Published in Agron. J. 97:349-363 (2005).
© American Society of Agronomy
677 S. Segoe Rd., Madison, WI 53711 USA
Symposium Papers
Carbon Storage in Soils of the North American Great Plains
Effect of Cropping Frequency
C. A. Campbella,*,
H. H. Janzenb,
K. Paustianc,
E. G. Gregoricha,
L. Sherrodd,
B. C. Liange and
R. P. Zentnerf
a Agric. and Agri-Food Canada, Cent. Exp. Farm, Ottawa, ON K1A 0C6, Canada
b Agric. and Agri-Food Canada Res. Cent., Lethbridge, AB T1J 4B1, Canada
c Dep. of Soil and Crop Sci., Colorado State Univ., Fort Collins, CO, 80523
d Great Plains Syst. Res. Unit, USDA-ARS, P.O. Box E, Fort Collins, CO, 80522
e Pollution Data Res., Environ. Canada, 851 St. Joseph Blvd., 9th Floor, Hull, QC K1A 0H3, Canada
f Agric. and Agri-Food Canada Res. Centre, Swift Current, SK S9H 3X2, Canada
* Corresponding author (campbellca{at}agr.gc.ca)
Received for publication December 15, 2003.
 |
ABSTRACT
|
|---|
Summer fallow (fallow) is still widely used on the North American Great Plains to replenish soil moisture between crops. Our objective was to examine how fallowing affects soil organic carbon (SOC) in various agronomic and climate settings by reviewing long-term studies in the midwestern USA (five sites) and the Canadian prairies (17 sites). In most soils, SOC increased with cropping frequency though not usually in a linear fashion. In the Canadian studies, SOC response to tillage and cropping frequency varied with climate—in semiarid conditions, SOC gains under no-till were about 250 kg ha–1 yr–1 greater than for tilled systems regardless of cropping frequency; in subhumid environments, the advantage was about 50 kg ha–1 yr–1 for rotations with fallow but 250 kg ha–1 yr–1 with continuous cropping. Specific crops also influenced SOC: Replacing wheat (Triticum aestivum L.) with lentil (Lens culinaris Medikus) had little effect; replacing wheat with lower-yielding flax (Linum usitatismum L.) reduced SOC gains; and replacing wheat with erosion-preventing fall rye (Secale cereale L.) increased SOC gains. In unfertilized systems, cropping frequency did not affect SOC gains, but in fertilized systems, SOC gains often increased with cropping frequency. In a Colorado study (three sites each with three slope positions), SOC gains increased with cropping frequency, but the response tended to be highest at the lowest potential evaporation site (where residue C inputs were greatest) and least in the toeslope positions (despite their high residue C inputs). The Century and the Campbell et al. SOC models satisfactorily simulated the relative responses of SOC although they underestimated gains by about one-third.
Abbreviations: Cont W, continuous wheat • F-W, fallow–spring wheat (rotation) • PET, potential evapotranspiration • SOC, soil organic carbon • W-Lent, spring wheat–lentil
|
INTRODUCTION
|
|---|
THE FACTOR MOST LIMITING to crop production on the North American Great Plains is available water. Summer fallow—leaving land unplanted for a season to replenish stored water—reduces the likelihood of crop failure from drought and has been widely used for more than a century (Janzen, 2001). Though its prevalence has waned in recent decades, there are still about 6 million ha of summer fallow land in Canada (Campbell et al., 2002) and 20 million ha in the USA (USDA-NRCS, 2000).
Summer fallow has some clear advantages compared with continuous cropping; however, it also has some serious shortcomings, one of which is that it results in soil C (organic matter) loss. The storage of C in soils is a function of C inputs and losses (Campbell et al., 2000a, 2000b). Inputs, primarily as photosynthates, are added either directly (e.g., crop residue) or indirectly (e.g., animal manure derived from plant C). Losses occur mostly as CO2 from decomposition. Summer fallow reduces C storage in several ways (Janzen et al., 1998). First, frequent summer fallow usually reduces inputs of photosynthetically derived C into soils because there are no plant C inputs (except via weeds) during the fallow phase. Second, it may enhance the rate of mineralization of soil organic matter to CO2 because it keeps the soil wetter (and perhaps warmer) for longer periods. Further, if tillage is used to control weeds during summer fallow, decomposition may be accelerated by disruption of soil aggregates and exposure of organic matter to microbial activity. Summer fallow may also accelerate soil C losses by erosion, but these losses are localized, often resulting in redistribution of C more than in its release to the atmosphere (Gregorich et al., 1998).
Our objective was to review data from long-term experiments in the U.S. and Canadian Great Plains to determine the magnitude of the effect of summer fallow on changes in SOC and to assess how these changes are modified by such factors as tillage, fertility, soil type, topography, and climate.
|
DATA SOURCES AND LIMITATIONS
|
|---|
Quantifying changes in soil C is a difficult task; annual changes per year are small compared with C already present, and its spatial variability is very large (Campbell, 1978; McGill and Bailey, unpublished, 1999). Thus, reliable estimates of C change depend on sampling randomly at test sites over many years or by sampling at specific locations at a site, repeatedly over time (Ellert et al., 2001). In the past, scientists studied soil C mostly from the perspective of soil fertility (Ridley and Hedlin, 1968; Janzen, 2001). Consequently, many of the studies focused on relative differences in soil C concentration among treatments; bulk density, needed to calculate soil C mass, was often not determined. Adding further to the difficulty are changes in analytical methods over time (e.g., the shift from wet digestion methods to the dry combustion methods now commonly used); in the absence of archived soil samples, these changes in methods can introduce artifacts in temporal trends (Monreal and Janzen, 1993).
Sampling depth is another complicating factor. In the semiarid prairies where tillage is usually shallow (no more than 15–20 cm) most of the SOC changes are expected to occur near the soil surface. Thus, sampling usually is in increments to 15- or 30-cm soil depth. Typically, absolute differences in SOC among treatments increase with the depth sampled, but statistical significance declines because total mass of C measured and its variability increase much more than the difference among treatments (Ellert et al., 2001). Significant differences are often observed only in the surface 7.5 or 10 cm (Table 1), if at all (Campbell et al., 2000b; Sherrod et al., 2003). We have therefore confined our discussion primarily to changes in SOC in the 0- to 15-cm depth in Canadian soils and 0- to 20-cm in U.S. soils.
