SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

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 Similar articles in ISI Web of Science
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 HighWire
Right arrow Citing Articles via ISI Web of Science (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.
Agricola
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.
Related Collections
Right arrow Global Change
Right arrow Plant and Soil Interactions
Right arrow Nitrogen
Published in Soil Sci. Soc. Am. J. 68:493-506 (2004).
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

DIVISION S-3—SOIL BIOLOGY & BIOCHEMISTRY

Emissions of N2O from Alfalfa and Soybean Crops in Eastern Canada

Philippe Rochette*, Denis A. Angers, Gilles Bélanger, Martin H. Chantigny, Danielle Prévost and Gabriel Lévesque

Soils and Crops Research and Development Centre, Agriculture and Agri-Food Canada, 2560 Hochelaga Blvd., Sainte-Foy, QC, Canada, G1V 2J3

* Corresponding author (rochettep{at}agr.gc.ca).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
There is considerable uncertainty relative to the emissions of N2O from legume crops. A study was initiated to quantify N2O fluxes from soils cropped to alfalfa (Medicago sativa L.) and soybean (Glycine max L.), and to improve our understanding of soil and climatic factors controlling N2O emissions from these crops. Measurements were made on three soils cropped to alfalfa, soybean, or timothy (Phleum pratense L.), a perennial grass used as a control. In situ soil-surface N2O emissions (FN2O) were measured 47 times during the 2001 and 2002 growing seasons. Soil water, NH4–N, NO3–N, and N2O contents, and soil temperature were also determined to explain the variation in gas fluxes. Emissions of N2O were small under the grass where very low soil mineral N content probably limited denitrification and N2O production. Soil mineral N contents under legumes were up to 10 times greater than under timothy. However, soil mineral N contents and FN2O were not closely related, thus suggesting that the soil mineral N pool alone was a poor indicator of the intensity of N2O production processes. Higher FN2O were measured under legume than under timothy in only 6 out of 10 field comparisons (site-years). Moreover, the emissions associated with alfalfa (0.67–1.45 kg N ha–1) and soybean (0.46–3.08 kg N ha–1) production were smaller than those predicted using the emission coefficient proposed for the national inventory of greenhouse gases (alfalfa = 1.60–5.21 kg N ha–1; soybean = 2.76–4.97 kg N ha–1). We conclude that the use of the current emission coefficient may overestimate the N2O emissions associated with soybean and alfalfa production in eastern Canada.

Abbreviations: DM, dry matter • WEOC, water-extractable organic carbon


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE ATMOSPHERIC concentration in N2O, a very potent greenhouse gas, is increasing at a rate of 0.25% annually (IPCC, 2001). In Canada, agricultural sources are estimated to emit approximately 80 Mg N2O-N yr–1 or 60% of the total anthropogenic emissions (Janzen et al., 1999). These N2O emissions are largely attributed to nitrification and denitrification of N that is added to the soil to sustain crop productivity. Soil N sources that can result in N2O production and emission include mineral fertilizer, manure, crop residues, and biological fixation of atmospheric N2 by legume crops (Bremner, 1997).

Symbiotic N2 fixation by legumes, such as soybean and pulse crops, is estimated to account for 22% of agricultural N2O emissions in Canada (Desjardins and Riznek, 2000). Despite the magnitude of this estimate, there is limited direct empirical data on the N2O production associated with symbiotic N2 fixation. The presence of legumes could stimulate N2O emissions by increasing N inputs into soils, thus providing additional substrates for nitrification and denitrification. Ta et al. (1986) showed that alfalfa contributes to raise N concentration in soils by excreting N compounds from nodules and by the decomposition of dead tissues in soils. Similarly, Mayer et al. (2003) reported that faba bean (Vicia faba L.) could release 13% of their fixed N as rhizodeposition. Denitrification by N-fixing bacteria can be another source of N2O in legume stands (O'Hara and Daniel, 1985). Breitenbeck and Bremner (1989) observed that free-living cells of soybean rhizobia (Bradyrhizobium japonicum) can denitrify nitrate under anaerobic conditions but concluded that the population of these bacteria is likely too small to influence the rate of denitrification in soils.

