| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Univ. of Arkansas, Fayetteville, AR 72701
* Corresponding author (kbrye{at}uark.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: ASPB, Arkansas Soybean Promotion Board • CBES, Cotton Branch Experiment Station • CT, conventional tillage • CTIC, Conventional Tillage Information Center • IPCC, Intergovernmental Panel on Climate Change • NT, no-tillage • NOAA, National Oceanic and Atmospheric Administration • OM, organic matter • PTBS, Pine Tree Branch Experiment Station • VWC, volumetric water content • WFPS, water-filled pore space
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Crop residues in general serve a number of beneficial functions, including soil surface protection from erosion, water conservation, and maintenance of soil organic matter (OM). Large amounts of residue on the soil surface have traditionally been viewed as a nuisance, and have been associated with mechanical planting difficulties, poor stand establishment, decreased efficacy of herbicides, release of growth-inhibiting allelopathic compounds and ultimately yield reductions (Sanford, 1982; Caviness et al., 1986; Hairston et al., 1987). Therefore, crop residues, particularly wheat residue, are commonly burned or plowed under followed by disking to prepare a seedbed for double-cropped soybean (Sanford, 1982; Heatherly et al., 1996; Kelley and Sweeney, 1998; Prasad et al., 1999).
Residue burning is a quick, labor-saving practice to remove residue that is viewed as a nuisance by producers. However, residue burning has several adverse environmental and ecological impacts. The burning of dead plant material adds a considerable amount of CO2 and particulate matter to the atmosphere and can reduce the return of much needed C and other nutrients to the soil (Prasad et al., 1999). The lack of a soil surface cover may also increase the loss of soil minerals via runoff. Crop residues returned to the soil maintain OM levels, and crop residues also provide substrates for soil microorganisms. As microbes use or decompose crop residues and soil OM, CO2 is given off as a by-product of soil respiration. Therefore, it is reasonable to believe that residue level might affect soil surface CO2 fluxes.
Microbial and root activity, which together constitute soil respiration, are simultaneously influenced by many factors, such as fluctuations in soil moisture and temperature (de Jong et al., 1974; Wildung et al., 1975; Hendrix et al., 1988; Grahammer et al., 1991; Fortin et al., 1996; Wagai et al., 1998). Environmental fluctuations of temperature and moisture are in turn affected by residue management practices, such as burning and tillage. For example, burning removes residue and the insulating effect of a residue cover on the soil surface causing temperatures to increase, which can stimulate microbial activity and soil respiration, and enhanced fluctuations in soil moisture and temperature. Managing crop residues by burning also removes the evaporation barrier the residue provides causing the soil to dry out quicker; thus also affecting microbial activity and soil respiration. However, little is known about the effects of crop residue burning on soil C and its potential release to the atmosphere through soil respiration (Lal et al., 1998). Tillage mixes and loosens the soil and has also been shown to stimulate microbial activity, OM oxidation, and soil respiration (Doran, 1980; Weil et al., 1993; Franzluebbers et al., 1995; Fortin et al., 1996; Lal, 1997; Reicosky, 1997; Dao, 1998; Kessavalou et al., 1998a; Reicosky et al., 1999; Curtin et al., 2000; Aulakh et al., 2001).
Elevated CO2 and particulate matter in the atmosphere have been recognized as contributing factors to the greenhouse effect, global warming, and general climate change (Houghton et al., 1990; Post et al., 1990; Wood, 1990; Curtin et al., 2000; Hulugalle, 2000). Consequently, burning is increasingly being looked on as an unacceptable residue management practice because of the additional CO2 and particulate matter that is expelled into the atmosphere upon combustion of crop residues. The 1995 Intergovernmental Panel on Climate Change (IPCC) determined that agriculture was directly responsible for approximately 20% of the annual anthropogenic emissions of greenhouse gases (IPCC, 1995). Agricultural soils can however actually serve as a significant C sink, rather than a C source, at least until the maximum capacity to store C is achieved, if improved residue management and reduced tillage systems are adopted (Paustian et al., 2000). Therefore, alternative residue management options must be developed to minimize or eliminate the tradition of residue burning and reduce soil respiration.
