| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Greenhouse and Processing Crops Research Centre, Agric. & Agri-Food Canada, Harrow, ON, Canada N0R 1G0
b Eastern Cereal and Oilseed Research Centre, Agric. and Agri-Food Canada, 960 Carling Avenue, Ottawa, ON, Canada K1A 0C6
* Corresponding author (DruryC{at}agr.gc.ca)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: CT, conventional moldboard plow tillage • FC, field capacity • NT, no-tillage • ZT, fall zone tillage
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The adoption of no-tillage for corn production on fine-textured soils in cool temperate regions is hampered by cold, wet soil conditions in the spring, which cause delayed emergence and early growth of seedlings, reduced final plant populations, and lower grain yields (Drury et al., 1999). Hence, alternative tillage systems, such as zone tillage, have been developed for these soils to improve seed bed conditions in the planting zone (typically 20 cm wide by 10–15 cm deep zone) while retaining the conservation tillage benefits of increased soil surface cover, reduced soil erosion, and reduced residue decomposition in the untilled interrow area, which typically comprises about 72.4% of the total field surface area (Drury et al., 2004a). Hence, zone tillage and similar conservation tillage practices provide farmers the potential for achieving much of the cost and environmental benefits of conservation tillage, while retaining the higher yield characteristics of conventional tillage.
In corn production, N fertilizer is often applied twice; once at planting and once as a "sidedress" when the corn is at the 6–8 leaf stage. This practice is especially popular in humid regions where significant losses of N fertilizer can occur through nitrate leaching below the crop root zone and efflux of N gases as a result of denitrification. Sidedress N fertilizer is typically injected 10 cm into the soil adjacent to the corn rows. As soil moisture typically increases with depth, the injected N is placed in wetter soil layers, which have lower oxygen concentrations than the surface soil layers, and consequently greater N fertilizer losses occur as a result of denitrification. These losses are further exacerbated when large amounts of soil C are available near the N fertilizer. Hence, the degree and depth of residue incorporation into the soil, and therefore the type of tillage system, may indirectly impact denitrification-based N losses by controlling the amount of C substrate available to denitrifying bacteria.
Large-scale adoption of conservation tillage practices can be achieved by demonstrating competitive yields, lower fuel and labor costs, and reduced soil erosion relative to conventional tillage. These advantages may be offset, however, if conservation tillage also increases global warming through enhanced generation of greenhouse gases. The objective of this research was therefore to determine the effect of tillage system (conventional moldboard plow tillage, zone tillage, no-tillage) and N placement depth (10 cm vs. 2 cm) on soil-borne greenhouse gas (N2O and CO2) emissions from a clay loam soil in Eastern Canada.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The N fertilizer sidedress treatments (160 kg N ha–1) included 10-cm placement depth (deep N placement) and 2-cm placement depth (shallow N placement). Moldboard plowing was to 15- cm depth, and zone tillage used a "Trans-till" (ZT, Row-Tech Inc., Stover, MI), which produced a 21 cm wide by 15 cm deep tilled zone centered on the crop row. The measurements reported here apply to the corn phase of the rotation because this crop requires the most N fertilizer, and thus provides the greatest potential for N losses through denitrification.
All winter wheat plots following the wheat harvest (19 July 2000, 17 July 2001, and 25 July 2002) were sprayed with 1.8 kg ai ha–1 glyphosate [N–(phosphonomethyl) glycine] plus 1.0 kg ai ha–1 2,4-D [(2,4-dichloro phenoxy acetic acid] in the fall (18 Sept. 2000, 1 Oct. 2001, and 26 Sept. 2002) to kill perennial weeds and volunteer wheat. At the time of corn planting, all plots were sprayed with 1.68 kg ai ha–1 pendimethalin [N–(1-ethylpropyl)-3,4-dimethyl-2,6-dinitro benzene-amine] plus 1.0 kg ai ha–1 atrazine (2-chloro-4-ethylamino-6-isopropylamino-s-triazine) pre-emergence to control newly emerging weeds. Corn (Syngenta Seeds hybrid N58D1) was planted (76 140 plants ha–1) into all treatments on 2 June 2000, 7 May 2001, and 23 May 2002 in 76 cm rows with a John Deere 6-row no-till "MaxEmerge" planter. One 46-cm ripple coulter and one bubble coulter were used to open the seed furrow, and residue managers were used to improve seedbed conditions. At planting, N (22 kg N ha–1), P (88 kg P2O5 ha–1), and K (44 kg K2O ha–1) were applied in a band 5 cm beside and 5 cm below the seed using 8–32–16 granular fertilizer. During the 6 to 8 leaf growth stage, ammonium nitrate (34–0–0) was applied as a sidedress fertilizer (granular, 160 kg N ha–1) at a lateral distance of 20 cm from the corn row using either deep N placement (10 cm depth) or shallow N placement (2 cm depth) on 20 June 2000, 13 June 2001, and 20 June 2002. Corn was harvested on 15 Nov. 2000, 9 Nov. 2001 and 23 Oct. 2002, and grain yield was determined.
