nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: an undisturbed...

8
Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study Borja Muñoz-Leoz a , Iñaki Antigüedad b , Carlos Garbisu c , Estilita Ruiz-Romera a, a Department of Chemical and Environmental Engineering, University of the Basque Country, UPV/EHU, E-48013 Bilbao, Spain b Department of Geodynamic, University of the Basque Country, UPV/EHU, E-48940 Leioa, Spain c Department of Ecosystems, NEIKER-Tecnalia, E-48160 Derio, Spain abstract article info Article history: Received 30 June 2010 Received in revised form 1 October 2010 Accepted 4 October 2010 Available online 27 November 2010 Keywords: Carbon dioxide Denitrication Nitrate leaching Nitrous oxide Riparian soil Wetland Riparian wetlands bordering intensively managed agricultural elds can act as biological lters that retain and transform agrochemicals such as nitrate and pesticides. Nitrate removal in wetlands has usually been attributed to denitrication processes which in turn imply the production of greenhouse gases (CO 2 and N 2 O). Denitrication processes were studied in the Salburua wetland (northern Spain) by using undisturbed soil columns which were subsequently divided into three sections corresponding to A-, Bg- and B2g-soil horizons. Soil horizons were subjected to leaching with a 200 mg NO 3 L 1 solution (rate: 90 mL day 1 ) for 125 days at two different temperatures (10 and 20 °C), using a new experimental design for leaching assays which enabled not only to evaluate leachate composition but also to measure gas emissions during the leaching process. Column leachate samples were analyzed for NO 3 concentration, NH 4 + concentration, and dissolved organic carbon. Emissions of greenhouse gases (CO 2 and N 2 O) were determined in the undisturbed soil columns. The A horizon at 20 °C showed the highest rates of NO 3 removal (1.56 mg NNO 3 kg 1 DW soil day 1 ) and CO 2 and N 2 O production (5.89 mg CO 2 kg 1 DW soil day 1 and 55.71 μgNN 2 O kg 1 DW soil day 1 ). For the Salburua wetland riparian soil, we estimated a potential nitrate removal capacity of 1012 kg NNO 3 ha 1 year 1 , and potential greenhouse gas emissions of 5620 kg CO 2 ha 1 year 1 and 240 kg NN 2 O ha 1 year 1 . © 2010 Elsevier B.V. All rights reserved. 1. Introduction The extensive use of agricultural chemicals has led to contamina- tion of surface water and groundwater with pesticides, nitrogen compounds, etc. Nitrate (NO 3 ), the principal N-bearing constituent of groundwater (Pauwels and Talbo, 2004), has been reported as a cause of methaemoglobinaemia in humans (Höring and Chapman, 2004) and eutrophication in aquatic ecosystems (Galloway et al., 2008). Riparian wetlands bordering intensively managed agricultural elds can act as biological lters that retain and transform chemicals (e.g., NO 3 ) as groundwater passes through these transition zones between terrestrial and aquatic ecosystems (Casey et al., 2004; Ranalli and Macalady, 2010). In fact, signicant reductions in groundwater NO 3 concentration have been reported in riparian wetlands (Casey et al., 2004; Hefting et al., 2006; Zhao et al., 2009). Processes responsible for this NO 3 removal from groundwater include plant uptake, denitrication, microbial assimilative NO 3 reduction, etc. (Rivett et al., 2008). Nonetheless, in riparian wetland soils receiving large amounts of N, denitrication appears to be the most important process for NO 3 removal (as compared to upland soils, riparian wetland soils are usually wetter and have higher contents of organic C, conditions which favour denitrication) (Ranalli and Macalady, 2010). Denitrication, a microbial process that takes place in soil under anaerobic water-saturated conditions, involves the reduction of NO 3 to N 2 (with NO 2 , NO, and N 2 O as intermediates) and the oxidation of organic matter, as described by Jørgensen et al. (2004): 5CH 2 O þ 4NO 3 þ 4H þ 2N 2 þ 5CO 2 þ 7H 2 OðΔ¼ 448kJmol 1 Þ ð1Þ A priori denitrication appears desirable as it can remove NO 3 from polluted ecosystems but, on the other hand, it can generate signicant amounts of greenhouse gases such as CO 2 and, above all, N 2 O(Liu and Greaver, 2009). In riparian soils, the rate of emission of greenhouse gases can be a highly spatial variable: these gases are often produced in soil microsites around particles of decaying organic matter which is unevenly distributed in the soil prole (Hill et al., 2000). Likewise, N 2 O and CO 2 emissions often follow a strong temporal pattern depending on environmental conditions (e.g., temperature, rainfall, water table, etc.) (Hefting et al., 2004). Previous studies have shown that the physicochemical properties of the Salburua wetland riparian soil are highly conducive to denitrication, Science of the Total Environment 409 (2011) 763770 Corresponding author. Department of Chemical and Environmental Engineering, School of Engineering, University of the Basque Country, UPV/EHU, Alameda Urquijo s/n, E-48013 Bilbao, Spain. Tel.: +34 94 601 4109; fax: +34 94 601 4179. E-mail address: [email protected] (E. Ruiz-Romera). 0048-9697/$ see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2010.10.008 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Upload: borja-munoz-leoz

Post on 12-Sep-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

Science of the Total Environment 409 (2011) 763–770

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil:An undisturbed soil column study

Borja Muñoz-Leoz a, Iñaki Antigüedad b, Carlos Garbisu c, Estilita Ruiz-Romera a,⁎a Department of Chemical and Environmental Engineering, University of the Basque Country, UPV/EHU, E-48013 Bilbao, Spainb Department of Geodynamic, University of the Basque Country, UPV/EHU, E-48940 Leioa, Spainc Department of Ecosystems, NEIKER-Tecnalia, E-48160 Derio, Spain