One change many producers are implementing throughout all agricultural regions, and especially in the mid-southern USA, is the conversion from conventional-tillage to conservation- and NT practices. Conventional tillage leaves <30% crop residue cover after planting, whereas conservation tillage leaves more than 30% residue cover. No tillage is a form of conservation tillage that excludes any tillage between harvest and planting. The soybean area planted in a double-crop system in Arkansas using conservation tillage practices has increased by almost 5000 ha yr–1 in the last 15 yr (Conservation Tillage Information Center [CTIC], 2003). The soybean area planted under NT in a double-crop system in Arkansas has increased by nearly 7000 ha yr–1 since 1989 (CTIC, 2003).
As management practices change, their effects on crop production and soil and environmental quality must be evaluated. However, few field studies have investigated the effects of alternative wheat residue management practices on soil surface CO2 losses during the soybean cycle of a wheat soybean double crop production system in the subhumid climate of the mid-southern USA. Therefore, the objectives of this study were to (i) evaluate the effects of wheat residue level, residue burning, and tillage on soil surface CO2 flux during the soybean growing season of a wheat soybean double-crop production system, and (ii) evaluate the role of soil temperature and moisture in controlling soil surface CO2 flux in a relatively warm, humid environment. We hypothesized that soil surface CO2 flux would be lower under NT and without residue burning than in the typical, conventional-tillage with residue burning combination. In addition, we hypothesized that CO2 fluxes would be lower under a low residue amount setting. We also hypothesized that soil surface CO2 flux is controlled more by soil temperature than soil moisture fluctuations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The climate within the region of the two study locations is warm and wet. The 30-yr mean annual temperature in the region, averaged across records from the National Oceanic and Atmospheric Administration (NOAA) observation station closest to each study location, is 15.6°C with a January minimum of –2.4°C and a July maximum of 32.8°C (NOAA, 2002). The 30-yr mean annual precipitation is 1282 mm (NOAA, 2002).
Experimental Design and Treatments
The experimental design was a split-strip plot with six blocks of eight wheat residue treatment combinations. Wheat residue management practices evaluated included NT and CT before planting soybean into high and low wheat residue levels with and without burning. There were a total of 48 experimental units at each of the two locations.
Field Management
Before wheat planting, the plot area was disked twice followed by landplaning and field cultivation at PTBS and disked twice followed by field cultivating at CBES. In addition, before wheat planting at CBES, but not at PTBS, a 224 kg ha–1 broadcast application of 9–23–30 blended fertilizer was applied. In Fall 2001, a single wheat cultivar (Coker 9663) was drill-seeded with a 0.15-m row spacing at a rate of 110 kg ha–1 at PTBS and 112 kg ha–1 at CBES. In Spring 2002, 3 by 6 m plots were established at both locations. The first N application of 101 kg N ha–1 as urea (46% N), was made to all 48 plots on March 1 and a split application of an additional 101 kg N ha–1 as urea was made on March 30 to one-half of the plots to achieve a higher wheat residue mass. Fertilizer N rates were not chosen based soil tests or on a particular wheat yield goal, but rather to achieve two different desired residue levels. Since different N rates were used to establish different desired residue levels, this treatment is hereafter referred to as the N rate/wheat residue level treatment.
Wheat was harvested in early-June 2002 at both locations. The plot combine used to harvest the wheat left a pile of residue at the end of each plot, therefore the residue pile had to be spread uniformly by hand back over the plot from which it came. Wheat residue was then mowed to the soil surface to facilitate soybean planting. After wheat harvest and mowing, the burning treatment was imposed followed by imposition of the tillage treatments. Approximately 14 d after wheat harvest, a single glyphosate-resistant soybean cultivar (Pioneer 95B32, maturity group 5.3) was drill-seeded with a row spacing of 0.19-m at a rate of 100 kg ha–1 at PTBS and 47 kg ha–1 at CBES. The higher soybean seeding rate at PTBS was necessary due to low soil moisture conditions. In 2002, soybeans were furrow-irrigated at CBES and flood-irrigated at PTBS on an as-needed basis three times throughout the growing season.