The fraction of the N2 fixed by legume crops that is converted to N2O is not well documented and the IPCC greenhouse gas inventory methodology proposes for this source the same emission factor as for mineral fertilizers (1.25%) (IPCC, 1997). Moreover, N2O emissions from legume species used as forage crops (e.g., alfalfa, clover [Trifolium sp.]) are not included in the IPCC methodology. Our limited knowledge of the contribution of legume crops to N2O emissions results in considerable uncertainty for the Canadian inventory of agricultural N2O emissions. Alfalfa and soybean are grown on 0.8 and 1.0 million ha, respectively in eastern Canada, and more accurate estimates of their contribution to N2O emissions are necessary. The objectives of this study were to quantify N2O fluxes during the growing season from soils cropped to alfalfa and soybean, and to improve our understanding of soil and climatic factors controlling N2O emissions from legume crops.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Field Sites
The study was conducted at the Chapais (2001) and Harlaka (2001 and 2002) research farms of Agriculture and Agri-Food Canada, located 3 and 20 km southeast of Québec City, respectively (46°46'N; 71°19' W). At the Chapais site, the soil was a St-Pacôme sandy loam (loamy, mixed, frigid Umbric Dystrochept) (Table 1) and the experimental design consisted of randomized complete blocks with crops of alfalfa, soybean, and timothy as treatments repeated four times in 0.9 by 3 m plots. Alfalfa (Medicago sativa L., cv. AC Caribou) and timothy (Phleum pratense L., cv. Champ) were established in 1998 as part of another study on winter survival of perennial forage crops. Soybean (Glycine max L., cv. Gentleman) was seeded on 22 May 2001 in plots where orchardgrass (Dactylis glomerata L.) had been grown during the three previous years. Seeds were treated with a commercial inoculum (HiStick + 532C strain, Becker Underwood, Ames, IA) and seeded at a rate of 85 kg ha–1 with 18-cm interrows. Soybean received 62.2 kg K ha–1 and 16.1 kg P ha–1 before planting and was harvested on 4 Oct. 2001. Timothy was used as a nonlegume control to estimate background N2O emissions; for this reason, it was not fertilized in 2001.


View this table:
[in this window]
[in a new window]
  		Table 1. Characteristics of the soils at the Chapais and Harlaka sites measured at the onset of the study.

 
At the Harlaka site, plots were initiated in 2001 on a St-André loam (coarse-loamy, mixed, frigid Sombric Fragiorthod) and a Kamouraska clay (fine, mixed, frigid Typic Humaquept) (Table 1) in fields located approximately 200 m apart. The experimental design on the two soils was similar at the Chapais site except for the larger plot dimensions (3.1 x 8 m). Alfalfa (cv. Apica; 12 kg ha–1; commercial inoculum: Nitragin [Balzac strain], Nitragin, Brookfield, WI) and timothy (cv. Champ; 10 kg ha–1) were seeded between 18 and 22 May 2001. Oat (Avena sativa L., cv. AC Baton; 80 kg ha–1) was underseeded with timothy. Soybean (cv. Gentleman; 85 kg ha–1; 18-cm interrows; commercial inoculum: HiStick + [532C strain], Becker Underwood, Ames, IA) was seeded on 21 May 2001 and 22 May 2002. Preplanting fertilization of all crops was 62.2 kg K ha–1 and 16.1 kg P ha–1. Timothy plots received 30 kg N ha–1 only after the harvest of 31 July 2001 to improve establishment. Weeds were controlled by manual weeding and by use of herbicides.

Soil-Surface Gas Flux
At the Chapais site, in situ soil–surface gas fluxes were measured 20 times from 23 May to 22 Oct. 2001 in the alfalfa and timothy plots, and from 27 June to 22 Oct. 2001 in the soybean plots. Measurements on the two soils at the Harlaka site were made 10 times from 4 July to 24 Oct. 2001. The later starting date for measurements at the Harlaka than at the Chapais site was to allow for a good crop establishment before inserting the chamber frames in the soil. In 2002, measurements were made only at the Harlaka site and covered the entire snow-free period from 17 Apr. to 23 Oct. (29 sampling dates). Measurements were made weekly except during the weeks following alfalfa and timothy harvests when fluxes were measured three times. Two clear acrylic frames (0.55 by 0.55 m; 0.14-m height; 6.35-mm wall thickness) were inserted to a depth of 0.10 m in each plot 24 h before the first measurement. The frames were left at the same locations for the duration of the experiment. The frames height was measured at regular intervals during the experiment (48 measuring points per frame) to account for variations in headspace due to soil settling.

Soil-surface N2O fluxes were measured by the nonflow-through nonsteady-state chamber method (Livingston and Hutchinson, 1995). At sampling time, the frames were covered by a vented insulated square acrylic plastic chamber (height: 0.31 m) covering the same area as the frames. During each N2O flux measurement, the chamber was fixed to a frame and air samples were taken through a rubber septum at regular intervals (0, 25, 50, and 75 min after deployment in 2001 and 0, 12, 24, and 35 min after deployment in 2002). Air samples were taken using a 20-mL polypropylene syringe (Becton Dickinson, Rutherford, NJ) and pressurized (200 kPa) into pre-evacuated vials (12-mL Exetainer, Labco, High Wycombe, UK) in which 3 mg of magnesium perchlorate were placed to absorb water vapor.

Measurements were always made between 1000 and 1200 h (Eastern Standard Time) and fluxes of N2O (FN2O) were calculated using the following equation (Rochette and Hutchinson, 2004):

where the rate of change of chamber N2O concentration ({partial}C/{partial}t) is determined in dry air samples, V is the chamber volume, A is the soil area covered by the chamber, Mm is the molecular weight (44 g), Mv is the molecular volume at predeployment air temperature (0.0224–0.024 m3 mol–1), ea (kPa) is the predeployment partial pressure of water vapor in air, and P (kPa) is the barometric pressure.

