isotopic character of nitrous oxide emitted from streams

Published: May 02, 2011

r 2011 American Chemical Society 4682 dx.doi.org/10.1021/es104116a | Environ. Sci. Technol. 2011, 45, 4682–4688

ARTICLE

pubs.acs.org/est

Isotopic Character of Nitrous Oxide Emitted from StreamsHelen M. Baulch*,†

Environmental and Life Sciences Graduate Program, Trent University, 1600 West Bank Drive, Peterborough ON K9J 7B8 Canada

Sherry L. Schiff and Simon J. Thuss‡

Department of Earth and Environmental Sciences, University of Waterloo, 200 University Avenue West, Waterloo ON N2L 3G1

Peter J. Dillon

Department of Environmental and Resource Studies, Trent University, 1600 West Bank Drive, Peterborough ON K9J 7B8 Canada

bS Supporting Information

’ INTRODUCTION

Concentrations of the biogenic greenhouse gas nitrous oxide(N2O) are increasing in the atmosphere. Agricultural nitrogen(N) fertilization is the dominant cause of anthropogenic N2Oemissions. The contribution of N2O to warming is modest inglobal terms (<10% of that of CO2

1). However, projectedincreases in fertilizer use2 suggest that atmospheric concentra-tions of N2O will continue to grow. Beyond its contribution toclimate change, N2O is also the largest current emission con-tributing to stratospheric ozone depletion.3

Soils are a major site of both natural and anthropogenic N2Oemissions. IPCC budget methods suggest that a large amount ofanthropogenic N applied to agricultural soils is leached to aquaticecosystems where it can lead to N2O production.4 Surface watersare frequently supersaturated with N2O, indicating they functionas net sources.5 The strength of anthropogenic emissions fromstreams and rivers has been estimated at more than 1 Tg ofN2O�N y�1,6 or the equivalent of 298 Tg CO2 y

�1 over a 100-year time span. However, uncertainty in emissions budgets is high.4

Isotopic characterization of N2O has contributed to ourunderstanding of global N2O dynamics by helping to constrainglobal N2O budgets.7 Global isotope mass balances have helpedcharacterize the importance of stratospheric N2O loss, associatedtransport of enriched stratospheric N2O to the troposphere,7 andconfirmed the increased strength of an isotopically depletedanthropogenic N2O source, largely from agricultural soils,8 but

possibly derived in part from streams and rivers. Low orderstreams are thought to be important sites of N cycling,9 suggest-ing they could be hotspots of N2O emissions and importantcontributors to global N2O budgets. Although emissions fromstreams and rivers are reported as agricultural emissions,4 if N2Oemitted from fluvial systems has a signature distinct from that ofagricultural soils or if observed rates of N2O fluxes were large, thiscould suggest that global isotopic N2O budgets require reassess-ment to separately account for river and stream emissions.

The isotopic character of N2O emitted from soils and aquaticecosystems is governed by fractionation associated with biologi-cal processing (nitrification, denitrification, N2O consumptionvia denitrification), the isotopic character of the substrates, andphysical fractionation (see Supporting Information (SI)). Inaquatic ecosystems, physical fractionation due to gas exchange10

can lead to differences between the character of dissolved andemitted N2O. In this paper we assess the isotopic character ofN2O emitted from a series of low order (second�fourth order)streams, and contrast the character of emissions to those of largerrivers, soils, and oceans.

Received: December 8, 2010Accepted: April 11, 2011Revised: March 28, 2011

ABSTRACT: Global models have indicated agriculturally impacted rivers and streams maybe important sources of the greenhouse gas nitrous oxide (N2O). However, there issignificant uncertainty in N2O budgets. Isotopic characterization can be used to helpconstrain N2O budgets. We present the first published measurements of the isotopiccharacter of N2O emitted from low (2�4) order streams. Isotopic character of N2O variedseasonally, among streams, and over diel periods. On an annual basis, δ18O of emitted N2O(þ47.4 to þ51.4%; relative to VSMOW) was higher than previously reported for largerrivers, but δ15N of emitted N2O (�16.2 toþ2.4% among streams; relative to atmosphericN2) was similar to that of past studies. On an annual basis, all streams emitted N2O withlower δ15N than tropospheric N2O. Given these streams have elevated nitrate concentrations which are associated with enhancedN2O fluxes, this supports the hypothesis that streams are contributing to the accumulation of 15N-depleted N2O in the troposphere.