⁎ Corresponding author. Department of Chemical anSchool of Engineering, University of the Basque Country, UE-48013 Bilbao, Spain. Tel.: +34 94 601 4109; fax: +34

E-mail address: [email protected] (E. Ruiz-Romera

0048-9697/$ – see front matter © 2010 Elsevier B.V. Aldoi:10.1016/j.scitotenv.2010.10.008

a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 June 2010Received in revised form 1 October 2010Accepted 4 October 2010Available online 27 November 2010

Keywords:Carbon dioxideDenitrificationNitrate leachingNitrous oxideRiparian soilWetland

Riparian wetlands bordering intensively managed agricultural fields can act as biological filters that retain andtransform agrochemicals such as nitrate and pesticides. Nitrate removal in wetlands has usually beenattributed to denitrification processes which in turn imply the production of greenhouse gases (CO2 and N2O).Denitrification processes were studied in the Salburua wetland (northern Spain) by using undisturbed soilcolumns which were subsequently divided into three sections corresponding to A-, Bg- and B2g-soil horizons.Soil horizons were subjected to leaching with a 200 mg NO3

−L−1 solution (rate: 90 mL day−1) for 125 days attwo different temperatures (10 and 20 °C), using a new experimental design for leaching assayswhich enablednot only to evaluate leachate composition but also to measure gas emissions during the leaching process.Column leachate samples were analyzed for NO3

− concentration, NH4+ concentration, and dissolved organic

carbon. Emissions of greenhouse gases (CO2 and N2O)were determined in the undisturbed soil columns. The Ahorizon at 20 °C showed the highest rates of NO3

− removal (1.56 mg N–NO3−kg−1DW soil day−1) and CO2 and

N2O production (5.89 mg CO2kg−1DW soil day−1 and 55.71 μg N–N2O kg−1DW soil day−1). For the Salburuawetland riparian soil, we estimated a potential nitrate removal capacity of 1012 kg N–NO3

−ha−1 year−1, andpotential greenhouse gas emissions of 5620 kg CO2ha−1 year−1 and 240 kg N–N2O ha−1 year−1.

d Environmental Engineering,PV/EHU, Alameda Urquijo s/n,94 601 4179.).

l rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

The extensive use of agricultural chemicals has led to contamina-tion of surface water and groundwater with pesticides, nitrogencompounds, etc. Nitrate (NO3

−), the principal N-bearing constituent ofgroundwater (Pauwels and Talbo, 2004), has been reported as a causeof methaemoglobinaemia in humans (Höring and Chapman, 2004)and eutrophication in aquatic ecosystems (Galloway et al., 2008).

Riparian wetlands bordering intensively managed agriculturalfields can act as biological filters that retain and transform chemicals(e.g., NO3

−) as groundwater passes through these transition zonesbetween terrestrial and aquatic ecosystems (Casey et al., 2004; Ranalliand Macalady, 2010). In fact, significant reductions in groundwaterNO3

− concentration have been reported in riparian wetlands (Casey etal., 2004; Hefting et al., 2006; Zhao et al., 2009). Processes responsiblefor this NO3

− removal from groundwater include plant uptake,denitrification, microbial assimilative NO3

− reduction, etc. (Rivett etal., 2008). Nonetheless, in riparian wetland soils receiving largeamounts of N, denitrification appears to be the most important

process for NO3− removal (as compared to upland soils, riparian

wetland soils are usuallywetter and have higher contents of organic C,conditions which favour denitrification) (Ranalli and Macalady,2010).

Denitrification, a microbial process that takes place in soil underanaerobic water-saturated conditions, involves the reduction of NO3

to N2 (with NO2−, NO, and N2O as intermediates) and the oxidation of

organic matter, as described by Jørgensen et al. (2004):

5CH2O þ 4NO−3 þ 4H

þ→2N2 þ 5CO2 þ 7H2OðΔG° ¼ −448kJmol−1Þ

ð1Þ

A priori denitrification appears desirable as it can remove NO3−

from polluted ecosystems but, on the other hand, it can generatesignificant amounts of greenhouse gases such as CO2 and, above all,N2O (Liu and Greaver, 2009). In riparian soils, the rate of emission ofgreenhouse gases can be a highly spatial variable: these gases areoften produced in soil microsites around particles of decaying organicmatter which is unevenly distributed in the soil profile (Hill et al.,2000). Likewise, N2O and CO2 emissions often follow a strongtemporal pattern depending on environmental conditions (e.g.,temperature, rainfall, water table, etc.) (Hefting et al., 2004).

Previous studies have shown that the physicochemical properties ofthe Salburuawetlandriparian soil arehighly conducive todenitrification,

Page 2: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

Table 1Physicochemical properties of the soil.

Horizon

A Bg B2g

Depth (cm) 0–20 20–70 70–100Corg (g Corgkg−1DW soil) 18.3 9.2 5.0N (g N kg−1DW soil) 2.3 1.2 0.9NO3

− (mg N kg−1DW soil) 0.55 0.24 0.12NH4

+ (mg N kg−1DW soil) 0.85 0.39 0.39pH 7.98 8.37 8.39Texture

Clay (%) 28 35 17Silt (%) 41 32 53Sand (%) 31 33 29

DW, dry weight.

764 B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

i.e. a noteworthy clay cover, a very low hydraulic gradient, and a highorganic matter content (García-Linares et al., 2003). Indeed, nitrateconcentrations exceeding 50 mg L−1 in groundwater entering thewetland are less than 10 mg L−1 at the outlet. The capacity of Salburuawetland riparian soil to remove NO3

− has already been reported bySánchez-Pérez et al. (2003) by means of denitrification potential assays.In Salburua soils, the rate of denitrification is higher in the upper soilprofile (with potential denitrification rates between 18.7 and20.3 mg Nkg−1DW soil day−1) and then decreases, together with thecontent of organic matter, with increasing depth to denitrification ratesof 0.1–1.1 mg Nkg−1DW soil day−1 at deeper soil profiles (Sánchez-Pérez et al., 2003). In any case, short-term microcosm tests usuallyoverestimate denitrification capacity as they are carried out underoptimal conditions (no carbon or nitrate limitation) and do not simulatethe natural conditions of the soil under study (Groffman et al., 2006).

