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


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (164)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by West, T. O.
Right arrow Articles by Post, W. M.
Right arrow Search for Related Content
PubMed
Right arrow Articles by West, T. O.
Right arrow Articles by Post, W. M.
Agricola
Right arrow Articles by West, T. O.
Right arrow Articles by Post, W. M.
Related Collections
Right arrow Agricultural Systems
Right arrow Soil Biochemistry
Right arrow Tillage
Soil Science Society of America Journal 66:1930-1946 (2002)
© 2002 Soil Science Society of America

DIVISION S-6—SOIL & WATER MANAGEMENT & CONSERVATION

Soil Organic Carbon Sequestration Rates by Tillage and Crop Rotation

A Global Data Analysis

Tristram O. West* and Wilfred M. Post

Environmental Sciences Division, Oak Ridge National Lab., P.O. Box 2008, Oak Ridge, TN 37831-6335

* Corresponding author (westto{at}ornl.gov)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION AND CONCLUSIONS
 REFERENCES
 
Changes in agricultural management can potentially increase the accumulation rate of soil organic C (SOC), thereby sequestering CO2 from the atmosphere. This study was conducted to quantify potential soil C sequestration rates for different crops in response to decreasing tillage intensity or enhancing rotation complexity, and to estimate the duration of time over which sequestration may occur. Analyses of C sequestration rates were completed using a global database of 67 long-term agricultural experiments, consisting of 276 paired treatments. Results indicate, on average, that a change from conventional tillage (CT) to no-till (NT) can sequester 57 ± 14 g C m-2 yr-1, excluding wheat (Triticum aestivum L.)-fallow systems which may not result in SOC accumulation with a change from CT to NT. Enhancing rotation complexity can sequester an average 20 ± 12 g C m-2 yr-1, excluding a change from continuous corn (Zea mays L.) to corn-soybean (Glycine max L.) which may not result in a significant accumulation of SOC. Carbon sequestration rates, with a change from CT to NT, can be expected to peak in 5 to 10 yr with SOC reaching a new equilibrium in 15 to 20 yr. Following initiation of an enhancement in rotation complexity, SOC may reach a new equilibrium in approximately 40 to 60 yr. Carbon sequestration rates, estimated for a number of individual crops and crop rotations in this study, can be used in spatial modeling analyses to more accurately predict regional, national, and global C sequestration potentials.

Abbreviations: CT, conventional tillage • IPCC, Intergovernmental Panel on Climate Change • NT, no-till • RT, reduced tillage • SOC, soil organic C • SOM, soil organic matter


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION AND CONCLUSIONS
 REFERENCES
 
ORGANIC C in agricultural soils contributes positively to soil fertility, soil tilth, crop production, and overall soil sustainability (Bauer and Black, 1994; Lal et al., 1997; Reeves, 1997). Changes in agricultural management can increase or decrease SOC. Optimizing agricultural management for accumulation of SOC can result in the sequestration of atmospheric CO2, thereby partially mitigating the current increase in atmospheric CO2 (Sampson and Scholes, 2000). In addition to the environmental benefits of soil C sequestration, consideration has also been given to the implementation of a C credit trading system which may provide economic incentives for C sequestration initiatives (Marland et al., 2001a; 2001b).

Changes in agricultural practices for the purpose of increasing SOC must either increase organic matter inputs to the soil, decrease decomposition of soil organic matter (SOM) and oxidation of SOC, or a combination thereof (Follett, 2001; Paustian et al., 2000). These practices include, but are not limited to, reducing tillage intensity, decreasing or ceasing the fallow period, using a winter cover crop, changing from monoculture to rotation cropping, or altering soil inputs to increase primary production (e.g., fertilizers, pesticides, and irrigation). Implementing practices that sequester C can reverse the loss of SOC that may have occurred under intensive cultivation thereby increasing SOC to a new equilibrium (Johnson et al. 1995).

A global analysis of soil C loss following cultivation of forests or grasslands indicated a 20% reduction of the initial SOC, or approximately 1500 g m-2 in the top 30 cm of soil (Mann, 1986). A similar analysis by Davidson and Ackerman (1993) estimated a 30% SOC loss from the entire soil column within 20 yr following cultivation, with the majority of this loss occurring within the first 5 yr.