4683 dx.doi.org/10.1021/es104116a |Environ. Sci. Technol. 2011, 45, 4682–4688

Environmental Science & Technology ARTICLE

’MATERIALS AND METHODS

Our study streams are located in southern Ontario, Canada(Figure S1) in an area of mixed land use with low urbandevelopment (<3% of catchment area), moderate wetland cover-age (9�25%), and significant agricultural land use (51�76% inannual and forage crops, idle lands, hay, pasture, and marginallands).11 Streams were sampled from March 2007 to October2008 excluding periods of ice cover, when gas exchange isexpected to be negligible. In Jackson Creek (JC), Layton Creek(LC), and the Black River (BR), we typically sampled a single siteper stream. In Mariposa Brook (MB) we sampled two sites(upper site designated as MB-U, lower is MB-L) to help charac-terize spatial variation. The distance between the two samplingsites (7 km) was sufficiently long that upstream gases hadminimal effect on downstream gases (<5%, based on first-orderloss equations; see SI). MB receives tile drain inputs which mayaffect N2O emissions. None of the other streams are known tohave tile drains affecting sampling sites, and none of the streamsreceived direct inflows of animal waste or sewage. MB, JC, andBR were fourth order (Strahler) streams at our sampling sites,andLCwas a second order stream.Mean discharge was <300 L s�1

in MB, BR, and LC (over the period where we calculated annualemissions, see Table S1). In JC our ability to characterizeemissions was limited to two dates, when mean discharge was1200 L s�1. Annual discharge from these streams is approxi-mately 100 times lower than the Bang Nara, Grand, or TamaRivers, where the isotopic character of N2O has been previouslyreported (Table S2). For clarity, we refer to the current studysystems (MB, BR, LC, and JC) as streams, and the past studiesfrom higher discharge ecosystems as rivers.

Samples for nutrient analysis were filtered (preashed, B-pureleached glass-fiber filters). Nitrate concentrations were measuredusing the red azo dye method after reduction to nitrite.12 Ammo-nium concentrations were measured using the indophenol bluemethod, in a buffered alkaline solution.12 Samples for determina-tion of N2O concentrations were obtained in 125-mL serumbottles, sealed with a prebaked (at 60 �C for 24 h) Vacutainerrubber stopper, and immediately preserved with 0.4% (v/v)saturated mercuric chloride. Samples were analyzed followingheadspace equilibration using a Varian CP-3800 gas chromato-graph equipped with an electron capture detector. Analytical error,estimated from triplicate equilibrations, was 4.3% (coefficient ofvariation).

N2O flux was calculated by multiplying the piston velocity(k, in units of m h�1) by the difference between measured andequilibrium N2O concentrations. Piston velocity was estimatedusing oxygen13 and tracer-based methods,14 normalized tostream temperature and N2O. Results were extrapolated todifferent flow conditions by calibrating empirical models usingdirect measurements (see SI).15 Annual estimates of the char-acter of emitted N2O are based on daytime sampling and arereported only for BR, LC, and two sites on MB (see SI).Reported N chemistry and N2O data (Table 1) reflect datesand locations for which we estimate the isotopic character ofN2O flux. In addition to routine (daytime) sampling, we perfor-med diel sampling twice at three locations (LC, BR, MB-L) andonce atMB-U. Diel sampling was performed under low dischargeconditions (Table S1).

Samples for isotopic analysis of dissolved N2O were obtainedin 500-mL media bottles (sealed with black lyophilizationstoppers) or 125-mL serum bottles (as for N2O concentration

Table 1. Range of δ15N, δ18O of Emitted N2O, N2O Flux, Percent Saturation, and Concentrations of Nitrate (NO3�) and

Ammonium (NH4þ)a

δ15N�N2O (%) δ18O�N2O (%) N2O (% saturation) N2O flux (μmol m�2 h�1) [NO3� ] (μM) [NH4

þ] (μM)

BR

routine �22.6 to þ1.6 [17] þ34.5 to þ60.4 [17] 143�219 0.02�0.8 17�59 0.4�3.7

diel: 3�4 July 2007 �17.7 to þ7.5 [6] þ37.2 to þ56.5 [6] 192�234 0.4�0.6 58�69 0.1�1.2

diel: 28�29 August 2007 �7.9 to �0.1 [6] þ51.4 to þ67.9 [6] 147�156 0.2�0.3 14�21 0.8�1.7

