Biol Fertil Soils (2003) 38:340–348 DOI 10.1007/s00374-003-0663-9
ORIGINAL PAPER
Abdirashid A. Elmi · Chandra Madramootoo · Chantal Hamel · Aiguo Liu
Denitrification and nitrous oxide to nitrous oxide plus dinitrogen ratios in the soil profile under three tillage systems Received: 9 December 2002 / Accepted: 7 July 2003 / Published online: 5 September 2003 Springer-Verlag 2003
Abstract There is a growing interest in the adoption of conservation tillage systems [no-till (NT) and reduced tillage (RT)] as alternatives to conventional tillage (CT) systems. A 2-year study was conducted to investigate possible environmental consequences of three tillage systems on a 2.4-ha field located at Macdonald Research Farm, McGill University, Montreal. The soil was a sandy loam (0.5 m depth) underlain by a clay layer. Treatments consisted of a factorial combination of CT, RT, and NT with the presence or absence of crop residue. Soil NO3--N concentrations tended to be lower in RT than NT and CT tillage treatments. Denitrification and N2O emissions were similar among tillage systems. Contrary to the popular assumption that denitrification is limited to the uppermost soil layer (0–0.15 m), large rates of N2O production were measured in the subsurface (0.15– 0.45 m) soil, suggesting that a significant portion of produced N2O may be missed if only soil surface gas flux measurements are made. The N2O mole fraction (N2O:N2O+N2) was higher in the drier season of 1999 A. A. Elmi ()) Department of Engineering, Nova Scotia Agricultural College, 219 Resource Stewardship Building, 20 Tower Road, P.O. Box 550, Truro, Nova Scotia, B2N 5E3, Canada e-mail:
[email protected] Tel.: +1-902-893-6592 Fax: +1-902-893-0335 C. Madramootoo Brace Center for Water Resources Management, McGill University, Macdonald Campus, 21, 111 Lakeshore Road, Ste.-Anne-de-Bellevue, Quebec, H9X 3V9, Canada C. Hamel Semiarid Prairie Agricultural Research Center, Airport Road, Box 1030, Swift Current, SKS9H 3X2, Canada A. Liu Department of Natural Resource Sciences, McGill University, Macdonald Campus, 21, 111 Lakeshore Road, Ste-Anne-de-Bellevue, Quebec, H9X 3V9, Canada
under CT than in 2000, with the ratio occasionally exceeding 1.0 in some soil layers. Dissolved organic C concentrations remained high in all soil depths sampled, but were not affected by tillage system. Keywords Denitrification · Nitrous oxide production · Mole fraction · Soil profile
Introduction Traditional farming systems for the intensive production of agricultural lands can degrade the quality of soil and water resources. Continuous use of conventional tillage (CT) systems can accelerate the depletion of soil organic matter and lead to the deterioration of soil structure, resulting in severe soil erosion (Hussain et al. 1999; Martel and MacKenzie 1980). Furthermore, eroded soil particles associated with runoff water may carry plant nutrients such as P and N which contaminate water and contribute to the eutrophication of aquatic ecosystems. Water quality measurements at 300 locations in major USA rivers showed that suspended sediments associated with runoff from agroecosystems are the most damaging nonpoint source pollutants (Smith et al. 1987). As a result, the use of water for drinking, irrigation and recreation may be affected. There is a growing trend for the adoption of conservation tillage systems, including no-till (NT) and reduced tillage (RT), primarily because they have been shown to be significantly more water efficient (Lindwall and Anderson 1981), to improve soil and water quality (Hussain et al. 1999), and to reduce production costs due to lower fuel and labor inputs (Uri et al. 1999; Lindwall and Anderson 1981). The continuing increase in acreage under NT and RT, however, raises new water and air quality concerns. The formation of macropores coupled with reduced surface runoff in NT or RT fields can increase the downward movement of water containing NO3--N and other agrochemicals to subsurface tile drains or groundwater. While
341 -