Loss of SOC can be reversed by ceasing cultivation and returning to the original land cover or other perennial vegetation. Average global C sequestration rates, when changing land use from agriculture to forest or grassland, were estimated to be 33.8 or 33.2 g C m-2 yr-1, respectively (Post and Kwon, 2000). Silver et al. (2000) estimated that reforestation of abandoned tropical agricultural land and pasture sequesters C in the soil at a rate of 130 g C m-2 yr-1 for the first 20 yr, and then at an average rate of 41 g C m-2 yr-1 for the following 80 yr.

Loss of SOC can also be reversed by using less intensive cultivation practices or by changing from monoculture to rotation cropping. In an analysis of 17 experiments (n = 38), Kern and Johnson (1993) concluded that a change from CT to NT sequesters the greatest amount of C in the top 8 cm of soil, a lesser amount in the 8- to 15-cm depth, and no significant amount below 15 cm. They also concluded that, unlike NT, no significant change in SOC was realized in response to reduced tillage (RT). Kern and Johnson (1993) assumed the duration of C sequestration to be between 10 and 20 yr. Paustian et al. (1997) compared 39 paired tillage experiments, ranging in duration from 5 to 20 yr, and estimated that NT resulted in an average soil C increase of 285 g m-2 with respect to CT. Using an average experiment duration of 13 yr implies an approximate C sequestration rate of 22 g m-2 yr-1.

In an analysis of 17 European tillage experiments, Smith et al. (1998) found that the average increase of SOC, with a change from CT to NT, was 0.73 ± 0.39% yr-1, and that SOC may reach a new equilibrium in approximately 50 to 100 yr. Analysis of some long-term experiments in Canada (Dumanski et al., 1998) indicated that SOC can be sequestered for 25 to 30 yr at a rate of 50 to 75 g C m-2 yr-1, depending on soil type, in well fertilized Chernozemic and Luvisolic soils cropped continuously to cereals and hay. Analysis of these Canadian experiments focused on crop rotations, as opposed to tillage, and is unique in that it considered rates of C sequestration with regard to soil type.

The Intergovernmental Panel on Climate Change (IPCC) has developed guidelines for accounting of greenhouse gases, including C sinks in forest and agricultural ecosystems (Houghton et al., 1997). The IPCC suggests using a multiplication factor of 1.1 for a change from CT to NT (Houghton et al., 1997; Land-use change & forestry section), essentially corresponding to a 10% increase in SOC. Moving from CT to RT, factors of 1.05 and 1.0 are recommended for agricultural lands in temperate and tropical climate regimes, respectively. The IPCC suggests these factors be applied to a depth of 30 cm and over a period of 20 yr. Additional factors are provided for residue management, soil inputs (e.g., mulching and manure), and fallow frequency. The IPCC approach has recently been applied in a national inventory of C in agricultural soils (Eve et al., 2001).

In an effort to integrate data from previous regional analyses and improve estimates of agricultural C sequestration rates, we developed a global data set based on a review of long-term experiments in the published literature that recorded the response of SOC to changes in agricultural management. Soil organic C measurements and auxiliary data were specifically compiled to (i) quantitatively estimate the response of SOC to changes in tillage intensity and crop rotation, (ii) quantitatively estimate the duration of C sequestration rates, and (iii) provide confidence intervals for estimates of C sequestration rates that could be used in policy and C cycle modeling analyses. This analysis was intended to provide increased accuracy over past estimates of potential C sequestration by increasing the number of experiments (sample size) and stratifying the analysis by cropping practice (e.g., continuous corn, soybean-sorghum [Sorghum spp.] rotation, etc.).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION AND CONCLUSIONS
 REFERENCES
 
Database Compilation and Organization
Experiments from the published literature that recorded the response of SOC to changes in tillage or crop rotation, and that were greater than 5 yr in duration, were used in this study. A total of 67 global, long-term agricultural experiment sites, consisting of 276 paired treatments, were compiled (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Agricultural experiments used in this study.

 
Measurements for SOC were recorded as mass per unit area (e.g., g m-2 30-cm depth-1). Reported estimates that did not include SOC measurements to a 30-cm depth were not normalized to a depth of 30 cm. In many cases, little or no change in SOC was found to occur between 20 and 30 cm; hence, extrapolating SOC measurements from higher in the soil profile to represent SOC changes in the lower profile would incorrectly inflate C sequestration estimates. Therefore, when experimental results indicated SOC measurements, for example, to a depth of 15 cm, it was assumed that changes in SOC below 15 cm were negligible. Values reported for SOM were converted to SOC through division by 1.72 (Soil Survey Division Staff, 1993).

