SSSAJ
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 29 May 2008
Published in Soil Sci Soc Am J 72:908-916 (2008)
DOI: 10.2136/sssaj2007.0353
© 2008 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Agricola
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Related Collections
Right arrow Isotopes
Right arrow Nitrogen
Right arrow Biogeochemical Processes
Right arrow Frozen Soils
Right arrow Field evaluation techniques

SOIL PHYSICS

Linking Nitrous Oxide Flux During Spring Thaw to Nitrate Denitrification in the Soil Profile

C. Wagner-Riddlea,*, Q. C. Hub, E. van Bochovec and S. Jayasundaraa

a Dep. of Land Resource Science Univ. of Guelph Guelph, ON Canada N1G 2W1
b Institute of Soil Science and Fertilizer Shanxi Academy of Agricultural Science Taiyuan, Shanxi P.R. China
c Soils and Crops Research and Development Center Agriculture and Agri-food Canada, 2560 Hochelaga Blvd., Sainte-Foy, QC, Canada, G1V 2J3

* Corresponding author (cwagnerr{at}uoguelph.ca).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The importance of spring thaw nitrous oxide (N2O) fluxes to the total N2O emission budget in cold climates has been recognized recently. Two mechanisms have been proposed to explain the burst in N2O fluxes due to soil freezing and thawing: enhanced microbial activity due to increased nutrient availability at spring thaw, and release of N2O trapped at depth during winter. The objective of this study was to determine whether increased surface N2O fluxes were due to physical release at spring thaw of N2O accumulated all winter at depth in the soil profile, or whether fluxes were due to rapid N2O production in the surface layer during the thaw process. Micrometeorological flux measurements and a chamber method applied to in situ soil columns receiving 15N tracer were used in Ontario, Canada during winters of 2003 and 2004. Labeled K15NO3 fertilizer (60% excess 15N) at the rate of 100 kg N ha–1 was applied to two layers, that is, surface layer 0 to 5 cm (SL) and deep layer 12 to 17 cm (DL) in nondisturbed soil columns placed in the field during the winter. The burst in N2O fluxes from the soil surface measured by both methods occurred within the same period of soil thawing. Denitrification was the main mechanism responsible for N2O production, and conditions conducive to N2O and N2 production occurred both in the SL and DL during thawing. Despite high 15N2O concentrations at depth, the burst in N2O fluxes from DL soil columns were 1.5 to 5 times lower than that from SL soil columns as more N2O from DL was converted to N2 before diffusing out of the soil profile. Comparison of N2O fluxes originating from SL and DL soil columns indicates that the source of N2O burst at spring thaw is mostly ‘newly’ produced N2O in the surface layer, and not the release of N2O trapped in the unfrozen soil beneath the frozen layers.

Abbreviations: DL, deep layer 12 to 17 cm • SL, surface layer 0 to 5 cm


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Atmospheric concentrations of N2O, a gas involved in the destruction of stratospheric ozone (Crutzen, 1970) and in the enhancement of the greenhouse effect (Wang et al., 1976), have increased from 285 to 310 ppbv during the last 200 yr (Stauffer and Neftel, 1988). Approximately 90% of total anthropogenic emissions are due to agricultural sources (Duxbury et al., 1993) which has motivated the interest in N2O production from agricultural management practices and mitigation of N2O emissions. In cold climates, Wagner-Riddle et al. (1997) and Röver et al. (1998) reported that N2O fluxes during the spring thaw accounted for up to 70% of the annual N2O emissions from soils. For snow-covered agricultural soils in eastern Canada, van Bochove et al. (2000, 2001) observed N2O fluxes during winter and spring that were comparable or higher than growing season fluxes.

Laboratory studies comparing {gamma}-ray sterilized and nonsterilized soils have indicated that microbial processes were responsible for N2O production in frozen and thawing soils (Röver et al., 1998). Denitrification has been identified as the main microbial process leading to high N2O production during freeze-thaw cycles, through the use of 15N tracers in laboratory incubations (Müller et al., 2002; Mørkved et al., 2006). Increased denitrification due to availability of easily degradable carbon (Sehy et al., 2004) and nitrate (Müller et al., 2002), and increased soil water content due to snow melting (Nyborg et al.,1997) are some of the factors that have been suggested as causing intensive N2O production under freeze/thaw conditions.