View this table:
[in this window]
[in a new window]
|
Table 1. Soil organic C in top 30-cm depth after 10 yr as influenced by cropping frequency in New Rotation Experiment at Swift Current (values in Mg ha–1 on mass/area basis).
|
|
In several cases, the initial soil C levels were not available and had to be estimated by assuming that all treatments in a study had the same initial value, an assumption sometimes leading to large discrepancies (Fig. 1; Campbell et al., 1995, 2000b). For example, in 1981, Campbell et al. (1995) initiated a no-till study at Swift Current on farmers' land that had been in a fallow–wheat conventional tillage system for the previous 70 to 80 yr. Since they had not measured SOC at the start of the experiment but wished to assess changes in SOC in the no-till study in later years, they thoroughly sampled the adjacent farmland, which was still managed as in the previous 70 to 80 yr, assuming that SOC in that system would be at steady state. However, as seen in Fig. 1 and by their estimates of C inputs and conversion to SOC, they concluded that the measured SOC in 1981 (22.3 Mg ha–1) could not have been the true initial value for the test site. They therefore used a linear extrapolation (dotted line, Fig. 1) to estimate the probable starting value.
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 1. Changes in soil organic C in a Swinton loam at Swift Current, SK, after land that had been in fallow–wheat for 70 to 80 yr, with minimal N inputs and conventional tillage (CT), was converted to a continuous wheat (Cont W), fertilized, no-tillage (NT) system. (Note: Estimated initial soil organic C using adjacent fallow–wheat area gives poor estimate.) (Adapted from Campbell et al., 1995.)
|
|
For many experiments, differences in soil C stocks among treatments are not always significant at P = 0.05, even after several decades. However, repeated sampling of the same treatments over time, along with supportive modeling efforts, may help to bolster the reliability of the conclusions (Campbell et al., 2000a, 2000b, 2002).
Our approach, in this paper, was to summarize all data from long-term (>10 yr) experiments on the Canadian Prairies and U.S. Great Plains in which soil C changes were measured and summer fallow frequency was a variable. In some cases, we compare measured vs. modeled changes in soil C. We provide median and mean rates of C storage as a function of type of crop, tillage, fertility, and soil zone (in Canada) and potential evapotranspiration (PET) and slope position (in USA). Many of our results are presented as coefficients, expressed in Mg C ha–1 yr–1, but with the emphatic caution that extrapolation of such rates beyond one or two decades is probably not justified because, excluding major changes in agronomy or weather, SOC will tend to approach a steady state after several decades (Janzen et al., 1997; West and Post, 2002).
By definition, our review considers only those studies that include at least one cropping system with summer fallow and one or more treatments with reduced fallow frequency. By default, these experiments are limited primarily to the semiarid regions of the U.S. and Canadian Great Plains where moisture conservation is a priority and summer fallow is thus commonplace. Though few of the studies are free of the deficiencies described earlier, we have filtered the available data by including mainly those experiments where measurements have been made for at least 10 yr and, with few exceptions, where bulk densities have been measured.
|
EFFECT OF FALLOW FREQUENCY ON SOC WHERE INITIAL SOC IS KNOWN OR CAN BE ESTIMATED
|
|---|
U.S. Great Plains
Only one suitable study was found in the U.S. Great Plains (Peterson et al., 1998; Sherrod et al., 2003). In this 12-yr experiment in Colorado, four crop rotations of differing fallow frequency under no-tillage management were studied at three sites differing in PET and, with treatments arrayed across a catenary sequence at each site, divided into three topographic positions (summit, sideslope, and toeslope). Four annual crop rotation treatments included spring wheat–fallow, wheat–corn (Zea mays L.)–fallow, wheat–corn–millet (Panicum miliaceum L.)–fallow, and continuous cropping {including corn/sorghum [Sorghum bicolor (L.) Moench], wheat, hay millet, and sunflower (Helianthus annuus L.)}. Each year-phase of the crop rotations was present each year and replicated twice. All three sites had been cropped to wheat–fallow or sorghum–fallow for more than 50 yr before the start of this experiment when no-till management was established (Ortega et al., 2002). Before this, tillage was mostly done with a Noble blade or tandem disk from 1950–1985 with possibly some moldboard plowing before 1950 (G.A. Peterson, personal communication, 2003).
Soil organic C in the 0- to 10- and 10- to 20-cm depths was measured at the start of the experiment in 1986 and in 1997 (Fig. 2a and 2b). Initial soil C values (1986) were calculated as the median values of samples taken from each field plot, summed across all treatment plots and replicates, within a slope position and site. Soil organic C content was determined by Walkley–Black (Nelson and Sommers, 1996), and inorganic C was determined by a modified pressure calcimeter method (Sherrod et al., 2002). In theory, after 50 yr of crop–fallow, SOC in these conventionally tilled systems should be approaching a steady state (Janzen et al., 1998), unless there was some active erosion (Gregorich et al., 1998). However, the adoption of no-tillage would tend to curtail SOC changes due to erosion, and one might expect SOC gains to be in direct proportion to cropping frequency (Campbell et al., 2000a, 2000b). Further, the gains should be greater in the low-PET sites than in the high-PET sites due to greater crop production (input C) although the rates of SOC mineralization should be higher (because of wetter soil conditions), thus tending to temper this advantage. Although Sherrod et al. (2003) report such trends based on SOC change in the top 10 cm of this soil (Fig. 2a), the results for the top 20 cm (Fig. 2b) were less convincing. Generally, there was a positive relationship between SOC change and cropping frequency. In the 0- to 10-cm depth, most treatments with cropping frequencies of 66% or more maintained or increased SOC; this was especially true at low PET. However, in the 0- to 20-cm depth, SOC gains were apparent only in some continuous cropping systems. Gains were generally greater at low PET (higher yields and residue C inputs).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2. Changes in soil organic C (SOC) in (a) 0- to 10- and (b) 0- to 20-cm depth over 12 yr as a function of site [potential evapotranspiration (PET)], slope, and cropping frequency in a no-till study in Colorado, USA. Soil organic C in the 0- to 20-cm depth in 1986 for the summit, sideslope, and toeslope positions was 21.9, 21.5, and 29.0 Mg ha–1, respectively, at Sterling; 25.1, 21.8, and 33.9, respectively, at Stratton; and 9.2, 12.6, and 21.4, respectively, at Walsh. (Note: In this and subsequent figures, cropping frequency for fallow–crop, fallow–crop–crop, fallow–crop–crop–crop, and continuous crop = 50, 66, 75, and 100%, respectively.) [From Sherrod et al. (2003) and unpublished data (2004).]
|
|
In the 0- to 20-cm depth SOC gains were generally lowest in the toeslope position even though these positions had the greatest production and C inputs (Sherrod et al., 2003). This suggests that in the toeslope, decomposition effects on C gains were greater than C input effects. Based on the 0- to 20-cm depth, the results suggest that cropping 2 yr in 3, or sometimes even 3 yr in 4, was not sufficient to ensure SOC gains after 12 yr of adopting no-tillage. This result was surprising and, as seen later, was unlike results obtained in western Canada. Generally, there was little difference between low- and medium-PET treatments regarding their effect on change in SOC, but there was a tendency for changes to be lower (more negative) at high PET (Fig. 2). This was partly because annual straw production (and C input) was lower for the high-PET treatment and SOC was directly associated with annual straw production (Sherrod et al., 2003). However, overall, the changes in SOC over 12 yr were small and generally no more than 200 kg ha–1 in 12 yr for continuous cropping systems at low PET (Fig. 2a).