JC

routine �2.4 to �0.2 [2] þ50.6 to þ58.0 [2] 192�205 0.9�1.7 34�106 0.5�2.1

LC

routine �7.4 to þ10.7 [12] þ28.8 to þ68.4 [12] 134�400 0.2�0.8 34�117 0.3�6.3


diel: 28�29 August 2007 þ2.0 to þ10.0 [7] þ52.8 to þ57.1 [7] 165�288 0.5�1.4 27�54 0.8�1.6

MB-U

routine �20.7 to þ4.5 [17] þ35.1 to þ59.2 [17] 200�1700 0.5�8.3 37�206 0.3�9.1


MB-L

routine �13.6 to þ3.5 [15] þ37.7 to þ59.0 [15] 129�419 0.1�5.2 12�163 0.9�3.6


diel: 28�29 August 2007 �8.9 to þ2.6 [7] þ51.4 to þ57.7 [7] 148�272 0.2�0.8 8�21 1.1�2.7aThe number of samples used to estimate isotopic character of emissions is in brackets.



analyses) and immediately preserved with a saturated mercuricchloride solution (final concentration 0.4% v/v). Samples forisotope analysis were prepared by purging the sample with ultrahigh purity helium, while using liquid N2 to trap N2O followingthe removal of water vapor (using magnesium perchlorate and anafion dryer) and CO2 (chemical removal via Carbosorb). Cryo-genic trapping was performed in pre-evacuated (∼ 0.1 mbar)round-bottomed vials packed with Pyrex wool and glass beads,and sealed with thick (13 mm) blue butyl rubber stoppers(Bellco) and crimp rings (Chromatographic Specialties). Sampleswere overpressurized with ultrahigh purity helium, and 4�10 nmolof N2O was injected into a GV Trace Gas preconcentrator system(which chromatographically separated N2O from any residualCO2), coupled to a GV Isoprime mass spectrometer.16

All isotope ratios are reported in delta notation (δ) in unitsof parts per thousand (%) relative to 15N/14N of atmosphericN2 and 18O/16O of Vienna Standard Mean Ocean Water(VSMOW) (see SI). There are currently no internationallyrecognized reference materials for isotopic analysis of N2O.However, there is little variation in the isotopic character oftropospheric N2O throughout the northern hemisphere.17 Basedon this, internal standards were calibrated using samples oftropospheric N2O obtained at the University of Waterloo,assuming tropospheric N2O had δ15N of 6.72% and δ18O of44.62%.17 Tropospheric N2O at the University of Waterlooshowed high reproducibility over months (standard deviationδ15N = 0.2%, δ18O = 0.5%16). Two internal standards wereused (δ15N = þ2.8%, δ18O = þ40.0%, and δ15N = �92.6%,δ18O = þ20.5%) and 15% of our samples were analyzed induplicate or triplicate. Mean error (deviation from the mean)was (0.34% for δ15N and (0.66% for δ18O. Isotope analyseswere performed in the University of Waterloo EnvironmentalIsotope Laboratory.

Isotopic ratios of dissolved N2O differ from emitted N2O dueto the effects of kinetic fractionation during invasion and evasion.The isotopic ratio of emitted N2Owas calculated using equationsdescribed in the SI. Estimates of the δ15N and δ18O of N2Oemitted from our study streams are subject to error associatedwith measurement of dissolved N2O concentrations and isotoperatios. We estimated error in the isotopic character of emittedN2O by bounding the data using our measured error terms(concentration: ( 4.3%, δ15N�N2O: ( 0.34% and δ18O�N2O:( 0.66%) and determining the maximum error calculated(multiplicative error resulting from error in analysis of concen-trations and isotopes). Reported data are restricted to sites anddates where this maximum error did not affect the calculatedmean emissions character by more than(5% for δ15N or δ18O.We then calculated mean annual (flux and time-weighted)isotopic composition of emitted N2O (see SI). Although sam-pling was generally performed at least monthly, there are gaps inthe time series when low concentrations precluded isotopicanalysis, or the error term exceeded 5%. Notably, for this reason,we can report the isotopic character of N2O emitted from JC foronly two dates. Annual estimates of isotopic character reflect ourperiod of most frequent measurements (April or May 2007�2008) and are not reported for JC.