Under field conditions at spring thaw, large N2O fluxes have been attributed to contrasting mechanisms: (i) physical release of N2O accumulated over winter and trapped in the unfrozen subsurface of frozen soils (Bremner et al., 1980; Burton and Beauchamp, 1994) or within liquid water films present in frozen soil (Teepe et al., 2001), and/or (ii) to N2O production due to increased biological activity in the soil's top layer that is specifically associated with spring thaw (Nyborg et al.,1997; Lemke et al., 1998). The physical release mechanism also involves microbial processes, but the processes occur over the entire time of freezing. Nitrous oxide accumulated at depth and/or in unfrozen water films over winter is only released once the ice barrier melts. However, the two proposed mechanisms are quite different with regards to which depth (surface or deep layer) predominantly contributes to increased N2O fluxes at spring thaw under field conditions. Studies contributing to increased understanding of these mechanisms are needed, particularly for process-based models of N2O emissions (e.g., Li, 2007).

The objective of this study was to determine whether increased surface N2O fluxes were due to physical release at spring thaw of N2O accumulated all winter at depth in the soil profile, or whether fluxes were due to rapid N2O production in the surface layer during the thaw process. We hypothesized that N2O surface fluxes could be traced to a precursor substrate using the stable isotope 15N placed at two soil depths, hence determining the depth signature of surface N2O fluxes.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Site
The experiment was conducted at the Elora Research Station (43°39' N, 80°25' W, 376 m elev.), Ontario, Canada, as part of a project initiated in May 2000. This project focused on two management practices, conventional (CP) and best management practices (BMP), and their role in environmental N losses (Jayasundara et al., 2007; Wagner-Riddle et al., 2007). The study reported here was performed in BMP plots in 2003, and both BMP and CP plots in 2004. However, the focus was not on a comparison of management practices, but rather on the soil processes leading to N2O fluxes at spring thaw. In this region minimum air temperatures below 0°C occur, on average, for 148 d over the November to April period, with a snow cover ( >5 cm) averaging 65 d over this same period (Environment Canada, 2007). The soil at the experimental site is a Guelph loam (fine loamy, mixed, mesic Glossoboric Hapludalf). The soil texture consists of 29% sand, 52% silt, and 19% clay, with a carbon content of about 30 g C kg–1 in the 0 to 10 cm soil layer. The two agricultural management practices, noted above, have been compared on four 1.5 ha plots (150 by 100 m, two replicates each). Briefly, CP involved spring or fall plowing and N fertilization using general provincial rate guidelines applied at planting. The BMP included no-tillage, N fertilization based on soil N test and the use of cover crops when possible. A corn (Zea mays L.)–soybean (Glycine max L.)–winter wheat (Triticum aestivum L.) rotation was initiated at the site in 2000. Plots studied in winter 2003 had been harvested in August 2002, and wheat stubble (about 30 cm) was left on the soil surface with red clover (Trifolium pretense L.) (BMP plot only), which had been under-seeded to wheat in April 2002. Nitrogen fertilizer rate had been applied at 60 kg N ha–1 to the wheat crop in early April 2002. Plots studied in winter 2004, were fertilized with urea at 150 kg N ha–1 broadcast to CP at planting in May 2003, and urea–ammonium–nitrate solution (28% N) injected as side-dress at 60 kg N ha–1 to BMP plots in June 2003 (at corn sixth leaf stage). After corn harvest in November 2003, CP plots were plowed, but 30-cm high crop stubble was left on the soil surface in BMP plots. A detailed description of agronomic practices used can be found in Jayasundara et al. (2007).

Soil Gas Sampling and Gas Flux Measurement
Winter 2003
Twelve undisturbed soil columns were collected from the two 1.5 ha BMP plots by pushing PVC tubing (10 cm i.d. by 20 cm long) into the soil through gently tapping a hammer on its edges and using a steel ring as a guide on 25 Feb. 2003, day of year (DOY) 56. As explained above, these plots had wheat stubble plus a dormant cover crop on the surface, and were only frozen to 2 cm depth. It was not possible to collect soil columns from CP plots due to extensive soil freezing in those plots at that time. The collected soil columns were removed and transported to the laboratory for 1 d, where they received the treatments explained below. During this period, columns where subjected to room temperature but did not thaw completely, and stubble and red clover were removed from the surface of the soil columns.

In the laboratory, soil columns were instrumented and received 15N treatments (Fig. 1 ). The treatments with four replicates were: (i) control (no fertilizer) (NF), (ii) 15N tracer applied to the 0- to 5-cm SL, and (iii) 15N tracer applied to the 12- to 17-cm DL. The 15N tracer was applied as 6 mL of a K15NO3 solution (60% excess 15N). Six 1 mL injections were made into each soil column as described below using a 1 mL syringe (Becton Dickinson, Franklin, NJ). The total N rate applied equaled 114 mg kg–1 or 100 kg N ha–1. The 1 mL injection volume was purged from the syringe over the 0 to 5 cm or 12- to 17-cm layer, after the needle was pushed into the soil to 5 cm (from top of soil column) or 12 cm (from bottom), respectively.


