Journal of Geophysical Research: Biogeosciences

Measurement of isotopomer signatures of N2O in groundwater

Authors

  • R. Well,

    1. Institute of Soil Science and Forest Nutrition, University of Göttingen, Gottingen, Germany
    Search for more papers by this author
  • H. Flessa,

    1. Institute of Soil Science and Forest Nutrition, University of Göttingen, Gottingen, Germany
    Search for more papers by this author
  • F. Jaradat,

    1. Institute of Soil Science, University of Göttingen, Gottingen, Germany
    Search for more papers by this author
  • S. Toyoda,

    1. Department of Environmental Chemistry and Engineering, Tokyo Institute of Technology, Tokyo, Japan
    2. Also at SORST Project, Japan Science and Technology Corporation.
    Search for more papers by this author
  • N. Yoshida

    1. Department of Environmental Chemistry and Engineering, Tokyo Institute of Technology, Tokyo, Japan
    2. Also at SORST Project, Japan Science and Technology Corporation.
    3. Also at Department of Environmental Science and Technology, Tokyo Institute of Technology, Tokyo, Japan.
    4. Also at Frontier Collaborative Research Center, Tokyo Institute of Technology, Tokyo, Japan.
    Search for more papers by this author

Abstract

[1] N2O flux from aquifers caused by leaching of agricultural N is a poorly known component of the global anthropogenic source of this greenhouse gas. We measured isotopomer signatures of N2O (intramolecular distribution of 15N as well as conventional nitrogen and oxygen isotope ratios) in the interface region between shallow groundwater and the atmosphere in order to evaluate this technique for determining fluxes, production, reduction and the isotopomer fingerprint of N2O originating from the saturated zone. 15N-site preference (difference in δ15N between central and peripheral N-position) measured in the shallow groundwater of a hydromorphic soil (29 to 81‰) was distinctly larger compared to surface emitted N2O. Local and global isotopic budget calculations confirmed that the groundwater derived N2O flux of the experimental site was low compared to measured surface fluxes and suggest that 15N-site preference might be useful for validating global estimates of groundwater emitted N2O.

1. Introduction

[2] N2O is an atmospheric trace gas contributing to global warming and stratospheric ozone depletion. Its major sources are nitrification and denitrification in soils and aquatic systems. Despite extensive studies on N2O fluxes and turnover processes in various environments the global N2O budget is still poorly constrained. Stable isotopic signatures of N2O have been used to study sink and source processes of N2O in terrestrial and aquatic systems and in the atmosphere [Stein and Yung, 2003] and to constrain the atmospheric N2O budget [Röckmann et al., 2003]. In the denitrification pathway, N2O occurs as an obligate intermediate during the reduction of NO3 to N2, whereas it is a by-product of nitrification. N2O production of both nitrification and denitrification yields N2O which is isotopically light in relation to its precursors, whereas reduction during denitrification results in an enrichment of 15N and 18O in the residual N2O [Barford et al., 1999]. 15N depleted N2O found in aerobic aquifers and in oceans has been attributed to nitrification [Ostrom et al., 2000]. 15N enrichment of N2O in lakes [Wada and Ueda, 1996; Boontanon et al., 2000], oceans [Naqvi et al., 1998; Popp et al., 2002] and emitted from soils [Tilsner et al., 2003; Wrage et al., 2004] has been explained by N2O reduction during denitrification.

[3] Recently, specific analysis of the terminal and central N-position of the linear N2O molecule was realized [Toyoda and Yoshida, 1999; Brenninkmeijer and Röckmann, 1999] and was used to characterize N2O from various environments. Theoretically, N2O production during nitrification and denitrification can cause 15N accumulation at both N-sites, depending on the type of NO-reductase catalyzing this reaction [Stein and Yung, 2003; Schmidt et al., 2004]. This was also demonstrated experimentally with various pure cultures of nitrifiers [Sutka et al., 2003, 2004a] and denitrifiers [Sutka et al., 2004b; Toyoda et al., 2005]. Because the reduction step consists of the cleavage of NO-bonds, it is expected to cause 15N-accumulation at the central N-position of the residual N2O [Yoshida and Toyoda, 2000; Toyoda et al., 2002; Popp et al., 2002; Schmidt et al., 2004]. In contrast to 18O and average 15N, the difference between central and peripheral 15N enrichment is considered to be independent of the isotopic signature of the precursor [Popp et al., 2002; Toyoda et al., 2002] and thus supplies clear process information even if isotopic signatures of additional N-species are lacking. Measurement of site-specific 15N signatures in soil emitted N2O [Pérez et al., 2001; Yamulki et al., 2001; Bol et al., 2003, 2004] resulted mostly in enrichment of the central N-position. This pattern is even more pronounced in stratospheric [Yoshida and Toyoda, 2000] and oceanic samples [Toyoda et al., 2002; Popp et al., 2002].

[4] In denitrifying aquifers, NO3 input from agriculture is partially or completely reduced [Böhlke, 2002]. N2O produced under these conditions can be transported to the atmosphere via upward diffusion or groundwater discharge to wells, springs and streams [Mühlherr and Hiscock, 1998; Heincke and Kaupenjohann, 1999]. From N2O accumulation in the saturated zone it was estimated that groundwater can be a significant source of atmospheric N2O [Mühlherr and Hiscock, 1998; Mosier et al., 1998]. More recently, this has been questioned by theoretical considerations [Groffman et al., 2000; Nevison, 2000] as well as by direct measurement of low N2O fluxes from drainage systems [Reay et al., 2003]. However, experimental evidence suitable for generalizing N2O-fluxes from the saturated zone and clarifying the controlling process kinetics is still lacking [Groffman et al., 2002; Well, 2002].

[5] Process kinetics of N2O production and reduction by denitrification have been measured in closed laboratory systems [Holtan-Hartwig et al., 2002] in order to investigate the control of the N2O yield of denitrification and to predict N2O emission from various environments. Groundwater N2O concentration and upward diffusive flux had been measured and simulated using a model of N2O production, reduction and diffusion in hydromorphic soils, where the groundwater level was at the soil surface [Well et al., 2001]. The poor fit of the model was attributed to insufficient knowledge of the process kinetics. Principally, these approaches can be extended by measurement of N2O isotopomer signatures in order to investigate N2O cycling in the saturated zone more successfully.