In November 2002, the same wheat cultivar as the previous year was planted to the study area by drill seeding with a 0.15-m row spacing at a rate of 112 kg ha–1 at PTBS and 110 kg ha–1 at CBES. The first N application of 101 kg N ha–1 as urea was made on 11 March and the additional 101 kg N ha–1 as urea was applied on 4 April to the high N rate/residue plots. Wheat was harvested and chopped in mid-June 2003 at both locations. The burn followed by the tillage treatments were again imposed immediately following wheat harvest. Two and thirteen days after wheat harvest at PTBS and CBES, respectively, a single glyphosate-resistant soybean cultivar (Pioneer 95B32; maturity group 5.3) was drill-seeded without tillage with a row spacing of 0.19 m at a rate of 90 kg ha–1 at PTBS and 107 kg ha–1 at CBES. In 2003, soybeans were again irrigated on an as-needed basis three times throughout the growing season. During both cropping cycles, weeds were controlled at both locations on an as-needed basis with no more than two applications of glyphosate during any single growing season.
Soil Physical and Chemical Properties
At wheat harvest in 2002, ten 2-cm diam. cores were collected and composited from the 0- to 10-cm depth of each plot (i.e., n = 48). Soil samples were oven dried at 70°C for 48 h, crushed, sieved through a 2-mm mesh screen, and used for particle-size analysis using the hydrometer method (Arshad et al., 1996). Dried and sieved soil was also extracted with Mehlich-3 extractant solution (Tucker, 1992) in a 1:10 soil/extractant solution ratio and analyzed for extractable phosphorus (P) and potassium (K) using inductively coupled argon-plasma spectrophotometry (CIROS CCD model, Spectro Analytical Instruments, MA). Soil pH and electrical conductivity (EC) were determined with an electrode on a 1:2 soil/water solution. Soil OM was determined by weight-loss-on-ignition after 2 h at 360°C (Schulte and Hopkins, 1996).
Approximately 8 wk after soybean planting in 2002 and 2003, a single 4.8-cm diam. soil core was collected from the 0- to 10-cm depth in each plot, oven dried at 70°C for 48 h, and weighed for bulk density determination. The soil-core sampling chamber was beveled to the outside to minimize compaction on sampling.
Wheat Residue
Before wheat harvest each year, plants in a 1-m section of one of the nonharvest rows were cut at the soil surface and collected for total aboveground biomass determination. The 1-m row samples were dried at 70°C for 3 d, weighed, and mechanically thrashed. Grain from the 1-m row samples was also collected after thrashing and weighed.
Following wheat harvest and residue burning each year, but before cultivation and soybean planting, soil surface wheat residue levels were measured by cutting and collecting the residue within a 0.25-m2 metal frame in each non-burned plot (i.e., n = 24; Brye et al., 2002). Residue samples were oven dried at 70°C for 48 h and weighed to express residue level on a kg ha–1 basis.
Soil Surface Carbon Dioxide Flux
Soil surface CO2 flux was measured four times between mid-July and mid-September throughout the 2002 soybean growing season and five times between late-June and the end of September throughout the 2003 soybean growing season. Soil surface CO2 flux was measured with a LI-6400 portable CO2 infrared gas analyzer (Li-Cor Inc., Lincoln, NE) equipped with a 10-cm diam. soil respiration chamber (LI-6400–09). This methodology provides a rapid, reliable estimate of soil surface CO2 flux that is comparable with other approaches (de Jong et al., 1974; Norman et al., 1997) and has been used extensively in a variety of ecosystems including forests (Haynes and Gower, 1995), grasslands (Norman et al., 1992; Wagai et al., 1998; Brye et al., 2002), and agroecosystems (Wagai et al., 1998; Brye et al., 2002). Measurements were conducted in four of the six blocks of treatment combinations for a total of 32 plots measured per location per measurement date.
Before the first measurement date in 2002 and 2003, one thin-walled (3.2 mm), polyvinyl chloride (PVC) ring (i.e., collar; 10-cm i. d. x 5 cm tall) was inserted approximately 2 cm into the soil at a random location in each plot. The collars were constructed such that the soil respiration chamber would fit snugly on to each collar. Once placed on top of the PVC collar, the CO2 in the head space of the chamber was scrubbed to approximately 25 mg L–1 below the target concentration by passing the gas sample through a cylinder of soda lime pellets. Once scrubbing ceased, the CO2 concentration in the head space was allowed to build up until it reached 10 mg L–1 below the target concentration. At this point, the actual measurement sequence began, such that the soil surface CO2 flux was determined from 10 mg L–1 below ambient, through the ambient CO2 concentration, to 10 mg L–1 above the ambient CO2 concentration, which took between about 20 s to several minutes depending on the soil temperature and moisture status at the time of measurement. The internal flow rate was 0.94 L min–1, in which the air flow is distributed through a manifold for thorough mixing so that any disturbance (i.e., artificial pressure gradient) imposed on the boundary layer that may effect the CO2 flux is negligible.