Figure 1
View larger version (41K):
[in this window]
[in a new window]

  		Fig. 1. Schematic diagram of undisturbed soil column with silicone gas probes used in field experiment: (a) control, no N fertilizer added (NF); (b) K15NO3 tracer applied to 0 to 5 cm surface layer (SL); (c) K15NO3 tracer applied to 12 to 17 cm deep layer (DL).

 
Gas sampling probes were installed at depths of 2.5, 7.5, and 15 cm (Fig. 1). These were made using 7 cm lengths of silicone tubing (16 mm i.d. by 20.8 mm o.d.-Cat No: A-06411–82, Cole-Parmer Instrument), with an internal volume of 15 mL, and were inserted through the PVC casing after predrilling a hole and extracting a small volume of soil. One end of the silicone tubing was sealed with a silicone rubber septum, and the other end was connected to 5 cm of polyethylene tubing (3/8#, Nalgene*489) using brass fittings (R1CJKA029b reducing union and R1CAS03108 male connector, Swagelock). A male lure lock stopper (80891, 00B-Vygon) was connected to PVC spaghetti tubing (1.5 m length by 1.27 mm i.d., Chapin Watermatics, Waterdown, NJ) with 2 cm of flexible tubing (3.18 mm i.d., Saint-Gobain Performance Plastics, Bridgewater, NJ).

On 26 February (DOY 57), 1 d after column removal from the field, when the soil column gas sampler installation was completed, the soil columns were returned to the field. Because the original holes where soil cores had been removed from the field were filled with snow, it was necessary to dig a large hole to place the soil columns and then fill-in soil around the columns. Sampling for soil gas samples and surface gas flux occurred during March and April, starting on DOY 77 in 2003 when air temperature warmed above 0°C, and every 2 to 3 d thereafter until DOY 119. Soil gas samples (12 mL) from each soil depth and from the column headspace (12 mL) were taken using a 30 mL syringe and injected into pre-evacuated 10 mL septum-capped serological tubes (Exetainer, Labco, High Wycombe, UK) (Stevens et al., 1993).

Surface gas fluxes were obtained with a closed chamber approach. The chamber consisted of a 5.0 cm high PVC cap with a 3.3 cm high PVC section with the same diameter as the PVC column (10 cm i.d.) glued into its interior. An O-ring (10 cm diam., 0.2 cm thick) was placed at the edge of the 3.3 cm high PVC section. During gas sampling, the chamber was placed on the 1.5 cm-high PVC wall of the soil column protruding above the soil surface, giving a chamber height of 5.0 cm. The total headspace volume of the chamber was 392.5 mL. A rubber septum (Z512117, Sigma-Aldrich) located at the top of the chamber permitted for gas sampling. Petroleum jelly (DIN 00365351, Lever Ponds) and a rubber bungee cord were used to ensure a good seal between chamber and soil column.

Gas samples (12 mL) were taken at 1200 h after a 2 h chamber placement according to the method of Stevens and Laughlin (1994, 2001), before soil gas samples were taken. Typical flux chamber deployment involves using multiple gas samplings measurements, usually taken over <1 h, to allow for derivation of the concentration gradient over time (Rochette and Hutchinson, 2005). Use of chambers in the 15N gas-flux method used here required 2 h deployment (Stevens and Laughlin, 1998) and due to cost constraints did not allow for multiple samplings.

The 15N enrichments of the N2O and N2 gas samples were determined using automated gas-phase continuous-flow isotope-ratio mass spectrometry at the Agriculture and Environment Science Division laboratory (AESD), Belfast, Ireland, according to Stevens et al. (1993). Concentrations of N2O and N2, and their associated fluxes were calculated according to Stevens and Laughlin (1998). Gas analyses were performed within 2 to 6 mo after sampling. This delay has implications for N2O concentration values obtained, as decreases in concentration of test mixtures due to diffusion losses and septum adsorption have been observed over time (~20% over 20 wk), however, enrichment values were not affected (Laughlin and Stevens, 2003). As our experiment involved comparison of N2O fluxes between treatments, conclusions obtained should not be affected by these losses.