Soil column experiments have been widely used to monitor thefate and mobility of pollutants (e.g. nitrate), as well as to evaluate soilhydrogeological properties and to study the kinetics of soil microbialcommunities (JØrgensen et al., 2004; Egiarte et al., 2006; Lewis andSjöstrom, 2010). Undisturbed column leaching assays allow testing ofsoil which is as close to actual field conditions as possible. However,traditionally, leaching assays have been designed to monitor leachatecomposition. Here, we proposed a new experimental design forleaching assays using undisturbed soil columns, which enables notonly to evaluate leachate composition but also to measure gasemissions resulting from groundwater flow simulation under differ-ent environmental conditions.

The objectives of this work were to study under laboratoryconditions: (i) the relative importance of differentmicrobial processesinvolved in NO3

− removal as groundwater passes through the Salburuawetland riparian soil; (ii) the rate of NO3

− removal and greenhousegases (CO2 and N2O) emissions; and (iii) the effect of temperature andsoil organic carbon (Corg) on these processes.

2. Materials and methods

2.1. Study site

The study was carried out in the Salburua wetland, a riparian zoneof the Alegría river (within the catchment of the Ebro river)surrounded by a large area of intensive agriculture. It is locatedwithin the East Sector of the Quaternary Aquifer of Vitoria-Gasteiz(Basque Country, northern Spain), which was declared in 1999 asVulnerable Zone to the pollution by nitrates of agricultural originaccording to Nitrates Directive 91/676/EEC. The thickness of thequaternary deposits in the area surrounding the wetland variesbetween 2 and 4 m, but can reach up to 10 m coinciding with theappearance of a furrow in the marly impermeable substratum. Thequaternary deposits form a spatially heterogeneous sequence com-posed, over the marls, of gravel (bottom), sand, and clay (top), whichappears homogenous in the order but not in the thicknesses. Thepresence of a top thick layer (up to 2 m) of clay results in an aquifer ofsemi-confined character. The main water entering the Salburuawetland (a natural discharge zone for the quaternary aquifer)comes from the various ditches that drain the croplands lying to thesouth and from the aquifer itself; at the eastern sector of the wetland,where this study was carried out, the only water entering the wetlandis groundwater (García-Linares et al., 2003). In this area, theconcentration of NO3

− in the water entering the wetland usuallyexceeds 50 mg NO3

−L−1, but is less than 10 mg NO3−L−1 (usually

below detection limit) in the water leaving the wetland.

2.2. Undisturbed soil column design

Nine undisturbed soil columns (1 m deep) were collected from theriparian zone located in the eastern sector of the Salburua wetland, an

area characterized by high levels of NO3− in the groundwater entering

the wetland. Before sampling, the vegetation was manually removedfrom the soil surface. Soil columns were sampled in PVC pipes (1 mlength, 45 mm inner diameter) using a percussion hammer (HM 1800model, Eijkelkamp). Immediately after sampling, the top and bottomof the columns were sealed with plastic lids to avoid drying. Columnswere kept at 4 °C and without delay transported to the laboratory.

Undisturbed soil columns were found to have 3 different soilprofile layers (horizons): (i) A-horizon: superficial (0–20 cm), ofbrownish-black colour, clay-sandy texture, and abundant fine sizeroots; (ii) Bg-horizon: sub-superficial (20–70 cm), of grayish-browncolour, clayey texture, presence of oxide reduction bands of iron andmanganese, and calcareous nodules; and (iii) B2g-horizon: sub-superficial (70–100 cm), of obscure brownish-red colour, clayey–siltytexture, and abundant oxide reduction stains of great size.

Threeof thecolumnsweredivided into three sections, correspondingto the three soil horizons (i.e., A, Bg, and B2g), sieved separately tob2 mm, air dried at 20 °C for 48 h, and finally subjected to physico-chemical characterization according to Sparks et al. (1996): pH, organiccarbon (Corg), total nitrogen (N), ammonium (NH4

+) and nitrate (NO3−)

contents, and clay mineralogy (Table 1).The other six columns were used for the leaching assays: for each

incubation temperature (10 °C versus 20 °C), three undisturbed soilcolumns were divided into their three constituent horizons (i.e., threeA-horizons, three Bg-horizons, and three B2g-horizons). Then, foreach soil horizon, a column leaching assay was carried out accordingto the scheme shown in Fig. 1: the bottom of each leaching columnhad a PVC adapter sealed with an inflow connection (22 cm siliconepipe) to a 12 channel peristaltic pump; the top of each leachingcolumnwas sealed with a 650 mL chamber made of PVC pipe (45 mminner diameter) and PVC adapters. The top of the chamber wasconnected to a polypropylene gas bag in order to compensate for theoverpressure resulting from leachate accumulation as well as tocollect gas samples. At the bottom of the chamber, a rubber septumsealed silicone pipe was placed so that leachates could be collected,using a polypropylene syringe, at each sampling time. The top andbottom of the leaching columns were filled with 5 cm of a washed seasand layer and glass wool, in order to minimize losses of soil as well asto avoid contamination of leachates with soil and sand particles.