After completing the set of measurements at one location, the collars were collected, inserted into the same plots at the second location, and allowed to equilibrate overnight. After completing the measurements at the second location the following day, the collars were repositioned in different locations within the same plots until the next measurement date. This protocol was followed throughout both measurement years such that systematic variations due to measurements being conducted at the same time of day at the same location each measurement interval were avoided. Measurements were generally conducted between mid-morning and mid-afternoon between 0900 and 1500 h.
Soil Temperature and Moisture Content
Soil temperature and volumetric water content (VWC) measurements were conducted along with each soil surface CO2 flux measurement. Soil temperature was measured within 5 cm of the outside perimeter of each collar at depths of 2.5 and 10 cm using a long-stem thermometer. Volumetric soil water content was measured in the 0- to 6-cm depth once inside the collar immediately after the CO2 flux measurement was completed using a Theta Probe (model TH2O, Dynamax, Houston, TX, USA), which records dielectric voltage readings and converts them to volumetric water contents using a soil-specific calibration equation. Water-filled pore space (WFPS) was calculated by dividing measured volumetric water contents by the total soil porosity for each plot. Total soil porosity was calculated using measured bulk densities and an assumed soil particle density of 2.65 g cm–3.
Statistical Analyses
Two-sample t tests were conducted to ascertain differences in initial soil physical and chemical properties among the two study locations (Minitab 13.31, Minitab Inc., State College, PA). Two-sample t tests were also conducted to ascertain differences in soil surface bulk density between tillage treatments and in soil surface residue levels between N rate/residue treatments at each location separately due to resulting differences in inherent soil fertility between the two locations.
Analysis of variance (ANOVA) was used in accordance with the experimental design to ascertain the effects of N rate/residue level, burning, tillage, and time on soil surface CO2 flux using SAS Version 8.1 (SAS Institute, Inc., Cary, NC). Separate ANOVAs were conducted for each study location due to inherent soil fertility differences between locations. Since there was a total of only nine sample dates over the 2-yr period and at least 7 d between consecutive measurements in the same plot, CO2 fluxes measured on the nine samples dates were assumed independent of each another.
Multiple linear regression was used to determine the contribution or relative importance of each measured or calculated soil variable (i.e., 2- and 10-cm soil temperature and 0- to -6-cm VWC and WFPS) toward explaining the variation in soil surface CO2 flux. The relative importance of each variable was determined as the percentage of the total sum of squares attributable to each variable's sequential sum of squares from a multiple regression model. The sequence of variables in the model was based on linear correlations, in which variables were added to the model from highest to lowest correlation.
Based on correlation significance, regression analyses were also conducted to determine the combination of measured variables with significant linear and quadratic terms that produced the highest goodness of fit (R2). All regression analyses were conducted for each location separately (n = 288 observation per location) and for data from both locations combined (n = 576 total observations).
Since soil temperature has been shown to confound the effects of soil moisture on soil surface CO2 (Grahammer et al., 1991; Wagai et al., 1998) and preliminary correlations using actual measured fluxes yielded many poor and nonsignificant relationships, all correlation and regression analyses presented were conducted using normalized soil surface CO2 fluxes that were adjusted to a standard temperature (T0) of 10°C using the following modified equation after Ryan (1991):
 |
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
For the NT treatment, preharvest total aboveground wheat biomass was generally unaffected by the applied N treatments, except in 2002 at PTBS where the high N rate/wheat residue treatment, achieved using a higher N rate (202 vs. 101 kg N ha–1), had a significantly higher residue mass than the low N rate/wheat residue treatment (Table 1). Similarly, soil surface wheat residue levels immediately before soybean planting were generally statistically unaffected by the applied N treatments, except in 2003 at CBES where the high N rate/wheat residue treatment had a significantly higher residue mass than the low N rate/wheat residue treatment (Table 1). Though not formally compared, the apparent two-fold increase in mean surface residue levels in the high N rate/residue level at CBES from 2002 to 2003 was likely the result of increased weed biomass that was stimulated by the additional N inputs. In general, soybeans were directly planted into more wheat residue at CBES than PTBS in both years, though formal statistical comparisons between locations were not made. Since there was a lack of a desired wheat residue difference between the high and low N rates in three out of 4-yr location combinations, the N rate/wheat residue treatment is hereby referred to only as the N rate treatment and used as such in all forthcoming statistical analyses.