Winter 2004
In 2004, the same treatments (NF, SL, DL) were studied on both BMP and CP plots. However, the procedure of soil column preparation was different. Columns were not brought back to the laboratory, but prepared in situ. In addition, the application of 15N fertilizer solution in the DL treatment occurred at the start of winter (early January), while the SL treatment received the fertilizer application closer to the time of thawing (mid-February). The in situ preparation avoided the disruption of the surrounding soil and difficulties with placement of the soil column back into its original position, which had occurred in 2003. A drawback of the in situ preparation was the impossibility of installing gas samplers in the soil columns as done in 2003. The earlier timing of fertilizer application in DL treatment was to ensure the potential build up of N2O concentrations deep in the soil profile over the winter. The later timing in the SL treatment was intended to maximize the containment of labeled NO3– in the surface layer, while minimizing leaching to lower depths in the profile.

The PVC tubing (26.5 cm length) was pushed into the soil as per previous year's procedure on 22 Nov. 2003 (DOY 326). The 15N fertilizer solution was injected into the 12 to 17 cm depth on 8 Jan. 2004, when the soil was frozen to a depth of 5 and 15 cm, respectively, in the BMP and CP plots. Six vertical holes were made in each soil column, using a 4 mm steel rod, and 1 mL of the fertilizer solution was injected into each hole as per the 2003 procedure. The small holes were re-filled with soil. The same procedure was followed when injecting the solution into the 0 to 5 cm depth on 12 Feb. 2004 (DOY 43). Surface gas fluxes were obtained with a closed chamber approach, and gas concentrations determined as described above for 2003. Sampling started on DOY 64, when air temperature increased above 0°C and some snow had melted, and every 2 to 3 d thereafter until DOY 99. Soil gas samples at various depths were not obtained in 2004.

Soil Supporting Data
On 29 Apr. (DOY 119) 2003, after conclusion of field measurements, the soil columns were removed from the field and transported to the laboratory then separated into 0- to 5-, 5- to 12-, and 12- to 17-cm sections. A similar procedure was followed in 2004, with sampling on 26 Apr. 2004 (DOY 117), and separation into these same layers and an additional section (17 to 25 cm). Soil NO3–N concentrations were determined using a 15 g wet subsample from each layer after thorough mixing of the soil, and extracting with 50 mL 2 M KCl and shaking for 1 h (Maynard and Kalra, 1993). The 15N enrichment of the NO3–N was determined using the diffusion procedure (Brooks et al., 1989).

Soil temperature profiles were measured with thermistors (Model #107, Campbell Scientific Inc., Edmonton, AB) inserted at 5, 25, and 55 cm depth. Hourly mean temperatures were automatically recorded with a datalogger (21X, Campbell Scientific Inc, Edmonton, AB). Snow depth on the ground was measured on each plot at several points (at least 5) using a ruler. Air temperatures were obtained from the Elora Weather Station located about 100 m from the experimental site.

Micrometeorological Nitrous Oxide Fluxes
The N2O fluxes from the CP and BMP plots during the experimental period were measured using a flux-gradient approach (Wagner-Riddle et al., 1996, 2007). Average hourly N2O concentration differences between two sampling heights were obtained for each plot with a tunable diode laser trace gas analyzer (TGA 100, Campbell Scientific Inc). The vertical fluxes of N2O from the four plots were calculated using the flux-gradient method (Wagner-Riddle et al., 1996). Cup and sonic anemometers were used to calculate the integrated eddy diffusivity between z1 and z2 as a function of the friction velocity, air intake heights and integrated similarity functions for heat for both sampling heights (Wagner-Riddle et al., 1996).

Statistical Analyses
Analyses of variance of the data were performed using the General Linear Model (SAS Institute, 1999) to determine the significance of difference between treatments. The multiple comparisons for the data were tested with the Least Significant Difference (LSD) (P ≤ 0.05).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Weather Conditions and Micrometeorological Nitrous Oxide Fluxes
Air temperature from January to February (DOY 1–59) averaged –9.7°C in 2003, and –8.5°C in 2004. Daily mean air temperature during these months remained much lower than 0°C, except for a brief 2 d period in early January of both years. Average snow depth on the ground was 16 cm for BMP plots in 2003, 15 cm for CP and 20 cm for BMP plots in 2004 over January to February. Over this period daily mean soil temperature at 5 cm depth was –0.1°C for BMP plots in 2003, –1.5°C for CP plots and 0.0°C for BMP plots in 2004. Mean daily N2O fluxes measured with the flux-gradient method over the same period were 2.8 ng N m–2 s–1 in 2003, and 4.5 ng N m–2 s–1 in 2004.