The entire column set-up was placed in darkness inside anincubator (half of the leaching columns at 10 °C and the other halfat 20 °C). As seen in Fig. 1, columns were operated during 125 days ina vertical position with rising flow (in order to simulate the naturalflow of water coming up from the groundwater towards the wetland)from the NO3

− solution reservoir: a constant flow of 90 mL day−1

(equivalent to 46 mm day−1 of water discharged during an irrigationevent on surrounding crop fields) of a helium flushed aqueoussolution of 200 mg NO3

−L−1 was used (i.e., the groundwater nitratecontent that, from the beginning of the 80s to the early 90s, wasfrequently measured in the East Sector of the aquifer) (Arrate et al.,

Page 3: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

Fig. 1. Design of leaching assay with undisturbed soil columns.

765B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

1997), in an attempt to reproduce the worst environmental condi-tions faced by the wetland.

2.3. Leachates analysis

Leachates were collected in the upper chamber of the column,filtered (0.45 μm Whatman filter) and stored at 4 °C in plastic vialsuntil analysis. The following parameters were analysed thrice a weekfor 18 weeks: NO3

− concentration, NH4+ concentration, and dissolved

organic carbon (DOC) according to APHA (2005). Rates of NO3−

removal were calculated according to the following equation:

NR =V⋅ NI−NE½ �⋅0:226

T⋅SWð2Þ

where NR=rate of NO3− removal (mg N–NO3

−kg−1DW soil day−1);V=volume of leachate sampled at each time; NI=influent nitrateconcentration (200 mg NO3

−L−1); NE=effluent NO3− concentration

(mg NO3− L− 1); T=time (days); SW=soil dry weight (kg);

0.226=nitrate to nitrogen–nitrate conversion factor.

2.4. CO2 and N2O emissions

CO2 and N2O were determined during the leaching assays (N2O isan intermediate product of denitrification that can be easily quantifiedand has more interest than N2 from an environmental point of view).Gas samples were collected in the upper chamber of the leachingcolumns with a 1 mL crystal syringe (Supelco) and then CO2 and N2Ogas composition was analyzed, at the same time than leachates,throughout the experiment with a gas chromatograph (KNK 3000HRGC; TCD thermal conductivity detector). The packed column usedwas a Porapak Q 80/100 3 m×1/8 in. (Sugelabor). Operation condi-tions were as follows: column temperature 25 °C, injection temper-ature 25 °C, detector temperature 150 °C, and He as carrier gasflowing at 16 mL min−1. The amount of CO2 and N2O dissolved in theliquid phase was calculated using Henry's Law and then corrected fortemperature (Tiedje, 1982).

2.5. Statistical analysis

Statistical analyses were performed using SPSS software (SPSS 12Inc., 2003). A two-way analysis of variance (ANOVA) was used to

compare means of the parameters measured here among incubationconditions (10 and 20 °C) and soil horizons. Fisher's PLSD-test wasused to establish the significance of the differences among means.Values were considered to be significantly different at a 95%confidence level (Pb0.05). Pearson correlation analyses were usedto study the relationships between rates of NO3

− removal, and CO2 andN2O emissions.

3. Results and discussion

3.1. Nitrate removal

A considerable removal of NO3− was observed in all tested columns

(Fig. 2). Higher rates of NO3− removal were found in the A-horizon, as

compared to Bg- and B2g-horizons, especially during the first 15 daysof the leaching assay, indicating a positive relationship between Corgand rate of NO3

− removal. Indeed, highest values of NO3− removal were

observed in the A-horizon: 3.99 (8th day) and 2.14 (10th day) mg N–NO3

−kg−1DW soil day−1 at 20 and 10 °C, respectively. Higher rates ofNO3

− removal were found at 20 than at 10 °C. This NO3− removal could

in theory be linked with the increase in soil NO3− content observed

after leaching in all soil horizons (Table 2), owing to physicalretention of the input nitrate solution in the soil matrix itself: theoriginal interstitial soil water with a low NO3

− content would bereplaced by the high NO3

− content (200 mg NO3−L−1) input solution.

In any event, this phenomenon appears to be somewhat relevant onlyfor B2g-horizons: in B2g-horizons, the observed increase in soil NO3

content was equivalent to 11 and 29% of the total amount of NO3−

removed during the leaching assay at 10 and 20 °C, respectively.Nevertheless, nitrate (a very soluble compound with a negativecharge which hinders its adsorption to soil particles) is easily leachedwith moving water; thus, this physical phenomenon of nitrateattenuation in groundwater is not a nitrate removal mechanismper se, but just a temporary nitrate retention in the soil.

Another possible reason for the observed NO3− removal could be

microbial NO3− assimilation (Davidsson et al., 1997). Unfortunately,

this phenomenon cannot be properly quantified when using undis-turbed soil columns, as the soil contains many roots and other organicresidues (N2 mm) which were excluded during the determination ofsoil physicochemical properties (see Materials and methods), result-ing in a degree of uncertainty in relation to their contribution to the Nbalance. Apart from the A-horizons at 20 °C (Table 2), the content of N

Page 4: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

Fig. 2. Evolution of NO3− removal rate, and leaching rates of NH4

+ and DOC in the horizon leachates throughout the assay (incubation temperatures: 10 and 20 °C; soil horizons: A, Bgand B2g). Mean values (n=3) for each soil horizon at each sampling time.

766 B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

decreased or remained constant in all tested columns, thus ruling outmicrobial NO3

− assimilation as a process responsible for the observedNO3

− removal. In consequence, in A-columns at 20 °C, where thevalues of N and Corg slightly increased (Table 2), the mineralization ofparticulate organicmatter present in the soil could be also providing Nfor the microbial community. Nevertheless, apart from situations

Table 2Increase (positive value) or decrease (negative value) in soil chemical properties at the end

Horizon

10 °C

A Bg B2g

Corg (g Corgkg−1DW soil) 0.15±0.02a −1.18±0.11b −0.2N (g N kg−1DW soil) 0.00±0.03a −0.20±0.00b −0.2NO3

− (mg N kg−1DW soil) 5.97±0.15a 6.61±0.21b 8.2NH4

+ (mg N kg−1DW soil) −0.30±0.03a −0.18±0.02b −0.2

Mean values±S.E. (n=3). Values followed by different letters or asterisks are significantlysame incubation temperature; asterisks: among incubation temperatures for each soil hori

where microbial biomass development is extensive (e.g., following arelease of readily biodegradable organic matter into the soil or duringvery active bioremediation), it is difficult to foreseemany cases wheresoil microorganisms would take up an amount of nitrate that wouldresult in significant changes in soil nitrogen content (Rivett et al.,2008).

of the leaching assay.