[6] In a previous study, denitrification, N2O concentration and N2O production were measured near the groundwater surface of a sandy aquifer using in situ and laboratory techniques [Well et al., 2003a, 2005]. High N2O concentrations of up to 100 μM could be explained by denitrification and suggested significant diffusive emission to the unsaturated zone and thus potentially to the atmosphere. The aim of the present study was to evaluate the use of isotopomer signatures of N2O as a tool for determining fluxes, production, reduction and the isotopomer fingerprint of N2O in groundwater at a local and global scale. This was conducted by measuring fluxes, production, reduction and isotopomer signatures of N2O in the interface region between groundwater and atmosphere in a hydromorphic soil.

2. Materials and Methods

2.1. Experimental Site

[7] Experiments were conducted in the shallow groundwater of a hydromorphic soil under agricultural management located in the northwest German lowland 30 km southwest of Oldenburg, Lower Saxony, Germany. The soil is a deep-ploughed fen (“German sand-mix culture”) over glacio-fluvial sand with groundwater level at 95 to 155 cm depth. The soil profile consists of a recent plough layer (0–30 cm), a deep-ploughed subsoil layer (30–80 cm) consisting of tilted layers of peat and sand, and an undisturbed sand layer below 80 cm (Figure 1). The site was an experimental plot (4.5 m × 16 m) of a fertilizer experiment cropped with silage corn and fertilized with 240 kg N ha−1 yr−1 (calcium ammonium nitrate) where nitrate leaching had been monitored. Annual leaching rates were between 45 and 220 kg N ha−1 yr−1 [Landwehr et al., 2001]. Subsoil properties are given in Table 1. Organic C and total N decreased between 100 and 200 cm from 6.8 to 1.6 g kg−1 and from 0.13 to 0.05 g kg−1, respectively, whereas pH slightly increased (4.4–4.7) and texture (sand) was constant with depth. Denitrification rates in the shallow groundwater (160 to 220 cm depth) as determined in laboratory and in situ studies were 104 to 152 μg N kg−1 d−1 and 14 to 119 μg N kg−1 d−1, respectively [Well et al., 2003b, 2005]. The field was drained by a ditch running approximately at 50 m distance from the sampling site.

Figure 1.

Study site and experimental set-up for determining fluxes and isotopomer signatures of N2O from the groundwater and the unsaturated zone. The soil is a deep-ploughed fen (“German sand-mix culture”) over glacio-fluvial sand with groundwater level (GWL) between 95 cm (December sampling, GWL12) and 155 cm depth (July sampling, GWL07). The profile consists of a recent tillage horizon (Ap, 0–30 cm), a deep-ploughed subsoil horizon with tilted bars of peat and sand (R, 30–80 cm), and an undisturbed sand layer with gleyic properties (Go and Gr-horizons, >80 cm depth). The N2O concentration gradients within the profile are used to calculate vertical diffusive fluxes within the groundwater (Fgw), from the groundwater across the groundwater surface including the capillary fringe (CF) to the unsaturated zone (Fgw,uz) and within the unsaturated zone (Fuz). The surface flux into the chamber atmosphere (Fch) is determined from N2O accumulation.

Table 1. Organic C (Corg), Total N (Nt), Texture, pH, and Denitrification Potential (DP) in the Experimental Site Between 1 and 2 m Deptha
Depth, cmCorg, g kg−1Nt, g kg−1Sand, %Silt, %Clay, %pHCaCl2DPlab,b μg N kg−1 d−1DPin situ,c μg N kg−1 d−1N
  • a

    Means ± standard deviation.

  • b

    DP measured in the laboratory by anaerobic incubation of soil slurries.

  • c

    DP measured in situ using a 15N injection-extraction technique [Well et al., 2003b].

100–1206.87 ± 1.580.13 ± 0.0195.71.72.64.4  4
120–1404.94 ± 1.180.11 ± 0.0496.01.12.94.5  4
140–1603.66 ± 1.920.09 ± 0.0595.91.52.74.8152 ± 5259 ± 1224
160–1801.89 ± 0. 760.05 ± 0.0296.41.52.04.7121 ± 4514 ± 144
180–2001.640.0595.71.52.84.7104 ± 66 2

2.2. Sampling

[8] Collecting gas and water samples for measuring concentration gradients and surface emission of N2O had been conducted several times during previous studies [Well et al., 2003b, 2005]. On 7 July and 6 December 2001 this was repeated and a total of 11 samples were selected for additional analysis of N2O isotopomers.

2.2.1. Water Sampling

[9] Probes for collecting groundwater samples from a defined depth described in detail earlier [Well et al., 2003b] were installed at varying depths depending on the groundwater level (July sampling: 160, 180, 220 cm; December sampling: 115, 135, 155 and 175 cm) with one probe per depth (Figure 1). The probes were positioned in a row in the center area of the plot with approximately 1 m distance between adjacent probes. Water samples from the drainage ditch were collected through a submersed groundwater probe. For isotopomer analysis of dissolved N2O, water was pumped through the probes into 115-mL serum bottles using a peristaltic pump. The bottles were then immediately sealed without trapping air bubbles using butyl rubber septa (Altmann, Holzkirchen, Germany) and crimp caps. The samples were stabilized with 0.1 mL of saturated HgCl2 solution. For N2O analysis by gas chromatography, samples were collected using partially evacuated (3 kPa) bottles as described earlier [Well et al., 2003b]. Thus approximately 70% of the bottle volume was filled with water leaving a headspace with atmospheric air in the residual volume. In the laboratory, gas samples were transferred from the headspace to evacuated septum-capped Exetainers™ (12 mL, Labco, Wycombe, U.K.) using a syringe after equilibrating gas and liquid phase at constant temperature (25°C).

2.2.2. Collecting Gas Samples From the Unsaturated Zone and From Surface Emission

[10] Gas probes for sampling the unsaturated zone were installed with two replicates at 60, 90 and 120 cm depth when the groundwater table was low (July) or at 20, 50 and 80 cm depth during high groundwater table (December, Figure 1). The probes consisted of a perforated brass cup (6 mm ID × 5 cm length), connected to a gas sampling port at the top (brass fitting with rubber septum) by 2-mm ID stainless steel tubing. Fifteen milliliters of soil gas was collected from each probe with a plastic syringe and was injected into evacuated Exetainers™. The probes were positioned in a row in the center area of the plot with approximately 1 m distance between adjacent probes and approximately 3 m distance from the row of groundwater probes. Flux chambers (25 cm diameter × 8 cm height) were used for collection of surface emitted gases. Ten-centimeter-long PVC-rings were pushed into the soil 5 cm deep. During gas collection, chambers were fitted to the rings with a rubber gasket. Gas samples were collected by connecting evacuated septum-capped 115-mL serum bottles to the gas sampling ports of the chambers using a double needle. Sampling time was 30, 60 and 90 min after closure. Four replicate chambers were placed on spots located between corn rows and evenly distributed over the plot.