Nitrous Oxide and Carbon Dioxide Measurements
Nitrous oxide and CO2 emissions were measured a total of 53 and 43 times, respectively, over the 3-yr period (from mid April until mid November), using 36 field chambers consisting of an acrylic ring (60 cm i.d. by 14 cm high) and a removable gas-tight lid. In three of the field replicates, each of the six treatments had two chambers in each plot. The fluxes were measured at about the same time of day over the 3 yr (between 0900 h and 1200 h) and the measurements were made one replicate at a time.
The acrylic rings partially overlapped the corn row, and were inserted 8 cm into the soil to ensure a gas-tight seal between ring and soil. The degree to which the rings overlapped the corn row was the same for all three tillage treatments, and it was set to accommodate the zone tillage treatment, that is, the percentage of tilled area inside the ring chamber matched the percentage of tilled area in the zone tillage plots (i.e., 27.6%).The head space volume above the soil surface in the chambers was determined periodically over each growing season. A wooden cover containing 20 holes (0.4 cm diam.) drilled in three concentric rings (eight holes in the outer ring, eight holes in the middle ring, four holes in the inner ring) was placed on the chamber, and then a metal rod fitted with an upward reading scale was lowered through the holes until contact was made with the soil surface. The head space volume (Vc) was then determined using,
 | [1] |
The N2O and CO2 flux chambers were removed and reinserted for planting, sidedress fertilizer application, and harvesting operations. The chamber lids had gasket foam (Volara Polyolefin foam, 2 pound density, Duraco Inc., Forest Park, IL) attached to the bottom and were clamped onto the chamber just before the N2O and CO2 gas flux measurements. The N2O fluxes were determined by collecting timed gas samples for later analysis in the laboratory, while the CO2 flux measurements were made directly in the field over two 20-s intervals using a modified lid (described above) attached to a LI-COR 6400 soil respiration system (LI-COR Inc., Lincoln, NE). The N2O gas samples (24 mL) were collected with a syringe at 10 min intervals over 30 min and these samples were injected into pre-evacuated 12 mL Exetainers (Labco, Buckinghamshire, UK). The resulting 101 kPa overpressure in the Exetainers would reduce contamination from outside air. Standards from a calibrated reference gas cylinder were injected into pre-evacuated Exetainers in a similar manner to the samples, and both samples and standards were injected (2.5 mL) into a Varian 3800 gas chromatograph fitted with a Combi-PAL automatic sampling system (Varian, Mississauga, Ontario). The samples and standards were initially injected into a vented 0.25-mL sampling loop connected to a 10-port valve so that atmospheric pressure was obtained. From there the N2O samples were injected into a precolumn and then into the main column. The precolumn was then backflushed to vent to remove water vapor from the gas sample. Nitrous oxide was separated in the main column using a 5.0-m long Porapak Q chromatography column with Ar (95%) and CH4 (5%) carrier gas flowing at a rate of 30 mL min–1 at 70°C. Nitrous oxide concentrations were then determined using an electron capture detector (ECD) at 350°C, and the change in concentration among the timed samples was used to infer N2O flux using a linear regression. Total growing season N2O and CO2 emissions (May 1 to October 1) were calculated as a linear interpolation between sampling date (Simpson rule for numerical integration). These growing season emissions were calculated separately for each chamber.
Soil Physical and Chemical Analysis
Volumetric soil water content measurements were made in situ in the corn row midway between plants three times per week at the 0- to 5-cm and 0- to 30-cm depths of each treatment using a Tektronix 1502B cable tester (Tektronix, Inc., Beaverton, OR; Topp et al., 1980; Topp, 1993). Soil temperature was measured hourly during each growing season using thermocouples inserted horizontally into the soil profile at the 5- and 10-cm depths, with two replicate thermocouples per treatment. Soil temperature and volumetric soil water content (0–10 cm) were also determined in situ at the same times and locations as the CO2 and N2O measurements using a HI93510 thermometer (Hanna Instruments Canada Inc., Laval, Quebec) and a Trime-FM TDR Moisture Meter (Mesa Systems Co., Medfield MA). One composite soil sample (0–30 cm depth) was collected five to eight times during the growing season from each treatment plot (n = 4) to determine inorganic soil N (NH4+ and NO3–), each composite consisting of five cores (2.5 cm diameter) collected at 5-cm intervals on a transect centered across the sidedressed N application band.