Data that were presented in terms of SOC percentage were converted to mass per unit area by multiplying the fraction of SOC by respective measurements of soil bulk density (g cm-3) and depth of soil sampled (cm). In experiments where SOC concentration was provided without data on soil bulk density, bulk density was calculated according to equations provided by Chen et al. (1998). Chen et al. (1998) provide regression equations for estimating soil bulk density with respect to tillage practice and soil depth, based on clay- and sand-particle fractions and the percentage of SOM.

A more direct approach for estimating soil bulk density (Adams, 1973) has been used by Post and Kwon (2000) in an analysis of C accumulation in forest and grassland ecosystems. A separate analysis was performed on the data compiled for use in this study, between SOC under CT and SOC under NT, using soil bulk density estimates based on both Chen et al. (1998) and Adams (1973). Analyses included soil samples at varying depths from experiments only where soil bulk density was measured (n = 202), thus allowing a comparison between changes in SOC with both measured and estimated bulk density values. A linear regression analysis (Fig. 1) indicated that calculated changes in SOC based on equations from Chen et al. (1998) had a slightly higher correlation (r2 = 0.87) with changes in SOC using actual soil bulk density measurements than did estimates based on Adams (1973) (r2 = 0.81).



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 1. Comparison between changes in soil organic C (SOC) based on measured soil bulk densities and soil bulk densities estimated using equations from Adams (1973) and Chen et al. (1998). Dashed line indicates 1:1 relationship.

 
Analysis of Experimental Data
Rates of Carbon Sequestration
Carbon sequestration rates were calculated for (i) a decrease in tillage intensity, and (ii) an enhancement of rotation complexity. A decrease in tillage intensity refers to a change in tillage practice that reduces soil disturbance and generally results in increased surface residue. In this analysis, CT included the use of a moldboard plow, RT consisted of practices that used tillage operations other than plowing (e.g., disking), and NT included practices that did not till the soil. Crop systems (e.g., continuous corn, corn–soybean, and wheat–fallow) were grouped separately in the tillage analysis to determine whether a statistically significant amount of C was sequestered within each group in response to a reduction in tillage intensity.

Enhancement of rotation complexity refers to (i) a change from monoculture to continuous rotation cropping, (ii) a change from crop–fallow systems to continuous monoculture or rotation cropping, and (iii) an increase in the number of crops used in a rotation cropping system. In this analysis, continuous cropping is a cropping system without a fallow season, monoculture is a system with only one crop grown, and rotation cropping indicates two or more crops rotated over time on the same unit of land. Most of the monoculture and rotation systems in this analysis are cropped continuously, with the exception of wheat–fallow systems.

Carbon sequestration rates were estimated by calculating the mean difference between the initial and alternative practices, using soil sample data from the latest year available (e.g., comparing C measurements for NT and CT in Year 20, as opposed to comparing NT in Year 20 to CT in Year 1). This method of analysis was intended to reduce the variability in C sequestration estimates caused by deviations in annual precipitation and temperature from the average annual mean, which has been shown to cause significant gain or loss of SOC in any particular year (Campbell et al., 2001b). For example, if changes in mean annual weather conditions cause changes in productivity and C input to the soil, these changes will be observed in both the CT and NT treatments in any given year and will counterbalance each other when the difference between the two treatments is calculated. In using this method, it is possible that a calculated change in SOC could be caused, not only by an increase in SOC under the new management, but by a decrease in SOC over time under the initial management regime. However, for this study we are primarily interested in the overall benefit of changing from conventional management to the new mode of operation, and this method satisfies that purpose.

A paired T-test, using Minitab v. 12.23 (Minitab, 1999), was used on all groups of paired treatments to determine whether the new management scenario was significantly different than the baseline scenario (e.g., NT vs. CT; corn–soybean–cotton [Gossypium hirsutum L.] rotation vs. continuous cotton). A 95% confidence interval of the mean was determined for all estimates of C sequestration, regardless of whether the mean difference in SOC between treatments was statistically significant. Treatments that were replicated using different fertilizer application rates or different crop sequence orders were averaged prior to statistical analyses, so multiple replications from one experiment would not overly influence the results or mask the results of other experiments. The effects of regional climate (i.e., annual temperature and precipitation) on C sequestration rates are not presented here, because preliminary analysis using all treatments indicated no significant correlation between climate variables and C sequestration rates.

