Spatial and Seasonal Variation of Dissolved Nitrous Oxide in Wetland Groundwater

Understanding the spatial and temporal pattern of dissolved nitrous oxide (N2O) in groundwater is essential to estimate the N2O emissions from groundwater to the unsaturated zone and to the atmosphere. In order to study the spatial distribution and seasonal change of dissolved N2O in wetland, a headwater wetland in Ichikawa, Chiba Prefecture, Japan, was chosen. Variations of nitrate (NO3), dissolved N2O and δN-NO3 indicated that the dissolved N2O in the groundwater of study wetland consists of two parts, one from denitrification within the wetland, and another from nitrification at upland. Principal component analysis (PCA) was used to assess the shallow groundwater parameters in the wetland. And t-test was conducted to find statistically significant differences of the variables between the ASW and NS, warm season and cool season. The concentrations of dissolved N2O increased from the upland to the zone of adjacent area between slope and wetland (ASW) and then decreased at the zone near the stream (NS). In sight of dissolved N2O associated nitrogen migration, groundwater in the study area can be divided into three stages: upland as the stage 1, ASW as the stage 2, and NS as the stage 3. Higher temperature results in higher denitrification rate, lower dissolved oxygen (DO) and oxidation-redox potential (ORP), yielding higher concentration of N2O in the warm season. Therefore, the seasonal change of dissolved N2O in study wetland can be mainly interpreted by the variation of temperatures of groundwater.


Introduction
Over the last few decades, much interest has been focused on specific natural systems, such as wetland (or riparian zone) which are vulnerable to improve water quality by physical, chemical and biological process that remove N from groundwater (García-García, Gómez, Vidal-Abarca, & Suárez, 2009;Groffman, Gold, & Simmons, 1992;Sabater et al., 2003).Wetlands offer an abundant organic C supply and dominated by inherently wet surface soil create anaerobic environment to consume nitrate via denitrification that is considered the most important reaction for nitrate removal in aquifer (Bastviken, Olsson, & Tranvik, 2003;Burgin & Hamilton, 2007;Whitmire & Hamilton, 2005).Especially in the shallow ground water of riparian areas, redox conditions are often favorable for intense denitrification processes (Ross, 1995).
The trace gas N 2 O is an obligate intermediate product of biological denitrification.And it is known to contribute to global warming and the destruction of stratospheric ozone.A significant amount of N 2 O emissions originates denitrification (Mathieu et al., 2006).Emissions from aquifers are most likely to occur from shallow aquifers, where N 2 O can be quickly transferred through the unsaturated zone to the atmosphere by diffusion (Rice & Rogers, 1993).N 2 O emission from wetland system has been estimated by numerous studies (Dhondt, Boeckx, Hofman, & Van Cleemput, 2004;Groffman, Gold, & Addy, 2000;Verhoeven, Arheimer, Yin, & Hefting, 2006).Understanding the spatial and seasonal pattern of dissolved N 2 O is essential to assess the indirect emission of N 2 O from groundwater (Geistlinger, Jia, Eisermann, & Florian Stange, 2010).Level of dissolved N 2 O in groundwater has been paid lots of attentions.For example, N 2 O concentration in groundwater was reported to exceed greatly those of atmospheric equilibration (with a mean value of 28.98 µg L -1 ) under aerobic condtion in Kanto district, Japan (Ueda, Ogura, & Yoshinari, 1993), and the maximum up to 30000 times of that in the ambient air (Heincke & Kaupenjohann, 1999).However, few studies estimated level of dissolved N 2 O in wetland groundwater.
www.ccsenAccording river, lake Dahlgren, of dissolve example, t when the n However, with the hi of dissolve manure a groundwat 1998; Spa at upland t Therefore, based on δ seasonal c convenien nitrificatio stage 3 wh a proceed

Site De
The study (Figure 1 ected go up r that ck to were measured in situ with sensors (HIROBA).All water samples were filtered (0.45 µm) before analysis for major ions by ion chromatography (Shimadzu CDD-6A and CDD-10Avp). 2 L water for each sample was collected for δ 15 N-NO 3 -analysis in March 2012.NO 3 -was collected by passing the water through pre-filled, disposable, anion exchanging resin columns in the field and then was eluted by 3 M HCl from the column.The nitrate-bearing acid eluant was neutralized with Ag 2 O, filtered to remove the AgCl precipitate, then freeze dried to obtain solid AgNO 3 , which was then combusted to N 2 in sealed quartz tubes for analysis by Integra CN mass spectrometer (Pdz Europa LTD) at Chiba University, Japan (Yingjie Cao, Tang, Song, Liu, & Zhang, 2012).All the samples were measured twice and the result showed the difference between the two measurements was less than ± 5%.Then the mean of two measurements was used as the value of δ 15 N-NO 3 -in this study.