20 °C

A Bg B2g

9±0.03c 0.29±0.04a⁎ −0.35±0.02b⁎ −0.47±0.05c⁎

0±0.02b 0.20±0.05a⁎ 0.00±0.03b⁎ −0.30±0.03c⁎

6±0.25c 3.47±0.11a⁎ 4.86±0.13b⁎ 7.38±0.17c⁎

3±0.03ab −0.40±0.02a⁎ −0.14±0.01b −0.16±0.02b

different (Pb0.05) according to Fisher's PLSD-test (letters: among soil horizons at thezon).

Page 5: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

767B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

On the other hand, dissimilatory reduction of NO3− to NH4

+ can becarried out by fermentative bacteria that use NO3

− as a source ofelectrons (Megonigal et al., 2004). As a matter of fact, in the presenceof abundant Corg, this dissimilatory reduction could occur jointly withdenitrification under reducing environments. Nonetheless, Yin et al.(1998) reported values b5% for this dissimilatory reduction in paddysoils of China. In our study, the presence of NH4

+ in the columnleachates was only relevant in the A-horizons at 10 and 20 °C, as wellas in the B2g-horizon at 10 °C (Fig. 2). In the A-horizon at 20 °C, adecrease in NH4

+ leaching rate was observed for the first 15 days of theleaching assay, which might be due to leaching of the NH4

+ originallypresent in the soil horizon, as reflected by the decrease in column soilNH4

+ content (Table 2). Later, in A-horizons at 20 °C, leachate NH4+

progressively increased until reaching amaximum value of 64.8 μg N–NH4

+kg−1DW soil day−1 (at 106th day), a fact that could beattributed, at least partially, to reduced use of NH4

+ for bacterialgrowth due to scarcity of substrate by the end of the assay (Hefting etal., 2004). On the whole, highest values of NH4

+ leaching rate wereobserved in A-horizons, as compared to Bg- and B2g-horizons,emphasizing the positive relationship between Corg and leachateNH4

+ concentration. If all the NH4+ present in the column leachates

were due to the abovementioned dissimilatory reduction, the amountof NH4

+ found in leachates of A-horizon at 20 °C (highest rates of NH4+

leachingwere detected in these columns)would be equivalent to only1.39% of the NO3

− removed from the 200 mg NO3−L−1 input nitrate

solution. Likewise, the NH4+ present in the column leachates could

come from microbial ammonification (conversion of soil organic N toNH4

+), which in theory should be higher in the A-horizon, ascompared to Bg- and B2g-horizons, thus making the contribution ofdissimilatory reduction even less important (practically, negligible).

The production of CO2 and N2O gas was followed throughout theleaching assays (Fig. 3). Highest values of cumulative CO2 productionwere observed in A-horizons: maximum values of cumulative CO2

Fig. 3. Cumulative production of CO2 and N2O throughout the leaching assay (incubation temhorizon at each sampling time.

production of 736.2 and 315.9 mg CO2kg−1DW soil were found at 20and 10 °C, respectively. In general, the pattern of CO2 productionfollowed the classical first-order kinetics (Monod, 1949): an initialphase of very active growth for the first 10 days (maximum slopes:10.08 and 7.51 mg CO2kg−1DW soil day−1 were found in A-horizonsat 20 and 10 °C, respectively), followed by slower growth untilreaching a plateau.

On the contrary, cumulative N2O production followed a linearpattern (Fig. 3) throughout the experimental period. Again, highestvalues of cumulative N2O production were observed in A-horizons:maximum values of cumulative N2O production of 6.67 and5.63 mg N–N2O kg−1DW soilwere found at 20 and 10 °C, respectively.

Significant correlations were found between rate of NO3− removal

and CO2 and N2O production (Table 3). We carried out an stechio-metric mass balance according to equation “5CH2O+4NO3

−+4H+→2N2+5CO2+7H2O”, in order to estimate the expected amount of CO2

production assuming denitrification as the only cause of NO3− removal

in our columns: estimated average CO2 production values agreed veryclosely with those measured in our columns (99.31 and 95.59% ofvalues measured at 20 and 10 °C, respectively). Although theseestimations cannot be done for N2O (as during denitrification NO3

may end up as both N2O and N2), taking into account N2O/N2 ratiospreviously reported (Well et al., 2003; Ruser et al., 2006) andaccording to our observed values of N2O production, we estimated aconversion of NO3

− toN2 of 64 and 78% for columns incubated at 20 and10 °C, respectively.

Rates of denitrification calculated from values of observed N2Oproduction were lower than those previously reported for theSalburua wetland by Sánchez-Pérez et al. (2003). However, Sánchez-Pérez et al. (2003) calculated potential denitrification rates using theacetylene blockage technique with no carbon or nitrate limitation fordenitrifying communities, while in the present work actual rates ofN2O productionwere quantified. Our data are in agreementwith those

peratures: 10 and 20 °C; soil horizons: A, Bg and B2g). Mean values (n=3) for each soil

Page 6: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

Table 3Linear correlation coefficients (r) between rates of NO3

− removal, and CO2 and N2Oemission.

NO3− CO2 N2O

NO3− 1

CO2 0.767⁎ 1N2O 0.686⁎ 0.753⁎ 1

⁎ Significant at 0.05 level of probability; n=43 measures; NO3−, NO3

− removal rate;CO2, CO2 emission rate; N2O, N2O emission rate.

