CQESTR: a model to estimate carbon sequestration in agricultural soils.
ABSTRACT: The PC-based model for predicting tillage and crop rotation effects on organic carbon decomposition and storage in the soil as organic matter (OM) uses crop residue and root biomass production, tillage type and timing, and average temperature from existing c-factor files, created with the Revised Universal Soil Loss Equation (RUSLE). Residue nitrogen content and soil layering information--including layer thickness, organic matter content, and hulk density--complete the input requirements for the model. Short-term trends of surface and remaining buried residue and long-term trends in soil organic matter content, are camputed for individual fields and cropping practices. The model was calibrated with soil organic carbon observations (converted to equivalent OM values) from plots with over 60 years of recorded management history including wheat/fallow with manure additions and wheat/fallow with stubble removal. Computations have a 95% confidence interval of +/- 3.3 g OM k[g.sup.-1] soil (0.33% OM). Com parison of OM calculations from the calibrated model with observed values from Lancaster, Wisconsin, and Lexington, Kentucky, produced mean residual errors ranging from -3.0 to + 0.2 g OM k[g.sup.-1] soil(- 0.3 to + 0.02% OM).
Keywords: Carbon storage, crop residue, greenhouse gasses, organic matter, sail amendments
The ability to predict and plan soil organic carbon (OC) storage and loss is becoming an increasingly important policy and management tool. The term 'organic' is used throughout this manuscript to mean 'of biological (plant or animal) origin' as opposed to describing an agricultural production practice or technique. Soil OC is an excellent indicator of soil quality and the change of organic carbon levels in agricultural soils is an indicator of conservation management. Soil OC is associated with numerous onsite and offsite production and environmental benefits including improved plant growth conditions, improved capacity of soil to retain agricultural chemicals and reduce their offsite movement, and improved soil-water relations that reduce the risks of floods or drought. Soil OC levels in conventionally managed agricultural soils have been depleted by an average of half of their native levels under grass or forest cover (Lal et al. 1998, Lal 1999). As such, there is a large capacity in most intensively tille d cropland soils to increase soil OC.
Conservation tillage practices that keep residues on the soil surface and utilization of cover crops, crop rotations, and organic amendments, usually maintain or increase the soil organic matter (OM) reservoir. Soil organic matter, or OM, and soil organic carbon, or OC, both refer to the same carbon in the soil. The difference between the two terms is that soil OC refers to carbon only. A kilogram (or pound) of soil OC is a kilogram (or pound) of elemental carbon. The term OM refers to the decomposed or composted organic substance that began as a crop residue or biomass added to the soil. A kilogram (or pound) of OM includes nor only the elemental carbon but also the oxygen, nitrogen, phosphorous, sulfur, and other elements that make up the formerly living substance that now resides in the soil. To convert from OM to OC, it is generally accepted that 1.70-1.72 kg of OM will contain 1 kg of OC.
Changes in soil OC could be a performance-based indicator for agricultural subsidies currently being considered for the next farm bill. An interest in soil OC storage is also emerging in international negotiations to reduce net greenhouse gas emissions. Carbon (C) emission trading in a free market is believed by many to be a cost-effective method to achieve emissions reduction goals. This "least cost scenario would also allow C emitters to offset their emissions by "buying" soil OC credits, created by soil C sequestration. Emitters will then have time to develop more efficient technology to reduce their emissions. In the meantime, soil OC can be restored, and conservation management practices underpinning sustainable agriculture can be established in agricultural systems. The injection of resources from the industrial sector into agriculture to sequester C, inexpensively relative to reducing emissions from many industrial sources with currently available technology, will allow agriculture to afford the risks of transition and improve the technology available for conservation systems that restore soil OC.
Restoring soil OC will improve soil quality, help achieve conservation goals, help agriculture adapt to climate variability and risk, help offset C emissions from other sources, and help reduce net greenhouse gas emissions. In order for soil OC storage to receive acceptance and credit in international or domestic policies or programs, including the market-based sale of C credits, quantitative methods to plan, predict, and measure soil OC changes must be established. Efforts to develop soil OC inventories are underway in the United States and in other nations (NRI; IGBP).