Duration of Carbon Sequestration Rates
Soil organic C is expected to reach a new equilibrium at some period following adoption of a new management practice (Johnson et al., 1995). The most accurate method to quantify the mean duration of C sequestration rates would be to calculate the change in SOC, following a new management practice, from those experiments that have a documented land-use history that coincides with the baseline condition or treatment control (e.g., land-use history is CT, as opposed to NT). Most experiments, however, fail to meet this criteria. Instead, we used all experiments that had three or more sampling times documented throughout the experiment duration (n = 42) and estimated the percentage change in the annual rate of SOC sequestration ({Delta}SOCR% yr-1) using the following equation:

[1]
where NT1 and 2 and CT1 and 2 is SOC under NT and CT during the first and second years in which SOC was measured, respectively; SOCR is the estimated annual rate of soil C sequestration; and t1 and t2 are the number of years following initiation of the experiment in which SOC was measured. Equation [1] was repeated for consecutive years in each experiment and results were plotted against respective experiment durations. As the percentage annual difference in SOC decreased with time, it was inferred that SOC was approaching a new equilibrium and C sequestration had either decreased significantly or ceased.

As noted by Huston (2001), the biological response (e.g., change in SOC) to a specific environmental condition (e.g., NT) can be difficult to detect because of the many additional conditions or confounding variables that also affect the biological response. In this study, it is possible that deficiencies in soil macro and micronutrients, fluctuations in mean annual temperature and precipitation, and other common agronomic variables act as limiting factors to C sequestration. This phenomenon is supported by experiments with different wheat rotations in Canada that indicate little or no C sequestration with a change to NT when crops were not adequately fertilized (Campbell et al., 2001a).

In an effort to reduce the effects of confounding variables, and thereby reduce the variability in estimates of C sequestration rates, a quantile regression procedure (Koenker and Bassett, 1978) was used. Quantile regression algorithms have previously been used to estimate the effect of limiting factors within ecological experiments (Cade et al., 1999). In the case of C sequestration rates, a nonlinear quantile regression algorithm will usually be most representative of the nonlinear nature of SOC dynamics over time. While such an algorithm has been developed (Koenker and Park, 1996), it is not yet available in common statistical packages. In following the theory of nonlinear quantile regression analysis, we calculated the 75% quantile of mean annual changes in sequestration rates, and subsequently applied a nonlinear regression algorithm. Algorithms that provided the best fit (highest correlation) with the data were identified using Sigma Plot v. 4 (SPSS, 1997). Both the nonlinear quantile regression equations and the traditional nonlinear regression equations are presented for comparison.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION AND CONCLUSIONS
 REFERENCES
 
Vertical Distribution of Sequestered Carbon
Vertical distribution of C sequestered in the soil profile was analyzed based on a change from CT to NT (Fig. 2) . Measurements of SOC best fit into four categories of sampling depth: 0 to 7, 7 to 15, 15 to 25, and 25 to 35 cm. A significant increase in SOC of 482 ± 87 g m-2 (P <= 0.000, {alpha} = 0.05, n = 59) and 73 ± 57 g m-2 (P <= 0.013, {alpha} = 0.05, n = 55) was found for the 0- to 7- and 7- to 15-cm depths, respectively. The 15- to 25-cm (n = 41) and the 25- to 35-cm depths (n = 19) did not show a significant increase in SOC when changing from CT to NT. Therefore, it is estimated that approximately 85% of the C sequestered in soil, with a change from CT to NT, occurs in the top 7 cm. Regression equations for the 0- to 7-cm depth (Eq. [2]) and 7- to 15-cm depth (Eq. [3]) are:

[2]

[3]
where SOCNT and SOCCT is SOC in grams per square meter under NT and CT, respectively.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2. Soil organic C (SOC) at different soil depths as a result of changing from conventional tillage to no-till. Dashed line indicates 1:1 relationship.

 
Influence of Tillage Intensity on Carbon Sequestration
A comparison of the amount of C sequestered was conducted between CT, RT, and NT to determine whether the three treatments resulted in significantly different amounts of SOC. Only experiments that included measurements for all three tillage treatments were included in the analysis (n = 29). Soil organic C levels under NT were significantly different from SOC levels under CT (P <= 0.002, {alpha} = 0.05) and RT (P <= 0.016, {alpha} = 0.05), while SOC levels under CT and RT were not significantly different from each other (P <= 0.142, {alpha} = 0.05). Therefore, CT and RT treatments were grouped together and referred to collectively as CT in all analyses hereinafter.