Statistical Analysis
Variables were tested using student t-test and principal component analysis (PCA), with SPSS 8.0 for Windows (SPSS, 1997, IL, USA).T-test was used to determine if two sets of data are significantly different from each other.The PCA is a data transformation technique that attempts to reveal a simple understanding structure that is assumed to exist within a multivariate dataset (Davis, 1986).

Basic Parameters and Dissolved N 2 O in Upland Shallow Groundwater
Samples were taken from W1 in July and November 2012, respectively.DO and ORP concentrations were higher in July (Table 1).pH values of groundwater were lower than 7 both in July and November.Groundwater temperature was little higher in July than that in November.NO 3 --N and N 2 O-N concentrations were both higher in July than that in November.Groundwater temperatures in the wetland ranged from 14.2 to 24.8 °C during the study period (Table 2).pH values of groundwater ranged from 6.53 to 7.97, indicating that the groundwater was alkaline except S14 which pH was lower than 7 during the warm season.DO concentrations ranged from 0.07 to 11.50 mg L -1 .It was lower than 4 mg L -1 , and as low as 0.07 mg L -1 in June at R2.At S4 and S14, the DO concentrations were lower than 5 mg L -1 in the warm season, but up to 11.5 mg L -1 in the cool season (S14-3 m in November).ORP values ranged from -244 to 303 mV.At 1 m and 2m depth of R2, ORP values were below about 0 mV in the warm season with the lowest value of -189 mV.However, ORP was up to 175 mV in March.At 3 m depth of R2, ORP was above 0 mV except in September (-244 mV).The NO 3 --N concentration changed from 0 to 114.0 mg L -1 in study sites.At S4 and S14, most NO 3 --N concentrations are clearly above the standard of the drinking water (10 mg L -1 ) set by United States Environment Protection Agency (Figure 3), whereas NO 3 --N concentration was extremely low for detection at R2. NH 4 + and NO 2 -were also measured with other major ions, and found below the detectable limit.
www.ccsenThe scores represent the influence of the component on the groundwater (Figure 4).It is possible to group the samples according to the axes of component 1 and 2. As a result, the samples are classified into four groups to showing seasonal and spatial patterns.The samples of NS are plotted at the down-left of the diagram for the warm season and the down-middle for the cool season.On the other hand, the samples of ASW are plotted at the upper-left of the diagram for the warm season and the upper-right for the cool season.
As the result of the groups from PCA, mean value, standard deviation and a t-test was conducted to find the difference and statistically significant differences of the variables between the ASW and NS, warm season and cool season (Table 5).NO 3 --N, DO, ORP and N 2 O-N in groundwater were significantly higher at ASW than those at NS, whereas there was no significantly difference of temperature and pH between ASW and NS.For N 2 O-N, NO 3 --N concentrations, and DO, variability was higher at ASW than those at NS on the basis of standard deviation.In contrast, variability of ORP was lower at ASW than it at NS.In addition, the mean N 2 O-N concentration at ASW was high (36.14 µg L -1 ), which was about 60 times of that in the ambient air.N 2 O-N concentration and temperature in groundwater were significant higher in warm season than those in cool season, and DO, ORP and pH were significantly lower in warm season.There was no significant difference of NO 3 --N concentrations between two seasons which seems to be the rule rather than the exception.

Source of Dissolved N 2 O of Shallow Groundwater
In order to estimate the concentration of N 2 O in groundwater, it is important to identify its source.Fertilizer and manure ammonium-nitrogen applied in the orchard are oxidized to nitrate-nitrogen and nitrous oxide in unsaturated zone of the upland.Nitrate leaches to the groundwater from unsaturated zone in the upland.δ 15 N-NO 3 -value of W1 is coincided with range of δ 15 N-NO 3 -(+4.5 ‰ to +8.5 ‰ ) in the area effected by mineral fertilizer (Cao, Sun, Xing, & Xu, 1991;Choi, Lee, & Ro, 2003;Choi, Han, Lee, Lee, & Yoon;Heaton, 1986;Singleton et al., 2007), indicating the dissolved N 2 O was produced via nitrification in the unsaturated zone of upland.DO concentrations were high at W1, indicating that denitrification could not occur.Nitrate and N 2 O transport from upland to wetland with groundwater consequently.N 2 O is difficult to denitrified to N 2 because the groundwater in upland is often assumed to have low biological activity due to low C content (Groffman, Gold, & Jacinthe, 1998).Geistlinger et al. (2010) found there will be a diffusive N 2 O flux from the deeper water to the capillary fringe.However the time scale of this process is very large i.e., for 10 cm travel distance, the N 2 O molecules need ≈ 230 d.Thus, diffusive loss to upward is considered to have little effect on N 2 O concentration in the groundwater during transporting from upland to wetland.
At the wetland, denitrification can enrich 15 N in the residual nitrate of groundwater (Cey, Rudolph, Aravena, & Parkin, 1999;Lehmann, Reichert, Bernasconi, Barbieri, & McKenzie, 2003).δ 15 N-NO 3 -in the residual nitrate enriched from 2.8 ‰ to 78.32 ‰ when the concentration of NO 3 --N decreased from 35.68 mg L -1 to 0.45 mg L -1 in a sand aquifer (Böttcher, Strebel, Voerkelius, & Schmidt, 1990), and from 6.4 ‰ to 24.8 ‰ when the NO 3 --N concentration decreased from 13.3 to 5.6 mg L -1 in a riparian zone (Cey et al., 1999) .In this study wetland, the δ 15 N-NO 3 -enriched by 9.81 ‰ or even higher when the NO 3 --N was no longer detectable.Therefore, dissolved N 2 O in the shallow groundwater of wetland consists of two parts, one from denitrification within the wetland, and another from the upland where nitrification is dominant.