Cumulative N2O-N losses were calculated by linearly interpolating N2O emissions between sampling dates, assuming that measurements made between 1000 and 1200 h provided a valid estimation of average daily FN2O. The IPCC procedure was used as recommended for pulse crops and for N fertilizer applications, and adapted for alfalfa to estimate annual N2O emissions: timothy = N fertilizer rate x 1.25%; alfalfa N fixation = harvested N x 1.25%; soybean N fixation = {grain N + [2.1 x grain dry matter (DM) x 0.03 kg N kg–1 DM]} x 1.25%; soybean crop residues = (2.1 x DM x 0.03 kg N kg–1 DM) x 1.25%, for a soybean residue/crop ratio of 2.1 (IPCC, 1997).

Concentrations of O2, CO2, and N2O in the soil profile were monitored simultaneously with flux measurements by taking soil air samples within 2 m of chamber frames using gas probes. The probes were made of plastic mesh cylinders (10-cm length; 3.5-cm diameter) containing 3-mm glass beads. The cylinders were inserted horizontally at a depth of 7.5 cm and connected to the surface using plastic tubes (Bev-a-line IV, Ryan Herco Industrial Plastics, Seattle, WA) (0.60-m length; 6.35-mm o.d.; 3.18-mm i.d.) with two-way Luer-type stopcock valve (Cole Parmer, Vernon Hills, IL) and a polyisoprene septum (male luer-lock stopper with injectable membrane, Vygon, Ecouen, France) at the surface end. Air samples were collected with a 20-mL polypropylene syringe after expelling 20 mL of air from the tubes to account for the dead volume of the tubes. Samples were stored at 200 kPa in vials and handled as the chamber air samples.

Chamber and soil air samples were analyzed for N2O concentrations within 2 wk of sampling by means of a gas chromatograph fitted to electron capture detector (Model 3800. Varian, Walnut Creek, CA) equipped with a headspace autoinjector (Combi Pal, CTC Analytics, Zurich, CH). Concentrations in O2 and CO2 were also determined in the soil air samples on the same instrument using thermal conductivity and flame ionization detectors (Rochette and Hutchinson, 2004).

Soil temperature and moisture were measured next to each frame at the time of the N2O flux measurements. Soil temperature at 7.5 cm was monitored using copper-constantan thermocouples. Average soil moisture in the top 10 cm was measured with 15-cm three-bar time domain reflectometry probes inserted in soil at a 45° angle and read at every sampling date with a cable tester (model 1502B, Tektonix Inc, Beaverton, OR). Water-filled pore space (WFPS) was calculated using measured bulk soil densities and assuming a mineral particle density of 2.65 g cm–3. Rainfall was measured daily with standard rain gages at the Harlaka site.

Soil Analyses
Soil samples (0–15 cm) were collected weekly and kept at 4°C until analyzed. Soil mineral N content was measured by shaking 25 g of field-moist soil samples with 100 mL of 1 M KCl for 60 min. The slurry was then centrifuged (3000 x g, 10 min) and filtered (Whatman no. 42). Ammonium concentration in the extracts was determined by the salicylate method by flow injection analysis on a flow injection analyzer (Model QuickChem 8000, Lachat Instruments Division, Zellweger Analytics, Inc., Milwaukee, WI), whereas NO3– was detected in the UV radiation at 214 nm using a liquid chromatograph (Model 4000i, Dionex Corp., Sunnyvale, CA). Water-extractable organic carbon (WEOC) was obtained by shaking 6.25 g of field-moist soil in 20 mL of distilled water for 30 min. The extracts were centrifuged at 16000 x g and filtered at 0.45 µm. Soluble C was determined using a Formaacs Combustion TOC Analyzer (Skalar Analytical, De Breda, The Netherlands). A fraction of the soil samples was air-dried and sieved at 2 mm for the determination of soil pH (CaCl2 0.01 M). Total C and N contents were measured by dry combustion (CNS-1000, LECO Corp., St. Joseph, MI) on finely ground (0.15 mm) soil samples. Soil N and C values were expressed per unit of soil area using measured bulk soil densities.