A comparison between all CT and NT paired treatments (n = 93) indicated that, on average, a move from CT to NT can sequester 48 ± 13 g C m-2 yr-1 (Table 2). Moving to NT in wheat–fallow rotations showed no significant increase in SOC and therefore may not be a recommended practice for sequestering SOC. Excluding wheat–fallow treatments from the analysis resulted in an average potential C sequestration rate of 57 ± 14 g C m-2 yr-1, when changing from CT to NT (Fig. 3) .


View this table:
[in this window]
[in a new window]
  		Table 2. Soil organic C (SOC) sequestered in response to changing from conventional tillage to no-till.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Comparison of soil organic C (SOC) between conventional tillage and no-till. This analysis includes all tillage experiments except those involving wheat-fallow rotation systems (see text for explanation). Dashed line indicates 1:1 relationship.

 
Mean C sequestration rates, with a change to NT, for rotation cropping systems were significantly greater than for continuous monocultures (P <= 0.087, {alpha} = 0.1) (Table 2). Corn–soybean rotations, with a change from CT to NT, resulted in the highest C sequestration (90 ± 59 g C m-2 yr-1) of all monoculture and rotation cropping systems. Since corn–soybean rotations make up the majority of the rotation corn and rotation soybean categories, analyses of rotation corn and rotation soybean data were also completed with the exclusion of corn–soybean treatments (Table 2).

Influence of Enhanced Rotation Complexity on Carbon Sequestration
Enhancing rotation complexity (i.e., changing from monoculture to continuous rotation cropping, changing crop–fallow to continuous monoculture or rotation cropping, or increasing the number of crops in a rotation system), did not result in sequestering as much SOC (15 ± 11 g C m-2 yr-1) on average as did a change to NT (Table 3). Changing from continuous corn to a corn–soybean rotation did not result in increased C sequestration. Continuous corn generally produces more residue and C input than a corn–soybean rotation system. The decrease in residue C input may be the cause of lower C sequestration rates or possible SOC loss, as indicated by correlations found between SOC and soil residue inputs (Clapp et al., 2000; Duiker and Lal, 1999; Rasmussen et al., 1980). An analysis of all rotation enhancements, not including a change from continuous corn to corn–soybean rotations (Fig. 4) , resulted in a mean C sequestration rate of 20 ± 12 g C m-2 yr-1 (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Soil organic carbon sequestered in response to enhancing crop rotation.{dagger}

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Comparison of soil organic C (SOC) between initial and enhanced rotation cropping systems, which include comparisons between (i) monoculture and continuous rotation, (ii) wheat–fallow and continuous monoculture or rotation cropping, and (iii) rotation with two crops and rotation with three or more crops. This analysis includes all rotation enhancement experiments except those involving a change from continuous corn to corn–soybean rotation (see text for explanation). Dashed line indicates 1:1 relationship.

 
Enhancing rotation complexity, while already using NT, did not result in a significant increase in SOC. It is possible that SOC under NT is closer to a maximum steady-state level than that of CT, and therefore stands to gain less SOC under a rotation enhancement. This explanation is supported by the observation that SOC accumulation in response to changes in agricultural management is likely to be greatest on lands with a history of intensive cultivation, because soils that have lost the most C stand to gain the greatest (see Johnson et al. 1995).

Many of the wheat experiments consisted of decreasing the fallow period (e.g., changing from a wheat–fallow rotation to a wheat–wheat–fallow rotation) or rotating wheat with one or more different crops (e.g., wheat–sunflower [Helianthus annuus L.] or wheat–legume rotations). These practices appeared to be more successful in sequestering C (51 ± 47 g C m-2 yr-1) than moving from a wheat–fallow rotation to continuous wheat (6 ± 8 g C m-2 yr-1). Therefore, while moving from wheat–fallow to continuous wheat may increase C residue inputs, it does not appear to increase SOC as effectively as a continuous cropping system that either rotates wheat with other crops or reduces the fallow period.