2.3. Laboratory Experiment of N2O Production and Reduction

[11] Soil samples from two depths of the saturated zone (140 to 160 and 180 to 200 cm) of the experimental site were incubated as slurries under denitrifying conditions in a closed system with four replicates per depth in order to determine the time course of N2O concentrations and isotopomer signatures under process conditions similar to the shallow groundwater of the experimental soil. One-hundred-twenty-five grams of aquifer material were supplemented with 100 mL of KNO3 solution (30 mg N L−1) and incubated anaerobically at 10°C in 1-L transfusion bottles sealed with rubber septa and aluminum screw caps. After closure, the bottles were flushed with N2. Prior to each sampling, dissolved and headspace N2O were equilibrated by vigorous shaking. Samples were then collected using syringes and transferred to evacuated vials. Depending on the expected concentration, 115 mL or 12 mL of gas was collected and stored for further analysis. An amount of N2 equivalent to the sample was injected after sampling in order to re-establish atmospheric pressure. Sampling was repeated at varying time intervals until N2O concentration approached zero.

2.4. Analytical Techniques

[12] N2O was analyzed using a gas chromatograph equipped with an electron capture detector and an autosampler as described earlier [Well et al., 2003b]. Dissolved N2O was calculated from the headspace gas concentrations of the water samples using the Bunsen absorption coefficient of N2O [Well and Myrold, 1999] and taking into account atmospheric N2O contained in the partially evacuated bottles before sampling.

[13] Isotopomer analysis of the field samples was conducted in a laboratory of the Tokyo Institute of Technology (Japan). Dissolved N2O was extracted and introduced into the preconcentration/gas chromatograph/isotope ratio mass spectrometer system [Toyoda et al., 2002]. The isotopomer ratios of 15Rbulk, 18R, and 15Rα were determined and 15Rβ was obtained by the relationship of 15Rbulk = (15Rα + 15Rβ)/2, where 15Rα = [14N15N16O]/[14N14N16O], 15Rβ = [15N14N16O]/[14N14N16O], 18R = [14N14N18O]/[14N14N16O] [Toyoda and Yoshida, 1999]. Isotopomer ratios of a sample (Rsample) are expressed as ‰ deviation from 15N/14N and 18O/16O ratios of the standard materials (Rstd), atmospheric N2 and standard mean ocean water (SMOW), respectively: δX = (Rsample/Rstd − 1) × 1000, where X = 15Nbulk, 15Nα, 15Nβ, or 18O. Typical analytical precision is 0.6, 0.9, 1.5, and 0.9‰ for δ15Nbulk, δ15Nα, δ15Nβ, and δ18O, respectively. The difference between the isotopomer ratios of N (δ15Nα − δ15Nβ) is referred to as 15N-site preference (SP, in ‰).

[14] N2O of the laboratory experiments was analyzed in a similar manner at the University of Göttingen using a Precon-DeltaS instrumentation (Thermo-Finnigan, Bremen). Gas samples were collected from the headspace of the incubation system as described above.

[15] To measure δ15N of added KNO3 in the laboratory experiment, the pure salt was weighed into tin capsules and analyzed using an elemental analyzer coupled to an isotope ratio mass spectrometer system (Thermo-Finnigan, Bremen). NO3 concentration of the water samples was analyzed by steam distillation [Bremner and Keeney, 1965].

2.5. Estimation of the Movement of Gaseous and Dissolved N-Species

[16] To estimate the vertical N2O flux at the site, diffusive N2O flux from the groundwater to the unsaturated zone and from the unsaturated zone to the soil surface (Fdiff) was calculated from concentration gradients within the soil profile [Burton and Beauchamp, 1994],

equation image

where Da and dCN2O/dx are apparent gas diffusion coefficient and vertical N2O concentration gradient, respectively. Da in the unsaturated zone was determined in 100-mL soil cores (5 cm diameter) in the laboratory using diffusion chambers [Frede, 1986]. In the saturated zone, Da was calculated from porosity (E) and tortuosity (T) and the N2O diffusion coefficient in water (D0 = 2.3 × 10−5 cm2 s−1): Da = D0 × E/T [Frede, 1986], where E was calculated from bulk density and T was assumed 1.3, which is a suggested default value for sand [Affek et al., 1998]. To calculate fluxes across the groundwater surface (Fsat/unsat, subscripts sat and unsat denote saturated and unsaturated zone, respectively) the following approximations were made. Because Da,unsat ≫ Da,sat, the diffusive resistance of the unsaturated part of the diffusion path can be neglected and N2O concentration within the unsaturated part is thus considered constant. Consequently, the concentration difference within the saturated path (dcsat) is considered equal to the concentration difference between sampling spots above and below the groundwater surface (dcsat,unsat). The length of the saturated diffusion path (dxsat) consists of the groundwater and the capillary fringe, where the thickness of the latter can be estimated from empirical tables (10 cm for medium sand [Kuntze et al., 1985]). Thus the flux across the groundwater surface can be calculated using a modification of equation (1): Fsat/unsat = Da,sat × dcsat,unsat/dxsat.

[17] Residence time of the groundwater within reactive zones is of interest because it can be used in conjunction with process rates in order to estimate the progress of denitrification, i.e., the concentration ratio of reduced NO3 to initial NO3. Estimation of groundwater age at the probe positions was conducted using the following approximation, which assumes that recharge occurs uniformly across the surface, that the change in groundwater level is small compared with aquifer thickness, and that hydraulic properties do not vary systematically with depth [Böhlke, 2002],

equation image

where n is porosity, Z is the total thickness of the saturated zone, ti is the age of a groundwater parcel at depth zi below the water table and r is the recharge rate. r was estimated from seepage rates continuously measured at a nearby monitoring site provided by W. Schäfer (unpublished data, 2004), and n was calculated from bulk density at 140 cm depth (1.6) and assuming mineral density of 2.65. Z is 25 to 150 m (G. Josupeit et al., unpublished data, 1991). Here zi was approximated by the distance between probe depth (115 to 220 cm below surface) and mean of summer and winter groundwater levels (125 cm below surface, Figure 1). Because the upper probes were located within the range of groundwater level fluctuation, calculated ti is not exact, but is considered as a semi-quantitative ranking of the residence time. For the probe position above mean groundwater level (115 cm depth), ti was estimated from the duration of saturation at this depth, which was calculated from the amplitude of the groundwater table between July and December (600 mm) and the time interval of 60 days between beginning of the seepage period (6 October; W. Schäfer, unpublished data, 2004) and sampling (6 December). The ratio of these numbers gives a rate of groundwater table rising of 10 mm d−1 if a constant rate is assumed. Then ti results by dividing the rising interval between 115 to 95 cm depth by the rising rate giving 200 mm/[10 mm d−1] = 20 d.