In Europe, the potential for C sequestration in soils has been estimated using several mathematical models and data from the Soil Organic Matter Network (SOMNET) (Smith et al. 1997) task 3.3.1 of the Global Change and Terrestrial Ecosystems (GCTE), a core project of the International Geosphere-Biosphere Programme (IGBP). Relationships between management practices and yearly changes in soil OC were developed and used to estimate changes in the total C in European soils.
In the United States, a research model named "Century," which includes a detailed soil carbon component, is currently being applied at the regional scale with the intention to apply this methodology to all regions of the United States, eventually scaling up to national inventory estimates (Kelly et al. 1997). A quantitative field level soil OC sequestration planning and prediction tool is also needed (Berc 1999). This tool must be sensitive to local soils, climate, crop and tillage management systems, crop rotations, fertilization, cover crops, and organic amendments. It must operate at a field level, utilizing readily accessible data sets, to effectively, efficiently, and reliably assist farm planning efforts to enhance soil OC sequestration. It is our goal to provide such a model-based tool for use in the United States, and in the international community. The model is named CQESTR. The name is a liberal phonetic condensation of the word sequester.
Methods and Materials
The C sequestration model, CQESTR, is based on the balance of organic C added to a soil and lost to microbial oxidation. Added organic C is any plant or animal material, including roots of crops, added to or left in a field. The loss of C, according to CQESTR, includes decomposition by microrganisms of all residues including the antecedent soil OM. Decomposition algorithms were taken from the existing residue decomposition model named 'D3R' (Douglas and Rickman 1992). The timing and amount of organic residue, including: roots added, location of residue as determined by timing and type of tillage, and local long term average temperatures are obtained from existing 'c-factor' data files created with the Revised Universal Soil Loss Equation (RUSLE) (Renard et al. 1997). These existing files are read by CQESTR to obtain required input data. Neither mechanical loss or deposition of OM or residues by such forces as wind or water erosion nor removal of residues by soil fauna are incorporated into CQESTR.
The model CQESTR computes residue decay, as does 'D3R', as a declining exponential function of accumulated heat, cumulative degree days (CDD), (computed as the sum of daily average air temperature in [degrees]C, using a 0[degrees]C base where all average temperatures < 0[degrees]C are replaced by 0), initial nitrogen content of the decomposing residue, an index of soil water content that depends on surface location or burial of residues, a rapid decay index, and a biomass type index. Only the addition of biomass type, which includes OM, manure or composted materials, roots and crop residues, as different decomposing biomasses differentiate the decomposition computations in D3R and CQESTR. The model 'D3R' has accurately predicted residue decomposition from data sets for a variety of crops (wheat, barley, corn, soybeans, peas, canola, red clover) from Alaska (Cochran 1991), Washington (Stott et al. 1990), Oregon (Douglas et al. 1980), Idaho (Smith and Peckenpaugh 1986), Missouri (Broder and Wagner 1988), Indian a (Stott et al. 1990), North Carolina (Buchanan and King 1993), Georgia (Thomaston 1984), Texas (Stott et al. 1990), Colorado (Ma et al. 1999), Canada (Moulon and Beckie 1993, 1994; Curtin et al. 1998), and Uppsula, Sweden (Berg et al. 1987). The decomposition computations with 'D3R' for all of these different residues and locations use the same decomposition coefficients. Management, temperature, and residue properties unique to each site provide criteria for selecting appropriate index values for either model.
The decomposition equation used for CQESTR contains only one more term (fB, a biomass or residue type factor) than D3R. Both models compute residue decomposition according to the following procedure. The terms in the equation are: a "fundamental" decomposition rate constant (k), which is independent of residue type or location; a residue nitrogen content factor (fN), which provides different decomposition rates for nitrogen rich (i.e., legume) and nitrogen poor (i.e., cereal) residues; a water availability factor that is determined by the location of the residue either buried or laying on the soil surface (fW); and a biomass type term (fB) (Table 1). Residue remaining (Rr in units of weight of residue per unit area) is computed from an initial residue amount (Ir, same units as Rr) and the amount of heat accumulated (CDD in units of [degrees]C with a base temperature of 0[degrees]C by equation 1
Rr = Ir*exp(k*fN*fW*fB*CDD) (1)
where exp is the exponential function. Soil texture is not currently a factor in the model.
Equation 2 contains the components of the total soil C budget as used in CQESTR. The weight of all organically based C containing compounds ([C.sub.OM]) is computed, using units of weight of crop residue or organic biomass per unit area (not just the weight of elemental carbon, which is referred to by OC).