Changes in SOC for the full 0- to 20-cm depth may be partially masked by carbonate C, which was substantially higher in the 10- to 20-cm depth, averaging 1.92, 3.42, and 5.08% of total soil mass at Sterling, Stratton, and Walsh, respectively. We suspect that this factor, together with the much greater variability of SOC in the 10- to 20-cm depth compared with the 0- to 10-cm depth (Sherrod et al., 2003), accounts for the surprising apparent failure of the adoption of no-tillage (after many years of conventional tillage) and more intensive cropping, to increase SOC storage in 0- to 20-cm depth, even after 12 yr. Of the 36 treatments, only 14 showed possible SOC gains in the 0- to 20-cm depth compared with 27 in the 0- to 10-cm depth. Obviously, the large variability in the SOC measurements in the 10- to 20-cm depth was a major contributor to this discrepancy.
Canadian Prairies
In Canada, there are several long-term experiments in which SOC is being measured, but in most of these, scientists either did not measure initial SOC levels or measured only SOC concentrations at the initial sampling (Janzen et al.,1997; Campbell et al., 1997; VandenBygaart et al., 2003). For this exercise, we mainly discuss two experiments conducted at Swift Current, SK, in which we are able to make fairly acceptable estimates of the initial SOC levels. One study, the "Old Rotation Experiment" initiated in 1967, has been much discussed (Campbell et al., 1997, 2000a) in the past, and the second study is the "New Rotation Experiment" initiated in 1987 (Campbell et al., 2000b).
In the Old Rotation Experiment, no SOC determinations were made in 1967; the first SOC measurements were made in 1976. Because the land on which the study is being conducted was managed as fallow–spring wheat (F-W; conventionally tilled with mostly P fertilizer applied) for the previous 60 yr, we assumed that SOC in the F-W rotation in 1976 was representative of the initial SOC levels for this site (Fig. 3a) (Campbell et al., 2000a). This assumption is not infallible, as we have discussed for the New Rotation Experiment (Campbell et al., 2000b) where the initial samples taken in the interrow area between plots in 1987 did not provide accurate estimates of the initial SOC levels of the adjacent plot areas, and thus we had to use a model and crop production data (1987–1990) to estimate the starting SOC value for each treatment (Campbell et al., 2000b).
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 3. Effect of cropping frequency on soil organic C (SOC) trends (1967–1999) (a) in four monoculture wheat rotations and (b) in two continuous cropping rotations in Swift Current Old Rotation study (0- to 15-cm depth). (Adapted from Campbell et al., 2001b.) F-W = fallow–wheat, oat (hay)-W-W = oat (hay)–wheat–wheat, Flx-W-W = flax–wheat–wheat, Cont-W = continuous wheat, and W-Lent = wheat–lentil.
|
|
The difficulties and advantages of monitoring SOC changes over time to estimate the influence of cropping frequency on SOC changes (i.e., small relative change and large spatial variability) are readily seen in results of the Old Rotation Experiment at Swift Current (Fig. 3a). The results suggest a generally constant level of SOC from 1967–1990 in F-W and F-W-W, an initial increase in SOC from 1967–1980 in the more frequently cropped systems and then a "leveling off" from 1980–1990, and all systems tending to increase in SOC during the more humid 1990s (Campbell et al., 2001b). The highly variable behavior of SOC in the continuous wheat (Cont W) system from 1981–1993 was confusing. The more uniform and predictable trends of SOC in the spring wheat–lentil (W-Lent) rotation, which should give results similar to Cont W (Fig. 3b), suggest that Cont W results are more variable and less reasonable in this period. One can easily surmise how variable the conclusions on SOC changes could be, depending on which dates of sampling one might choose to use when making comparisons for the Cont W system. Perhaps a more reasonable picture of the effect of cropping frequency on SOC changes in the Old Rotation Experiment was provided by using the model derived by Campbell et al. (2000a) to estimate the trends in SOC (Fig. 4). The model output showed a small increase in SOC in F-W from 1967 to 1999, the positive influence of cropping frequency on SOC, and showed no evidence of a decrease in SOC in Cont W occurring between 1981 and 1993. This latter is further evidence that the response for Cont W (N + P) between 1981 and 1993 was spurious.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4. Effect of cropping frequency on soil organic C (SOC) trends (1967–1999) in 0- to 15-cm depth in Swift Current Old Rotation Study as estimated by model of Campbell et al. (2000a). Cont-W = continuous wheat, F-W-W = fallow–wheat–wheat, and F-W = fallow–wheat.
|
|
We calculated the 33-yr mean rates of gain in SOC (1967–1999) for all treatments in the Old Rotation Experiment (Fig. 5). The measured rates for F-W, F-W-W, F-W-W-W-W-W, and Cont W were 85, 179, 248, and 291 kg ha–1 yr–1, respectively. Our model (Campbell et al., 2000a) estimates of changes that should have occurred in this study were 24 to 39% lower than those calculated from the measured values and were especially low for F-W and F-W-W, but the trends over time were similar to the measured data (Fig. 5).
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 5. Change in soil organic C (SOC) (1967–1999) in 0- to 15-cm depth in Swift Current Old Rotation study, as measured and as estimated by the model of Campbell et al. (2001b). (All treatments receive N and P.) (Assumed SOC in 1967 = 30.5 Mg ha–1 as measured in F-W, which was rotation on test site during previous 70 yr.) (a) Effect of cropping frequency. (b) Effect of crop type. F-W = fallow–wheat, Cont-W = continuous wheat, F-Ry-W = fallow–rye–wheat, W-Lent = wheat–lentil, and F-Flx-W = fallow–flax–wheat.
|
|
Replacing wheat with other crops may result in greater or lesser amounts of SOC storage (Fig. 5). For example, SOC stored was similar whether lentil or wheat was grown in the rotation. However, the high-lignin, low-yielding flax residue, which was often blown from the plots, diminished SOC storage compared with wheat (Campbell et al., 2000a) while fall-seeded rye, which is an efficient user of N (Campbell et al., 1984), and which protects the soil against erosion, was more effective than spring wheat in storing SOC (Fig. 5).
The rate of SOC storage is highly dependent on weather because of its influence on net primary production (C input) and decomposition (Campbell et al., 2001b). We attempt to demonstrate this point using the results for the Old Rotation Experiment and by estimating the rate of SOC increase relative to F-W (control) during three periods: an initial changeover period (1967–1976), a droughty period (1976–1989), and a period of above-average precipitation (1990–1999) (Table 2). In comparing systems according to cropping frequency, we excluded the fallow–rye–wheat rotation because its fallow period is much shorter than for the spring-seeded cereal systems. We included the 6-yr rotation with the frequently cropped systems because it was continuously cropped during the first 18 yr of the experiment (Zentner and Campbell, 1988).