We determined whether the character of emittedN2O (annualmean δ15N and δ18O) differed between this current study whichfocused on low order streams (LC, BR, mean of 2 sites on MB),and the mean value for rivers (based on three rivers: Bang Nara,Grand, Tama; δ15N�N2O: �8.4%, δ18O: þ36.6%5,16,18). Weextended the analysis to determine whether the δ18O�N2O

across our study streams differed from the maximum annualvalue previously reported (δ18O�N2O: þ41.1%; in the TamaRiver5). Using one-sample (two-tailed) t-tests, we also deter-mined whether δ15N and δ18O of N2O emitted from fluvialsystems differed from tropospheric values.17 This analysis usedannual means from our study streams (LC, BR, and mean valueof two sites for MB) and reported values for rivers (Bang Nara,Grand, Tama5,16).

We used 2-way analysis of variance (ANOVA) to assesswhether δ15N and δ18O varied spatially and seasonally (bet-ween summermonths of June�July�August and othermonths).Full ANOVA tables are presented in the SI. We first tested fordifferences between MB-U and MB-D, then combined data forthe two MB sites and tested for differences among streams andseasons. We determined which pairs of streams differed usingBonferroni-corrected multiple pairwise contrasts. One outlierwas apparent for both δ15N and δ18O and was excluded fromanalyses (absolute value of studentized residual >3.5). We usedcorrelation analysis to determine relationships between % satura-tion and character of emitted N2O, as well as δ

15N and δ18O ofemitted N2O. Finally, we used paired t-tests to contrast daytime(nearest solar noon) and nighttime (farthest from solar noon)samples across all sampling efforts. All statistical analyses wereperformed in Systat 13.0.19 A level of significance of 0.05 wasused in all analyses and data were tested for adherence to theassumptions of each analysis.

’RESULTS AND DISCUSSION

Although oceans were one of the first frontiers for isotopiccharacterization of N2O, and research on the character of N2Oemitted from soils has expanded rapidly, work in fresh water isonly starting. Data are available from only three rivers, two ofwhich are temperate rivers with significant sewage inflows(Grand River,16 Tama River5) while the third is a tropical riverin a forested, swamp catchment (Bang Nara River16,18).

Our data almost uniformly indicate that stream-emitted N2Ohas lower δ15N (range: �22.6 to þ10.7%; Table 1,Figures 1�3) than tropospheric N2O (δ15N: þ6.72%17). Thisis consistent with past studies of rivers.5,16 The annual meanδ15N of N2O emissions from the 6 fluvial systems where δ15N ofN2O has now been reported is lower than tropospheric N2O(one sample t-test: p = 0.009, df = 5). Emitted δ15N�N2O didnot vary between our study streams and the mean of past reportsfor rivers (one sample t-test, p = 0.3, df = 2). Although thesystems with sewage inputs (Tama and Grand Rivers) might beexpected to differ inN2O character based on potential differencesin N cycling and N substrates (see SI), δ15N of N2O emissionsfrom sewage-influenced systems was similar to that of systemsthat did not have direct inflows (Figure 3).

All of our study streams (BR, LC, JC, MB) have elevatedNO3

� concentrations (Table 1; estimated predevelopmentconcentration <11 μM20). N enrichment of fluvial ecosystemsis a widespread phenomenon occurring in many regions of theglobe.6 Based on the observation that fluvial systems emit 15N-depleted N2O relative to tropospheric N2O, and the widelyaccepted view that N enrichment is contributing to increasedN2O fluxes,4 we suggest that the N-enrichment of streams iscontributing to the observed accumulation of 15N-depleted N2Oin the troposphere. On a global scale, the relative importance offluvial emissions remains highly uncertain. However, current



IPCC emissions factors for fluvial ecosystems are lower thanproposed in early models.6,21

Interestingly, δ18O emitted N2O from our study streams isdistinct from past studies of fluvial systems (annual time and flux-weighted range: þ47.4 to þ51.5% among streams; Figure 3;one sample t-test against annual mean of past studies: p = 0.01,df = 2; one sample t-test against maximum of past studies

p = 0.03, df = 2). Although error in our estimation of gas transfer(Table S1) may contribute to error in estimation of the characterof annual emissions (see SI), this does not affect daily orinstantaneous data. Our finding of a distinct δ18O�N2O is sup-ported in seasonal and diel data (Figures 1 and 2). Across allmeasurements (across streams, dates, and including diel sampl-ing) emitted δ18O�N2O was less than 41.1% (the highest

Figure 1. Seasonal variation in δ15N (plots a�d) and δ18O (plots e�h) of emitted N2O (daytime sampling, 2007�2008). Data from LC (whitetriangles) are plotted on the same panels as data from JC (gray circles). Error bars show error in isotopic character of emitted N2O based on analyticalerror terms.