Experimental Design and Statistical Analysis
A split plot randomized complete block design was used with the main plot units being tillage treatment and the subplots units being N application depth. There were four replicates and each plot was 20 m long and 6.1 m wide. The agronomic measurements (yield, soil nitrate analysis) were conducted on all four replicates; however N2O and CO2 emissions and corresponding soil temperature and soil water content measurements were measured on only three of the replicates with two chambers per plot (subsamples). Statistical analysis was performed using the Mixed model (Littell et al., 1996). When significant interactions occurred, preplanned treatment comparisons were evaluated with the least significant means procedure (p = 0.05).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
27°C on August 8th and then gradually decreased to 8 to 9°C by the end of October.
|
|
|
|
In 2002, precipitation was well above the 30 yr average in April and May, then substantially below the 30-yr average in July through September (Table 1). In fact, the July through September period had only 126 mm of rainfall, which is only about 50% of the corresponding 30-yr average (251.4 mm). As a result, the near-surface (0- to 10-cm depth) soil water contents decreased from an average of about 30% during May-June, to about 7% in late September (Fig. 1). In general the CT treatment had lower soil moisture contents than the ZT and NT treatments in the first three sampling periods in 2002 (Fig. 1). The average seasonal soil water content was the lowest for the CT treatment and the highest for the NT treatment in both the 0- to 5-cm and 0- to 30-cm depths (Table 2). Unlike the previous 2 yr, the soil temperature was similar among all treatments for both the first two sampling periods and the entire growing season (Fig. 2).
Corn grain yields were comparable between CT and ZT in 2000 and 2002 (Drury et al., 2004). Corn grain yields were lower with both ZT and NT in 2001 compared with CT as a result of the drought conditions in the critical growing months of June and July of 2001. Averaged over 3-yr and tillage treatments, the deep N placement (10 cm) produced about 4% greater yields (7.48 Mg ha–1) than the shallow N placement (2 cm) treatment (7.18 Mg ha–1). There were no significant interactions between N placement method and tillage treatment on corn grain yields.
Nitrous Oxide Emissions
In year 2000, the N2O emissions from both N placement depths formed a "spike" pattern, with initially low fluxes (<5 g N ha–1 d–1) during April through early June, then a dramatic increase after N sidedress application on June 20, and then a rapid return for low fluxes (<100 g N ha–1 d–1) by the second week in July (Fig. 3
). Deep N placement produced both greater peak fluxes (Fig. 3) and greater total growing season emissions (Table 3) than shallow N placement. Although there was no significant difference in seasonal N2O emission amongst the three tillage treatments for shallow N placement, the pattern for deep N placement followed the order CT > NT > ZT (p < 0.05, Table 3).
|
|
In 2002, the N2O fluxes and total emissions formed a pattern similar to those of 2000, but with a lower magnitude (Fig. 3, Table 3). Spikes in N2O fluxes occurred after sidedress N application, and deep N placement produced both larger spikes and greater total growing season emissions than shallow N placement. The post-sidedress spikes in N2O fluxes (about 50–240 g N ha–1 d–1) were substantially lower than those of 2000, however, and the total growing season N2O emissions (1.50–3.95 kg N ha–1) were a factor of two to four lower than the 2000 values (Table 3). The total growing season emission of N2O in 2002 followed the pattern CT > NT > ZT for both N placement depths (Table 3).
The emissions in 2001 were probably lower than those in 2000 because of drier soil conditions, especially during late June and July (Table 1 and Fig. 1). In 2001, the NT treatment had greater N2O emissions at the end of May and in early June as compared with the CT and ZT treatments for both N placement depths, which was a different pattern than observed for 2000 where NT emissions were intermediate between CT and ZT for deep N placement.
In 2002, the N2O emissions were greater than those in 2001 but still lower than those in 2000. The N2O emissions in late April until mid June 2002 were low for all treatments but peaked through late June to early July, following sidedress N application. The N2O emissions in August through October period were very low for all treatments as a result of the dry and warm soil conditions (Fig. 1
–3). For deep N placement, the CT treatment had the greatest total emissions (3.95 kg N ha–1) over the growing season, followed by NT (2.00 kg N ha–1), and then ZT (1.60 kg N ha–1) (Table 3). Although the pattern was similar for the shallow N placement treatments (i.e., CT > NT > ZT), the emissions were much lower than the corresponding deep N placement treatments (Table 3).