Duration of Carbon Sequestration
An analysis of the annual rate of change in SOC in response to a change to NT, reveals that the majority of SOC change occurs within the first 10- to 15-yr following implementation of NT (Fig. 5) . The highest correlation (r2 = 0.12) between {Delta}SOCR% yr-1 and time, using a traditional nonlinear regression analysis, was obtained using the following nonlinear regression equation:

[4]
where y is {Delta}SOCR% yr-1 and x is the number of years the land has been in NT. Applying quantile regression techniques resulted in a higher correlation between {Delta}SOCR% yr-1 and time (r2 = 0.92), using the following nonlinear regression equation on the 75% quantile of mean values:

[5]
where y is {Delta}SOCR% yr-1 and x is the number of years the land has been in NT.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 5. The percentage change in annual soil organic C (SOC) sequestration rates under NT, relative to CT. Solid line represents data (solid circles) using a nonlinear regression equation (see Eq. [4] in text). Dashed line represents the 75% quantile of mean values (open squares) using a nonlinear regression equation (see Eq. [5] in text). A data point at Year 8 and 1236% has been excluded from the graph, for easier visual interpretation, but was included in the analysis.

 
The correlation between {Delta}SOCR% yr-1 and time in response to enhancing rotation complexity (Fig. 6) was less explicit than that for a reduction in tillage intensity (Fig. 5). This was expected since the average C sequestration rates for rotation improvements were lower than those for a reduction in tillage intensity (Tables 2 and 3). The data for rotation enhancements were more variable than those for a reduction in tillage intensity, and there was no indication of a delayed response to sequestering SOC. In terms of nonquantile regression analysis, the following hyperbolic decay equation best fit the data (r2 = 0.02):

[6]
where y is {Delta}SOCR% yr-1 and x is the number of years in which an enhanced rotation system has been in use. The quantile regression technique again resulted in a higher correlation between {Delta}SOCR% yr-1 and time (r2 = 0.30), using the following hyperbolic decay equation on the 75% quantile of mean changes in C sequestration rates:

[7]
where y is {Delta}SOCR% yr-1 and x is the number of years in which an enhanced rotation system has been in use. With a relatively low correlation coefficient, little can be concluded regarding the duration of C sequestration rates for improvements in rotation management. However, it can be speculated that while C sequestration rates are lower for an enhancement in rotation complexity, as compared with a decrease in tillage intensity, the rate of sequestration may continue for a longer period of time (~40–60 yr).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 6. The percentage change in annual soil organic C (SOC) sequestration rates under enhanced rotation, relative to monoculture or rotation with a lesser number of crops (see text for further explanation). Solid line represents data (solid circles) using a hyperbolic decay regression equation (see Eq. [6] in text). Dashed line represents the 75% quantile of mean values (open squares) using a hyperbolic decay regression equation (see Eq. [7] in text).

 

   DISCUSSION AND CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION AND CONCLUSIONS
REFERENCES
 
The results presented here indicate that a change from CT to NT can sequester an average 57 ± 14 g C m-2 yr-1, excluding a change to NT in wheat–fallow systems. This average value is higher than previous estimates of 24 to 40 g C m-2 yr-1 (Lal et al., 1999) and 10 to 50 g C m-2 yr-1 (Lal et al., 1998), and is within the higher range of values (approximately 10–60 g C m-2 yr-1) cited in a recent review by Follett (2001). An enhancement in rotation complexity, with the exception of a change from continuous corn to corn–soybean, can sequester an average 20 ± 12 g C m-2 yr-1. This value is similar to that of 10–30 g C m-2 yr-1, with an average of 20 g C m-2 yr-1, estimated by Lal et al. (1998)(1999) for an improvement in rotation management. The IPCC provides a multiplication factor of 1.10, corresponding to a 10% increase in SOC, to be used in calculations of C sequestration with a change to NT (Houghton et al., 1997). Results from this study indicate that a factor of 1.16 and 1.07 may be more appropriate for a change to NT and an enhancement in rotation complexity, respectively.