768 B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

of Casey et al. (2004) for denitrification activity in macropores of ariparian wetland, and Davidsson and Ståhl (2000) in wetland soils.

3.2. Effect of temperature and Corg on denitrification

The activity of denitrifying microorganisms is known to stronglydepend on the presence of available Corg. Actually, Dahl et al. (2007)proposed the utilization of soil Corg as an indicator of denitrificationpotential when other parameters, such as humidity and presence ofNO3

−, were not limiting. Similarly, the activity of denitrifying micro-organisms is temperature-dependent (Dhondt et al., 2004). Inour study,the only parameters varying between columns were soil Corg andincubation temperature. In this respect, for all soil horizons, during thefirst 2–3 weeks of leaching, values of DOC leaching rate, a parameterrelated to both soil Corg and incubation temperature, were higher at 10than at 20 °C (Fig. 2).Maximumvalues of DOC leaching ratewere foundin B2ghorizons: 35.41 and27.09 mg DOC kg−1DW soil day−1 at 10 and20 °C, respectively. Then, unlike most parameters here measured, thehigher the amount of Corg present in the soil horizon, the lower the DOCleaching rate. McCarty and Bremner (1992) attributed this phenome-non to DOC leaching from the topsoil to deeper soil layers (in our case,this phenomenon would have occurred in the wetland itself before soilcolumn sampling). Regarding temperature, the higher values of DOCleaching rate at 10 versus 20 °C can be explained by lower microbialactivity at lower temperature, as indicated by Zsolnay and Steindl(1991).

Likewise, the higher the amount of Corg present in the soil horizon,the higher the average rates of NO3

− removal and CO2 emission(Fig. 4). Highest and lowest average rates of N2O emission were foundin A- and Bg-horizon, respectively.

At 20 °C, the average rate of NO3− removal was 62 and 70% higher

in A-horizons versus Bg- and B2g-horizons, respectively. At 10 °C, theaverage rate of NO3

− removal was 50 and 74% higher in A-horizonsversus Bg- and B2g-horizons, respectively. Finally, the average rate ofNO3

− removal was 42, 24 and 51% lower in A-, Bg- and B2g-horizons,respectively, at 10 versus 20 °C. Furthermore, it was found that thehigher the average rate of NO3

− removal, the lower the amount of soil

Fig. 4. Effect of soil horizon on average rates of NO3− removal and CO2 and N2O emission rate

different for each parameter (Pb0.05) according to Fisher's PLSD-test (letters: among soil hofor each soil horizon).

NO3− content present in the column (Table 2); this phenomenon was

more accentuated at 10 than at 20 °C.The effect of temperature was more pronounced in A-horizons, as

compared to Bg- and B2g-horizons: average CO2 emission rates were57, 30 and 36% lower in A-, Bg-, and B2g-horizons, respectively, at 10than at 20 °C.

Regarding N2O, a different pattern was observed (Fig. 4): forinstance, at 20 °C, average rates of N2O emission in A-horizons were58 and 35% higher than in Bg- and B2g-horizons, respectively. The lowsolubility of N2O gas in water might negatively affect its diffusion andtransport from the column soil to the leachate-gas chamber. Šimek etal. (2004) found that values of N2O concentration in the pores ofcolumn soil were higher than values of N2O emission, suggesting anaccumulation of this gas in the soil pores. In our study, Bg-horizonswere longer than A- and B2g-horizons (A-horizons: 20 cm; Bg-horizons: 50 cm; B2g-horizons: 30 cm) and, consequently, subjecteda priori to more diffusion barriers, which could explain the lowervalues of N2O emission in Bg-horizons versus B2g-horizons, despitehaving higher values of Corg and average NO3

− removal rates. Besides,the solubility of N2O in water is 25% higher at 10 than at 20 °C, whichcould also influence the observed values of N2O production at 10 and20 °C. Finally, some authors (Avalakki et al., 1995) have found that,during denitrification, the lower the temperature the higher the N2O/N2O+N2 ratio, as the reduction of N2O to N2 has a higher activationenergy than the production of N2O.

3.3. Application to Salburua wetland

Our results are in agreement with the nitrate removal ratesreported for Salburua by other authors (García-Linares et al., 2003;Sánchez-Pérez et al., 2003) and highlight the key role that denitrifi-cation plays in the NO3

− removal observed in the Salburua riparian soil.Considering the rate of NO3

− removal observed in B2g-horizon at10 °C (in our study area, average annual groundwater temperature is10.2 °C; piezometric level=85 cm), and a soil bulk density of1200 kg m−3, we estimated a maximum nitrate removal capacity of1011.8 kg N–NO3

−ha−1 year−1, which is slightly higher than thosepreviously reported by other authors for riparian soils (Leonardson etal., 1994; Hefting et al., 2006; Wang et al., 2009).

The groundwater nitrate loss taking place in Salburua indicatesthat the natural nitrate removal capacity of the wetland exceedsnitrate inputs; thus, despite NO3

− content in the groundwater enteringthe wetland being N50 mg NO3

−L−1, values of NO3− content in the

water leaving the wetland are frequently below detection limit.However, nitrate removal by denitrification results in the

emission of significant amounts of greenhouse gases. According torates for B2g-horizon at 10 °C, we estimated maximum emissions of5620 kg CO2ha−1 year− 1 and 240 kg N–N2O ha− 1 year−1. N2Oemissions agreed with those reported by Machefert et al. (2004)

s. Mean values±S.E. (n=43) followed by different letters or asterisks are significantlyrizons at the same incubation temperature; asterisks: among incubation temperatures

Page 7: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

769B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

in two riparian ecosystems draining different agricultural fields. CO2

emissions, however, were significantly lower than those observed byTufekcioglu et al. (2004) in vegetated riparian buffer zones. In ourstudy, N2O emissions were one order of magnitude lower than CO2

emissions. In terms of equivalent emissions of CO2 and according tothe N2O–CO2 conversion factor proposed by IPCC (2001), our N2Oemissions correspond to 113.07 t CO2equivalent ha−1 year−1.

