands.North Wyke (United Kingdom):Instituteof Grasslandand EnvironmentalResearch. SteinhartGS, LikensGE, GroffmanPM. 2000. Denitrificationin stream sedimentsof five northeastern(USA) streams.Verhandlungender InternationalenVereinigungfir Theoretischeund AngewandteLimnologie 27: 1331-1336. TartowskiS, HowarthRW.2000. Nitrogen,nitrogencycling.Encyclopedia of Biodiversity4: 377-388. TilmanD, FargioneJ,WolffB,D'AntonioC,DobsonA, HowarthR,Schindler D, SchlesingerWH, SimberloffD, SwackhamerD. 2001. Forecasting agriculturallydrivenglobalchange.Science292:281-284. ValielaI, McClellandJ, HauxwellJ, Behr PJ,Hersh D, ForemanK. 1997. Macroalgalblooms in shallowestuaries:Controlsand ecophysiological and ecosystem consequences. Limnology and Oceanography 42: 1105-1118. vanAardenneJA,DentenerFJ,KlijnGoldewijkCGM,LelieveldJ,OlivierJGJ. 2001.A 1oX 10 resolutiondatasetof historicalanthropogenictracegas emissionsfor the period 1890-1990, GlobalBiogeochemicalCycles15: 909-928. van BreemenN, et al. 2002.Wheredid all the nitrogengo?Fateof nitrogen inputs to largewatershedsin the northeasternUSA. Biogeochemistry 57-58: 267-293. van EgmondND, BresserAHM, BouwmanAF 2002. The Europeannitrogen case.Ambio 31: 72-78. VitousekPM,HowarthRW.1991.Nitrogenlimitationon landandin the sea: How can it occur?Biogeochemistry13:87-115. VitousekPM,HowarthRW,LikensGE,MatsonPA,SchindlerD, Schlesinger WH, TilmanGD. 1997.Human alterationof the global nitrogencycle: Causesand consequences.Issuesin Ecology 1:1-17. WolfeA, PatzJA.2002.Nitrogenand humanhealth:Directand indirectimpacts.Ambio 31: 120-125.
© Copyright 2026 Paperzz