3. Results

3.1. Field Study

3.1.1. Variability of Isotopomer Signatures

[18] N2O in gas and water samples of the field study contained both atmospheric and process derived N2O. In the following, isotopomer signatures of the latter fraction are reported, which were calculated using a solution of a common mixing equation, where concentrations and isotopomer signatures of the ambient air and of the samples are the known variables.

[19] The average δ15N signal of N2O (δ15Nbulk) was slightly positive in the drainage ditch samples (4 and 8‰) and negative in the soil emission sample and in the unsaturated zone (−11 and −32‰, Table 2, Figure 2). In the groundwater, δ15Nbulk ranged widely between negative and positive values (−42 to +86‰). Site preference (SP = δ15Nα − δ15Nβ) was always positive and highly variable, ranging between 2 and 81‰. SP of the other samples (drainage ditch, unsaturated zone and surface emission) ranged between 2 and 17‰ and was thus less variable and lower compared to the groundwater samples. The δ18O ranged between 21 and 84‰ and exhibited a positive, highly significant relationship with SP (Table 2, Figure 2b), with δ18O = 0.77 × (δ15Nα − δ15Nβ) + 10.3 (R2 = 0.98).

Figure 2.

Comparison of isotopomer signatures measured in this study with published data from soil emission studies and of tropospheric average (1, Bol et al. [2003]; 2, Bol et al. [2004]; 3, Pérez et al. [2001]; 4, Yoshida and Toyoda [2000]; 5, Well et al. [2004]; 6, Yamulki et al. [2001]). (a) Site preference and δ15Nbulk versus air-N2 (‰). (b) Site preference and δ18O versus SMOW (‰).

Table 2. Groundwater NO3 Concentrations, N2O Concentrations of Water and Gas Samples, N2O Fluxes, and Isotopomer Signatures of Process-Derived N2Oa
DateSampleTypebDepth, cmNO3, mg N L−1N2O, mg N L−1Flux,c g N ha−1 d−1δ15Nα, ‰δ15Nβ, ‰δ15Nbulk, ‰δ15Nα − δ15Nα (SP). ‰δ18O, ‰ti,d days
  • a

    Isotopic composition of oxygen (δ18O) and nitrogen (central N-position: δ15Nα, peripheral N-position: δ15Nβ, average: δ15Nbulk, site preference [SP] = δ15Nα − δ15Nα); n.a., not applicable; n.d., no data; surface and subsurface fluxes are means of four replicates (± standard deviation) and two replicates (difference in brackets), respectively; all other data are individual observations; variation coefficients of groundwater N2O and NO3 concentrations during earlier sampling events (n = 4 [Well et al., 2003b]) ranged from 97 to 236% and from 9.7 to 142%, respectively. Isotopomer signatures of soil derived N2O were calculated using concentrations and signatures of the samples and of ambient air, see text.

  • b

    Ch, UZ, GW, and Di denote chamber gas, unsaturated zone, groundwater, and drainage ditch, respectively.

  • c

    Calculated from accumulation (chamber) or concentration gradients within the soil profile (see equation (1) and section 3.1.2).

  • d

    Groundwater residence time within the flow path preceding the sample location calculated using equation (2).

July 2001Chn.a.n.a.0.090 ± 0.0711006−7.3−14.3−11.36.732.1n.a.
July 2001UZ60n.d0.564 (0.288)       
July 2001UZ90n.d0.755 (0.161)73−28.6−35.9−32.37.320.7n.a.
July 2001UZ120n.d0.219 (0.395)−58      
July 2001GW1600.50.094−1126.845.586.181.389.8148
July 2001GW18025.91.504410.7−53.1−21.263.766.9235
July 2001GW2000.90.011−458.8−17.220.876.083.6324
Dec 2001UZ80 0.028 (0.005)      n.a.
Dec 2001GW1152.50.4752−19.0−55.9−37.536.937.120
Dec 2001GW1351.91.3952−26.8−56.3−41.629.523.442
Dec 2001GW15527.32.7233−1.2−49.5−25.448.348.2127
Dec 2001GW17521.90.107−712.2−51.2−19.563.476.7213
Dec 2001Din.a.1.60.005n.a.4.52.43.52.143.3n.a.
Dec. 01Din.a.0.40.018n.a.16.6−1.37.618.046.6n.a.

3.1.2. Contribution of Groundwater N2O to Surface Emission

[20] Vertical N2O fluxes at the soil surface and within the profile of the field site were determined using closed chambers at the soil surface and vertical N2O concentration gradients on 8 March 2001 and on the dates of isotopomer sampling (July and December), giving the following results (in Table 2, only the results of the July sampling are given): Surface fluxes (n = 4) were 23 ± 17, 1005 ± 793 and 5 ± 8 g N ha−1 d−1, respectively; mean concentrations of the unsaturated zone were 31 ± 36 (n = 8), 453 ± 367 (n = 6) and 14 ± 9 μg N L−1 soil air (n = 6), respectively; and mean concentrations in the saturated zone were 220 ± 207 (n = 4), 432 ± 679 (n = 3) and 835 ± 748 μg N L−1 groundwater (n = 4), respectively. It can be seen that groundwater N2O concentrations were constantly high, whereas the surface fluxes and concentrations in the unsaturated zone exhibited a seasonal pattern with low values in March and December and high values in July. The diffusive fluxes within the unsaturated zone (−58 to 73 g N ha−1 d−1) are much larger than the fluxes in the saturated zone, although the concentration gradients are in the same order of magnitude. The surface flux calculated from chamber gas concentrations was much larger (1005 g N ha−1 d−1) than the subsurface fluxes.

[21] Diffusive fluxes within the saturated zone and across the groundwater surface in July and December were 1.2 to 3.1 g N ha−1 d−1 (Table 2), i.e., 2 to 3 orders of magnitude lower compared to fluxes within the unsaturated zone or at the soil surface obtained for the July sampling, but were in the same order of magnitude as the December surface fluxes (5 g N ha−1 d−1; see above). Despite high N2O concentration, the N2O flux from the saturated zone was low and thus its contribution to the surface flux was negligible. In December, the flux from the groundwater may have significantly contributed to the low surface flux. However, the source data are not accurate enough for quantifying the fraction of groundwater derived flux exactly.