[C.sub.OM] = (OM - [D.sub.OM]) + (R - [D.sub.R]) + (A - [D.sub.A]) (2)
The weight of OM is the amount present in the soil at some starting time plus that added periodically from sufficiently composted or decomposed crop residue (R) or organic amendments (A). Carbon loss is from the daily decomposition of OM ([D.sub.OM]), decomposition of R ([D.sub.R]), and decomposition of A ([D.sub.A]). Net loss or gain of OM is determined by the cumulative daily loss ([D.sub.OM]) and the periodic (in the model) contributions of (R - [D.sub.R]) and (A - [D.sub.A]). The amount of each residue or amendment still remaining after it has decomposed for approximately four years is added to the existing soil OM. The term [C.sub.OM] is a dynamic value, varying with the month to month additions of residue and decomposition losses. The soil OM is a relatively static variable. The daily amount of decomposition, [D.sub.OM] ,is small, but omnipresent, and both R - [D.sub.R] and A - [D.sub.A] are small after the four year composting time used by CQESTR.
CQESTR tracks each crop biomass (residue) addition to a soil separately. The "decomposition" of these 'fresh' residues, discussed here, refers to the computed change in residue mass predicted by CQESTR. For the first 1000 degree days for each biomass, the decomposition rate is computed using the initial nitrogen content of the biomass to determine a value for fN, location of the biomass in the soil (either buried or on the soil surface) to determine a value for fW, and biomass type for a value for fB. After 1000 degree days of decomposition, the nitrogen factor (fN) is reduced to its minimum value (fN0) for all future decomposition of that biomass regardless of its initial N content. The D3R model provided the proving ground for this process to allow for differing decomposition rates of residues with different initial nitrogen contents. The value for the water factor (fW) is determined by the location of the biomass, either buried or on the soil surface as controlled by the timing and type of tillage, and the presence or absence of sufficient water for maximum decomposition.
In the arid US west and southwest there are extended periods as a crop is growing and following harvest that decomposition is slowed by a lack of water. Decomposition continues at a maximum rate in these arid conditions only after the arrival of fall rains, usually in mid October. In CQESTR an average calendar date may be set to control the annual beginning of decomposition when projecting future effects of management changes. In the U.S. central plains, southern, and eastern states, with summer rainfall patterns, fW will be the larger of its possible values for the entire summer. Each tillage operation conducted during a crop rotation buries additional amounts of each biomass. As a part of tracking for each biomass addition, the fraction remaining on the surface and the fraction buried is cataloged separately. The buried and surface amounts of each biomass are decomposed using the rate appropriate for their location. Existing soil OM is also tracked as a separate biomass entity for each soil depth that is be ing considered.
After a residue has decomposed for an extended time, it becomes incorporated into the soil and loses its identity as an added residue. After that time, the residue is classified as OM and is decomposed with the fB value for OM, not the fB value of its original residue type. The time when this transition occurs was found to be between 3 1/2-4 yr for wheat straw. We have little data for other crop residues to determine whether this transition time should be residue type dependent, but preliminary validations indicate no change is required for non-wheat residues. The position of the biomass (surface or buried) did not change the time for conversion from residue to OM for wheat straw. Values for the residueto-OM transition time and the biomass type factors were determined by interactive manual fitting of the model to OM contents from the ON, 90 kgN [ha.sup.-1], and added manure treatments of the long term crop residue experiment at the Pendleton Research Center, as shown in Figures 1 and 2 (Rasmussen et al. 1989) . Nitrogen content of the strawy manure added to the plots was estimated to be 0.75%. Wheat straw nitrogen content of 0.40% was used. The crop residue experiment contained, among other treatments, wheat/fallow rotations, including no added N and burning of crop residues, chemical fertilization at 90 kgN [ha.sup.-1] and addition of manure as the N source under conventional moldboard plow and clean fallow tillage. Clean fallow tillage refers to a primary inversion tillage that leaves less than 15% residue cover on the soil surface followed by multiple mixing tillages (such as disc, harrow, or rod weeder) throughout the fallow period.