|
Effects of Wheat Residue Management Practices on Soil Surface Carbon Dioxide Flux
Soil surface CO2 fluxes measured over the 2-yr period of this study varied from a low of 0.46 µmol CO2 m–2 s–1 in mid-August 2002 from the CT-low N rate/residue-burned treatment combination to a high of 21.03 µmol CO2 m–2 s–1 in mid-July 2002 from the CT-low N rate/residue-non-burned treatment combination at CBES (Fig. 1
). At PTBS, soil surface CO2 flux varied from a low of 1.54 µmol CO2 m–2 s–1 at the end of September 2003 to a high of 13.8 µmol CO2 m–2 s–1 at the end of June 2003 from the CT-low N rate/residue-burned treatment combination (Fig. 1).
|
|
There was a significant burn x time (P = 0.032) interaction at PTBS (Fig. 1). Depending on sample date, the mean soil surface CO2 flux from the burned treatment was either higher (sample date: 9/11/02) or lower (sample date: 8/12/03) than that from the unburned treatment at PTBS.
Both locations also had a significant tillage x time (P < 0.0001) interaction (Table 2). Depending on sample date, the mean soil surface CO2 flux from CT was either higher (sample dates: 7/17/02, 6/23/03, and 7/24/03) or lower (sample dates: 9/4/02, 7/18/03, and 9/20/03) than that from NT at PTBS (Fig. 1). At CBES, the mean soil surface CO2 flux from CT was higher than that for NT on every sample date except for in mid-August 2002 when the CO2 flux from NT was higher and in mid-September 2002, late-July 2003, and mid-August 2003 when mean soil surface CO2 flux from CT and NT was the same (Fig. 1).
In addition to the interactions over time, there was a significant N rate x burning interaction (P = 0.016) at PTBS, but not at CBES (Fig. 2 ). At PTBS, soil surface CO2 flux from the burned treatment was 13% higher from the low than from the high N rate treatment. However, there was no difference in mean soil surface CO2 flux from the high or low N rate treatment when the residue was left at PTBS. There was also a significant (P = 0.014) burning x tillage x time interaction at CBES, but not PTBS (Table 2).
|
|
0.004), but 10-cm soil temperature explained 20.3% compared to only 0.1% for 2.5-cm soil temperature of the variation in temperature-normalized soil surface CO2 flux. Though not significant in the model, WFPS explained 15.0% of the variation in temperature-normalized soil surface CO2 flux at CBES. For data from both locations combined, only 10-cm soil temperature was significant in the model (P < 0.001) and alone explained 23.1% of the variation in temperature-normalized soil surface CO2 flux.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Soil OM oxidation is generally more pronounced in the southeastern USA due to the thermic soil temperature regime and extended duration of high temperatures (Reicosky et al., 1999). However, Wood et al. (1991) concluded that, even in the organic-matter-depleted southeastern USA, increased soil C can still be achieved in thermic-udic regimes with the proper combination of conservation tillage and cropping practices.
The significant effects of tillage and N rate on soil surface CO2 flux were observed at CBES, but not at PTBS (Table 2). Averaged across all treatments and measurement date, soil surface CO2 flux was 5.04 and 5.57 µmol CO2 m–2 s–1 at CBES and PTBS, respectively. Though sand and clay content, soil pH, and EC, and extractable P and K were significantly different between locations, there is little evidence to suggest that these inherent differences contributed to the difference in treatment effects between locations. However, PTBS had a higher initial soil OM concentration than CBES, which generally supports the average CO2 flux difference between locations over the course of the study. Since substrate availability (i.e., soil OM level) is known to affect soil respiration (Robinson et al., 1994), this may have contributed to the lack of significant treatment effects at PTBS because the slightly higher background levels of OM may have masked any effects of imposed wheat residue management practices. In contrast, the lower native soil OM level at CBES likely resulted in a more responsive soil environment to imposed wheat residue management practices than at PTBS.