View this table:
[in this window]
[in a new window]
|
Table 2. Rate of increase of soil organic C (SOC) (0- to 15-cm depth) relative to summer fallow–spring wheat in three periods in Swift Current Old Rotation Experiment.
|
|
Relative to F-W, the fallow-containing treatments (excluding the rotation with fall rye) gained SOC at an average rate of 127 kg ha–1 yr–1 in the changeover period while the well-fertilized, frequently cropped systems (including the 6-yr F-5W rotation) gained SOC, on average, at 613 kg ha–1 yr–1. During the droughty period (Campbell et al., 2001b), average SOC gains (relative to F-W) for the other fallow-containing systems (75 kg ha–1 yr–1) were less than those for the 1967–1976 period as were gains for the frequently cropped systems (255 kg ha–1 yr–1). Then, in the more humid decade of the 1990s, SOC gains in the fallow-containing systems (relative to F-W) continued to decrease (36 kg ha–1 yr–1) while gains for the frequently cropped systems were much higher than for the droughty 1980s but less than for the initial changeover period (470 kg ha–1 yr–1).
The advantages of including the fall-seeded rye crop in the rotation and the negative impact of inadequate fertilization (poor residue production and C inputs) were also demonstrated in the Old Rotation Experiment (Table 2). These results conformed to our understanding of how certain factors influence SOC changes. The fallow-containing systems should now be approaching a steady state because this land was in F-W for 60 or more years before initiation of the study. The change to continuous cropping would initially increase C inputs and reduce SOC decomposition (drier soil for longer periods), thus resulting in marked SOC gains compared with previous F-W. Droughty periods will tend to reduce C gains (reduced C inputs slightly counterbalanced by reduced SOC decomposition due to dry conditions), and wet conditions should increase SOC gains (increased C inputs slightly counterbalanced by increased SOC decomposition due to higher soil moisture).
In the New Rotation Experiment at Swift Current (Campbell et al., 2000b), initiated in 1987 on land that was F-W for decades previously, we did not measure SOC in each test plot at the start. Instead, we measured SOC in the pathways between adjacent plots (to cut down on work and analyses), taking 30 samples (10 per replicate). However, analyses later showed that these pathway initial values did not adequately represent the initial SOC values of each individual plot. We therefore used the model of Campbell et al. (2000a) to estimate the starting SOC values for each plot based on SOC measured in 1990, 1993, and 1996 and residue C inputs for the period 1987 to 1990 (Campbell et al., 2000b). Using these estimated initial SOC values and the SOC measured in 1996, we calculated the SOC changes over the 10-yr period (1987–1996) for F-W-W, F-W-W-W, and Cont W treatments receiving N and P in a minimum-till, snow-managed system (Table 3). The gains in SOC were high in all treatments, possibly because growing season precipitation was above average in this 10-yr period (Campbell et al., 2001b). Thus, for F-W-W, F-W-W-W, and Cont W, the measured gains in SOC were 390, 580, and 670 kg ha–1 yr–1, respectively, almost the same as values estimated with the Campbell et al. (2000a) model (Table 3). These results support the findings in the Old Rotation Experiment showing the strong effect of weather conditions in influencing gains in SOC. They also suggest that net primary production, and thus input C, may play a bigger role in influencing SOC gains on the Canadian prairies than rate of decomposition of SOC, at least over periods of several years or decades.
View this table:
[in this window]
[in a new window]
|
Table 3. Soil organic C (SOC) change in 0- to 15-cm depth in 10 yr (1987–1996) as a function of cropping frequency in New Rotation Experiment at Swift Current.
|
|
|
SUMMARY OF STUDIES IN WHICH EFFECT OF CROPPING FREQUENCY ON SOC CHANGES CAN BE ESTIMATED USING SOME GENERAL ASSUMPTIONS
|
|---|
In many of the studies in which SOC has been assessed, SOC was only measured several years after the experiment was started, often only once after, and in other instances, periodically after. In such cases, the goal was usually to compare effects of various management treatments (e.g., effect of tillage, fertilizers, or manure addition) on SOC. Usually the assumption is made that all treatments at a site started at the same level of SOC though this is often not so. When estimating rates of change in SOC, scientists often divide net change by years (i.e., they assume the change is linear over time). Most researchers suggest that changes in SOC due to changes in management are curvilinear, increasing or decreasing rapidly at first before approaching an asymptote or steady state (Janzen et al., 1997). These inadequacies in assumptions notwithstanding, we have summarized the related studies that have been conducted on the Great Plains of North America using the aforementioned assumptions. We caution that these rates cannot necessarily be extrapolated into the future.
U.S. Great Plains
We only found five studies that were appropriate (i.e., had a summer fallow component, varying cropping frequency, and where SOC was expressed on a mass basis) in the U.S. Great Plains (Table 4). These studies were summarized by location, soil type, rotation, tillage, fertility, and the SOC in the 0- to 20- or 0- to 15-cm depth at the end (last date of sampling) of the study. We expressed SOC on a mass/volume basis (Mg ha–1), and the change in SOC was expressed as percentage change in stocks and as average annual rates (kg ha–1 yr–1), both relative to the SOC in the continuously cropped, well-fertilized no-till treatment. Most of the studies were short term (12 yr or less), no-tilled, and well fertilized with N; consequently, we used the data from the most comprehensive of these studies (Sherrod et al., 2003) to demonstrate the relationship between change in SOC as influenced by cropping frequency, using a "box-and-whiskers" type of plot (Fig. 6).
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6. Annual change in soil organic C (SOC) relative to continuous cropping with no-till on U.S. Great Plains (0- to 20-cm depth) (from Sherrod et al., 2003). (Note: The box plots show the median, quartiles, and extreme values in the data set. The box represents the interquartile range containing 50% of the values. The whiskers are lines that extend from the box to the highest and lowest values excluding outliers. The thick line across the box is the median.)
|
|
The results for this U.S. Great Plains study show the direct relationship between relative rate of change in SOC and cropping frequency. There was considerable variability and range in the SOC changes, and the median values were similar for the fallow–crop–crop and fallow–crop–crop–crop rotations, with their rate of change in SOC being about 125 kg ha–1 yr–1 less compared with continuous cropping, whereas for the fallow–crop rotation, the median rate of change was about 260 kg ha–1 yr–1 less compared with continuous cropping.
Of the other sites (Table 4), no significant responses to cropping frequency were found at Akron although the short time period (5 yr) precludes any definite conclusions. At Mandan, SOC stocks in the continuous crop rotations averaged about 3.7 Mg C ha–1 more than under spring wheat–fallow. There was significant interaction with tillage: SOC increased as tillage intensity decreased under continuous cropping, but there was no significant response to tillage treatments in the wheat–fallow system. For the experiments at Bushland, C stocks tended to be somewhat higher for Cont W compared with the fallow-containing wheat systems, but the trends were not significant.