Spatial and Seasonal Pattern of Dissolved N 2 O in Shallow Groundwater of the Wetland
The previous section suggested that the source of dissolved N 2 O of groundwater in wetland comes from nitrification in upland and denitrification in wetland.In the study wetland, denitrification controls the behavior of dissolved N 2 O.Because N 2 O is an intermediate product of denitrification that is producted when nitrate is reduced and is consumed by reduction to N 2 .Denitrification is considered to be related to many factors (DO, ORP, T, pH and NO 3 -).For example, the highest concentrations of N 2 O were found in the aerobic section of a limestone aquifer with the DO concentration below 4.00 mg L -1 and in a phreatic aerobic aquifers with the DO concentration below 3.15 mg L -1 (Deurer et al., 2008;Ronen, Magaritz, & Almon, 1988).However, the optimal maximum DO concentration for nitrogen removal was determined to be around 2.0-2.5 mg L -1 in the laboratory experiments (Yoo et al., 1999).According the early study, Nelson and Knowles (1978) reported that the startup of denitrification can be inhibited while the oxygen level is as low as 0.13 mg L -1 in a dispersed-well sludge reactor.In the laboratory experiments, as the ORP drops below 0 mV, the nitrate begins to be converted to nitrite and nitrite accumulates continuously for ORP ranging from 0 to -225 mV.From -225 to -400 mV, the accumulated nitrite is converted to N 2 .As the ORP below -400 mV, the nitrate is firstly converted first to nitrite then the nitrite is converted immediately to N 2 without accumulation (Lee et al., 2000).It also reported that ORP below about 200 to 300 mV were found to be conducive to denitrification, and the maximum N 2 O were found at a ORP value of 0 mV (Kralova, Masscheleyn, Lindau, & Patrick Jr, 1992).Therefore, the optimum value of DO and ORP for N 2 O accumulation is not consistent with the value of the optimum for denitrification due to the N 2 O is an intermediate product.For nitrate, DeSimone and Howes (1998) studied that kinetics of denitrification at nitrate concentrations >1 mg-N L -1 is zero order and even small amount of nitrate (lower than 2 mg-N kg -1 ) leached was sufficient to create a large amount of N 2 O in groundwater (Müller, Stevens, Laughlin, & Jäger, 2004).Many studies suggested that high concentration of NO 3 --N inhibits the N 2 O reductase yielding the higher concentration of N 2 O (Blackmer & Bremner, 1978;Deurer et al., 2008;Heisterkamp, Schramm, de Beer, & Stief, 2012).At ASW, the DO (m = 5.22 mg L -1 ) and ORP (m = 237 mV) values were both higher than the optimum values respectively, as well as high concentrations of NO 3 --N which were conducive to N 2 O accumulation (m = 36.14µg L -1 ) (Table 5).However, the mean value of DO concentrations (2.02 mg L -1 ) and ORP were much lower (-5 mV) at NS. Additionally, NO 3 --N is low or undetectable throughout the study.Under these conditions, the N 2 O is used as an electron acceptor instead of nitrate in denitrification process (Ishii, Ohno, Tsuboi, Otsuka, & Senoo, 2011), resulting in the lower concentration (m = 13.87 µg L -1 ).Therefore, ASW and NS can be considered as in the stage 2 and stage 3, respectively.In addition, the average flux of N 2 O was found to be higher at ASW than it at NS (Li, Tang, Han, Cao, & Zhang, 2013) which is consistent with the trend of dissolved N 2 O.
Seasonal changes of dissolved N 2 O are most associated with NO 3 -concentration and water temperature (Bouwman, Boumans, & Batjes, 2002;Hinshaw & Dahlgren, 2013;Velthof, Oenema, Postma, & Van Beusichem, 1996).The T-test indicates that the concentrations of NO 3 --N had no significant difference between the two seasons, which suggests NO 3 --N is not the limited factor for denitrification rate in study wetland (Table 5).Temperature affected the dissolved N 2 O directly by controlling the denitrification rate (Nowicki, 1994;Pfenning & McMahon, 1997;Saunders & Kalff, 2001).The threshold temperature for controlling the rate of denitrification was 20 °C (Halling-Sørensen & Jorgensen, 