[22] In July, isotopomer signatures of all compartments (groundwater, unsaturated zone surface flux, ditch) had been measured, whereas only groundwater and ditchwater were analyzed at the December sampling. Site preference (SP) was a suitable variable for evaluating the upward N2O flux from the saturated zone of the experimental site because (1) the groundwater values of SP were much higher than the values of the other samples and (2) potential reduction of N2O during diffusive transport to the surface would possibly increase but not lessen SP (see above [Yoshida and Toyoda, 2000; Schmidt et al., 2004]). The fraction of groundwater derived N2O in surface emitted N2O of the July sampling event can be estimated using the mixing calculation

equation image

where F is N2O flux and subscripts ch, gw and uz denote chamber, groundwater and unsaturated soil, respectively. SP of the groundwater samples (median of 76‰), of N2O produced in the unsaturated soil layers (assumed to be within the range of reported values of surface emission, i.e., −5 to +20‰; see Figure 2), the mixture of both sources (process derived N2O in the chamber gas = 6.8‰), and the surface flux (Fgw + Fuz) derived from the chamber which was analyzed for isotopomers (363 g N ha−1 d−1) are used as known quantities in this calculation. The resulting fraction of groundwater derived N2O [Fgw/(Fgw + Fuz)] ranges between 0 (SPuz ≥ 6.8‰) and 0.14 (SPuz = −5‰).

3.1.3. NO3 and N2O Concentrations

[23] NO3 concentrations in the groundwater varied between 0.5 and 27.3 mg N L−1. The highest values were within the range of seepage water concentrations of this site (10 to 50 mg N L−1 [Landwehr et al., 2001]). N2O concentrations in the groundwater were between 0.005 and 2.7 mg N L−1 and the N2O-to-NO3-ratios were between 0.005 and 0.73.

3.2. Laboratory Study

[24] N2O concentrations during anaerobic laboratory incubation of sample 1 from 140 to 160 cm depth (Figure 3a) almost linearly increased to a maximum of 0.5 mg N kg−1 at 30 days after T0 and then decreased until day 90 reaching 0.05 mg N kg−1 at the final sampling date.

Figure 3.

Time course of concentrations and isotopomer signatures (δ15Nbulk versus air-N2, site preference [SP] and δ18O versus SMOW) of N2O during anaerobic incubation of two aquifer samples supplemented with NO3 solution. (a) Sample 1, 140 to 160 cm depth. (b) Sample 2, 180 to 200 cm depth.

[25] The δ15Nbulk starts with a value approximately 30‰ below initial δ15N of NO3 (3‰) and then continuously increases during the entire experiment. SP initially drops from 24 to 10‰ during days 0 to 14, stays at this low level until day 30, and then continuously increases, reaching 39‰ at day 70. δ18O is relatively constant at approximately 37‰ until day 30 and then steadily increases reaching 73.4‰ at day 70.

[26] In sample 2 (180 to 200 cm depth, Figure 3b), N2O concentration at the final sampling was still high (0.4 mg N kg−1), indicating that reaction progress and thus denitrification activity was lower compared to sample 1. The temporal trends of δ15Nbulk and δ18O of both samples were similar. This was also the case for SP of the initial period when SP was measured in both samples. SP values of sample 2 after day 40 are missing owing to instrumental failure. The consistent trends of both samples demonstrate that the fractionation pattern of sample 1 was not a singularity.

[27] Maximum N2O concentrations per mass of soil (sample 1: 0.41 mg N2O-N kg−1 soil; sample 2: 0.51 mg N2O-N kg−1 soil) were in the same order of magnitude as in the field study (Table 2: 2.7 mg N L−1, giving 0.7 mg N kg−1 at bulk density of 1.6 and porosity of 0.4). However, owing to the large headspace (900 mL) and the lower solid-to-liquid mass ratio of the laboratory experiment (100 g dry soil per 100 mL), the liquid phase concentration was much lower in the laboratory experiments (0.05 and 0.06 mg N L−1).

4. Discussion

4.1. N2O Concentrations

[28] N2O concentrations in the groundwater of the field site (0.005 to 2.7 mg N L−1) and the N2O-to-NO3-ratios (0.005 to 0.73) were relatively high compared to other aquifers [McMahon et al., 2000; Mühlherr and Hiscock, 1998; Mosier et al., 1998] since maximum values of our study were by approximately one order of magnitude higher. The finding of high denitrification potential of our earlier studies [Well et al., 2003b, 2005] suggests that denitrification was the dominating N2O source process. This was also confirmed by N2O accumulation during anaerobic incubation of aquifer samples from the same site. Maximum concentrations of N2O per mass of soil was comparable to the field observations. This clearly demonstrates that the magnitude of observed N2O accumulation can be explained by denitrification in the reactive material near the groundwater surface.

[29] N2O concentrations during anaerobic laboratory incubation of sample 1 from 140 to 160 cm depth (Figure 3a) almost linearly increased to a maximum of 0.5 mg N kg−1 at 30 days after T0 and then decreased until day 90 reaching 0.05 mg N kg−1 at the final sampling date. Earlier studies using similar closed incubation systems [Holtan-Hartwig et al., 2000, 2002; McConnaughey et al., 1985] demonstrated that incubation times needed for complete reduction were almost identical for NO3 and N2O. From the nearly complete N2O reduction at day 90 of our experiment it can be thus assumed that the NO3 pool was probably almost exhausted; that is, reaction progress was close to unity. Thus the mean overall denitrification rate, i.e., rate of NO3 reduction or N2O + N2 production, can be roughly estimated by dividing N-input (24 mg N kg−1) by incubation time (90 days) giving 0.27 mg N kg−1 d−1. This is near the upper end of in situ denitrification rates measured earlier at the same depth (0.05 ± 0.15 mg N kg−1 d−1). The time course of concentrations of total denitrification products (N2O + N2) can be roughly estimated by multiplying these rates with residence time (data not shown). N2O concentrations observed in the laboratory and in the field were always much lower than estimated N2O + N2, demonstrating that a large fraction of produced N2O was reduced to N2.

4.2. Comparison of Isotopic Fingerprints of N2O

[30] The δ15Nbulk emitted from fertilized soils is predominantly negative, which is explained by 15N depletion during N2O production by nitrification and denitrification [Pérez et al., 2000, 2001; Stein and Yung, 2003; Bol et al., 2003]. Positive δ15Nbulk was observed in relatively few cases and was attributed to ongoing N2O reduction during denitrification [Wrage et al., 2004]. The large range of δ15Nbulk between negative and positive values (−42 to +86‰) of our groundwater data thus suggests that process dynamics were highly variable and both production and reduction occurred.