Information required by the model includes: crop biomass, including roots, applied to or remaining in a field after harvest, dates of all residue or amendment additions and tillage operations, fraction of pre-tillage residue weight remaining on the soil surface after each tillage, depth of tillage, nitrogen content of residues at decomposition initiation, average daily air temperature expected throughout the time of interest, number and thickness of soil layers, organic matter content and bulk density of each layer, and an approximate date for the first significant rain after harvest to initiate decomposition. Most of this information is automatically extracted from the c-factor, crop, and operation files that are created for the Revised Universal Soil Loss Equation (RUSLE) (Renard et al. 1997). One need only provide the RUSLE c-factor (*.rus), croplist (croplist.dat) , and oplist (oplist.dat) files that describe rotations and management practices of interest for the field sites of concern. CQESTR will automa tically extract the data it needs when the file names and directory locations are provided. The availability of the monthly mean temperatures from the c-factor files eliminates the need to search for independent sources of temperature data. Daily average air temperature is estimated by fitting the annual temperature trend provided by the average monthly temperatures. Daily values are estimated using the mean temperature as the estimate for the 15th of the month. A temperature for each day following or preceding the 15th was estimated using the weighted average of the rate of temperature change between preceding and following months. Weighting was based on the number of days from the date to the preceding or following month.
Only residue nitrogen content with the thickness, starting GM content, and bulk density of soil layers of interest are additional required inputs to the CQESTR model. Up to nine soil layers of any depth may be tracked with CQESTR. Normally, the number and depth of layers will be determined by the availability of soil OM measurements for comparison. Tillage depths, provided in RUSLE oplist files, are used to incorporate surface residues into appropriate depths. Complete mixing is assumed. Currently bulk density changes are not included in CQESTR computations. A small database of possible ranges of nitrogen contents for some common crops is provided as a part of the model. An internet source of nutrition information for specialty crops is provided by Speedy and Waltham (1998). It lists crude protein content of hundreds of plant species produced under a wide variety of growing conditions. Nitrogen content for these plants may be estimated by assuming that protein is 16% N (Lehninger 1970).
Soil layering and OM content must be provided by the operator. The soil series present, their natural horizon depths, and the expected ranges of OM content by horizon are available in the county by county national soil survey (USDA NRCS 1997), or may be assessed by field sample analysis. Specific use-dependent OM data is preferable for individual field projections. General OM contents are usable for trend projections. Local NRCS offices may provide more recent references to current local OM data as they become available.
Results and Discussion
The model was calibrated to the OM observations from three treatments, (90 kg [ha.sup.-1] N, 0 kg [ha.sup.-1] N with stubble burned, and 22 Mg [ha.sup.-1] manure added biennially) of the long term residue management plots from the Pendleton Agricultural Research Center (Rasmussen et al. 1989)(Figures 1 and 2). Using the residuals from the 167 computed and observed pairs of OM values the mean squared residual error for the model was 2.8. The mean residual error was 0.026 g OM [kg.sup.-1] soil (0.0026% OM). Based on the mean squared residual error, the model estimates have a 95% confidence interval of +/- 3.3 g OM [kg.sup.-1] soil (0.33% OM).
The crop rotation for each treatment, including year to year variation in yields and treatment-specific tillage, was used to create RUSLE c-factor files that described the 70 yr of each treatment. Values for the biomass type factor were adjusted to provide the best fits for soil OM for all three treatments at two soil depths (Figures 1 and 2). The final values for the biomass type factor, fB, for biomass types of OM, manure, roots, and crop residue were 0.017, 0.60. 0.40, and 1.00, respectively for the best fit calibration of CQESTR (Table 1). These calibrated values were then used for all validation computations by the model. Only location long term temperature, added crop residue amounts, tillage practices and timing, residue nitrogen content, and initial soil properties (layers, %OM, and bulk density) were provided to CQESTR for computing OM at other locations. A beginning soil OM value usually has to be adjusted to provide a match to the earliest possible OM observation at a site. Once the appropriate beg inning OM value for each site is determined, local rotation and tillage values allow CQESTR to estimate OM trends for specific practices.
Calibration was based on the decomposition of wheat residues exposed to the arid summer environment of the Pacific Northwest. To evaluate the calibrated model at other sites, information on treatment management practices, yields and soil OM were obtained from Lexington Kentucky, and Lancaster Wisconsin. The Lexington data were obtained from Paul et al. (1997). Part of the Lancaster data were also from Paul et al. (1997) but soil samples from the low and high fertility treatments of the continuous corn rotation were collected during August of 1998 and analyzed for organic C at the Pendleton research center. The Lancaster data and the model prediction of OM for the 0-15 and 15-30 cm layers are shown in Figure 3.