Crop Dry Matter and Nitrogen Yields
The 4-yr old alfalfa and timothy were harvested three times (11 June, 16 July, and 6 Sept. 2001) at the Chapais site. At the Harlaka site, alfalfa and timothy were harvested once in the seeding year (31 July 2001) and three times in 2002 (12 June, 16 July, and 28 Aug.). Alfalfa and timothy were harvested in each plot with a Carter plot harvester on a 0.9 by 3 m area at the Chapais site and a 1.8 x 8 m area at the Harlaka site at a cutting height of 5 cm. At Harlaka in 2002, the DM yield was also measured late in the fall (21 Oct.) by sampling two 0.55 x 0.55 m quadrats in each plot at a 5-cm height. A subsample of approximately 500 g was dried at 55°C in a forced-draft oven for 3 d, ground to pass a 1-mm screen in a Wiley mill, and stored at room temperature in plastic jars. Four rows of soybean, each 3 m in length, were hand-harvested on 4 Oct. 2001 at the Chapais and Harlaka sites, and on 9 Sept. 2002 in the loam and on 16 Sept. 2002 in the clay at the Harlaka site. Whole-plants of soybean were dried at 55°C in a forced-draft oven for 3 d, and then threshed manually. Soybean grains were ground to pass a 1-mm screen in a Wiley mill and stored at room temperature in plastic jars. Nitrogen concentration of forage and grain samples was determined on dry ground samples using a flow injection analyzer (Model QuickChem 8000) after the H2SO4–H2O2 digestion method of Kjeldahl (adapted from Richards, 1993). Nitrogen yield was calculated as the product of DM yield and N concentration.

Statistical Analysis
One-way analysis of variance on the effects of crop types on soil N2O concentration, soil–surface N2O flux, soil temperature, and soil water, WEOC, and mineral N content was performed for each sampling date using the General Linear Models (GLM) procedure of SAS (SAS Institute Inc., 1990 p.1686). All experimental error variances were tested for homogeneity using Bartlett's test and values of N2O flux and soil atmospheric composition were log-transformed to achieve normality. Treatment effects were considered statistically significant at P < 0.05.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Soil Temperature and Water Content
Mid-summer soil temperatures at 7.5 cm remained between 15 and 25°C in all plots in 2001 and 2002 (data not shown). Differences in soil temperature between crop types were only occasionally >2°C even if often statistically significant. The clay was generally slightly cooler than the loam and sandy loam soils. A very dry spring resulted in low soil moisture conditions early in the 2001 season at the Chapais site (Fig. 1 and 2) . Large rainfalls restored soil water reserves in early June and WFPS remained between 40 and 60% during the rest of the growing season. At the Harlaka site, soil moisture remained relatively constant except for two slightly drier periods around 8 Aug. and 17 Sept. 2001. Values of WFPS were 10 to 20% higher in the clay than in the loam soil during both years, and 10 to 20% lower with soybean than with perennial crops, likely as a result of a lower soil bulk density following seedbed preparation. Lowest values of WFPS occurred in August and early September 2002 when only 15 mm of rain were received over a 40-d period (Fig. 1). Total rainfall from May to October was 483 mm in 2001 and 559 mm in 2002, representing 75 and 87% of the long-term average in Québec City.



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. Daily rainfall during the study at the Harlaka site near Québec City, Canada.

 


View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Soil water-filled pore space (7.5-cm depth) under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 
Crop Dry Matter and Nitrogen Yields
The seasonal DM yields of alfalfa in the seeding year at the Harlaka site in 2001 (4.1– 4.9 Mg DM ha–1) and in production years at the Chapais site in 2001 (11.6 Mg DM ha–1) and at the Harlaka site in 2002 (11.1–11.2 Mg DM ha–1) are typical of those obtained in eastern Canada (Table 2) (Bélanger et al., 1999; Bélanger and Richards, 2000). The seasonal DM yields of timothy in the production years at the Chapais site in 2001 and at the Harlaka site in 2002 are much lower than those usually reported for timothy in eastern Canada (Bolinder et al., 2002; Kunelius et al., 2003); this is explained by the absence of N fertilization. The soybean grain yield on the clay loam at Harlaka and at Chapais (3 Mg DM ha–1), an area with 2300 to 2500 corn (Zea mays L.) heat units, was approximately 1 Mg DM ha–1 less than that reported for an area with 2700 to 2900 corn heat units in southern Québec (Tremblay et al., 2002). The low soybean grain yield (1.7 Mg DM ha–1) observed on the loam at Harlaka in 2002 is probably related to a water deficiency; it is likely that soybean was more affected by the August drought period in the well-drained loam soil than in the clay soil.


View this table:
[in this window]
[in a new window]
  		Table 2. Annual dry matter (DM) and N yields of timothy (Phleum pratense L.), alfalfa (Medicago sativa L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites in 2001 and 2002.

 
Soil Mineral Nitrogen and Water-Extractable Organic Carbon
Soil mineral N (NH4+ + NO3–) contents in the 0- to 15-cm layer under timothy were generally <10 kg N ha–1 during the experiment (Fig. 3 and 4) . Low soil mineral N contents are often observed in soils under perennial grasses because plant N uptake is limited by net mineralization rates (Christensen, 1983; Piper, 1993; Beauchamp et al., 1996; Gil and Fick, 2001). Net soil N mineralization rates of the grass plant tissues are usually reduced by their high C/N ratio and high lignin and polyphenols contents (Robertson et al., 1993; Jarvis et al., 1996). Relatively greater soil mineral N contents were measured at the Harlaka site shortly after N fertilization in August 2001 (30 kg N ha–1) and early in the 2002 growing season (22 kg N ha–1) (Fig. 3 and 4). These values are similar to those reported by Bélanger and Richards (1997) (around 25 kg N ha–1) before spring fertilization of timothy in New Brunswick. Relatively high soil mineral N contents early in the growing season in perennial grasses are likely the result of favorable conditions for the decomposition of labile soil organic matter and plant material that did not survive winter.