The reported significant effect of N rate on soil surface CO2 flux must be tempered by the fact that, though differential total applied N rates were intended to produce differential levels of surface wheat residue into which the subsequent soybean crop was to be planted, a significant difference in wheat residue level was realized in only one of four year–location combinations (Table 1). This result indicates that carry-over N may have existed for use by the subsequent soybean crops. Substantial carry-over N in the plots that received the higher N rate at the time of soybean planting may have contributed to the observed treatment differences, or lack thereof, in soil surface CO2 flux in ways that were not addressed in this study. Readily available inorganic N that may have carried-over from the previous wheat crop could have stimulated microbial activity due to a slightly lower soil C to N ratio and may have resulted in potentially higher soil surface CO2 fluxes in the high than in the low N rate plots. Though, in theory, microbial activity may have increased some in the presence of readily available inorganic N, it is unlikely that the effects of slightly more inorganic N would have dominated over other environmental factors, such as temperature and moisture, that are known to control microbial activity.
In a related study conducted adjacent to the current study at both locations during the same time period, neither nitrate-N, ammonium-N, nor total inorganic N (NO3–N + NH4–N) in the top 10 cm differed among plots fertilized at the same N rates as those used in the current study at either location in 2002 or 2003 following wheat harvest, but before soybean planting. Results of this related study indicated that no carry-over N existed in plots before soybean planting following wheat fertilization at 101 and 202 kg N ha–1. In addition, in three of four year–location combinations, total aboveground wheat biomass, and surface residue mass into which the subsequent soybean crop was planted were numerically higher in the high than low N rate treatment (Table 1). Thus, evidence exists suggesting a possible wheat-residue effect on soil surface CO2 flux despite the lack of statistically significant residue-level differences.
The results of this study are similar to those of previous studies of CO2 flux in crop rotations involving wheat. In west-central Minnesota on a clay-loam soil, Reicosky and Lindstrom (1993) and Reicosky (1997) demonstrated that soil surface CO2 flux was higher under moldboard plowing compared with conservation tillage practices such as chisel plowing (CP) and NT. Mean annual CO2 flux was 20 to 25% less under NT than CT following at least 13 yr of NT and CT of continuous wheat and a wheat-fallow rotation on silt-loam soil in southwestern Saskatchewan (Curtin et al., 2000). Lower CO2 fluxes under NT than CT were reported in wheat in south-central Texas (Franzluebbers et al., 1995) and in a corn (Zea mays L.)–soybean– crimson clover (Trifolium incarnatum L.) cover crop rotation on loamy sand in central Alabama (Reicosky et al., 1999). Similarly, Dao (1998) reported CO2 fluxes that were two-fold higher under CT than NT within a 60-d fallow period following wheat grown on a silt-loam soil in Oklahoma. Kessavalou et al. (1998a) concluded that annual interrow CO2 emissions from a wheat-fallow rotation were generally higher under tilled than nontilled loam in Nebraska. In a laboratory incubation experiment using silt-loam soil, 38% of wheat residue-C was emitted as CO2 with incorporation of residue into the soil compared to only 13% when wheat residue was surface-applied (Curtin et al., 1998).
However, the results of this study are in contrast to several that have reported higher CO2 loss under NT than CT practices. Over an 18-mo period in Georgia, higher CO2 fluxes were recorded from 5- and 6-yr-old NT than from CT soils in both grain sorghum and soybean (Hendrix et al., 1988). Wagai et al. (1998) demonstrated higher annual CO2 loss under NT than CP in corn grown on a silt-loam soil in south-central Wisconsin.
Though not examined in this study, tillage has also been shown to have a seasonal effect on soil surface CO2 flux. Wagai et al. (1998) reported CO2 fluxes were higher in corn under CP than NT early in the growing season, but that CO2 fluxes were higher under NT than CP late in the growing season. A similar trend was reported by Fortin et al. (1996) for barley and oat grown under CT and NT on a silt-loam soil in Ottawa, ON.
Wheat residue mass has also been shown to affect CO2 flux (Dao, 1998; Duiker and Lal, 2000). Duiker and Lal (2000) reported few wheat residue mass treatment differences on soil surface CO2 flux in a NT setting, except that the soil surface CO2 flux was most often higher from 16 Mg ha–1 of wheat residue than from no wheat residue. Dao (1998) reported a tillage x residue level interaction effect on soil surface CO2 evolution, where there was hardly any effect of wheat residue mass (0 and 4 Mg ha–1) on cumulative CO2 evolution under NT, but considerably more cumulative CO2 loss with wheat residue than without under moldboard plowing.