The model was provided with the Lancaster, Wisconsin starting OM content for 1966. Computations for the trend in OM after that time are dependent only on the cropping practices and temperature of the location as described in the c-factor files that were created for the site. Corn residue was assumed to contain 1.0% N. The water factor was set to "wet" for all decomposition in both Wisconsin and Kentucky. The mean residual error for the Lancaster data was - 1.4 g OM [kg.sup.-1] soil (- 0.14% OM), indicating an average underestimation of observed values. The mean squared residual error was 4.5.
Soil OM observations, temperature records, and production and cultivation practices from Lexington, KY were available for 20 yr of continuous corn production with four fertility levels using both conventional and no-tillage. Unfortunately, no pretreatment OM observations were available for the data set. A beginning soil OM content was chosen that caused CQESTR to match the observed OM content of the low fertility no-till treatment. The same starting soil OM content was used for all other computations for the Lexington data. The data point from the low fertility no-till treatment was not used in the computation of validation error for Lexington. Figures 4 and 5 show the observed and computed OM trends for 20 yr of conventional and no-till management, respectively,. The model predicted no difference among the fertility treatments of the conventionally tilled plots, as the same initial nitrogen content was used for the corn residue of all treatments. The nitrogen content of the residues, which must have differed between the 0-336 kg N [ha.sup.-1] treatments, was not available. This contributed to a mean residual error of - 3.0 g OM [kg.sup.-1] (- 0.3 g %OM) and a mean squared residual error of 13.5. Where a description of the tillage differences among no-till treatments was available for input, the model predicted a range of OM contents similar to those observed. The mean residual error for the no-till treatments was 0.2 g OM [kg.sup.-1] (0.02 %OM) with a mean squared residual error of only 3.1. The apparent lack of sensitivity to observed differences in OM in the surface soil layer among the conventional treatments and the second layer in all treatments requires further investigation.
Two possible contributions to lack of matching of reported soil OM contents in continuous corn fertility treatments at Lexington include the presence of an unknown contribution of biomass by a winter rye cover crop in all rotations and a failure of the current version of the program to incorporate any buried residue into the second layer of soil even though the first layer may be thinner than the depth of tillage. Residue incorporation to the proper depth by tillage will be corrected before the program is released. Comparisons will continue with other locations where more complete data sets with management history, residue nitrogen content, and initial soil OM observations can be obtained. Note that there were no changes made to any of the model decomposition or OM conversion parameters for application to the Lancaster or the Lexington sites. Only the site specific initial soil OM contents with crop production, tillage, and weather histories are provided to the model in order to compute the trends in soil OM content. Selectable layer depths with tillage dependent depth of residue incorporation are being incorporated into the model.
Several environmental factors that can influence OM content are not included in the CQESTR model. Loss of surface soil to erosion by wind and water is not considered. For water erosion the RUSLE model, which was developed specifically for estimating soil loss, should provide estimates of soil loss from the same rotations and management options that are described by the c-factor files used in CQESTR. These soil losses could be combined with the predictions from CQESTER to provide improved estimates of trends in actual field OM content.
Worms, insects, and small mammals may consume a fraction of residues. This non microbial consumption depends strongly on local fauna population, climate, type and amount of surface residue, and time residue has been on the soil surface. CQESTR does not attempt to predict this consumption. If non microbial consumption is significant, independent estimates of it must be subtracted from the residue mass input to CQESTR for improved OM predictions.
The information generated from this model could be used to assist producers in their effort to sequester C in their soils, and could be incorporated into geographical information systems for use by producers or government agencies. It could be used as a basis for private contracts or public sector agreements to increase OM at the field level. If proven applicable nationally, it could also be added to national statistical resource inventory protocols to track regional and national soil C stocks. Such a tool could help policy and program development for C sequestration, as soil loss equations have been used to develop and evaluate erosion control policy objectives. Software is currently being developed and tested for the application of CQESTR on personal computers that could be used by farmers, farm consultants, or government agency staff.