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 3. Soil ammonium content (7.5-cm depth) under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 4. Soil nitrate content (7.5-cm depth) under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 
Soil mineral N contents under alfalfa in the seeding year (Harlaka site in 2001) and in the first-year soybean (Chapais and Harlaka sites in 2001) were slightly greater than under timothy but the differences were rarely statistically significant (Fig. 3 and 4). Beauchamp et al. (1996) and Burton et al. (1997) also measured low soil N contents under alfalfa during the seeding year. In contrast, soil mineral N under second-year legumes (Harlaka site in 2002) was up to 40 kg N ha–1 or 10 times greater than under timothy. These results are in agreement with reports of greater soil N inputs by alfalfa during the year following seeding than during the seeding year (Kelner et al., 1997) and of greater soil N mineralization when a legume, particularly alfalfa, is present in a rotation (Carpenter-Boggs et al., 2000). Consequently, two to three times more soil mineral N was measured under alfalfa than under nonfertilized grass monocultures (Gil and Fick, 2001). Interestingly, the soil mineral N content under the fourth-year alfalfa at the Chapais site in 2001 was low (<10 kg N ha–1). Schertz and Miller (1972) and Rasse et al. (1999) also reported low soil N contents (<10 kg N ha–1) under living alfalfa stands. Roots of growing alfalfa plants have been shown to release considerable amounts of N that was recently fixed by rhizobia (Ta et al., 1986) and the transfer of significant amounts of N from legumes to grasses has been reported when both plant types were grown in a mixture (Elgersma and Hassink, 1997). The absence of accumulated soil N at the Chapais site suggests that the N release by alfalfa roots and by N mineralization occurred at a rate that did not exceed those of plant uptake and other soil N losses such as immobilization, denitrification and leaching.

In the second-year soybean, greater soil N content earlier (before July 2002) than later in the growing season indicates that N accumulation was more the result of the decomposition of the previous-year crop residues than from the actively growing crop roots (Fig. 3 and 4). The large increase in soil mineral N contents occurred shortly after seedbed preparation, which involved two superficial passes (7 cm) of a rototiller on 22 May 2002. Soil disturbance induced by these tillage operations likely favored N mineralization from the soil and crop residues (Calderón et al., 2001). Similar soil N values under the first-year soybean and the nonfertilized grass in 2001 also indicate that the net contribution of the soybean roots to soil N accumulation during the growing season was small. Considering that 13% of the total faba bean N can be lost annually as rhizodeposition (Mayer et al., 2003), our results suggest that root-derived N was either reabsorbed by plant roots and soil microbes, or lost through leaching and denitrification.

Highest soil mineral N contents under the second-year alfalfa and soybean (Harlaka site in 2002) occurred early in the growing season for both legume crops and also after each alfalfa harvest (Fig. 3 and 4). Soil mineral N was generally greater under alfalfa than under soybean. This can be attributed to the larger belowground biomass in perennial than in annual crops (Bolinder et al., 1997; 2002) and to the greater amounts of N2 fixation by alfalfa than by soybean (Table 2) (Vance, 1997). Under alfalfa, the early season increase in soil mineral N was mostly as NH4–N (Fig. 3), indicating that mineralization of organic N was occurring at a higher rate than those of nitrification and plant uptake. This increase coincided with the early season soil warming and the major source of ammonium N was likely the mineralization of above and belowground plant tissues that did not survive winter. This hypothesis is reinforced by the similar NH4–N increase under the perennial grass (Fig. 3).

Following the alfalfa harvests, both NH4–N and NO3–N tended to increase (Fig. 3 and 4). Harvesting alfalfa modifies several source-sink relationships, and consequently could impact on soil N contents. Following harvest, N2 fixation and soil mineral N uptake are considerably reduced, and N is remobilized from taproots to new shoots (Kim et al., 1991). Ta et al. (1986) measured an increase in N exudates from the nodules of alfalfa after plants were harvested and Vance et al. (1979) reported a senescence of alfalfa nodules following harvest that may contribute to N release from the root system. Belowground alfalfa tissues are rich in N (Walley et al., 1996) and their decomposition following harvest of the aboveground biomass likely led to an increase in soil mineral N.

The agronomic effects of soil residual N during the year after cultivation of soybean have been well documented. For example, Staggenborg et al. (2003) estimated that the apparent supply of N to a wheat crop following soybean was approximately 21 kg N ha–1. The high contents of soil NO3–N (up to 30 kg N ha–1) early in the growing season under second-year soybean agree with these estimates of the contribution of soybean residues to the N requirements of the following crop in the rotation.