Figure 2. Diel variation in δ15N (plots a�g) and δ18O (plots h�n) of emitted N2O. Black line over plots indicates period between sunset and sunrise.Gray lines indicate time and flux-weighted character N2O. Error bars show error in isotopic character of emitted N2O based on analytical error terms.



annual estimate in past studies, see Figure 3) in fewer than 10% ofcases. There are a number of reasons δ18O�N2O could varybetween streams and rivers. If denitrification rates are higher insmall streams due to their morphometry,9 δ18O�N2O may bemore 18O-enriched than in systems where pelagic N cyclingprocesses are more important (see SI). Small streams may alsoreceive significant N2O inputs from ground waters and tiledrains22 which can have a high δ18O�N2O.

23 O-exchangeprocesses are another key determinant of δ18O�N2O (see SI;refs 24,25). Although δ18O of stream-emitted N2O differs fromthe few available measurements of rivers, these values lie withinthe bounds of past values reported for soils.26,27 The soleexception is LC, where emitted N2O is more enriched than mostsoil emissions, but within the range of values observed inoceans.28 Across all fluvial systems studied to date, the δ18O ofemitted N2O did not differ from tropospheric N2O (one samplet-test: p = 0.5, df = 5).

MB sites did not differ significantly in δ15N or δ18O�N2O(Figures 1 and 3; 2-way ANOVA, Tables S3, S4), despite theirspatial separation, but did show evidence of seasonal changes inδ18O and δ15N (2-way ANOVA, Tables S3, S4, p < 0.01). Theδ15N of emitted N2O varied across streams and seasons (2-wayANOVA, p < 0.001, Table S5) and δ15N�N2O emissions diffe-red among all pairwise combinations of streams (LC, BR, MB;Bonferroni, p< 0.05). The BR emitted themost 15N-depletedN2O(Figure 1). LC emitted N2O with the highest δ15N (Figures 1 and3). N2O emitted from JC on the twooccasions onwhich it could becharacterized was within the range of the other study streams(Figure 1). δ18O also varied among streams (2-way ANOVA,Table S6, p = 0.049), but only BR and LC showed significantdifferences in pairwise contrasts (Bonferroni, p = 0.046).

Emitted N2O generally had higher δ15N and δ18O in summer(June�September) than other seasons (Figure 1), although thistrend was statistically significant for δ15N (2-way ANOVA, TableS5, p = 0.046) and not for δ18O (2-way ANOVA, Table S6, p =0.06). Similar patterns observed in the Grand River, Ontario,were thought to be due at least in part to greater kinetic isotopeeffects associated with lower N2O production rates in cooler

weather.16 Another possible factor is the increase in N2O reduc-tion with temperature29 which could result in isotopic enrich-ment of N2O in summer. Changes in composition of the endmembers may also contribute. For example, δ18O of H2O waslikely highest during dry summer months.30 δ18O�O2 is expec-ted to undergo the largest degree of diel variation in summermonths when aquatic productivity is highest (see SI). N2Opercent saturation was not related to δ15N or δ18O of emittedN2O within a stream (seasonal data). δ15N and δ18O of emittedN2O were correlated (BR: r = þ0.66, p = 0.02; LC: r = þ0.63,p = 0.03; MB: r = þ0.83, p < 0.0001). This may reflectfractionation associated with denitrification (Table S7, S8) orgas exchange.

Major diel changes in N cycling processes have been shown insome systems,31 which could contribute to changes in thecharacter of emitted N2O. The extent of day�night variationin the isotopic character of emitted N2O actually nears the extentof seasonal variation in some streams (Figures 1 and 2); however,consistent day�night patterns were not apparent (Figure 2; alsosee SI). Contrasting pairs of samples closest and furthest fromsolar noon showed no significant difference for either isotope(paired t-test: δ15N p = 0.9, δ18O p = 0.83, df = 6). This isprobably due, at least in part, to variable rates of gas transferamong streams. Streams with slower gas exchange will show alonger lag period before a change in the character of newlyproduced N2O (e.g., due to a change in pathway or substratecomposition) significantly affects the character of emitted N2O.While the average gas transfer coefficient was 5.3 per day acrossstreams during diel sampling, the upstream site in MB had muchlower rates of gas exchange (1.5 per day), which helps explain therelatively stable isotopic composition at this site. The largemagnitude of variation in δ18O of emitted N2O suggests changeswere occurring either in the processes (nitrification vs denitrifi-cation), the importance of O-exchange reactions, or δ18O of theNO3

� end member. δ18O�H2O is expected to be relativelystable during diel sampling, and clear relationships with δ18O�O2 were not apparent (see SI).