Canadian Prairies
Many more studies were conducted on the Canadian Prairies than in the USA that include summer fallow and were appropriate for use in assessing the influence of cropping frequency on SOC changes (Table 5). Further, several of these studies were long term (>30 yr), which allowed treatment effects to be more easily differentiated. We summarized the results of these studies in a manner similar to those for the U.S. Great Plains studies, except that in this case, we have used the more frequently included continuously cropped, conventionally tilled, fertilized treatments as the control to derive relative SOC change. Studies varied in duration from 6 to 60 yr and were conducted on soils from the semiarid Brown and Dark Brown Chernozems to the subhumid Black Chernozems and Gray Luvisols (Table 5).
We again used the box-and-whiskers plot to analyze most of the data in Table 5, partitioning them into data for the semiarid soils (Fig. 7a) and those for the subhumid soils (Fig. 7b). In this case, because the control was the tilled continuous cropping treatment, most no-tillage treatments in the Brown and Dark Brown Chernozems showed positive rates of C change relative to the control. In contrast, and not surprisingly, all tilled systems except the control (by definition) showed rates of SOC change that were negative compared with the control. As observed for the Colorado experiment, there was a positive relationship between rate of change in SOC and cropping frequency in both the tilled and no-till treatments in the semiarid soils (Fig. 7a). In the tilled treatments, the median rates of change in SOC relative to the control were between –150 and –100 kg ha–1 yr–1 for the 50 to 85% cropping frequencies while for the no-till systems, values were between +50 and +100 kg ha–1 yr–1 for 50 to 66% cropping frequency and about +250 kg ha–1 yr–1 for continuous cropping. Fewer studies were conducted in the subhumid soils (Fig. 7b), and the trends in SOC response to cropping frequency for tilled treatments were similar to those observed for the Brown and Dark Brown soils (aridic and typic Haploboroll) while for no-till treatments, the main difference, compared with results found for the semiarid soils, was that the presence of summer fallow resulted in decreases, not increases, in SOC relative to the control.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 7. Annual change in soil organic C (SOC) relative to continuous cropping on conventional tillage as control: (a) Brown and Dark Brown Chernozems and (b) Black Chernozems and Gray Luvisols, in the Canadian Prairies.
|
|
In the Canadian prairies, the influence of cropping frequency on the mean rates of change in SOC in the 0- to15-cm depth was also assessed (Fig. 8). As in Colorado, the rate of SOC gains increased with cropping frequency in semiarid and subhumid soils, but the relationship appears to be curvilinear. The rates of increase in SOC were generally similar for fallow–crop, fallow–crop–crop, or fallow–crop–crop–crop but increased sharply at frequencies > 75%. In the semiarid prairie, rates of SOC gain were about 225 to 250 kg ha–1 yr–1 greater for no-till than for tilled systems irrespective of cropping frequency. In the subhumid soils, the difference in SOC gain due to adopting no-tillage was about 50 to 75 kg ha–1 yr–1 favoring no-tillage at cropping frequencies < 75%, but at 100% frequency, the difference was similar to that in semiarid prairie (about 250 kg ha–1 yr–1). The advantage of adopting no-tillage on SOC gains was therefore generally less under subhumid conditions than under semiarid conditions. This tends to support observations in eastern Canada (which has more humid conditions than the prairies) where adoption of no-tillage rarely results in SOC gains (Angers et al., 1997). Perhaps under such humid conditions, the relative influence of additional available water favors increased mineralization of organic matter more so than increased SOC due to greater crop residue C inputs.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 8. Effect of cropping frequency on mean annual rate of change in soil organic C (SOC) relative to well-fertilized continuous cropping, tilled treatments (control) in Canadian Prairie studies (developed from data in Table 5).
|
|
An interesting question concerns the curvilinear relationship between cropping frequency and soil C in the summary figure of data from the Canadian sites (Fig. 8) and to what extent this response is driven by differences in C inputs to soil vs. other factors. Campbell et al. (2004) analyzed data for nine of the experiments listed in Table 5 and found that grain yield (kg ha–1 yr–1) was directly proportional to cropping frequency (%). In the semiarid prairies, or under drought conditions, yield = 556 + 8.02 x cropping frequency (r2 = 0.99), and in the subhumid prairies, yield = 761 + 10.81 x cropping frequency (r2 = 0.99). Since crop residue yield (and thus C inputs) and grain yield tend to be directly proportional (Campbell et al., 1997), increases in cropping frequency represent a proportional increase in C inputs. Theory suggests that (all else being equal) soil C contents should increase linearly as a function of C inputs (Paustian et al., 1997). Hence, the data suggest that while C input rate is a major contributor to SOC response, this is not the only factor concerned. The lack of linearity in this response in some cases suggests that other factors (such as the effects of temperature and water on decomposition and/or disturbance effects on soil C stabilization) may be more important than C inputs in some situations. The data suggest that there is little difference between fallow every other year and fallow once every 3 yr and that a cropping frequency of >75% is needed before soil C responds to the higher rates of C input.
Interaction of Fertilization and Cropping Frequency on SOC
The application of fertilizers to crops may enhance crop production, increase crop residues returned to the soil, and thereby increase SOC (Nyborg et al., 1995; Solberg et al., 1997; Campbell et al., 2001a). The magnitude of the increase in SOC is much larger in soils with inherently low C than in ones where C is already high (Janzen et al., 1997). The influence of fertilization on SOC also depends on the frequency of summer fallow. As shown by findings at Swift Current and Indian Head, SK, even after 24 to 30 yr, fallow-containing rotations showed very little C increase due to fertilization; only Cont W showed a significant increase due to fertilization (Table 5 and Fig. 9).
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 9. Effect of fertilizer N and P and cropping frequency on soil organic C (SOC) in 0- to 15-cm depth, measured in a Black Chernozem at Indian Head, SK, in 1987 after 30 yr of conventional tillage and again in 1997, 7 yr after conversion to no-tillage management and increased rates of N (from Campbell et al., 2001a). F-W = fallow–wheat, F-W-W = fallow–wheat–wheat, and Cont W = continuous wheat.