Water-extractable soil organic C contents were three to four times greater in the clay than in the loam and sandy loam soils (Fig. 5) , likely as a result of the higher C content of the clay soil (Table 1). There was no clear temporal variation pattern in WEOC during the growing season, indicating that neither harvesting nor crop growth patterns had a major impact on WEOC contents. Differences in WEOC between crop types were small in both soils at the Harlaka site in 2001, but the WEOC contents under alfalfa and timothy were consistently greater than under soybean in 2002. Contents of WEOC in soils are usually related to vegetation type and amounts of plant litter returned to the soil (Chantigny, 2003). The higher WEOC contents in soils under perennial crops likely reflect their greater root biomass and belowground C inputs compared to annual crops (Bolinder et al., 1997, 2002).



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 5. Water-extractable organic carbon (WEOC) (7.5-cm depth) under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 
Nitrous Oxide Concentration in Soil
Accumulation of soil N2O at a given location occurs when the rates of N2O reduction and net transport are lower than those of N2O production, mostly by nitrification and denitrification. Soil N2O concentrations were low during most of the study under all crops, except for occasional increases under the legumes, particularly under the fourth-year alfalfa (Chapais site in 2001) where the concentration reached 6µmol mol–1 after the first harvest (Fig. 6) . Burton et al. (1997) also reported generally low soil N2O concentrations under first-year alfalfa except for two short periods of higher concentrations (up to 200 µmol mol–1 at 10 cm) following precipitation events. They attributed the higher N2O concentration under alfalfa than under fallow to the contribution of the legume crop to soil C and N throughout the soil profile; this input of N and C substrates at depth resulted in increased N2O concentrations when the infiltration of rainfall water stimulated denitrification and increased resistance to gas transfer.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 6. Nitrous oxide concentration in the soil gas phase (7.5-cm depth) under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 
As for soil mineral N contents, lowest soil N2O concentrations were found under timothy (<1 µmol mol–1), thus suggesting that the small rates of N2O production were limited by NO3–N availability (Beauchamp et al., 1996). There was, however, no clear relationship between soil mineral N content and soil N2O accumulation in our study. For example, the largest increase in soil N2O at the Harlaka-clay site was observed late in 2001 under soybean (1.7 µmol mol–1) following the largest rainfall of the season (Fig. 6). This accumulation likely resulted from a combination of reduced gas diffusivity and greater denitrification rate of soybean crop residue N that had mineralized and nitrified after crop senescence. However, there was no simultaneous rise in soil mineral N or WEOC (Fig. 3, 4, and 5). A similar situation was observed under the fourth-year alfalfa at the Chapais site where the highest soil N2O concentrations were also recorded in absence of measurable increases in soil N and C pools. Moreover, the opposite situation occurred at Harlaka in 2002 when very high soil mineral N contents were reached under legumes with no increase in N2O accumulation even after high rainfalls. These observations are in agreement with the statement that variations in the size of soil N pools are imperfect indicators of the rates at which N transformations occur in soils (Jarvis et al., 1996).

Some very small differences in soil N2O concentration between crop types (15–50 nmol mol–1) were found to be statistically significant at sampling dates when N2O concentrations were near ambient levels (Harlaka site in 2002) (Fig. 6). On those sampling dates, coefficients of variation of mean soil N2O concentrations were often <3% (data not shown) and resulted in statistical significance of very small differences between crop types. The lowest N2O concentrations were never found under alfalfa on any of those sampling dates and the highest values were nearly always found under legume crops (21 of the 24 dates) (Fig. 6). These results suggest that the presence of legumes induced slightly higher background soil N2O concentrations than timothy either directly by rhizobia denitrification (O'Hara and Daniel, 1985) or indirectly by increasing inputs of N to the soil (Ta et al., 1986). However, the ecological significance of this small increase in N2O concentration for the N cycling in the soil–plant–atmosphere system remains to be demonstrated.

Soil N2O concentrations under alfalfa were increased by approximately two-fold following the first two harvests in 2002 at the Harlaka site (Fig. 6). Increases in soil N2O often occurred shortly after alfalfa harvests, thus suggesting that crop harvest had an effect on sources of N2O in the soil–alfalfa system. However, these values were not significantly different from those under the other two crops because of a high variability among replications on those sampling dates. We suggest that the large spatial heterogeneity of N2O formation and consumption processes in soils is such that statistically significant differences were more difficult to obtain in periods of increased N2O dynamics compared with periods of lower activity.