Low air temperatures in January to February were followed by milder air temperatures in early March, with a period of rapid snow melting occurring between DOY 75 and 80 in 2003 (Fig. 2a ), and DOY 60 to 75 in 2004 (Fig. 3a , 4a ). Once the snow had melted, several days of mean air temperature >0°C resulted in increased soil temperatures above 0°C, first at 5 cm depth around DOY 85 in both years, followed by warming of the soil layer at 25 cm (Fig. 2a, 3a, and 4a). Soil temperatures at 5 cm depth warmed more rapidly in CP plots from DOY 84 to 85 in 2004 when compared to BMP plots (from –0.2 to 2.5°C vs. –0.2 to 0.1°C) probably due to the lack of crop residue on the surface and the difference in albedo of the plot surfaces (Fig. 3a and 4a).


Figure 2
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 2. Fluxes and environmental conditions for plots subjected to best management practices (BMP) in 2003: (a) Snow depth, daily mean air, and soil temperature at 5, and 25 cm, and N2O flux measured using the flux-gradient (FG) technique for March and April; (b) N2O fluxes and enrichment and (c) N2 fluxes measured using a closed chamber approach from field soil columns receiving surface layer (SL) and deep layer (DL) applied 15N tracer treatments. Bars indicate the standard errors of means. Gas samples were taken on Days 77, 79, 81, 83, 85, 87, 89, 91, 93, and 100. Dates with missing values are due to gas concentrations below the detection limit.

 

Figure 3
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 3. Fluxes and environmental conditions for plots subjected to conventional practices (CP) in 2004: (a) Snow depth, daily mean air, and soil temperature at 5 and 25 cm and N2O flux measured using the flux-gradient (FG) technique for March and April; (b) N2O fluxes and enrichment and (c) N2 fluxes measured using a closed chamber approach from field soil columns receiving surface layer (SL) and deep layer (DL) applied 15N tracer treatments. Bars indicate the standard errors of means. Gas samples were taken on Days 64, 66, 73, 79, 84, 85, 87, 88, 89, 96, and 99. Dates with missing values are due to gas concentrations below the detection limit.

 

Figure 4
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 4. Fluxes and environmental conditions for plots subjected to best management practices (BMP) in 2004: (a) Snow depth, daily mean air, and soil temperature at 5 and 25 cm and N2O flux measured using the flux-gradient (FG) technique for March and April; (b) N2O fluxes and enrichment and (c) N2 fluxes measured using a closed chamber approach from field soil columns receiving surface layer (SL) and deep layer (DL) applied 15N tracer treatments. Bars indicate the standard error of means. Gas samples were taken on Days 64, 66, 73, 79, 84, 85, 87, 88, 89, 96, and 99. Dates with missing values are due to gas concentrations below the detection limit.

 
Peak N2O fluxes were associated with increases in air temperature above 0°C during the snow melt period, with the main thaw emission period occurring on DOY 74 to 84 in 2003 and 77 to 89 in 2004 (Fig. 2a, 3a, and 4a).

Chamber Nitrous Oxide and Dinitrogen Fluxes
In 2003 the application of 15NO3– to the SL resulted in significantly higher N2O fluxes than when it was applied to the DL treatment (Fig. 2b). Similar trends occurred in 2004 for the CP plots (Fig. 3b). Peak N2O fluxes were 202 ng N m–2 s–1 for SL and 43 ng N m–2 s–1 for DL soil columns in BMP plots during 2003 (Fig. 2b), and 83 ng N m–2 s–1 for SL and 44 ng N m–2 s–1 for DL soil columns in CP plots during 2004 (Fig. 3b). Fluxes were not significantly different between SL and DL treatments of the BMP plots in 2004 with maxima of 34 and 24 ng N m–2 s–1, respectively (Fig. 4b). Nitrous oxide fluxes from soil columns not receiving 15N tracer were not detected due to N2O concentration in headspace being lower than the mass spectrometer detection level (1 ppm N2O).

During the peak flux period the 15N enrichment of the evolved N2O ranged from 10 to 54 atom% (Fig. 2b, 3b, and 4b) under both the SL and DL treatments. This indicates that anywhere from 17 to 90% of the N2O emitted derived from the applied 15N-fertilizer (60 atom%). In 2003 (BMP plots) the maximum 15N enrichment of the evolved N2O from the SL and DL treatments was 53 and 49 atom%, respectively (Fig. 2b). However, in 2004 the 15N enrichment of the N2O was lower, peaking at 36 and 25 atom%, respectively, for the SL and DL treatments in the CP plot (Fig. 3b), and only ~10 atom% for both the SL and DL treatments in the BMP plot (Fig. 4b).