|
|
Soil scientists suggest that after many years of fertilizing a crop, and perhaps increasing its soil C content, such systems may approach a steady state in C content. However, if the rate of fertilization is changed, and this influences production (C inputs), the SOC gains may also change. Campbell et al. (2001a) tested this theory at Indian Head by resampling, in 1997, plots that had been sampled in 1987 and in which, in 1990, the method of tillage management on these plots was changed from conventional tillage to no-tillage and the fertilizer rates applied were increased in anticipation of greater available soil water than when they were conventionally tilled (Campbell et al., 2001a). This change in management resulted in higher yield responses for the crops seeded on fallow and on stubble land and greater crop residue C returned to the soil for the fertilized systems after 1990. Concomitantly, there was an increase in SOC in both Cont W and F-W-W systems in response to the change in fertilizer–tillage management (Fig. 9). There was a positive effect of cropping frequency on SOC in the fertilized treatments but no effect in the unfertilized treatments. Fertilized F-W, F-W-W, and Cont W gained about 4, 5, and 2 Mg C ha–1 between 1987 and 1997. During this period, C emissions from the manufacture and transport of the N fertilizer was estimated to be 0.28, 0.53, and 0.90 Mg ha–1 for these three rotations, respectively (Campbell et al., 2001a). Thus, there was a net positive effect of frequent cropping on C sequestration but only for adequately fertilized systems. Although we cannot be certain (because both tillage and fertilization regimes were changed at the same time in this study), we suspect that fertilization was the main contributor to the increased crop residue yields between 1987 and 1997 and that the SOC gains observed were directly related to the crop residue responses of the treatments. These results suggest that without adequate fertility, conversion of conventional tillage to no-tillage may not always result in an increase in SOC. Further, as shown in Table 5, in very fertile soils such as the thick Black Chernozem at Melfort, SK, fertilization will not increase SOC substantially, regardless of crop rotation or cropping frequency.
|
ESTIMATING CHANGES IN SOC OVER 15-YEAR PERIOD DUE TO CHANGE IN A COMBINATION OF MANAGEMENT PRACTICES ON THE CANADIAN PRAIRIES
|
|---|
Many studies have been conducted on research plots in which the impact of various cultural practices (tillage, fertilizers, cropping frequency, etc.) on SOC on the Canadian Prairies have been measured (Table 5). Many of these studies are single-factor experiments, but scientists are interested in estimating the net change in SOC due to multifactored changes in cultural practices. This type of exercise can be accomplished by the use of simulation models such as Century (Parton et al., 1987) or EPIC (Williams, 1995). However, McConkey et al. (1999) proposed a simple model based on "expert opinion" and a coefficient-based technique as another approach in achieving this goal. In this approach, one is able to estimate SOC gains as a function of cropping frequency, tillage, fertilization, soil texture and landscape position, or any combination of those factors. This approach was used by the Canadian Fertilizer Institute to estimate the maximum potential of the agricultural sector in Saskatchewan to sequester C (Canadian Fert. Inst., 2002).
Estimating Changes in SOC Using Deterministic Model Such as Century
The Century model (version 4.2) was used to simulate SOC in the 0- to 15-cm depth for three well-fertilized rotations (F-W, F-W-W, and Cont W) from the Old Rotation Experiment at Swift Current, SK, for the period 1967–1998. In the initialization time block, wheat–fallow with a low-yield wheat was run from 1910 to 1966. Initial C pools were calibrated so that SOC in 1967 would match that of the measured (estimated) values at the experimental site at 15-cm depth. Weeds were introduced in the summer fallow years only. The tillage events destroyed and incorporated weeds that developed. Actual fertilizer application values from the experiment were used in the simulations. Monthly weather data from the Swift Current weather station were used. A low-yield wheat was used in the second time block (1967–1977). A medium-yield wheat was used in the third time block (1978–1987). A higher-yield wheat, which reflects current yields, was used in the last time block (1988–present).
The simulation results (Fig. 10) suggest SOC in F-W would remain constant throughout the 33 yr; SOC would increase by 171 g m–2 (=1.7 Mg ha–1) in F-W-W and by 500 g m–2 (=5 Mg ha–1) in Cont W with the increase being generally linear over time. Thus, the rates of gain in SOC would be 0, 53, and 156 kg ha–1 yr–1, respectively. These compare with rates of 55, 124, and 203 kg ha–1 yr–1 estimated by the Campbell et al. (2000a) model (Fig. 5). One apparent shortcoming of the Century model simulation was its failure to identify the differential rate of SOC gain during the 1990s, which was simulated satisfactorily by the Campbell et al. (2000a) model (Fig. 4). The reason for this weakness in Century was not related to its ability to simulate yields since it did this quite acceptably (data not shown). This aspect requires further research. Generally, however, both of these models performed reasonably well in estimating the relative trends in SOC changes and could be used for estimating the effect of cropping frequency on SOC changes.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 10. Effect of cropping frequency on soil organic C in 0- to 15-cm depth in the Old Rotation Experiment at Swift Current, SK (1967–1999)—simulation with Century Model. F-W = fallow–wheat, F-W-W = fallow–wheat–wheat, and Cont W = continuous wheat.
|
|
|
ESTIMATING THE IMPACT OF CHANGE IN FALLOW FREQUENCY ON CARBON GAINS IN THE CANADIAN PRAIRIES
|
|---|
Campbell et al. (2001b) used the data from the Old Rotation Experiment at Swift Current (1967–1999) to relate rate of gain in SOC in the 0- to 15-cm depth to fallow frequency. We converted their relationship to relate SOC change to cropping frequency (cropping frequency = 1 – fallow frequency) as shown in Fig. 11. This revealed a positive relationship but with a decreasing rate of soil C accumulation with increasing cropping frequency for these well-fertilized treatments. Figure 11 shows both measured values and values estimated by the model of Campbell et al. (2000a) for the 33-yr period (1967–1999) in which the last decade experienced above-average growing season precipitation. The measured values were higher than the model estimates, but we have more confidence in the model estimates for reasons previously discussed. The model estimates suggest that in the semiarid prairie, F-W (N + P) will sequester about 50 kg SOC ha–1 yr–1, F-W-W (N + P) about 125 kg ha–1 yr–1, and W-Lent (N + P) about 200 kg ha–1 yr–1. Under more favorable weather conditions (higher moisture), the rates of SOC gains would be higher (Campbell et al., 2001b).
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 11. Effect of cropping frequency on rate of change in soil organic C (SOC) (0- to 15-cm depth) in well-fertilized treatments of the Old Rotation Experiment at Swift Current, SK, during 1967–1999 (adapted from Campbell et al., 2001b).
|
|
In 1976, fallow frequency was 48% (cropping frequency 52%) in the Brown soil zone of the Canadian prairies and 42% (cropping frequency 58%) in the Dark Brown soil zone (Campbell et al., 2002). By 1998, the cropping frequency values had increased to 62 and 80%, respectively, in these two soil zones. Assuming that the rate of SOC change estimated by the Campbell et al. (2000a) model for the 1967–1999 period provides the most realistic estimate of SOC change for this region (Fig. 11), then in 1976, at 52% cropping frequency in the Brown soil zone, the rate of SOC change would be 60 kg ha–1 yr–1, and at 58% cropping frequency in the Dark Brown soil zone, it would be about 100 kg ha–1 yr–1. Similarly, by 1998, these rates would be about 105 kg C ha–1 yr–1 for the Brown soil zone and 175 kg C ha–1 yr–1 for the Dark Brown soil zone.