There was a significant site-year effect on the seasonal N2O emissions (Table 3). Averaged over all treatments, the N2O emissions in 2000 (6.30 kg N ha–1) were 2.9 times greater than in 2002 (2.21 kg N ha–1) and 4.3 times greater than in 2001 (1.47 kg N ha–1). The higher N2O emissions in 2000 relative to 2001 and 2002 were likely due to the higher precipitation and the correspondingly greater average near-surface soil water contents during the growing season (Table 1 and Fig. 1). Nitrogen placement depth, the interaction between N placement depth and site-year, and the interaction between N placement depth and tillage, significantly affected N2O emissions (Table 3). Shallow N placement had significantly lower N2O emissions than deep N placement in both 2000 and 2002, but no differences were evident in 2001, the year with the longest drought. Averaged over the 3 yr, shallow N placement (2.83 kg N ha–1 yr–1) had 26% lower N2O emissions than deep N placement (3.83 kg N ha–1 yr–1). The increase in N2O emissions with deep N fertilizer placement was likely the result of increasing soil water content with increasing depth, which would promote more frequent and/or more extreme anaerobic conditions and thereby enhanced nitrate conversion to N2O through denitrification.
The significant tillage by N placement depth interaction was due to the significant effect of tillage treatments when N was placed deeper in the soil profile. With deep placement, 3-yr average N2O emissions were 38% lower with ZT (2.98 kg N ha–1 yr–1) than CT (4.81 kg N ha–1 yr–1). This reduction was also evident in each year as ZT reduced emissions by 33% in 2000, 13% in 2001, and 60% in 2002 compared with CT. Averaged over all 3 yr, NT had 23% lower N2O emissions (3.71 kg N ha–1 yr–1) than CT. The pattern of ZT < NT < CT for N2O emissions when N2O was placed at the 10-cm depth is very interesting. So the question is why did ZT consistently result in the lowest N2O emissions? Tillage operations affect soil physical properties (aeration, bulk density, aggregation, soil temperature, and moisture) and soil chemical properties (organic C content and distribution) in part due to the degree of incorporation of plant residues (Drury et al., 2004). Since the differences in N2O emissions between these tillage treatments was most pronounced when nitrate was placed at the 10-cm depth, it is likely that the effect of tillage on the amount of available C, available nitrate, and oxygen caused differences in denitrifier activity. Denitrification in the Brookston clay loam soils has been found to be controlled by the amount of available soil C (Drury et al., 1991). Since both ZT and NT have similar amounts of plant residue on the soil surface (which is often an order of magnitude greater than CT), the lower N2O emissions from these two conservation tillage treatments could result from less available C at depth (Drury et al., 2003). Further, the improved soil structure, especially in the tilled zone of the ZT treatment, produced drier soil in the 0- to 30-cm depth than NT (Table 2). Hence the reduction in N2O emissions from ZT relative to CT and NT may have stemmed from a combination of relatively low C inputs and relatively dry soil conditions in the N application zone. Lower germination rates and plant populations in NT compared with ZT treatments (Drury et al., 2003) would also result in lower nitrate uptake rates and greater levels of available nitrate in NT soils vs. ZT soils.
When N was placed at the 10-cm depth, the N2O emissions accounted for 2.64% of the applied fertilizer N for CT, 2.04% for NT, and 1.64% for ZT. Similar percentages (1–3% of applied fertilizer N) were obtained for corn grown on a silt loam soil in Tennessee (Thornton and Valente, 1996), and on a clay loam soil in southwestern Quebec (MacKenzie et al., 1998). MacKenzie et al. (1998) also noted that N2O emissions increased with N fertilizer application rate, and that 2.0 to 5.7 kg N ha–1 was emitted over a 7.5-mo period. On the other hand, a denitrification study on a clay loam soil in Argentina (Palma et al., 1997) produced both similarities and differences with this study. The main differences included lower overall N2O emissions and greater emissions from NT (0.35 kg N ha–1) than from CT (0.19 kg N ha–1) in the Argentina study. The main similarity was that most of the N2O losses (50–76%) in the Argentina study occurred shortly after fertilizer application. A possible reason for the differences is that the N fertilizer rate in the Argentina study (60 kg N ha–1) was much lower than in this study (182 kg N ha–1). In a tillage study conducted over 3 yr in the Canadian Prairies, the N2O emissions from no-tillage over the spring and summer period were equal or lower than those from intensive tillage systems (Lemke et al., 1999). The N2O emissions under dryland conditions in the Canadian Prairies were much lower than those reported here for soils in the more humid regions of Canada, however the decrease in N2O emissions with conservation tillage was observed in both studies.