Despite nonburning of crop residues being advocated as a management practice to reduce agriculturally related CO2 emissions to the atmosphere (Prasad et al., 1999; Hulugalle, 2000), few have investigated the effects of residue burning on CO2 evolution from the pedosphere in an agricultural setting (Lal et al., 1998). Similar to the results of this study, burning had little effect on subsequent soil surface CO2 fluxes in Brazilian soils subject to slash-and-burn agricultural practices (Anderson and Poth, 1998) and in the tallgrass Konza Prairie in Kansas (Tate and Striegl, 1993).
The lack of consistent significant treatment effects and consistent effects between locations was likely the result of several study limitations. The N fertilization regimes used failed to produce statistically different wheat residue levels. The low CO2 flux measurement frequency likely missed some of the extreme variations in soil moisture and temperature that may have made correlations more significant. Similarly, spatial variability in soil CO2 flux may not have been adequately captured with only one chamber measurement per plot.
Factors Controlling Soil Surface Carbon Dioxide Flux
Both soil temperature and soil moisture content have generally been shown to control soil surface CO2 flux in a variety of climate regimes (de Jong et al., 1974; Wildung et al., 1975; Hendrix et al., 1988; Grahammer et al., 1991; Fortin et al., 1996; Wagai et al., 1998). However, the results of this study conducted in the relatively warm, subhumid climate of the mid-southern USA deviate somewhat from this generality.
Temperature-normalized CO2 flux was significantly, though negatively, correlated with variations in 2- and 10-cm soil temperature at both locations (Fig. 3 ), with 0- to 6-cm VWC (Fig. 3) and WFPS at CBES, and with all four variables when data from both locations were combined (Table 3). However, when added to a multiple regression model based on significance and by order of highest correlation, 10-cm soil temperature was the only variable that was both significant in the model and explained more than 20% of the variation in temperature-normalized CO2 flux at either location separately or for data from both locations combined (Table 4). This result is similar to others who have observed that soil temperature controlled soil surface CO2 flux more than soil moisture parameters (Hendrix et al., 1988; Fortin et al., 1996; Wagai et al., 1998).
|
The results of this study are in contrast to de Jong et al. (1974) who reported the addition of soil moisture to a multiple regression model did not improve the fit for cereal or fallow cover crops in Saskatchewan. Similarly, in contrast to the negative correlation between temperature-normalized CO2 flux and soil temperature and soil moisture parameters in this study, Kessavalou et al. (1998a, 1998b) reported that CO2 flux increased significantly with both soil temperature and WFPS in the top 7.6 cm of a loam cropped to a wheat–fallow rotation in Nebraska. Differences in significant model parameters between locations are likely a partial result of varying inherent soil properties between locations.
With a significant tillage x time interaction at both locations (Table 2), the alternating moist-dry cycles related to irrigation events and periodic rainfall may have masked potential significant effects of soil moisture parameters among tillage treatments on soil surface CO2 flux. Curtin et al. (2000) demonstrated reduced soil surface CO2 emissions from soils subjected to alternative wet-dry cycles than those subjected to continuously moist soil conditions. Overall, including 2.5- and 10-cm soil temperature, VWC, and WFPS in a multiple regression model explained only 29 to 35% of the variation in temperature-normalized soil surface CO2 flux (i.e., 0.29 < R2 < 0.35; Table 4). Consequently, there must be other soil and/or environmental factors controlling soil surface CO2 flux during the soybean growing season of a wheat–soybean rotation in the mid-southern USA than the ones measured or derived in this study.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Results of this study indicate that tillage and N fertilization, potentially through an effect of surface residue levels, but not burning, affect the loss of C as CO2 from the soil. Specifically, conversion to NT has the potential to reduce soil surface CO2 fluxes. The combination of nonburning residue followed by NT planting is a feasible set of alternative wheat residue management practices that will likely result in the build-up of soil OM and increased C sequestration, both of which will improve the long-term sustainability of the wheat–soybean rotation in the mid-southern USA.
Results also indicate that there are additional soil and/or environmental factors, other than near-surface soil moisture and temperature that contribute to controlling soil surface CO2 flux during the soybean growing season of a wheat-soybean rotation in the mid-southern USA. Though not evaluated in this study, soil microbial biomass dynamics may be a significant factor controlling soil surface CO2 fluxes.
Received for publication May 12, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