the likelihood of NO3 -N leaching is increased under NT and RT systems, published studies of NO3--N losses from NT and different forms of RT have given somewhat divergent results (Kanwar et al. 1997; Randall and Iragavarapu 1995). Grain-corn residues contain large amounts of N, thus contributing to soil N pools. Further, soils under NT and/or RT retain greater soil moisture than those under CT, which may enhance denitrification. Denitrification is a major pathway by which N2O enters the atmosphere. N2O emissions have been a topic of increasing concern because N2O has a well-documented role in stratospheric O3 depletion and contributes to the atmospheric greenhouse gas effect (MacKenzie et al. 1998; Mummey et al. 1998; Burton et al. 1997; Rice and Smith 1982). Subsoil denitrification has been suggested as an important mechanism for the removal of excess NO3--N which has accumulated in the soil profile before leaching to groundwater or discharging to surface aquifers via subsurface drainage (Sotomayor and Rice 1996). Although it is often assumed that denitrification in the top soil layer (0.15 m) is truly representative of the overall rate of denitrification processes, we hypothesize that denitrification can occur in subsoils if soluble C is not limited. An interesting feature of denitrification in subsurface soil is that it is unlikely to add to concerns over global atmospheric N2O concentrations due to the further reduction of N2O to N2 during diffusion up the soil profile. Bergsma et al.(2002) recognized the scarcity of information about denitrification in subsoil environments. Whether NO3- removal by denitrification is beneficial to the environment depends on the partitioning of denitrification into N2O and N2. To our knowledge, this is the first field study to consider N2O:N2O+N2 ratio production in surface and subsurface soil environments as impacted by different tillage systems. The main objectives of this study were to assess the effects of long-term tillage practices on: (1) NO3--N in the soil profile, (2) denitrification and N2O emissions in subsurface (0–0.45 m) soils, and (3) to estimate the ratio of N2O:N2O+N2 produced during two growing seasons.
Materials and methods
for grain corn, and the other half without residue and planted to corn harvested as silage. Plots were 18 m80 m, and drained by a subsurface drainage system to a depth of 1.0 m below the soil surface. In this study, only plots with residue were included. The grain-corn plots were harvested with a combine that removed only grain, leaving all residues on the plots. Some coarse chopping or crushing of aboveground residue occurred during harvest. Corn (Funk 4120 hybrid) was planted in rows spaced 0.76 m apart. All plots received: at seeding, diammonium phosphate (1846-0), banded 50 mm below and 50 mm laterally from the seeds to provide 40 kg N ha-1 and 102 kg P2O5 ha-1. The field was seeded on 6 May in 1999 and on 8 May in 2000. NH4NO3 (34-0-0) and muriate of potash (0-0-60) were top-dressed 2–3 weeks later to provide an additional 140 kg N ha-1 and 148 kg K2O ha-1. The second application occurred on 4 June in 1999 and on 9 June in 2000. Measurements of N2O and mineral N in soil To assess the relative proportion of N2O and N2 emissions in the surface and subsurface soil, three incremental depths (0–0.15, 0.15– 0.30, 0.3–0.45 m) were sampled simultaneously in pairs. Due to the heavy labor commitment required by the core sampling and time constraints related to incubating cores, it was only feasible to collect one pair of samples from each treatment plot for each sampling date. Soil cores were sampled in non-wheel-tracked rows. Samples were never taken from the same location more than once within a growing season. Care was taken to avoid cross-contamination between sampling depths with careful cleaning of each drilling depth site. Denitrification and N2O production rates were measured using the core incubation method, in the presence and absence of C2H2, respectively. Undisturbed soil core samples (50 mm diameter150 mm long) were incubated in 2-l Mason jars. For denitrification measurements, 100 ml headspace air was replaced by C2H2 (5% v/v) to inhibit N2O reduction. The second sample was incubated without C2H2 to estimate N2O emissions. C2H2 concentrations of 1–10% in air (v/v) are usually sufficient for the inhibition to occur, causing only N2O to be evolved (Granli and Bøckman 1994). N2O:N2O+N2 ratios were estimated by dividing the N2O fluxes measured with and without C2H2 present (Eq. 1). N2 O N without C2 H2 N2 O ¼ N2 O þ N2 N2 O N with C2 H2
ð1Þ
Soil samples (three samples per plot) for NO3–N analysis were taken prior to planting in the spring (April) and shortly after harvest in fall (October) from 0–0.25 m, 0.25–0.50 m, and 0.50–0.75 m depth increments using a hand-held auger sampling probe. Replicate samples were then thoroughly mixed and moist subsamples of 10 g were shaken with 100 ml of 1 M KCl for 60 min. The soil suspensions were filtered through Whatman no. 5 filter papers. NO3--N was quantified using a Lachat flow injection autoanalyzer (Lachat Quickchem, Milwaukee, Wis.) according to Keeney and Nelson (1982). The detection limit was 0.05 mg l-1.