Summary and Conclusion
The CQESTR model can serve as the basis for a field-level soil organic matter (OM) planning and prediction tool. It may also have applications for statistically based national inventories. Preliminary tests show it predicts C levels within 5% of observed values on the average with a negative bias of 1.0 g OM [kg.sup.-1] soil (0.10 %OM). It runs on available data and can be easily adapted for use at other locations inside and outside of the United States.
The model CQESTR provides soil OM and residue estimates for various times during a rotation or after many cycles of rotation. The trend of C storage, or loss, is provided when several rotation cycles are computed on a time scale of decades. Short-term patterns of crop biomass decomposition and surface cover can be estimated at intervals of just a few weeks. This short-term capability is based on the inclusion of features from the residue decomposition model, D3R (Douglas and Rickman 1992). CQESTR describes the effects of climate, crop rotation, and tillage management practice on the C sequestration of cropland soils. The model uses existing RUSLE databases for the majority of its input. Required nitrogen content of crop biomass additions are available from a provided data file or from a Food and Agriculture Organization (FAO) internet page. The soil layer and starting OM values are available from national county soil surveys or field sampling and testing.
The authors want to thank Dale Wilkins for collecting the Lancaster, Wisconsin, soil samples.
Berc, J. 1999. NRCS roles and needs. Paper presented at the U.S. Department of Agriculture, Agricultural Research Service workshop on soil carbon, January 12, 1999 at Baltimore Maryland.
Berg, B., M. Mullerm, and B. Wessen. 1987. Decomposition of red clover (Trifolium pratense) roots. Soil Biology Biochemistry 19:589-593.
Broder, M.W. and G.H. Wagner. 1988. Microbial colonization and decomposition of corn, wheat and soybean residue. Soil Science Society of America Journal 52:112-117.
Buchanan, M. and L.D. King. 1993. Carbon and phosphorus losses from decomposing crop residues in no-till and conventional till agroecosystems. Agronomy Journal 85:631-638.
Cochran, V.L. 1991. Decomposition of barley straw in a subartic soil in the field. Biology and Fertiliry of Soils 10:227-232.
Curtin, D., F. Selles, H. Wang, CA. Campbell, and V.0. Biederbeck. 1998. Carbon dioxide emissions and transformations of soil carbon and nitrogen during wheat straw decomposition. Soil Science Society of America Journal 62:1035-1041.
Douglas, C.L. Jr., R.R. Ailmaras, P.E. Rasmussen, RE. Ramig, and N.C. Roager, Jr. 1980. Wheat straw composition and placement effects on decomposition in dryland agriculture of the Pacific Northwest. Soil Science Society of America Journal 44:833-837.
Douglas, C.L., Jr. and RW. Rickman. 1992. Estimating crop residue decomposition from air temperature, initial nitrogen content, and residue placement. Soil Science Society of America Journal 56:272-278.
Kelly, RH., W.J. Parton, G.J. Crocker, P.R. Grace, J. Klir, M. Korsehens, P.R. Poulton, D.D. Richter. 1997. Simulating trends in soil organic carbon in long term experiments using the century model. Geoderma 81:75-90.
Lal, R., J.M. Kimble, R.F. Follett, and C.V. Cole. 1998. The greenhouse process. Chap. 2. Pp 3-13. In: The potential of U.S. cropland to sequester carbon and mitigate the greenhouse effect. Chelsea: Ann Arbor Press.
Lal, R., Dec/Jan 1999. Conservation tillage for mitigating greenhouse effect, National Conservation Tillage Digest: 6 (6):18-21.
Lehninger, A.L. 1970. Biochemistry. NewYork: Worth Publishers.
Ma, L., G.A. Peterson, L.R Ahuja, L. Sherod, M.J. Shaffer, and K.W. Rojas. Decomposition of surface crop residues in long term studies of dryland agroecosystems. Agronomy 91:401-409.
Moulin, A.P. and H.J. Beckie. 1993 Predicting crop residue decomposition. Pp 104-109 in the proceedings of the Saskatchewan Soils and Crops Workshop: Crop Quality held February 25-26, 1993 at the University of Saskatchewan.
National Resources Inventory (NRI). 1997 Survey. Created for the U.S. Department of Agriculture, Natural Resources Conservation Service (NRCS). Available from http://www.nhq.nrcs.usda.gov/NRI.
Paul, E.A., K. Paustian, E.T. Elliot, and C.V. Cole, 1997. Soil organic matter in temperate agroecosystems: Long term experiments in North America. New York: CRC Press, LLC.