Diel variation in the isotopic character of emitted N2O hasbeen investigated in two past studies. The first was a study ofN2O emissions from a urine-amended soil. Changes in thecharacter of emitted N2O occurred over short time periods,but did not appear to be related to day�night cycles.32 Thesecond study addressed diel changes in the eutrophic GrandRiver. Emissions of N2O from the Grand River during thedaytime had higher δ15N�N2O than at night. This was likelydue to diel changes in the isotopic composition of the substratesas well as switching of the major N2O production pathways.16

In conclusion, streams and rivers, like soils, tend to emit N2Owith δ15N that is lower than tropospheric N2O. Our resultssuggest not only that increased N2O emissions from N-enrichedstreams are contributing to the accumulation of 15N-depletedN2O in the atmosphere, but also indicate that the exclusion offluvial sources from global isotope mass balances7,28 could lead tothe overestimation of soil sources because of their similarity inδ15N, and the exclusion of fluvial systems from past work.Relatively small fluxes from our study streams in the context ofemissions from agricultural soils in the catchment (streamemissions estimated at 0.2�12% of emissions from agriculturalsoils þ streams33) indicate the error may be small compared tothe uncertainty in emissions budgets.4 However, relatively fewlong-term estimates of N2O flux exist for fluvial systems, andaccumulating evidence suggests that sewage and tile-drain

Figure 3. Annual (time and flux-weighted) isotopic character of N2Oemitted from the BR, LC, and MB (April or May 2007�8), flux-weighted data from the Grand River,16 Bang Nara River,16 and TamaRiver.5 Dashed gray lines show bounds of N2O emitted from theagricultural stream Strawberry Creek during a winter thaw.36



impacted systems may have high emissions.34,35 A more com-plete assessment of the role of fluvial systems in global N2Obudgets requires additional measurements across a broad spec-trum of ecosystems and should include isotopic characterization.

’ASSOCIATED CONTENT

bS Supporting Information. Additional information onstudy sites and stream discharge, calculations (isotopic notation,bulk N2O flux calculations, equations for calculation of isotopiccharacter of N2O flux, first order loss equation), statistics(ANOVA tables), and the processes affecting the isotopic signa-ture of dissolved and emitted N2O (Figure S1; Tables S1�S9).This information is available free of charge via the Internet athttp://pubs.acs.org.

’AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected]; phone: 608-262-3088; fax:608-265-2340.

Present Addresses†Center for Limnology, University of Wisconsin, 680 North ParkStreet, Madison, WI 53706.‡Golder Associates, 309Exeter Rd., Unit 1, London,ON,N6L 1C1.

’ACKNOWLEDGMENT

We thank members of the Schiff and Dillon laboratories forvaluable discussions, and help in the field and lab. Special thanksgo to R. Elgood, J. Venkiteswaran, H. Broadbent, J. Findeis, andthe University of Waterloo Environmental Isotope laboratory.Research was funded by a NSERC Strategic grant to S.L.S., aDiscovery grant to P.J.D., and scholarships to H.M.B.

’REFERENCES

(1) Forster, P.; Ramaswamy, V.; Artaxo, P.; Berntsen, T.; Betts, R.;Fahey, D. W.; Haywood, J.; Lean, J.; Lowe, D. C.; Myhre, G.; Nganga, J.;Prinn, R.; Raga, G.; Schulz, M.; Van Dorland, R. Changes in atmosphericconstituents and in radiative forcing. In Climate Change 2007: ThePhysical Science Basis. Contribution of Working Group I to the FourthAssessment Report of the Intergovernmental Panel on Climate Change;Soloman, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K. B.,Tignor, M., Miller, H. L., Eds; Cambridge University Press: Cambridge,NY, 2007.(2) Galloway, J. N.; Dentener, F. J.; Capone, D. G.; Boyer, E. W.;