1993) or even below 17 °C (McCutchan & Lewis, 2008;Nowicki, 1994).A study in coarse sandy soils found that the denitrification activity was low at 10 °C and completely inhibited at 2 and 5 °C because lower temperature may regulate metabolic rates for denitrifying bacteria (Vinther & Søeberg, 1991).Temperature also influences the solubility of oxygen, the rates of aerobic respiration of bacteria and the ORP change in groundwater, all of which in turn limit dissolved N 2 O indirectly.For example, the oxygen solubility is 14.60 mg L -1 at 0 °C , about double at 30 °C (7.54 mg L -1 ) (Weiss, 1970).Oxygen consumption by aerobic respiration increases when the temperature increases (Thamdrup, Hansen, & Jørgensen, 1998).When the temperature increased from 15 °C to 25 °C, the average ORP decreased from +40 mV to -60 mV (Zhu, Ndegwa, & Luo, 2002).In warm season, denitrification rate supposed not to be inhibited by temperature (m = 20.5 °C).The lower DO and ORP of groundwater could be assumed as a response to the higher temperature in the warm season.The characteristics of these factors resulted in the higher N 2 O concentration in the warm season (m = 34.19µg L -1 ) than it in cool season (20.19 µg L -1 ).In addition, the decrease of pH was interpreted as a sign of intense denitrification (Ilies & Mavinic, 2001).Mean value of pH is lower in the warm season (m = 7.11) than it in the cool season (m = 7.49), which also can explain the higher dissolved N 2 O concentrations in the warm season.The seasonal change of dissolved N 2 O coincides with N 2 O flux measured in the study wetland.In fact, the average monthly N 2 O flux ranged from 0.019 to 0.286 mg N m -2 h -1 with the highest value in the warm season and the lowest flux appeared in the cool season (Li et al., 2013).-decreased continuously from upland to the wetland.Along the groundwater flow, the dissolved N 2 O was produced through nitrification at the upland and denitrification in the wetland, which is supported by the variations of δ 15 N-NO 3 -in the shallow groundwater.The mean value of dissolved N 2 O-N increased from 11.42 µg L -1 at upland to 36.14 µg L -1 at the ASW and then decreased to 9.27 µg L -1 at NS.The dissolved N 2 O in the ASW zone is expected to be composed of two parts.One is transported from the upland and the other is produced from denitrification in the wetland.As a result, the dissolved N 2 O in the groundwater can be classified into the stage 1 for the upland, the stage 2 for ASW and the stage 3 for NS in the study area.Seasonally, the N 2 O concentration was higher in the warm season (m = 34.19µg L -1 ) and lower in the cool season (m = 20.19 µg L -1 ).Temperature and pH are main factors to control the dissolved N 2 O in groundwater of study area.Higher temperature results in higher denitrification rate by elevating metabolic rates for denitrifying bacteria directly, and creating the lower DO and ORP environment that affects the N 2 O concentration indirectly in the warm season.In addition, lower pH in the warm season also may explain the higher dissolved N 2 O concentrations because the decrease of pH is interpreted as a sign of intense denitrification.

Conclusions
This study put forward an understanding of spatial distributions of dissolved N 2 O from upland (agricultural area) which related the materials transformation to groundwater flow system.Temperature is considered as the main driver to seasonal change of dissolved N 2 O in wetland groundwater.

N 2 O
concentrations, denitrification related factors (NO 3 -, DO, ORP, pH and T) and δ 15 N-NO 3 -values were investigated in a typical headwater wetland and watershed.The main findings and conclusions are as follows: Spatially, NO 3 -, DO and ORP are main factors to control the dissolved N 2 O in groundwater of study area.DO, ORP and NO 3

Table 1 .
Basic parameters and dissolved N 2 O of upland shallow groundwater in July and November 2012

Table 5 .
Mean (m)and standard deviation (parentheses) of N 2 O, DO, ORP, NO 3 ** The difference between mean values is significant (p < 0.05)