We can use these estimated rates of change in SOC together with the area of arable land to estimate SOC gained due to changes in cropping frequency over the past 20 yr in the semiarid Canadian prairies. For example, for the period 1976–1998, there was about 6.5 million ha of arable land in the Brown soil zone and 8.5 million ha in the Dark Brown (Campbell et al., 2002). On this basis, the increase in cropping frequency in the semiarid prairies over this 22-yr period may have resulted in 1.75 times as much C being gained in the 0- to 15-cm depth of soil as was the case in 1976 (Table 6). These results emphasize the considerable advantage in SOC gains that may be achieved by crop intensification on the Canadian prairies and this achieved even without the adoption of no-tillage but assuming proper fertilization was conducted. However, there would have been some cost in C used for increased fertilization, as we have previously discussed. We caution that these estimates may be high since most farmers in the semiarid prairies do not fertilize their crop adequately (Campbell et al., 1986).
|
CONCLUSIONS
|
|---|
Many appropriate studies (that included summer fallow treatments) were found in the Canadian prairies, but only a few in the USA Great Plains; thus, most of our conclusions are based on the Canadian experience.
The accuracy of our conclusions are constrained by the many pitfalls encountered when trying to quantify changes in SOC (e.g., SOC changes are small and occur slowly, there is large spatial variability, and there are many inadequacies in sampling and analyzing for SOC). Nonetheless, we were able to make several useful observations from our analysis of the research data that were compiled.
Changes in SOC depend on the degree to which the soil has been degraded: the greater the previous degradation, the greater the likelihood that a change in management will reverse the process. Most of the experiments reviewed were initiated on land in which the soil had already been degraded to some extent by frequent use of summer fallow.
In most soils, SOC increased with cropping frequency; generally, this relationship was not linear. Soil organic C gains in no-till systems were greater than in tilled systems. In the semiarid Canadian prairie, SOC gains under no-till were about 250 kg ha–1 yr–1 greater than for tilled systems; in subhumid environments, the advantage was about 50 kg ha–1 yr–1 for fallow–crop and fallow–crop–crop rotations but 250 kg ha–1 yr–1 for continuously cropped rotations. These gains require several years to accumulate (10–15 yr). Under conventional tillage, the relationship of SOC to cropping frequency was unaffected by soil zone.
The type of crop grown in the rotation sometimes influenced the rate of SOC change. Gains in SOC were similar for W-Lent rotation as for Cont W. Flax is low yielding, and thus when it replaces wheat in rotation, SOC gains are much lower. Fall rye is an efficient user of N and protects the soil against erosion more so than spring wheat; thus, when fall rye replaces wheat in rotation, SOC gains are greater.
Fertilization influences SOC gains. In unfertilized systems, an increase in cropping frequency had little effect on SOC gains. However, in systems fertilized according to soil tests, SOC gains were directly proportional to cropping frequency (except in the high-organic-matter, thick Black Chernozems at Melfort, SK, where there was no measurable effect).
In the semiarid soils in Colorado, USA, a 12-yr no-tillage study on land that had been cropped to fallow–crop under conventional tillage for the previous 75 yr showed a positive effect of cropping frequency on SOC. Highest SOC gains were obtained at low-PET sites and lowest gains in the toeslope positions even though they had highest residue C inputs. The latter suggests that in more humid environments, the negative impact of mineralization on SOC gains may override the positive impact due to increased crop production and C inputs.
Both the Century model and the Campbell et al. (2000a) model, though providing lower values, were effective in simulating the relative effect of management on SOC changes. The Campbell et al. (2000a) model was more effective in simulating system responses to major weather effects.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank R. Riznek, B. Grant, W. Smith, B. Helgason, and L. Cramer for technical assistance.
|
REFERENCES
|
|---|
- Angers, D.A., M.A. Bolinder, M.R. Carter, E.G. Gregorich, C.F. Drury, B.C. Liang, R.P. Voroney, R.R. Simard, R.G. Donald, R.P. Beyaert, and J. Martel. 1997. Impact of tillage practices on organic carbon and nitrogen storage in cool, humid soils of eastern Canada. Soil Tillage Res. 41:191–201.
- Bowman, R.A., M.F. Vigil, D.C. Nielson, and R.L. Anderson. 1999. Soil organic matter changes in intensively cropped dryland systems. Soil Sci. Soc. Am. J. 63:186–191.[Abstract/Free Full Text]
- Bremer, E., H.H. Janzen, and A.M. Johnston. 1994. Sensitivity of total, light fraction and mineralizable organic matter to management practices in a Lethbridge soil. Can. J. Soil Sci. 74:131–138.
- Bremer, E., H.H. Janzen, and R.H. McKenzie. 2002. Short-term impact of fallow frequency and perennial grass on soil organic carbon in a Brown Chernozem in southern Alberta. Can. J. Soil Sci. 82:481–488.
- Campbell, C.A. 1978. Soil organic carbon, nitrogen, and fertility. p. 173–271. In M. Schnitzer and S.U. Khan (ed.) Soil organic matter. Elsevier, Amsterdam, the Netherlands.
- Campbell, C.A., K.E. Bowren, M. Schnitzer, R.P. Zentner, and L. Townley-Smith. 1991. Effect of crop rotations and fertilization on soil organic matter and some biochemical properties of a thick Black Chernozem. Can. J. Soil Sci. 71:377–387.
- Campbell, C.A., S.A. Brandt, V.O. Biederbeck, R.P. Zentner, and M. Schnitzer. 1992. Effect of crop rotations and rotation phase on characteristics of soil organic matter on a Dark Brown Chernozemic soil. Can. J. Soil Sci. 72:403–416.
- Campbell, C.A., R. deJong, and R.P. Zentner. 1984. Effect of cropping, summer fallow and fertilizer nitrogen on nitrate-nitrogen lost by leaching on a Brown Chernozem loam. Can. J. Soil Sci. 64:61–74.
- Campbell, C.A., H.H. Janzen, and N.G. Juma. 1997. Case studies of soil quality in the Canadian Prairies: Long-term field experiments. p. 351–397. In E.G. Gregorich and M.R. Carter (ed.) Soil quality for crop production. Elsevier Science Publ., Amsterdam, the Netherlands.
- Campbell, C.A., B.G. McConkey, R.P. Zentner, F.B. Dyck, F. Selles, and D. Curtin. 1995. Carbon sequestration in a Brown Chernozem as affected by tillage and rotation. Can. J. Soil Sci. 75:449–458.
- Campbell, C.A., B.G. McConkey, R.P. Zentner, F. Selles, and D. Curtin. 1996a. Tillage and crop rotation effects on soil organic matter in a coarse-textured Typic Haploboroll in southwestern Saskatchewan. Soil Tillage Res. 37:3–14.
- Campbell, C.A., B.G. McConkey, R.P. Zentner, F. Selles, and D. Curtin. 1996b. Long-term effects of tillage and crop rotations on soil organic C and N in a clay soil in southwestern Saskatchewan. Can. J. Soil Sci. 76:395–401.