[31] The range of site preference (SP = δ15Nα − δ15Nβ = 2 and 81‰) was far more variable than SP of soil derived N2O that had been reported so far (−2 to 14‰ [Yamulki et al., 2001; Pérez et al., 2001; Bol et al., 2003, 2004]). SP was >29‰ in all of our groundwater samples, but was 2 to 17‰ in the other samples (drainage ditch, unsaturated zone and surface emission), which is close to the range of the earlier soil emission data. SP of the laboratory experiment was somewhat lower (10 to 39‰, Figure 2) compared to the field data but was relatively high compared to soil emission data.

[32] High SP was also found in a recent set of groundwater samples collected between 3 and 40 m below the groundwater surface from two denitrifying sandy aquifers in northwest Germany (mean, median, minimum, maximum, 25th percentile, 75th percentile of SP were 37.6, 41.5, 6.0. 68.9, 20.5, 48.3‰, respectively, n = 18; R. Well et al., unpublished data, 2004). SP up to 61‰ had also been reported from a laboratory study with submerged tidal sediment [Bol et al., 2004]. In oceanic environments, maximum SP was also high (35‰ [Toyoda et al., 2002], 25‰ [Popp et al., 2002]). From the observation of high SP in the laboratory and field experiments as well as from the various reports of high SP in aquatic systems with presumed reducing conditions, we hypothesize that high SP is characteristic for denitrifying aquatic environments, where N2O reduction is favored because N2O fluxes are restricted by low gas diffusivity (see section 3.1).

[33] Because the isotopic fingerprints of the ditch samples (δ15Nbulk = 3.5 to 7.6‰, SP = 2.1 to 18‰, δ18O = 43.3 to 46.6‰) do not resemble any of the signatures of the other samples, it can be concluded that the major part of the N2O in the ditch came from sources with signatures different from the experimental site and that isotopic signatures within the ditch catchment were highly variable. A complete interpretation of the ditch signatures would require characterizing the isotopic fingerprints of the N2O sources of the entire catchment, which was beyond the scope of our study.

4.3. Fractionation Pattern

4.3.1. The δ15Nbulk and δ18O

[34] To interpret the observed isotopic signatures it is necessary to consider the control of isotope effects that cause specific signatures in the sequential reaction members of the denitrification pathway, i.e., NO3, NO2, NO, N2O, N2. For NO3it is known that isotope fractionation during movement into or out of cells is small or negligible [Bryan et al., 1983] and the magnitude of fractionation during denitrification largely depends on the relative rates of uptake and reduction within the cell [Ostrom et al., 2002]; the relative importance of these rates depends on the relative concentrations of electron acceptors (NO3) within and external to the cell as well as on the factors affecting the reduction rate, i.e., concentrations of electron donors (organic C) and enzymes (NO3reductase). If NO3is non-limited in relation to the reduction capacity, then NO3reduction is incomplete, resulting in isotopic enrichment of NO3leaving the cells. In contrast, when the supply of NO3is less than its reduction rate, then little or no NO3escapes the cells to express an isotope effect. Thus a substantial isotope effect results if NO3supply is high in relation to reduction capacity of the system, whereas the effect is low or negligible if the opposite is the case. Principally, the same fractionation control applies to the other N species subject to reduction during the further progress of denitrification, i.e., NO2, NO, and N2O [Barford et al., 1999]. For these species, however, the situation is even more complex because the concentration of electron acceptors of each reaction step of denitrification depends on the rate of the previous step. Furthermore, some microbes are lacking enzymes of some of the reduction steps [Stein and Yung, 2003] which implies that transport within denitrifying species is a precondition of further reduction. The isotopic signature of N2O as an intermediate is thus resulting from two processes, i.e., production during NO reduction and consumption during N2O reduction to N2. N2O that is instantaneously produced is depleted in 18O and 15Nbulk in relation to its precursor. The δ18O is also affected by oxygen exchange with water, where the exchange ratio varies among microbial species [Casciotti et al., 2002]. Reduction to N2 causes increasing δ15Nbulk in the residual N2O [Barford et al., 1999].

[35] To explain the pattern of observed isotopomer signatures, the temporal and spatial dynamics of the investigated system also need to be considered. It is assumed that each groundwater parcel represents a nearly closed system; that is, diffusion during groundwater flow is small compared to production and consumption of N-gases. Within a denitrifying layer, the concentration of NO3 decreases with residence time or with depth below the groundwater surface (see equation (2)), where the completeness of NO3 reduction can be described by the reaction progress (RP = [denitrified NO3-N]/[NO3-N of recharge water]; [Böhlke, 2002]). Close to the groundwater surface, δ15NO3 can be assumed to be close to the NO3 signature of the unsaturated zone, because residence time and thus RP are low. The δ15Nbulk of instantaneously produced N2O (δ15Nibulk) is depleted with respect to the NO3 signature. With ongoing RP, both δ15NO3 and δ15Nibulk must increase. δ15Nbulk of residual N2O is further increased by N2O reduction to N2. The δ15NO3 of seepage water can be assumed to be close to zero or, due to denitrification in the unsaturated zone, slightly positive [Kendall and Aravena, 1999], i.e., not more negative than the NO3 source. Thus δ15Nbulk should be initially negative and then continuously increase during the reaction progress. This pattern is roughly reflected by the groundwater data (Table 2), which exhibited lowest δ15Nbulk for the samples with lowest residence time (December samples at 115 and 135 cm depth: δ15Nbulk = −38 to −42‰) and a large range of δ15Nbulk at a higher level (−21 to +86‰) for the samples of longer residence time. The δ18O exhibit a similar pattern, i.e., lowest values for the 115 and 135 cm samples (23 to 37‰), and higher values for the other samples (48 to 77‰). The correlation between residence time and δ18O was significant (R2 = 0.62, P > 0.95). One sample (July, 160 cm, ti = 148 days) strongly deviates from the temporal trend of δ15Nbulk and δ18O since both values are highest among the data set (86 and 89‰, respectively) while ti is only 148 days, i.e., less than half of the maximum ti (324 days). This anomaly could be explained by a “hot spot“ of elevated denitrification activity within the flowpath preceding this sample: Isotopic signatures increase with reaction progress (RP, see above) which is the product of ti and denitrification rate. Hence, for a given ti, RP is directly related to denitrification rate. When this deviating sample is excluded, closer correlations between ti and δ15Nbulk (R2 = 0.81, P > 0.95) and between ti and δ18O (R2 = 0.89, P > 0.99) are obtained.