These estimated greenhouse gas emission rates highlight theenvironmental interest of the Salburua wetland (and similar riparianwetlands): the riparian zone of the Salburua wetland removes, viadenitrification, a considerable amount of NO3

− from the pollutedgroundwater (NO3

− pollution of agricultural origin) with a concomitantlocal beneficial effect on the quaternary aquifer of Vitoria-Gasteiz;unfortunately, during the process of NO3

− removal, a significantproduction of greenhouse gases (CO2 and N2O) does occur, withnegative consequences at a global scale (i.e., climate change). Thisrepresents a paradoxical example of how large-scale wetland andriparian restoration efforts to reduceNO3

− delivery towater streams andimprove water quality, may significantly affect regional and global CO2

and N2O budgets.

4. Conclusions

The new experimental design for leaching assays with undisturbedsoil columns here proposed has proven most useful when ground-water flow (rising flow to simulate the natural flow of water comingup from the groundwater towards thewetland)wants to be simulatedunder different environmental conditions (temperature, soil Corg) andthe composition of leachates and gases, as well as changes in soilphysicochemical properties, needs to be determined.

It was concluded that the Salburua wetland riparian zone canremove considerable amounts of NO3

− from polluted groundwaterthrough denitrification. Higher denitrification rates were found at 20than at 10 °C. It was observed that the higher the amount of Corg in thesoil horizon, the higher the NO3

− removal rate (higher values of NO3−

removal were found in A-horizons versus Bg- and B2g-horizons).During denitrification, considerable amounts of N2O and CO2 wereemitted to the atmosphere, with concomitant negative consequencesfor the environment. When considering the beneficial effect providedby Salburua wetland in terms of NO3

− removal from pollutedgroundwater (NO3

− pollution of agricultural origin) via denitrification,it is essential to always take into consideration the environmentally-negative emission of greenhouse gases coming from denitrification.

Further research on the transformation processes of pollutants ofagricultural origin occurring in different ecosystems is needed. In thisrespect, more sustainable agricultural practices must be urgentlypromoted in an attempt to find a compromise between crop yield andenvironmental health.

Acknowledgements

This study was financially supported by the Spanish Ministry ofScience and Technology through project CGL2006-06485, and by theBasque Government through “Consolidated Groups” program (ITE-516-10) and ETORTEK BERRILUR II (IE06-179). Borja Muñoz-Leoz isthe recipient of a fellowship from the Spanish Ministry of Science andTechnology. The authors sincerely thank the anonymous reviewers fortheir thorough revision and invaluable comments which have helpedto significantly improve this paper.

References

APHA (American Public Health Association, American Water Works Association andWater Pollution Control Federation). In: Eaton AE, Clesceri LS, Rice EW, GreenbergAE, editors. Standard methods for the examination of water and wastewater. 21stedition. Washington, DC: American Public Health Association; 2005.

Arrate I, Sánchez-Pérez JM, Antigüedad I, Vallecillo MA, Iribar V, Ruiz M. Groundwaterpollution in quaternary aquifer of Vitoria-Gasteiz (Basque Country, Spain). EnvironGeol 1997;30:257–65.

Avalakki UK, Strong WM, Saffigna PG. Measurement of gaseous emissions fromdenitrification of applied nitrogen− 15. II. Effects of temperature and added straw.Aust J Soil Res 1995;33:89–99.

Casey RE, Taylor MD, Klaine SJ. Localization of denitrification activity in macropores of ariparian wetland. Soil Biol Biochem 2004;36:563–9.

DahlM,NilssonB, Langhoff JH, Refsgaard JC. Reviewof classification schemes andnewmulti-scale typology of groundwater–surface water interaction. J Hydrol 2007;344:1-16.

Davidsson TE, Stepanauskas R, Leonardson L. Vertical patterns of nitrogen transformationsduring infiltration in two wetland soils. Appl Environ Microbiol 1997;63:3648–56.

Davidsson TE, Ståhl M. The influence of organic carbon on nitrogen transformations infive wetland soils. Soil Sci Soc Am J 2000;64:1129–36.

Dhondt K, Boeckx P, Hofman G, Van Cleemput O. Temporal and spatial patterns ofdenitrification enzyme activity and nitrous oxide fluxes in three adjacent vegetatedriparian buffer zones. Biol Fertil Soils 2004;40:243–51.

Egiarte G, Camps Arbestain M, Ruíz-Romera E, Pinto M. Study of the chemistry of anacid soil column and of the corresponding leachates after the addition of ananaerobic municipal sludge. Chemosphere 2006;65:2456–67.

Galloway JN, Townsend AR, Erisman JW, Bejunda M, Cai Z, Freney JR, et al.Transformation of the nitrogen cycle: recent trends, questions, and potentialsolutions. Science 2008;320:889–92.

García-Linares C, Martínez-Santos M, Martínez-Bilbao V, Sánchez-Pérez JM, AntigüedadI. Wetland restoration and nitrate reduction: the example of the peri-urbanwetland of Vitoria-Gasteiz (Basque Country, North Spain). Hydrol Earth Syst Sci2003;7:109–21.

Groffman PM, Altabet MA, Böhlke JK, Butterbach-Bahl K, David MB, Firestone MK, et al.Methods for measuring denitrification: diverse approaches to a difficult problem.Ecol Appl 2006;16:2091–122.

Hefting M, Clément JC, Dowrick D, Cosandey AC, Bernal S, Cimpian C, et al. Water tableelevation controls on soil nitrogen cycling in riparian wetlands along a Europeanclimatic gradient. Biogeochemistry 2004;67:113–34.