The temporal patterns of N2 fluxes were similar to that of the N2O fluxes during soil thawing (Fig. 2c, 3c, and 4c). The highest N2 fluxes ranged from 475 to 940 ng N m–2 s–1 and coincided in time with the highest N2O fluxes. However, the magnitude of the N2 fluxes from the SL treatment was only higher than the DL treatment in 2003 (Fig. 2c).

Nitrous Oxide and Dinitrogen Soil Concentrations at Different Depths
Nitrous oxide concentrations at 2.5 and 7.5 cm from the SL treatment in BMP 2003, were higher than at 15 cm depth during the main thaw period (DOY 77–83; Fig. 5a ). The peak in concentration at these depths occurred on DOY 79, which was earlier than the occurrence of the highest surface fluxes on DOY 81 for SL columns (Fig. 2b). Significantly higher soil profile N2O concentrations in the DL treatment were found at 15 cm when compared to other depths (Fig. 5b), peaking 4 d before the maximum surface N2O flux (DOY 83; Fig. 2b). The treatment with the highest concentration of N2O in the soil profile, DL (Fig. 5b), was not the treatment with the highest surface flux, which was SL (Fig. 2b). The 15N enrichment of the N2O peak concentrations in the soil profile was approximately 53 atom% (Fig. 5a, 5b) confirming that most of the N2O present was derived from the 15N labeled NO3– added at 2.5 cm for SL and 15 cm for DL treatment. For the soil layers which did not receive 15N tracer, enrichment levels were as high as 50 atom% at 2.5 and 7.5 cm depths for DL on DOY 83 (Fig. 5b), and ranged from 20 to 40 atom% at 7.5 and 15 cm for SL treatment (Fig. 5a).


Figure 5
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 5. Gas concentrations at different soil depths for plots subjected to best management practices (BMP) in 2003: (a) 15N2O concentrations and enrichment at 2.5, 7.5, and 15 cm of soil column receiving 15N tracer at 0 to 5 cm depth (SL, surface layer); (b) 15N2O concentrations and enrichment at 2.5, 7.5, and 15 cm of soil column receiving 15N tracer at 12 to 17 cm depth (DL, deep layer); (c) 15N2 concentration at 2.5, 7.5, and 15 cm of SL soil column; (d) 15N2 concentration at 2.5, 7.5, and 15 cm of DL soil column Bars indicate the standard error of means. Gas samples were taken on Days 77, 79, 81, 83, 85, 87, 89, 91, 93, and 100. Dates with missing values are due to gas concentrations below the detection limit.

 
The highest 15N2 concentrations in the SL treatment were measured at 2.5 and 15 cm depths on DOY 81 and 83 (Fig. 5c), respectively, while they occurred on DOY 81 at 15 cm depth in the DL treatment (Fig. 5d). These peak 15N2 concentrations occurred approximately 2 d after the highest N2O concentrations (Fig. 5a, 5b). The timing of peak 15N2 concentrations at 2.5 cm coincided with the peak in N2 fluxes for the SL treatment (Fig. 5c vs. 2c), but 15N2 concentrations at 15 cm depth preceded the maximum N2 fluxes in the DL treatment by 4 d (Fig. 5d vs. 2d).

Nitrate Tracer in Soil Columns
At the end of April 2003 (DOY 120), significantly higher NO3– contents were found in the soil depths where N fertilizer had been injected (Fig. 6a ). Nitrate content and its 15N enrichment in the 5 to12 cm layer of SL and DL treatments were significantly higher than values for the corresponding layer in the control, indicating that some vertical movement of the 15N tracer had taken place (Fig. 6a, 6b). The movement of labeled nitrate in the soil profile in 2004 was larger than in 2003 (Fig. 7 ). Leaching of 15N tracer from the 0 to 5 cm layer in the SL treatment distributed NO3– to lower depths (Fig. 7a, 7b), so that similar 15N enrichment levels were found in these layers (Fig. 7c, 7d). Likewise, NO3– moved from the 12 to 17 cm layer to the adjacent depths in the DL treatment. Overall, except for CP plots in 2004 (Fig. 7b, 7d), the applied 15N tracer was mostly present in the intended layer, even 20 d after the experiment concluded.


Figure 6
View larger version (7K):
[in this window]
[in a new window]

  		Fig. 6. Characterization of in situ soil columns used in plots subjected to best management practices (BMP) after the experimental period on 29 Apr. 2003 (DOY 119): (a) NO3– content and (b) 15N enrichment of NO3– in the 0 to 5, 5 to 12, 12 to 17 cm layers. Soil columns were obtained from BMP plots at the Elora Research Station and received the following treatments on 26 Feb. 2003 (DOY 54): Control, no 15N tracer; SL, K15NO3 applied to 0 to 5 cm surface layer; DL, K15NO3 applied to 12 to 17 cm deep layer (shown with no-fill bars). Bars indicate the standard error of means.