Howarth, R. W.; Seitzinger, S. P.; Asner, G. P.; Cleveland, C. C.; Green,P. A.; Holland, E. A.; Karl, D. M.; Michaels, A. F.; Porter, J. H.; Townsend,A. R.; Vorosmarty, C. J. Nitrogen cycles: Past, present, and future.Biogeochemistry 2004, 70 (2), 153–226.(3) Ravishankara, A. R.; Daniel, J. S.; Portmann, R.W.Nitrous Oxide

(N2O): The dominant ozone-depleting substance emitted in the 21stcentury. Science 2009, 326 (5949), 123–125.(4) IPCC. N2O Emissions from managed soils, and CO2 emissions

from lime and urea application. In 2006 IPCC Guidelines for NationalGreenhouse Gas Inventories. Volume 4 Agriculture, Forestry and Other LandUse. Chapter 11; Prepared by the National Greenhouse Gas InventoryProgram, 2006.(5) Toyoda, S.; Iwai, H.; Koba, K.; Yoshida, N. Isotopomeric analysis

of N2O dissolved in a river in the Tokyo metropolitan area. RapidCommun. Mass Spectrom. 2009, 23 (6), 809–821.

(6) Seitzinger, S. P.; Kroeze, C. Global distribution of nitrous oxideproduction and N inputs in freshwater and coastal marine ecosystems.Global Biogeochem. Cycles 1998, 12 (1), 93–113.

(7) Rahn, T.; Wahlen, M. A reassessment of the global isotopicbudget of atmospheric nitrous oxide.Global Biogeochem. Cycles 2000, 14(2), 537–543.

(8) Bernard, S.; Rockmann, T. R.; Kaiser, J.; Barnola, J. M.; Fischer,H.; Blunier, T.; Chappellaz, J. Constraints on N2O budget changes sincepre-industrial time from new firn air and ice core isotope measurements.Atmos. Chem. Phys. 2006, 6, 493–503.

(9) Alexander, R. B.; Smith, R. A.; Schwarz, G. E. Effect of streamchannel size on the delivery of nitrogen to the Gulf of Mexico. Nature2000, 403 (6771), 758–761.

(10) Inoue, H. Y.;Mook,W. G. Equilibrium and kinetic nitrogen andoxygen-isotope fractionations between dissolved and gaseous N2O.Chem. Geol. 1994, 113 (1�2), 135–148.

(11) Ontario Ministry of Natural Resources. Southern Ontario Inter-im Land Cover Dataset; Peterborough, Ontario, 2006.

(12) Ministry of the Environment. The Determination of AmmoniumNitrogen and Nitrate Plus Nitrite Nitrogen in Water and Precipitation byColourimetry; Ministry of Environment Laboratory Services BranchQuality Management Unit, Dorset, Ontario, 2001.

(13) Venkiteswaran, J. J.; Wassenaar, L. I.; Schiff, S. L. Dynamics ofdissolved oxygen isotopic ratios: A transient model to quantify primaryproduction, community respiration, and air-water exchange in aquaticecosystems. Oecologia 2007, 153 (2), 385–398.

(14) Kilpatrick, P. A.; Rathbun, R. E.; Yotsukura, N.; Parker, G. W.;Delong, L. L. Determination of stream reaeration coefficients by use oftracers. Chapter A18. In U.S. Geological Survey Techniques of Water-Resources Investigations Book 3; USGS: Washington, DC, 1987.

(15) Moog, D. B.; Jirka, G. H. Analysis of reaeration equations usingmean multiplicative error. J. Environ. Eng. 1998, 124 (2), 104–110.

(16) Thuss, S. J. Nitrous oxide production in the Grand River,Ontario, Canada: New insights from stable isotope analysis of dissolvednitrous oxide. M.Sc. Thesis, University of Waterloo, Waterloo, ON, 2008.

(17) Kaiser, J.; Rockmann, T.; Brenninkmeijer, C. A. M. Completeand accurate mass spectrometric isotope analysis of tropospheric nitrousoxide. J. Geophys. Res.-Atmos. 2003, 108 (D15), 4476; DOI: 10.1029/2003JD003613.

(18) Boontanon, N.; Ueda, S.; Kanatharana, P.; Wada, E. Intramo-lecular stable isotope ratios of N2O in the tropical swamp forest inThailand. Naturwissenschaften 2000, 87 (4), 188–192.