Nitrous Oxide Emissions
Seasonal variations in FN2O were similar to those in soil N2O concentrations (Fig. 6 and 7) . Fluxes were very small (<0.1 mg N2O m–2 h–1) under timothy except on the first 2001 sampling date in the clay soil at the Harlaka site. Our values under timothy compare well with average fluxes from nonfertilized grasslands reported by Christensen (1983) (0.044 mg N2O m–2 h–1), Velthof et al. (1996) (0.044 mg N2O m–2 h–1), and Wagner-Riddle et al. (1997) (0.013 mg N2O m–2 h–1). Low FN2O values under timothy are in agreement with the tight N cycle in this system as indicated by the constant low soil mineral N during the growing season (Fig. 3 and 4) and the low timothy DM and N yield (Table 2).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 7. Soil-surface N2O fluxes under alfalfa (Medicago sativa L.), timothy (Phleum pratense L.), and soybean (Glycine max L.) at the Chapais and Harlaka sites. Arrows indicate alfalfa, timothy, and soybean harvests. Bars represent the LSD (P = 0.05) values for the comparison of crop types at each sampling date; LSD values are only presented when the analysis of variance indicated a significant (P < 0.05) effect of the crop types.

 
Nitrous oxide fluxes were also generally low under the legume crops during the study with only six sampling dates when values under alfalfa and/or soybean were higher than 0.2 mg N2O m–2 h–1 (Fig. 7). These results are in agreement with the small FN2O measured by Wagner-Riddle et al. (1997) in first- and second-year alfalfa fields in southern Ontario but contrast with relatively large emissions reported by Duxbury et al. (1982) in northeastern USA. Significantly greater values under alfalfa than under timothy occurred mostly during the first half of the 2001 growing season in the fourth-year stand at the Chapais site (8 of the first 11 sampling dates). These higher emissions occurred at a time when soil mineral N contents were relatively low (Fig. 3 and 4), suggesting (i) that there was a significant turnover of N in this soil but that the N sinks (nitrification, denitrification, plant and microbial uptake, leaching) were using N at a rate equal or greater than that at which it was mineralized and/or (ii) that rhizobia denitrification (O'Hara and Daniel, 1985) was contributing to FN2O. In the sandy loam at the Harlaka site (2002), FN2O were also increased on several sampling dates under the second-year alfalfa compared with timothy. Interestingly, these differences occurred even if fluxes were very low, suggesting that the soil–alfalfa system was a weak but constantly greater source of N2O than the soil–timothy system. Robertson et al. (2000) also reported greater FN2O in alfalfa than in adjacent successional plots but the differences were greater than in our study.

There are few reports of N2O emissions from soybean in the literature. Wagner-Riddle et al. (1997) measured low average emissions during the growing season (0.025 mg N2O m–2 h–1) in Ontario. In southern Québec, MacKenzie et al. (1997) also reported low FN2O under continuous soybean except for two dates at the end of June when fluxes reached 0.94 mg N2O m–2 h–1. Jacinthe and Dick (1997) also measured a similar seasonal pattern of emissions in southern Ohio with highest values of 0.38 mg N2O m–2 h–1. In our study, N2O emissions under soybean were very low in the coarser-textured soils with most values smaller than 0.1 mg N2O m–2 h–1 and rarely greater than under timothy (Fig. 7). Relatively larger values were observed on the clay soil at the Harlaka site. Periods of increased emissions on that soil did not occur during the period of crop growth (July and August) but after harvest in 2001 and early in the 2002 season. These observations suggest that the N2O source in that system was not rhizobia denitrification but rather soil N biological transformations stimulated by the N mineralization associated with the decomposition of crop residues. Legume crop residues usually decompose faster than residues from nonlegume crops (Bremer et al., 1991; Trinsoutrot et al., 2000), and accordingly, higher N2O production and emission have been reported following incorporation of legume residues (Kaiser et al., 1998; Larsson et al., 1998; Baggs et al., 2000; Shelp et al., 2000). We hypothesize that the high postharvest 2001 emissions resulted from increased denitrification following the input of available C and N substrates at a time when the soil water content was high (Fig. 2). Again, close coupling between sources and sinks of C and N may explain the absence of variations in WEOC and soil mineral N pools during that period (Fig. 3, 4, and 5). High postharvest emission episodes were not seen in the coarse-textured soils at Harlaka and Chapais, suggesting that soil type played a major role in determining the N2O emissions during the transformation of soybean crop residue N.