4.3.2. Site-Specific N Fractionation

[36] Owing to the asymmetry of the N2O molecule, fractionation at the central (15Nα) and peripheral N-position (15Nβ) is not equal. It is generally accepted that the stability of the Nα-O bond governs isotope fractionation during N2O reduction resulting in increasing site preference (SP) in the residual N2O [Toyoda et al., 2002; Stein and Yung, 2003; Schmidt et al., 2004]. This has recently been confirmed in laboratory experiments with surface soils [Ostrom et al., 2004]. The N2O production step is more complex: The formation of the N-N-bond during NO-reduction to N2O can proceed via sequential or parallel binding of two NO-molecules to the NO reductase, depending on the type of this enzyme. There is disagreement on the extent of associated site-specific N fractionation, which is reflected by SPiP (SP of N2O that is instantaneously produced, i.e., that is not affected by partial reduction). Stein and Yung [2003] propose positive and low SPiP during sequential and parallel binding mechanisms, respectively, whereas others assume positive SPiP for the parallel binding [Toyoda et al., 2002; Schmidt et al., 2004], and Schmidt et al. [2004] proposed variable SPiP for the sequential mechanism. The variability of SPiP among denitrifiers was recently confirmed experimentally, giving SPiP close to zero for P. chlororaphis and P. aureofacines [Sutka et al., 2004b] and SPiP = 23.3 and 5.1‰ for P. fluorescens and P. denitrificans, respectively [Toyoda et al., 2005].

[37] SP was positive in all of the laboratory and groundwater samples. Since production and reduction of N2O have occurred during the reaction progress (see section 4.1), SP in the residual N2O results from fractionation during both partial processes. Site-specific fractionation factors of the partial processes given by the difference in δ15N between substrate and instantaneous product [Barford et al., 1999] cannot be determined from our data because both processes proceeded in parallel. Given the relatively large values of SP, it could be concluded that both processes probably contributed to positive SP of residual N2O. However, the possible effect of negative SPiP could theoretically also be masked by fractionation during N2O reduction.

[38] Similar to δ18O and δ15Nbulk, SP increased with ti (R2 = 0.63, P > 0.95) and the regression between ti and SP improved if the supposed “hot spot” sample (July, 160 cm, see above) is excluded (R2 = 0.96, P > 0.999). This clearly demonstrates that all components of the isotopic fingerprint, i.e., δ15Nbulk, δ18O and SP, reflect reaction progress. Furthermore, anomalies of the temporal pattern might be used to identify variation of process rates within the investigated system.

[39] The control mechanisms of isotope effects associated with denitrification given above (section 4.3.1) can be used to discuss possible explanation for the discrepancy between observations of high SP of residual N2O in our groundwater samples and relatively low SP of soil emitted N2O. High N2O concentration and/or low N2O reduction rate would be theoretical explanations, since isotope effects can increase with electron acceptor concentration-to-reduction rate ratio. However, the observed groundwater concentrations (0.005 to 2.7 mg N L−1) were within the range of reported pore space concentration (up to 8 mg N L−1 [Heincke and Kaupenjohann, 1999]). Furthermore, the low N2O concentration of samples with high SP (Table 2) demonstrates that reduction capacity was not rate limiting. This suggests that these parameters were not the main drivers. A fundamental difference between soil and groundwater lies in the fate of produced N2O: Owing to high diffusivity in the unsaturated zone a large fraction bypasses further reduction to N2 and is emitted to the atmosphere. In contrast, gases are more or less trapped in the groundwater. Thus there is a relatively high probability that N2O with positive SP leaving denitrifying cells is taken up by N2O reducing bacteria, where partial reduction may result in further increase SP of residual N2O. Consequently, multiple partial reduction cycles might be responsible for the observations of high SP in groundwater N2O. However, further evidence is needed to prove this hypothesis.

[40] In contrast to the observed field trends, SP in the laboratory experiment did not increase continuously but fluctuated with time. This could indicate temporal variation of site specific fractionation which might be explained by microbial adaptation during the initial phase of the experiments. It is possible that reduction was initially inhibited by a limited density of denitrifying enzymes resulting in relatively large isotope fractionation during N2O production as well as reduction (see section 4.3.1). Establishing anaerobic conditions and adding NO3 probably induced production of NO and N2O reductases which may in turn have resulted in decreasing site specific fractionation during the initial phase. After 30 days, SP of the laboratory incubation follows a positive trend (Figure 3a) similar to that observed in the field which could indicate the end of microbial adaptation. The change in fractionation pattern at this time is also reflected in increasing δ18O.

[41] The final SP and δ15Nbulk of the laboratory experiment was approximately half the highest field values. The following characteristics of the laboratory conditions are possible explanations for the observed differences: (1) N2O concentrations were lower. (2) Between the sampling events, the headspace gas containing most of the N2O was separated from the denitrifying environment, i.e., the submersed aquifer material, by a diffusion barrier consisting of the slurry (approximately 7 mm in height) and the supernatant (approximately 17 mm in height). This is unlike the in situ conditions, where maximum diffusion length is governed by the pore size distribution and is thus several orders of magnitude lower compared to the laboratory system. Furthermore, (3) the liquid-to-solid ratio was higher than in the field, (4) periodical shaking may have increased availability of organic substrates, and (5) the microbial community may have been altered because samples were not collected aseptically. The first and fourth enumerated impacts both imply a decrease in the electron acceptor–to-reduction rate ratio which potentially lessens isotope fractionation and could thus explain the observed differences.

[42] The positive relationship between SP and δ18O (Table 2, Figure 2b) demonstrates that the stability of the Nα-O bond was responsible for position-specific N-fractionation and O-fractionation at this site, which is in agreement with earlier assumptions [Yoshida and Toyoda, 2000; Popp et al., 2002; Toyoda et al., 2002; Schmidt et al., 2004]. The large growth of δ18O shows that either oxygen exchange with water played a minor role or the fractionation factors for oxygen are large compared to SPiP and SPiR.

4.4. Using the Isotopic Fingerprint of N2O in the Saturated Zone to Estimate Emission of Groundwater Derived N2O

[43] Large differences in SP between groundwater and soil emissions were observed in our local data set (Table 2). The isotopomer mixing model which was applied to the field site (section 3.1.2, equation (3)) suggested that the fraction of groundwater derived N2O contributing to the surface flux of 363 g N ha−1 d−1 was relatively small (0 to 0.14). Furthermore, this range demonstrates that a precise estimation of this fraction is not possible with our data set because SP of N2O produced in the unsaturated zone was not well defined. However, it confirms the low level of diffusive N2O fluxes calculated from concentration gradients in the saturated zone (Table 2), which were only a minor fraction of the surface flux. Thus both approaches, i.e., isotopomer signatures as well as N2O concentration gradients, prove that N2O emission measured at the soil surface mainly originated from the unsaturated zone.