(19) SYSTAT Software Inc. SYSTAT 13.0; 2009.(20) Smith, R. A.; Alexander, R. B.; Schwarz, G. E. Natural back-

ground concentrations of nutrients in streams and rivers of theconterminous United States. Environ. Sci. Technol. 2003, 37 (14),3039–3047.

(21) IPCC. IPCC Guidelines for National Greenhouse Gas Inventories;Prepared by the National Greenhouse Gas Inventory Program, 2006;Volume 1 General Guidance and Reporting, Chapter 7 Precursors andIndirect emissions; p 7.1.5.

(22) Clough, T. J.; Bertram, J. E.; Sherlock, R. R.; Leonard, R. L.;Nowicki, B. L. Comparison of measured and EF5-r-derived N2O fluxesfrom a spring-fed river. Global Change Biol. 2006, 12 (2), 352–363.

(23) Well, R.; Flessa, H.; Jaradat, F.; Toyoda, S.; Yoshida, N.Measurement of isotopomer signatures ofN2O in groundwater. J. Geophys.Res.-Biogeo. 2005, 110 (G2), G02006; DOI: 10.1029/2005JG000044.

(24) Kool, D. M.; Wrage, N.; Oenema, O.; Harris, D.; Van Groenigen,J. W. The O-18 signature of biogenic nitrous oxide is determined by Oexchange with water. Rapid Commun. Mass Spectrom. 2009, 23 (1),104–108.

(25) Snider, D. M.; Schiff, S. L.; Spoelstra, J. N-15/N-14 and O-18/O-16 stable isotope ratios of nitrous oxide produced during denitrifica-tion in temperate forest soils. Geochim. Cosmochim. Acta 2009, 73 (4),877–888.

(26) Perez, T.; Trumbore, S. E.; Tyler, S. C.; Matson, P. A.; Ortiz-Monasterio, I.; Rahn, T.; Griffith, D. W. T. Identifying the agricultural



imprint on the global N2O budget using stable isotopes. J. Geophys. Res.-Atmos 2001, 106 (D9), 9869–9878.(27) Bol, R.; Toyoda, S.; Yamulki, S.; Hawkins, J. M. B.; Cardenas,

L. M.; Yoshida, N. Dual isotope and isotopomer ratios of N2O emittedfrom a temperate grassland soil after fertiliser application. Rapid Com-mun. Mass Spectrom. 2003, 17 (22), 2550–2556.(28) Kim, K. R.; Craig, H. N-15 and O-18 characteristics of nitrous-

oxide - A global perspective. Science 1993, 262 (5141), 1855–1857.(29) Holtan-Hartwig, L.; Dorsch, P.; Bakken, L. R. Low temperature

control of soil denitrifying communities: Kinetics of N2O productionand reduction. Soil Biol. Biochem. 2002, 34 (11), 1797–1806.(30) Kendall, C.; Coplen, T. B. Distribution of oxygen-18 and

deuterium in river waters across the United States. Hydrol. Process 2001,15 (7), 1363–1393.(31) Laursen, A. E.; Seitzinger, S. P. Diurnal patterns of denitrifica-

tion, oxygen consumption and nitrous oxide production in riversmeasured at the whole-reach scale. Freshwater Biol. 2004, 49 (11),1448–1458.(32) Yamulki, S.; Toyoda, S.; Yoshida, N.; Veldkamp, E.; Grant, B.;

Bol, R. Diurnal fluxes and the isotopomer ratios of N2O in a temperategrassland following urine amendment. Rapid Commun. Mass Spectrom.2001, 15 (15), 1263–1269.(33) Baulch, H. M. Nitrous oxide emissions from streams. Ph.D.

Disseration, Trent University, Peterborough, ON, 2009.(34) Beaulieu, J. J.; Shuster, W. D.; Rebholz, J. A. Nitrous oxide

emissions from a large, impounded river: The Ohio River. Environ. Sci.Technol. 2010, 44 (19), 7527–7533.(35) Hasegawa, K.; Hanaki, K.; Matsuo, T.; Hidaka, S. Nitrous oxide

from the agricultural water system contaminated with high nitrogen.Chemosphere 2000, 2, 335–345.(36) Rempel,M. NO3

� andN2O at the Strawberry creek catchment:tracing sources and processes using stable isotopes. M.Sc. Thesis,University of Waterloo, Waterloo, ON, 2008.

isotopic character of nitrous oxide emitted from streams

Documents