Hefting M, Beltman B, Karssenberg D, Rebel K, Van Riessen M, Spijke M. Water qualitydynamics and hydrology in nitrate loaded riparian zones in the Netherlands.Environ Pollut 2006;139:143–56.

Hill AR, Devito KJ, Campagnolo S, Sanmugadas K. Subsurface denitrification in a forestriparian zone: interactions between hydrology and supplies of nitrate and organiccarbon. Biogeochemistry 2000;51:193–223.

Höring H, Chapman D. Nitrates and nitrites in drinking water. World HealthOrganization drinking water series. London: IWA Publishing; 2004.

IPCC (Intergovernmental Panel on Climate Change). Climate change 2001: the scientificbasis. In: Houghton JT, Ding Y, Griggs DJ, Noguer M, Van der Linden PJ, editors.Contribution of Working Group I to the Third Assessment Report of theIntergovernmental Panel on Climate Change. United Kingdom and New York:Cambridge University Press; 2001.

JØrgensen PT, Urup J, Helstrup T, Jensen MB, Eiland F, Vinther FP. Transport andreduction of nitrate in clayey till underneath forest and arable land. J ContamHydrol 2004;73:207–26.

Leonardson L, Bengtsson L, Davidsson T, Persson T, Emmanuelsson U. Nitrogenretention in artificially flooded meadows. Ambio 1994;23:332–41.

Lewis J, Sjöstrom J. Optimizing the experimental design of soil columns in saturated andunsaturated transport experiments. J Contam Hydrol 2010;115:1-13.

Liu L, Greaver TL. A review of nitrogen enrichment effects on three biogenic GHGs: theCO2 sink may be largely offset by stimulated N2O and CH4 emission. Ecol Lett2009;12:1103–17.

Machefert SE, Dise NB, Goulding KWT, Whitehead PG. Nitrous oxide emissions fromtwo riparian ecosystems: key controlling variables. Water Air Soil Poll Focus2004;4:427–36.

McCarty GW, Bremner JM. Availability of organic carbon for denitrification of nitrate insubsoils. Biol Fert Soils 1992;14:219–22.

Megonigal J, Hines M, Visscher P. Anaerobic metabolism: linkages to trace gases andaerobic processes. In: Schlesinger WH, editor. Biogeochemistry. United Kingdom:Elsevier-Pergamon; 2004.

Monod J. The growth of bacterial cultures. Annu Rev Microbiol 1949;3:371–94.Pauwels H, Talbo H. Nitrate concentration in wetlands: assessing the contribution of

deeper groundwater from anions. Water Res 2004;38:1019–25.Ranalli AJ, Macalady DL. The importance of the riparian zone and in-stream processes in

nitrate attenuation in undisturbed and agricultural watersheds — a review of thescientific literature. J Hydrol 2010;389:406–15.

Rivett MO, Buss SR, Morgan P, Smith JWN, Bemment CD. Nitrate attenuation ingroundwater: a review of biogeochemical controlling processes. Water Res2008;42:4215–32.

Ruser R, Flessa H, Schilling R, Beese F, Munch JC. Effect of crop type-specific soilmanagement and N fertilization on N2O emissions from a fine-loamy soil. Nutr CyclAgroecosys 2006;59:177–91.

Sánchez-Pérez JM, Iribar A, Martínez M, García-Linares C, Antiguedad I. Eliminación denitratos por desnitrificación en la zona no saturada del suelo en un humedal delcinturón peri-urbano de Vitoria-Gasteiz. In: Álvarez-Benedí J, Marinero P, editors.Estudios de la Zona no Saturada del Suelo VI. Valladolid; 2003. p. 385–90.

Šimek M, Elhottová D, Klimeš F, Hopkins DW. Emissions of N2O and CO2, denitrificationmeasurements and soil properties in red clover and ryegrass stands. Soil BiolBiochem 2004;36:9-21.

Sparks DL, Page AL, Helmke PA, Loeppert RH, Soltanpour PN, Tabatabai MA, et al.Methods of soil analysis: part 3 chemical methods. Soil Science Society of AmericaBook Series, Madison, WI; 1996.

Page 8: Nitrogen transformations and greenhouse gas emissions from a riparian wetland soil: An undisturbed soil column study

770 B. Muñoz-Leoz et al. / Science of the Total Environment 409 (2011) 763–770

Tiedje JM. Denitrification. In: Page AL, Miller RH, Keeney DR, editors. Methods of soilanalysis part 2. Madison, WI: ASA and SSSA; 1982. p. 1011–26.

Tufekcioglu A, Raich JW, Isenhart TM, Schultz RC. Soil respirationwithin riparian buffersand adjacent crop fields. Plant Soil 2004;229:117–24.

Wang X, Yang S, Mannaerts CM, Gao Y, Guo J. Spatially explicit estimation of soildenitrification rates and land use effects in the riparian buffer zone of the largeGuanting reservoir. Geoderma 2009;150:240–52.

Well R, Augustin J, Meyer K, Myrold DD. Comparison of field and laboratorymeasurement of denitrification and N2O production in the saturated zone ofhydromorphic soils. Soil Biol Biochem 2003;35:783–99.

Yin SX, Shen Q, Tang Y, Cheng LM. Reduction of nitrate to ammonium in selected paddysoils of China. Pedosphere 1998;8:221–8.

Zhao T, Xu H, He Y, Tai C, Meng H, Zeng F, et al. Agricultural non-point nitrogenpollution control function of different vegetation types in riparian wetlands: a casestudy in the Yellow River wetland in China. J Environ Sci 2009;21:933–9.

Zsolnay A, Steindl H. Geovariability and biodegradability of the water-extractableorganic material in an agricultural soil. Soil Biol Biochem 1991;23:1077–82.