[44] Similar to our local data set, the known range of low SP of soil emitted N2O as well as the presumed high SP of groundwater N2O could be used in a global isotopomer mixing model in order to investigate the potential impact of groundwater derived N2O on the isotopic signature of tropospheric N2O. If a significant effect could be expected, then the isotopic fingerprint of N2O from the saturated zone might be an important tool to judge the order of magnitude of so-called indirect N2O emissions from agriculture caused by transporting nitrogen from agricultural fields into aquatic systems via leaching and run-off [Mosier et al., 1998] which is still controversial [Groffman et al., 2000; Nevison, 2000] (see section 1). Using IPCC (International Panel on Climate Change) default emission factors, Mosier et al. [1998] calculated a global indirect N2O flux of 1.6 Tg N yr−1 which would be a considerable fraction of total N2O sources (16.4 Tg N yr−1 [Intergovernmental Panel on Climate Change, 2001]). From theoretical considerations, Nevison [2000] assumed that the IPCC emission factor for groundwater (EF5-g) of 0.015 kg N2O emitted per kg N leached is at least 1 order too high. This is in agreement with recent findings of relatively low N2O-to-NO3 ratios in groundwater [Sawamoto et al., 2005]. However, there is also evidence of relatively high N2O emission from riparian buffers which suggest that emission factors for nitrate loaded groundwater passing these areas before being discharged to surface waters might be even higher than EF5-g [Hefting et al., 2003].

[45] In the following, N2O isotopomer signatures in groundwater of this study are compared to the known range of surface emitted N2O (Figure 2). The potential impact of indirect N2O emissions from groundwater on site preference of the global N2O source (SPs) could be estimated using a simple mixing calculation, given the fluxes and associated signatures were known,

equation image

where F is N2O flux and subscripts s, sd, si denote total source, direct sources (i.e., total sources excluding the indirect source) and indirect source, respectively. SPs can be calculated from the values for central and terminal δ15N relative to tropospheric N2O determined by Röckmann et al. [2003] (δ′β = −12.1‰, δ′α = −18.5‰) and the values of tropospheric average published by Toyoda et al. [2004]Tβ = −2.9‰, δTα = 15.8‰), giving 12.3‰. The source data cited above are used for the fluxes, giving Fs = 16.4 Tg N a−1, Fsi = 0.16 to 1.6 Tg N a−1and Fsd = Fs − Fsi. A best estimate of the possible range of SPsi is derived from the range of SP in groundwater and flooded soils known so far. From the time course of SP and N2O concentration of our measurements it can be expected that the flux weighted SP of N2O from the saturated zone is at the lower end of the range of observed SP because maximum N2O concentrations occur during the early phase of the process cycle (Figure 3, Table 2). SP at maximum concentration was approximately 20 and 50‰ for the laboratory and field experiments, respectively. Mean SP of two flooded soils [Bol et al., 2004] and of our recent groundwater data set (R. Well et al., 2004, unpublished data) was 30.8, 40.1 and 37.6‰, respectively. Thus we assume SPsi to range between 20 and 50‰. Solving equation (4) for SPsd results in SPsd = 8.2 to 11.5‰ for Fsi = 1.6 Tg N and SPsd = 11.9 to 12.2‰ for Fsi = 0.16 Tg N. The share of indirect N2O emission with respect to SPs (12.3‰) is defined by SPsi × Fsi/Fs giving 2.0 to 4.9‰ for Fsi = 1.6 Tg N yr−1 and 0.2 to 0.5‰ for Fsi = 0.16 Tg N yr−1. Röckmann et al. [2003] measured site-specific 15N signatures of the pre-industrial source from N2O profiles in antarctic firn air. Calculating the shift of SPs since the pre-industrial era gives 0.6‰. Because significant NO3 leaching is a result of modern agriculture, pre-industrial indirect N2O emissions from the groundwater can be assumed to be negligible. Thus comparing the historical shift of SPs with the calculated range of SPsi × Fsi/Fs leads to the following conclusions. (1) The lower estimate of Fsi (0.16 Tg N) could explain a significant fraction of the historic shift of SPs, (2) the high estimate of Fsi (1.6 Tg yr−1) would imply a significant historical shift of SPsd since SPsi × Fsi/Fs is much larger than the historical shift of SPs, and (3) modern SPs could be used to constrain Fsi if SP of the other sources (industrial, direct soil emissions) was better constrained.

[46] It must be noted, however, that the indirect and direct fluxes and associated isotopomer signatures used in the mixing calculation are still highly uncertain. In particular, the available groundwater isotopomer data are from few locations and riparian systems are currently not included. Consequently, much more data on both direct and indirect source signatures are needed for isotopomer budget calculations suitable to confirm the order of magnitude of estimated indirect emissions.

5. Conclusions

[47] The difference in N2O isotopomer patterns between N2O from groundwater and N2O emitted at the soil surface of the experimental site proves that the contribution of the saturated zone to the total N2O flux from the system was small and thus confirms the fluxes calculated from vertical concentration gradients. The agreement of high 15N site preference (SP) observed in our field and laboratory study with the results from flooded soils and suboxic oceanic samples demonstrates that this pattern is typical for suboxic or anoxic systems where a large fraction of produced N2O is reduced via denitrification. SP of these environments is clearly distinct from the relatively low or negative SP of soil derived N2O. The maximum observed SP (81‰ δ15N) of our field study is to our knowledge the highest value that has been ever observed in N2O from biological processes. Comparing the potential contribution of N2O from the saturated zone to tropospheric SP with the historical shift of tropospheric SP demonstrates that SP may principally be used to validate estimates of indirect N2O emission caused by agricultural N-leaching to the groundwater.

[48] The observed spatial and temporal pattern of 18O and average and site-specific 15N signatures can be explained by the reaction progress of denitrification with associated isotopomer fractionation. A linear relationship between 18O and SP confirms that both quantities were governed by the same process, i.e. fractionation during breakage of NO-bonds by reduction of N2O to N2.

Acknowledgments

[49] We thank the Deutsche Forschungsgemeinschaft (DFG) for financial support, Manfred Kayser and Jürgen Müller for technical assistance and providing experimental sites, Susanne Richter and Martin Grönmeier for technical assistance, Reinhard Langel and Lars Szwec for isotope analysis, and Nathaniel Ostrom for constructive comments on an earlier version of this manuscript.

Ancillary