Rasmussen, P.E., H.P. Collins, and R.W. Smiley. 1989. Long term management effects on soil productivity and crop yield in semiarid regions of Eastern Oregon. Pendleron: U.S. Department of Agriculture, Agricultural Research Service. Oregon Agricultural Experiment Station Bull. No. 675.
Renard, K.G., G.R Foster, G.A. Weesies, D.K. McCool, and D.C. Yoder. 1997. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). Washington, D.C.: U.S. Department of Agriculture. Agriculture Handbook No. 703.
Smith, P., J.U. Smith, D.S. Powlson, W.B. McGill, J.R.M. Arah, O.G. Chertov, K. Coleman, U. Franko, S. Frolking, D.S. Jenkinson, L.S. Jensen, RH. Kelly, H. Klein-Gunnewiek, AS. Komarov, C. Li, J.A.E. Molina, T. Mueller, W.J. Parton, J.H.M. Thornley, A.P. Whirmore. 1997. A comparison of the performance of nine soil organic matter models using datasets from seven long term experiments. In: P. Smith, D. S. Powlson, J. U. Smith, E. T. Elliott. (eds). Evaluation and comparison of soil organic matter models. Special issue: Geoderma 81 (1/2):153-225.
Smith, J.H. and R.E. Peckenpaugh. 1986. Straw decomposition in irrigated soil: Comparison of twenty-three cereal straws. Soil Science Society of America Journal 5 0:928-932.
Speedy A. and N. Waltham. 1998. Database from tropical feeds. B. Gohl (ed). Food and Agriculture Organization of the UN (FAO) publication. 1998. Available from www.fao.org/livestock/agap/frg/tropfeed.htm.
Steffen, W.L., Noble, I., Canadell, J., Apps, M., Schulze, E.-D., Jarvis, P.G., Baldocchi, D., Clais, P., Cramer, W., Ehleringer, J., Farquhar, G., Field, C., Lambin, E., Linder, S., Mooney, H.A,, Murdiyarso, D., Post, W.M., Prentice, I.C., Raupach, M.R., Schimel, D.S., Shvidenko, A. and Valentini, R. (International Terrestrial Carbon Working Group- IGBP) (Eds). 1998. The terrestrial carbon cycle: implications for the Kyoto proecol. Science 280:1393-1394.
Stott, D.E., H.F. Stroo, L.F. Elliott, R.I. Papendick, and P.W. Unger. 1990. Wheat residue loss from fields under no-till management. Soil Science Society of America Journal 54:92-98.
Thomaston, SW. 1984. Crop residue decomposition as affected by soil erosion and tillage. Master's thesis, University of Georgia.
U.S. Department of Agriculture, Natural Resources Conservation Service, Soil Survey Division. 1997. National MUIR database. Available from www.statlab.iastate.edu/soils/muir.
[Figure 1 omitted]
[Figure 2 omitted]
[Figure 3 omitted]
[Figure 4 omitted]
[Figure 5 omitted]
Table 1 Values of parameters in the decomposition equation used in CQESTR. Factor Description and Varaiable Name Value Grouping Fundamental rate constant k -.0004 Nitrogen 0 rate (fN) fN0 0.8354 Nitrogen 1 rate (fN) fN1 1.2635 Nitrogen 2 rate (fN) fN2 1.9770 Nitrogen 3 rate (fN) fN3 3.4040 Nitrogen 4 rate (fN) fN4 3.4040 Surface dry (fW) sc 0.21 Surface wet (fW) sf 0.32 Buried dry (fW) bc 0.80 Buried wet (fW) bf 1.00 Crop residue (fB) frf 1.00 Roots (fB) rf 0.40 Composted or digested (fB) cdf 0.60 Soil organic matter (fB) omf 0.017
|Printer friendly Cite/link Email Feedback|
|Author:||Rickman, R.W.; Douglas, C.L. Jr.; Albrecht, S.L.; Bundy, L.G.; Berc, J.L.|
|Publication:||Journal of Soil and Water Conservation|
|Article Type:||Statistical Data Included|
|Date:||Jun 22, 2001|
|Previous Article:||Stocking rate effect on soil carbon and nitrogen in degraded soils.|
|Next Article:||A phosphorus budget for Wisconsin cropland.|