Deep N placement resulted in considerably more N2O emissions than would be estimated using the IPCC coefficient of 0.0125 for applied fertilizer N. This was also found in New Zealand by Choudhary et al. (2002) who compared N2O emissions from CT (9.2 kg N ha–1 yr–1) and NT (12.0 kg N ha–1 yr–1) in a double cropped (winter oats/corn) silt loam soil. They additionally found no significant difference between the tillage treatments, however both tillage treatments produced significantly greater N2O emissions than an adjacent permanent pasture (1.66 kg N ha–1 yr–1). Although the N2O emissions in the New Zealand study were considerably higher than the 3-yr averages reported in this study, we did find N2O emissions as high as 8.99 kg N ha–1 with CT in 2000, the wettest growing season.
Although the annual and growing season precipitation (Table 1) and soil water contents (Table 2, Fig. 1) indicate that the wettest and driest years of the study were 2000 and 2002, respectively, a very different pattern occurs for the near-surface (0–10 cm) presidedress (April–June) soil water contents (Fig. 1). During this period, year 2000 produced the lowest average presidedress soil water content (26.4%), while 2001 produced the highest average water content (30.9%), and 2002 was intermediate (29.5%). In addition, the presidedress soil water contents in the near-surface were frequently at or above the field capacity (FC) water content (32.9%) in 2001, but only occasionally at or above the FC value in 2002, and always well below the FC value in 2000. This situation, which occurred because the frequency of small precipitation events during April–June followed the pattern 2001 > 2002 > 2000 (data not shown), may explain the dramatically different post-sidedress patterns in N2O emissions (Fig. 3). In 2001, the degree and frequency of wet soil conditions in the near-surface soil matrix may have been sufficient to cause frequent anaerobiosis with the resulting development of reductase enzymes capable of reducing N2O to N2. This would cause production of N2 rather than N2O, which is consistent with the distinct dip in N2O emissions in 2001 at the time of N sidedress application (Fig. 3). In addition, gas diffusion would be slower in wetter soil, which would provide more time for N2O to be reduced to N2 (Drury et al., 1992). In 2000, on the other hand, the near-surface soil matrix likely remained dry enough that production of N2O was favored over N2, resulting in a dramatic spike in N2O emissions after N sidedress application (Fig. 3), and presumably much less N2 production. Year 2002, being intermediate in near-surface soil wetness between 2000 and 2001, produced a much smaller spike in N2O emissions than 2000 (Fig. 3), but presumably greater N2 emissions than 2000. This proposed explanation is consistent with Dendooven et al. (1996), who observed in a laboratory study that intact cores of a clay loam soil (collected from the 0–10 cm depth) produced greater N2O fluxes, greater total N2O emissions, and much smaller N2O/N2 ratios when the cores were saturated for a short period of time (6 h), as opposed to a long period of time (96 h), before gravity drainage to equilibrium. Furthermore, the difference in antecedent (equilibrium) soil water content between the two sets of cores was only 3.2% points, which is similar to the 4.5% point difference in average presidedress soil water content between 2000 and 2001. It consequently appears that denitrification mechanisms and losses of N2O and N2 are at least partially dependent on the antecedent water content in the near-surface soil at the time of N fertilizer application.
Soil Nitrate Levels
In all 3 yr of the study, soil nitrate levels were very low at planting (<5 mg N kg–1 soil), then peaked shortly after sidedress N fertilizer was applied in mid-late June (55–110 mg N kg–1 soil), and then declined toward preplant levels throughout the rest of the year (Fig. 4
). In 2000, the soil nitrate levels decreased to preplant levels by fall harvest as a result of plant uptake, denitrification losses, and leaching below the 30-cm depth. In 2001 and 2002, dry soil conditions during the summer reduced both N uptake and denitrification/leaching losses; and as a result, higher residual nitrate levels persisted in the soil at harvest time. There were no consistent effects of tillage treatment on soil nitrate levels, although peak soil nitrate levels were lower with shallow N placement than deep N placement in 2001 and 2002 (Fig. 4).
|
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 4, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Rochette, D. A. Angers, M. H. Chantigny, and N. Bertrand Nitrous Oxide Emissions Respond Differently to No-Till in a Loam and a Heavy Clay Soil Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1363 - 1369. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Sey, J. K. Whalen, E. G. Gregorich, P. Rochette, and R. I. Cue Carbon Dioxide and Nitrous Oxide Content in Soils under Corn and Soybean Soil Sci. Soc. Am. J., May 29, 2008; 72(4): 931 - 938. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