 

Figure 7
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 7. Characterization of in situ soil columns used in plots subjected to best management practices (BMP) and conventional practices (CP) after the experimental period on 26 Apr. 2004 (DOY 117): (a) NO3– content and (b) 15N enrichment of NO3– in the 0 to 5, 5 to 12, 12 to 17, and 17 to 25 cm layers. Soil columns were obtained from fields at the Elora Research Station and received the following treatments on 8 Jan. 2004 (DOY 8): Control, no 15N tracer; DL, K15NO3 applied to 12 to 17 cm deep layer; and on 12 Feb. 2004 (DOY 43): SL, K15NO3 applied to 0 to 5 cm surface layer. Bars indicate the standard error of means.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Peak micrometeorological N2O fluxes were associated with some thawing of the soil surface layers, although soil layers with temperature below 0°C still existed in the soil profiles (Fig. 2a, 3a, and 4a). This period of intense N2O fluxes only lasted a few days, emphasizing the importance of continuous flux measurements. Frequent chamber flux measurements encompassed the main N2O flux period as determined with the micrometeorological technique (Fig. 2b, 3b, and 4b). Differences between magnitude and times of maxima in fluxes measured with these techniques were most likely related to the addition of 15NO3– tracer, and/or to the effect of PVC wall heating or removal of crop residue (in BMP only) in accelerating soil thawing in columns.

The increase in N2O fluxes at spring thaw was linked to N2O produced in the surface soil layer, rather than N2O produced at depth in the soil profile, as fluxes from the SL soil columns were 1.5 to 5 times higher than DL in two of the three cases studied (Fig. 2b, 3b, 4b). To our knowledge, this is the first time that surface N2O fluxes during spring thaw were attributed to depth of production in the soil profile, through the use of chambers and micrometeorological methods. The 15N enrichment of SL peak fluxes confirmed that NO3– added to the 0 to 5 cm depth of the SL treatment was the precursor to N2O production through denitrification. For the DL treatment, the addition of 15N tracer to the 12 to 17 cm depth only had a limited influence on the N2O fluxes during soil thawing, despite high concentrations of N2O at 15 cm depth (Fig. 2b, 3b, and 5b). The lower 15N enrichment of the N2O flux for the DL treatment also indicated that the NO3– applied was not contributing as significantly to the surface flux as in the SL treatment.

Some labeled N2O (~20–40 atom%) was observed at depths not receiving 15N tracer in each treatment (7.5 and 15 cm in SL; 2.5 and 7.5 cm in DL) (Fig. 5a and 5b). This indicated movement of labeled N2O gas and/or 15NO3– in the soil columns. Nitrate concentration and 15N enrichment in the soil profile at the end of April suggested some movement of the tracer (Fig. 6 and 7), particularly in 2004. It is likely that most of this movement occurred after soil thawing. Hence, it appears that the experimental requirement of high 15NO3– present at the time of thawing of the surface or deep layers was met in the SL and DL treatments, respectively.

Our results show conditions were conducive to denitrification in both the 0 to 5 cm and 12 to17 cm depths (Fig. 5). As soil thawed gradually from the surface down the profile (Fig. 2a, 3a, 4a), snow melt water combined with reduced drainage due to the presence of a frozen soil layer might have presented anaerobic conditions as suggested by Nyborg et al. (1997). At depth, heat transfer from lower unfrozen soil horizons to the frozen layer generates a thawing front beneath the frozen layer (Ferguson et al., 1964). The high water content in this thawing front probably triggered N2O production beneath the frozen layer, and accumulation because diffusion to the surface was impeded. But, as shown here, the highest N2O concentrations in deep soil layers (Fig. 5b) did not result in the highest surface N2O fluxes during soil thawing (DL in Fig. 2b). This is in contrast to the proposal that increased N2O fluxes were due to the release of N2O trapped in the unfrozen layer beneath the frozen layers (Bremner et al., 1980; Burton and Beauchamp, 1994). Van Groenigen et al. (2005) also observed that high N2O concentrations in the soil did not result in high fluxes during winter. They suggested that high water-filled pore space in the soil would have caused a barrier to gas diffusion, leading to complete denitrification. In our study, a large fraction of the N2O observed in the DL treatment was reduced to N2 as evidenced by high 15N2 concentrations at this depth (Fig. 5d). At peak fluxes of both gases, the ratios of N2O–N to (N2O+N2)–N varied from 0.061 to 0.176 for the SL treatment and from 0.047 to 0.061 for the DL treatment. The slightly higher gas ratios in the SL treatment indicate that more complete denitrification occurred in the DL treatment than in the SL treatment. The N2O produced in the surface depth could have diffused more readily to the atmosphere, while the N2O produced in the deeper layer was further reduced to N2 as it diffused out of the soil profile. Our result is supported by the finding of Clough et al. (1999) where the ratio of N2O–N to (N2O+N2)–N decreased with diffusion from the denitrification zone toward the soil surface.