- Campbell, C.A., F. Selles, G.P. Lafond, and R.P. Zentner. 2001a. Adopting zero tillage management: Impact on soil C and N under long-term crop rotations in a thin Black Chernozem. Can. J. Soil Sci. 81:139–148.
- Campbell, C.A., R.P. Zentner, S. Gameda, B. Blomert, and D.D. Wall. 2002. Production of annual crops on the Canadian prairies: Trends during 1976–1998. Can. J. Soil Sci. 82:45–57.
- Campbell, C.A., R.P. Zentner, J.F. Dormaar, and R.P. Voroney. 1986. Land quality trends and wheat production in western Canada—a review. p. 318–353. In A.E. Slinkard and D.B. Fowler (ed.) Proc. Canadian Wheat Prod. Symp., Saskatoon, SK, Canada. 3–5 Mar. 1986. Div. of Ext. and Community Relations, Univ. of Saskatchewan, Saskatoon, SK, Canada.
- Campbell, C.A., R.P. Zentner, B.C. Liang, G. Roloff, E. Gregorich, and B. Blomert. 2000a. Organic C accumulation in soil over 30 years in semiarid southwestern Saskatchewan—effect of crop rotation and fertilizers. Can. J. Soil Sci. 80:179–192.
- Campbell, C.A., R.P. Zentner, F. Selles, V.O. Biederbeck, B.G. McConkey, and P.J. Jefferson. 2000b. Quantifying short-term effects of crop rotations on soil organic carbon in southwestern Saskatchewan. Can. J. Soil Sci. 80:193–202.
- Campbell, C.A., R.P. Zentner, F. Selles, V.O. Biederbeck, B.G. McConkey, R. Lemke, and Y.T. Gan. 2004. Cropping frequency effects on yields of grain, straw, plant N, N balance and annual production of spring wheat in the semiarid prairie. Can. J. Plant Sci. 84:487–501.
- Campbell, C.A., R.P. Zentner, F. Selles, B.C. Liang, and B. Blomert. 2001b. Evaluation of a simple model to describe carbon accumulation in a Brown Chernozem under varying fallow frequency. Can. J. Soil Sci. 81:383–394.
- Canadian Fertilizer Institute. 2002. Potential sequestration of carbon in the agricultural soils of Saskatchewan: Cost-effective early action options to help meet Canada's commitment to the Kyoto protocol. CFI, Ottawa, ON, Canada.
- Ellert, B.H., H.H. Janzen, and B.G. McConkey. 2001. Measuring and comparing soil carbon storage. p. 131–146. In R. Lal et al. (ed.). Assessment methods for soil carbon. Lewis Publ., Boca Raton, FL.
- Gregorich, E.G., K.J. Greer, D.W. Anderson, and B.C. Liang. 1998. Carbon distribution and losses: Erosion and deposition effects. Soil Tillage Res. 47:291–302.
- Halvorson, A.D., B.J. Wienhold, and A.L. Black. 2002. Tillage, nitrogen, and cropping system effects on soil carbon sequestration. Soil Sci. Soc. Am. J. 66:906–912.[Abstract/Free Full Text]
- Izaurralde, R.C., W.B. McGill, J.A. Robertson, N.G. Juma, and J.J. Thurston. 2001. Carbon balance of the Breton Classical Plots over half a century. Soil Sci. Soc. Am. J. 65:431–441.[Abstract/Free Full Text]
- Janzen, H.H. 2001. Soil science on the Canadian prairies—peering into the future from a century ago. Can. J. Soil Sci. 81:489–503.
- Janzen, H.H., C.A. Campbell, E.G. Gregorich, and B.H. Ellert. 1997. Soil carbon dynamics in Canadian agroecosystems. p. 57–80. In R. Lal, J. Kimble, R. Follett, and E.A. Stewart (ed.) Soil processes and the carbon cycle. CRC Press, Boca Raton, FL.
- Janzen, H.H., C.A. Campbell, R.C. Izaurralde, B.H. Ellert, N. Juma, W.B. McGill, and R.P. Zentner. 1998. Management effects on soil C storage on the Canadian prairies. Soil Tillage Res. 47:181–195.
- Larney, F.J., E. Bremer, H.H. Janzen, A.M. Johnston, and C.W. Lindwall. 1997. Changes in total, mineralizable and light fraction soil organic matter with cropping and tillage intensities in semiarid southern Alberta, Canada. Soil Tillage Res. 42:229–240.
- McConkey, B.G., B.C. Liang, and C.A. Campbell. 1999. Estimating gains of soil carbon over 15-yr period due to changes in fallow frequency, tillage system and fertilization practices for the Canadian prairies (an expert opinion). SPARC Misc. Publ. 379M0209. Agric. and Agri-Food Canada, Swift Current, SK.
- McConkey, B.G., B.C. Liang, C.A. Campbell, D. Curtin, A. Moulin, S.A. Brandt, and G.P. Lafond. 2003. Crop rotation and tillage system impact on carbon sequestration in Canadian prairie soils. Soil Tillage Res. 74:81–90.
- Monreal, C., and H.H. Janzen. 1993. Soil organic carbon dynamics after 80 years of cropping a Dark Brown Chernozem. Can. J. Soil Sci. 73:133–136.
- Nelson, D.W., and L.E. Sommers. 1996. Total carbon, organic carbon and organic matter. p. 961–1010. In D.L. Sparks et al. (ed.) Methods of soil analysis. Part 3. 3rd ed. SSSA Book Ser. 5. SSSA and ASA, Madison, WI.
- Nyborg, M., E.D. Solberg, S.S. Malhi, and R.C. Izaurralde. 1995. Fertilizer N, crop residue, and tillage alter soil C and N content in a decade. p. 93–100. In R. Lal, J.M. Kimble, R.F. Follett and B.A. Stewart (ed.) Soil management and greenhouse effect. Adv. Soil Sci. Lewis Publ., CRC Press, Boca Raton, FL.
- Ortega, R.A., G.A. Peterson, and D.G. Westfall. 2002. Residue accumulation and changes in soil organic matter as affected by cropping intensity in no-till dryland agroecosystems. Agron. J. 94:944–954.[Abstract/Free Full Text]
- Parton, W.J., D.S. Schimel, C.V. Cole, and D.S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grassland. Soil Sci. Soc. Am. J. 51:1173–1179.[Abstract/Free Full Text]
- Paustian, K., H.P. Collins, and E.A. Paul. 1997. Management controls on soil carbon. p. 15–49. In E.A. Paul, K. Paustian, E.T. Elliott, and C.V. Cole (ed.) Soil organic matter in temperate agroecosystems: Long-term experiments in North America. CRC Press, Boca Raton, FL.
- Peterson, G.A., A.D. Halvorson, J.L. Havlin, O.R. Jones, D.J. Lyon, and D.L. Tanaka. 1998. Reduced tillage and increasing cropping intensity in the Great Plains cons
TOP
ABSTRACT
INTRODUCTION