Site description and management This study, undertaken in 1999 and 2000, was conducted on a 2.4-ha site at McGill University’s research farm on Macdonald Campus, Ste-Anne-de-Bellevue, Quebec. The soil was mostly of the St Damase series (Typic Endoaquent; Humic Gleysol). The upper soil layer (about 0.28 m) was a sandy loam, underlain by a sand layer (mean thickness about 0.18 m), with clay beginning at a mean depth of 0.46 m (Burgess 2000). The study consisted of three tillage systems: NT, RT, and CT, factorially combined with two residue treatments: with and without. CT plots were moldboard-plowed to 0.20 m in the fall and spring, RT plots were offset-disked to 0.15 m in the fall and spring, and NT plots were not tilled at any time. The field has been under this management practice since 1991. Treatments were laid out in a randomized complete block design. The study site consists of 18 plots, half (nine plots) with residue and planted to corn harvested
Dissolved organic C Three soil samples were collected from each plot from 0–0.15, 0.15–0.30, and 0.30–0.45 m depth increments using a hand-held soil auger before planting and after harvest. Samples from within each treatment plot were combined to make a composite samples. To extract water-soluble organic C (i.e., dissolved organic C; DOC), a 10-g field-moist subsample was shaken in 100 ml deionized distilled water for 1 h, centrifuged for 10 min at 1,638 g, and then filtered through Whatman no. 5 paper. Samples were analyzed for organic C using a Shimadzu TOC-5000A total organic C analyzer (Shimadzu Scientific Instruments, Columbia, Md.).
342 Soil parameters
Results and discussion
Soil moisture content was determined by oven drying soil cores at 105C for 48 h. Soil bulk density (Mg m-3) was determined by the core method and total porosity was calculated assuming a particle density of 2.65 Mg m-3. Percent water-filled pore space (%WFPS) was calculated as:
Climatic data
WFPS ¼
%H2 O 100 porosity
ð2Þ
where, %H2O is the percentage of volumetric soil water content. Statistical analysis ANOVA was performed for each sampling date and depth. Differences among means were evaluated using Sheffe’s multiple comparison test. Data was analyzed as a randomized complete block design. All statistical analyses were conducted using the General Linear Model procedure of the Statistical Analysis System, (SAS Institute, Cary, N.C.).
Table 1 Mean monthly precipitation and air temperature during the 1999 and 2000 growing seasons measured at Macdonald Campus Research Weather Station, compared to the long-term mean (1961–1991) a
Month
May June July August September October Mean Total
Total seasonal (May–October) rainfall in 1999 was near the 30-year norm (Table 1). May and August were the driest months during the 1999 growing season, each receiving only slightly more than half the normal rainfall. Rainfall in the 2000 growing season was 40 mm (8%) greater than the norm, with May being the wettest month followed by August, and October was the driest. It is interesting to note that the two driest months in 1999 (May and August) corresponded to the wettest months in 2000 (Table 1). The average mean monthly temperature was 2C and 0.7C higher than the norm in 1999 and 2000, respectively (Table 1).
Air temperature (C)
Precipitation (mm)
1999
2000
1961–1991
1999
2000
1961–1991
15.3 20.8 21.6 19.1 17.1 7.6 17.0 –
13.0 16.7 19.2 19 13.7 8.6 15.0 –
12.9 18 20.8 19.4 14.5 8.3 15.7 –
40.8 111 100 55 100.1 90.6 – 497
133.3 86.0 81.2 125.5 84.0 29 – 539
68 83 86 100 87 75 – 499
a Long-term averages were not available for the Macdonald Weather Station. Values for the long-term averages were obtained from Dorval International Airport, about 10 km east of the field site
Fig. 1 NO3--N concentrations (mg kg-1 soil) in the soil profile under conventional tillage (CT), reduced tillage (RT) and notillage (NT) systems in a spring 1999, b fall 1999, c spring 2000, and d fall 2000. For each depth, bars with different letters are significantly (P