Soil C sequestration rates, with a change to NT practices, can be expected to have a delayed response, reach peak sequestration rates in 5 to 10 yr, and decline to near zero in 15 to 20 yr, based on regression analyses (Fig. 5). This agrees with a review by Lal et al. (1998), based on results from Franzluebbers and Arshad (1996), that there may be little to no increase in SOC in the first 2 to 5 yr after a change in management practice, but will be followed by a large increase in the next 5 to 10 yr. Campbell et al. (2001b) concluded that wheat rotation systems in Canada will reach an equilibrium, following a change to NT, after 15 to 20 yr, provided that average weather conditions remain constant. Lal et al. (1998) estimate that rates of C sequestration may continue over a period of 25 to 50 yr. While this estimate does not coincide with our estimates of sequestration duration with a change to NT, the extended C sequestration projected by Lal et al. (1998) may be consistent with projected sequestration rates for an enhancement in rotation complexity (Fig. 6).

While estimated changes in SOC are due to either an increase in C inputs or a decrease in CO2 efflux from the soil, it is not possible from this study, nor is it the intention, to determine which factor is responsible for the change in SOC. The position taken here is that all flows of C to and from the soil are inherently accounted for by the change in SOC (West and Marland, 2002). However, it is noted that SOC can be transported by erosional forces and deposited elsewhere in the watershed. Ignoring displacement and redistribution of SOC by erosion may lead to smaller estimates of C sequestration than actually exist.

Data used in this analysis was stratified separately with regard to a change in tillage or a change in crop rotation. In practice, these changes could occur simultaneously. It can be inferred from our results that if a decrease in tillage and an enhancement in rotation complexity occur simultaneously, the short-term (~15–20 yr) increase in SOC will primarily be caused by the change in tillage and subsequent decrease in the rate of SOC decomposition, while the long-term (~40–60 yr) increase in SOC will be primarily caused by the rotation enhancement and subsequent change in residue input and composition.

When assessing the potential for C sequestration in agricultural soils, it is particularly important to consider the crop rotation being used, in addition to tillage operations and inputs to production. Soil C sequestration rates estimated from this analysis provide increased resolution over previous analyses because of the disaggregation of data by crop type, estimates of sequestration duration, and the inclusion of confidence intervals. Estimates of C sequestration rates and the delineation of rates between cropping systems presented here may have a substantial impact on estimates of potential C storage at regional and global scales.


   ACKNOWLEDGMENTS
 
We thank Con A. Campbell, Warren A. Dick, and Joseph L. Pikul for sharing unpublished results for use in this analysis; Michael Huston and T.J. Blasing for discussions regarding data analysis; Jesse Miller for initial literature review and data compilation; and three anonymous reviewers. Research performed as part of the Consortium for Research on Enhancing Carbon Sequestration in Terrestrial Ecosystems, and sponsored by the U.S. Department of Energy's Office of Science, Biological, and Environmental Research. Oak Ridge National Laboratory is managed by UT-Battelle, LLC, for the U.S. Dept. of Energy under contract DE-AC05-00OR22725.

Received for publication October 16, 2001.


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




This article has been cited by other articles:


Home page
 Agron. J.Home page
M. Mazzoncini, C. Di Bene, A. Coli, D. Antichi, M. Petri, and E. Bonari
Rainfed Wheat and Soybean Productivity in a Long-Term Tillage Experiment in Central Italy
Agron. J., September 8, 2008; 100(5): 1418 - 1429.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
M. Kaiser, R. H. Ellerbrock, and H. H. Gerke
Cation Exchange Capacity and Composition of Soluble Soil Organic Matter Fractions
Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1278 - 1285.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
D. A. Angers and N. S. Eriksen-Hamel
Full-Inversion Tillage and Organic Carbon Distribution in Soil Profiles: A Meta-Analysis
Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1370 - 1374.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
F. Dou, A. L. Wright, and F. M. Hons
Sensitivity of Labile Soil Organic Carbon to Tillage in Wheat-Based Cropping Systems
Soil Sci. Soc. Am. J., August 20, 2008; 72(5): 1445 - 1453.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
Y. Liang, H. T. Gollany, R. W. Rickman, S. L. Albrecht, R. F. Follett, W. W. Wilhelm, J. M. Novak, and C. L. Douglas Jr.
CQESTR Simulation of Management Practice Effects on Long-Term Soil Organic Carbon
Soil Sci. Soc. Am. J., August 20, 2008; 72(5): 1486 - 1492.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
A. D. Halvorson, S. J. Del Grosso, and C. A. Reule
Nitrogen, Tillage, and Crop Rotation Effects on Nitrous Oxide Emissions from Irrigated Cropping Systems
J. Environ. Qual., June 23, 2008; 37(4): 1337 - 1344.
[Abstract] [Full Text]