Cumulative N2O-N Emissions
Cumulative N2O-N emissions under the nonfertilized timothy were small and varied little among sites (0.28–0.38 kg N ha–1) (Table 3). These values are slightly greater than those reported by Wagner-Riddle et al. (1997) for nonfertilized Kentucky bluegrass in southern Ontario (–0.07–0.26 kg N ha–1) but smaller than those measured by Velthof et al. (1996) in the Netherlands (0.5–1.2 kg N ha–1) in nonfertilized mowed grasslands. Cumulative N2O emissions under alfalfa and soybean were always equal or greater (1.3–4.6 times) than those under the nonfertilized timothy, with maximum values of 1.45 kg N ha–1 in alfalfa and 3.08 kg N ha–1 in soybean (Table 3). These emissions are in the range of values reported for alfalfa by Robertson et al. (2000) (1.9 kg N ha–1), Duxbury et al. (1982) (2.3–4.2 kg N ha–1), Wagner-Riddle et al. (1997) (1.0 kg N ha–1), MacKenzie et al. (1997) (1.18 kg N ha–1), and for soybean by Wagner-Riddle et al. (1997) (1.4 kg N ha–1) and MacKenzie et al. (1997) (1.23–2.16 kg N ha–1). Our results are in agreement with the suggestion that legumes may increase N2O emissions compared with non N2–fixing crops (Duxbury et al., 1982). Contributions of legume crops to N2O emissions (alfalfa 0.67–1.45 kg N ha–1, mean = 0.48 ± 0.33% of fixed N (n = 5); soybean 0.46–3.08 kg N ha–1, mean = 0.39 ± 0.27% of fixed N [n = 5]) are, however, much smaller than those calculated using the coefficient proposed by the IPCC methodology for soybean (2.76–4.97 kg N ha–1) and alfalfa (1.60–5.21 kg N ha–1) (Table 3). Interestingly, there were small differences in N2O emissions between the lightly fertilized timothy (30 kg N ha–1) and the legume plots in 2001 at the Harlaka site. Moreover, N2O emissions under the lightly fertilized timothy on the loam soil (Harlaka in 2001) were even greater than under alfalfa and soybean where amounts up to 145 kg N ha–1 were uptaken by the crop.


View this table:
[in this window]
[in a new window]
  		Table 3. Measured and estimated cumulative N2O emissions during the growing season (including postharvest emissions in soybean).

 
Our results suggest that the assumption that legume cropping produces N2O emissions as high as when N is added as mineral fertilizers (IPCC, 1997) may often not be valid in eastern Canada. However, we may have underestimated the cumulative N2O losses because emissions may have occurred during spring thaw when chambers could not be used (Christensen and Tiedje, 1990; Flessa et al., 1995). Wagner-Riddle et al. (1997) reported high spring thaw N2O emissions in Southern Ontario in cropping systems where organic amendments were added during the previous fall. However, they observed no significant increase in N2O emissions during spring thaw under alfalfa. It is also unknown to what extent the below-average rainfall in 2001 and 2002 may have decreased soil denitrification activity and N2O production compared with a year with near or above average rainfall.

For greenhouse gas accounting purposes, the other source of N2O emissions associated with soybean production is through the decomposition of crop residues (IPCC, 1997). The large N2O emissions (approximately 3 kg N2O-N ha–1) measured after soybean harvest (Harlaka-clay 2001) and early in the following growing season indicate that soybean crop residues are a source of C and N substrates that can induce significant N2O production in agricultural soils (Table 3). However, the contribution of soybean crop residues to N2O emissions could not be detected in the other four situations investigated, thus suggesting a strong interaction between N2O emissions, soil, and climatic conditions.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Emissions of N2O were small under the unfertilized grass where very low soil mineral N contents probably limited denitrification and N2O production. Soil mineral N contents under legumes were up to 10 times greater than under grasses. However, soil mineral N contents and FN2O were not closely related, thus suggesting that the soil mineral N pool was a poor indicator of the intensity of N2O production processes. Higher N2O fluxes were measured under legumes than under grasses in only 6 out of 10 field comparisons. Moreover, the emissions associated with N2–fixing crops were considerably smaller than those predicted using the emission coefficient proposed for the national inventory of greenhouse gases (1.25% of fixed N) (IPCC, 1997). We conclude that the use of the IPCC coefficient may overestimate the N2O emissions associated with soybean and alfalfa production in eastern Canada.


   ACKNOWLEDGMENTS
 
We thank D. Mongrain, I. Morasse, N. Bertrand, J.-P. Soucy, A. Toussaint, N. Bissonnette, P. Jolicoeur, and B. Gagnon for assistance in soil–surface gas flux measurements, soil analysis, plot maintenance, and statistical analysis. This project was partly funded by the Pollution Data Branch of Environment Canada.

Received for publication May 5, 2003.


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




This article has been cited by other articles:


Home page
 Agron. J.Home page
N. Z. Lupwayi and A. C. Kennedy
Grain Legumes in Northern Great Plains: Impacts on Selected Biological Soil Processes
Agron. J., November 6, 2007; 99(6): 1700 - 1709.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
R. L. Lemke, Z. Zhong, C. A. Campbell, and R. Zentner
Can Pulse Crops Play a Role in Mitigating Greenhouse Gases from North American Agriculture?
Agron. J., November 6, 2007; 99(6): 1719 - 1725.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
P. Rochette, J. Dionne, Y. Castonguay, and Y. Desjardins
Atmospheric Composition under Impermeable Winter Golf Green Protections
Crop Sci., June 20, 2006; 46(4): 1644 - 1655.
[Abstract] [Full Text] [PDF]

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 Similar articles in ISI Web of Science
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 HighWire
Right arrow Citing Articles via ISI Web of Science (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.
Agricola
Right arrow Articles by Rochette, P.
Right arrow Articles by Lévesque, G.