Conditions in the surface layer had a significant effect on N2O fluxes depending on soil management (BMP vs. CP) and year (2003 vs. 2004). For the SL treatment in 2004, BMP plots had significantly lower N2O fluxes than CP (Fig. 3b and 4b), a trend also seen in the micrometeorological fluxes (Fig. 3a and 4a). Wagner-Riddle et al. (2007) using data for the same site, including years 2003 and 2004, showed that no-tillage used in BMP plots significantly reduced N2O emissions during thaw by decreasing soil freezing due to the insulating presence of the deeper snow cover plus corn and wheat residue during winter. Denitrifiers have been shown to easily metabolize the organic compounds released from microbes killed as a result of freezing (Christensen and Tiedje, 1990; Koponen and Martikainen, 2004; Sehy et al., 2004) and from disintegrated aggregates (Christensen and Christensen, 1991; van Bochove et al., 2000) during subsequent thawing. The lack of difference in N2O fluxes between SL and DL treatments in BMP plots during 2004 (Fig. 4b) could be related to less freezing resulting in lower N2O production in the surface layer of BMP plots.

There were two possible sources for the increased N2O fluxes originating from the surface soil layer: (i) the release of trapped N2O in unfrozen water films as proposed by Röver et al. (1998) and Teepe et al. (2001), and/or (ii) ‘new’ N2O produced in the thawed layers above the frozen layers. Trapped N2O in unfrozen water films would lead to high N2O gas concentrations in soil, but we did not detect 15N enriched N2O in samples taken from the frozen 0 to 5 cm depth (SL treatment) during the second freezing cycle on DOY 91 and 93 of 2003 (Fig. 5a), when the depth of the 0°C isotherm descended to 5 cm (Fig. 2a). But, evolved N2O in the SL treatment increased on DOY 100 when soil thawed again (Fig. 2b). In addition, the 15N enrichment of evolved N2O for SL soil columns on DOY 100 and 102 was close to the applied tracer enrichment (Fig. 2b). This suggests that ‘newly’ produced N2O was the source of increased surface fluxes.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The peak N2O fluxes measured by both micrometeorological technique and chamber method occurred within the same period of soil thawing. Through the use of in situ soil columns and 15N tracer, we confirmed that denitrification was the main mechanism responsible for N2O production. The patterns of N2 fluxes were similar to those of the N2O fluxes during soil thawing. Conditions conducive to N2O and N2 production occurred both in the surface layer and at depth in the profile during thawing. Highly 15N enriched N2O concentrations in the deep unfrozen layer during soil thawing were attributed to the development of a thawing front beneath the frozen layer by ground heat. Despite highly 15N enriched N2O concentrations at the 12 to 17 cm depth, the peak N2O fluxes from the DL treatment were 1.5 to 5 times lower than those from SL treatment as subsurface produced N2O was converted to N2 while diffusing upward. This indicates that the source of the N2O peaks seen during spring thaw is mostly ‘newly’ produced N2O in the surface layer, and not the release of N2O trapped in the unfrozen soil beneath the frozen layers.


   ACKNOWLEDGMENTS
 
This project was funded by the Biological Greenhouse Gas Source and Sink (BGSS) program managed by Canadian Agri-Food Research Council (CARC) and Agriculture and Agri-Food Canada (AAFC) under the Canadian Climate Action Fund Initiative, and by the Ontario Ministry of Agriculture, Food and Rural Affairs. Special thanks to R.J. Stevens and R.J. Laughlin at Agriculture and Environment Science Division laboratory, Belfast, Ireland, for gas sample analyses.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication September 27, 2007.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Agricola
Right arrow Articles by Wagner-Riddle, C.
Right arrow Articles by Jayasundara, S.
Related Collections
Right arrow Isotopes
Right arrow Nitrogen
Right arrow Biogeochemical Processes
Right arrow Frozen Soils
Right arrow Field evaluation techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome