HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (23) Citing Articles via Google Scholar Google Scholar Articles by Izaurralde, R.C. Articles by Thurston, J.J. Search for Related Content PubMed Articles by Izaurralde, R.C. Articles by Thurston, J.J. GeoRef GeoRef Citation Agricola Articles by Izaurralde, R.C. Articles by Thurston, J.J. Related Collections Other Crop Management Biomass Accretion Other Cropping Systems

DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION

Carbon Balance of the Breton Classical Plots over Half a Century

^a Pacific Northwest National Laboratories, 901 D Street S.W., Ste. 900, Washington, DC 20024-2115
^b Dep. of Renewable Resources, University of Alberta, 4-42 Earth Sciences Bldg., Edmonton, AB T6G 2E3, Canada

Corresponding author (cesar.izaurralde{at}pnl.gov)

	ABSTRACT

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

 
We related C input and management to soil organic C (SOC) dynamics over 51 yr (1939–1990). We used two rotations from the Breton Classical Plots at Breton, Canada, on a Typic Cryoboralf: (i) wheat (Triticum aestivum L.)–fallow (WF) and (ii) wheat–oat (Avena sativa L.)–barley (Hordeum vulgare L.)–hay (primarily alfalfa, Medicago sativa L.)–hay (WOBHH), in factorial combination with three fertility levels: no added fertilizer [Nil], N-P-K-S fertilizers [F], and farmyard manure [M]. Net aboveground C productivity (NAGCP, kg ha^-1 yr^-1) averaged 576 in WF–Nil and 1078 in WF–F and SOC decreased in both, but NAGCP averaged 1208 in WF–M, where SOC increased. A NAGCP of 853 in WOBHH–Nil maintained SOC, while both 1831 in WOBHH–F and 1714 in WOBHH–M increased SOC. After 51 yr, WOBHH–M had 25 Mg ha^-1 more SOC than did WF–Nil. Because of contrasting decay rates and root/shoot ratios, C input needed to maintain the original SOC was twofold greater in WF than in WOBHH, which required a fourfold increase in NAGCP to attain these inputs. A three-compartment model fitted to the data suggested loss of C from the active compartments and gain of C by the passive compartments. Inputs of C that maintained SOC over 51 yr would lead to a steady state of 2.9 times more C than in 1939, and 26% higher than the native SOC content. Return of 30% of the crop C as manure would sustain SOC sequestration in all WOBHH rotations with NAGCP > 400 kg ha^-1 yr^-1 and in those WF rotations with NAGCP > 1000 kg ha^-1 yr^-1.

Abbreviations: F, N-P-K-S fertilizers treatment • M, farmyard manure treatment • NAGCP, net aboveground C productivity • Nil, no fertilizer treatment • WF, wheat–fallow • WOBHH, wheat–oat–barley–hay–hay

	INTRODUCTION

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

 
THE GLOBAL LOSS of SOC as CO₂ to the atmosphere has been historically important. Cole et al. (1996) estimated the loss of SOC at 55 Pg, including 11 Pg from wetland soils. However, there is increasing evidence that loss of SOC can be reversed when soil management (Janzen et al., 1998) or land use (Poulton, 1996) is changed to increase input of C or reduce soil C loss.

Long-term field studies provide the foundation to understand and the record to test models of SOC dynamics (Smith et al., 1997; Lal et al., 1998a, 1998b; Paul et al., 1997). In Canada, sites remaining from continuous field experiments aid understanding of SOC dynamics in soils and climates characteristic of the northern prairie region (Campbell et al., 1990). The Breton Plots are one such site. They were started in 1930 near Breton, Alberta, to find "a system of farming suitable for the wooded soil belt" (Wyatt et al., 1930). The Breton Plots research site has had several long-term crop rotation and tillage experiments (Wani et al., 1994), with the Breton Classical Plots, started in 1930, holding the longest agronomic and soil records for this soil type in North America.

There is a need to improve quantitative understanding of C inputs to and retention by soil under diverse combinations of soil, climate, and management. This understanding is required both for management practices that improve soil quality and for auditing soil C sequestration. Documentation of SOC changes in long-term field experiments has become an essential task to improve quantitative understanding of the C cycle. The objectives of this study were (i) to document the changes in SOC during a 51-yr (1939–1990) period on the Breton Classical Plots by using historical records of yield, management, and soil C measurements; (ii) to examine the influence of C inputs and manure on SOC content over diverse management; and (iii) to test the utility of a simple model to describe SOC content and its distribution among kinetic soil compartments of increasing stability. For a test of a complex ecosystem model using these data, see Grant et al. (2000).

	MATERIALS AND METHODS

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

 
Location and History
The Breton Classical Plots are located near Breton, AB, Canada, at latitude 53° 06' N, longitude 114° 26' W, and altitude 854 m above sea level. The long-term annual precipitation at the experimental site is 547 mm, while the annual air temperature averages 2.1°C. The soil is an Orthic Gray Luvisol (Typic Cryoboralf) developed on glacial till parent material under boreal forest vegetation. Cryoboralfs occur on 15% of the cultivated area in Alberta, have low fertility status, and may have excess acidity in both their surface and subsurface (Robertson and McGill, 1983).

The Breton Classical Plots were initiated in 1930 on an area that was cleared from its native forest vegetation about 1919. There is no record of the method of clearing, but based on methods used at the time, we believe that the area was cleared by hand and seeded to pasture crops.

Plot Topographic Maps and Treatments
An original map of the Breton Classical Plots was digitized and the data transferred to a file in GRASS format (United States Army Corps of Engineers, 1993) (Fig. 1A) . Topographic data were transferred to GRASS format and used to generate a contour-line map (Fig. 1B) and a slope-class map (Fig. 1C).

View larger version (29K): [in this window] [in a new window]   Fig. 1. (A) Plot layout, (B) contour lines spaced at 1-m vertical interval, and (C) slope classes of the Breton Classical Plots

Plant Yields and Calculations of Additions or Removals of Carbon
Data were obtained from logbooks, reports (Bentley et al., 1971; Wyatt, 1945), CanSIS (1999), and Izaurralde et al. (1996). We extracted the following variables: year (1939–1994); series (A–F) (Fig. 1); fertility treatment (Nil: Plots 1 and 5, M: Plot 2, and F: Plot 3) (Fig. 1); crop (spring wheat [SWHT], oat [OATS], barley [BARL]; grass–legume mixture [HAY₁ and HAY₂]), grain yield (GRNYLD, kg ha^-1), and straw yield (STRYLD, kg ha^-1). Total aboveground dry matter (TADM) for the grass–legume mixture was

(1)

and for the cereal crops

(1a)

When only grain yields were recorded, straw yields were estimated from grain yield data

(2)

where SGR is the straw/grain ratio measured at this site (1.5 for spring wheat, 1.55 for oat, and 1.00 for barley).

Total aboveground C (TAC) was calculated as

(3)

Root dry matter (RDM) estimates used literature values for root/shoot ratios (RSR) (Table 1). A RSR of 0.24 was used for nonfertilized cereal crops. A RSR of 0.12 was used for cereals receiving either commercial fertilizer or cattle manure. These estimates were obtained by Izaurralde et al. (1992) in field experiments on a Typic Cryoboralf that measured above- and belowground spring wheat productivity under a combination of amendments and erosion levels. The RSR for hay crops was 0.43 during the first year of growth and 0.53 during the second, regardless of fertility treatment. The RDM was calculated as

(4)

(5)

(6)

An amount equivalent to 10% of TADM (Eq. [1]) was estimated to enter the soil because most plant biomass is removed from the plots as forage, grain, or straw, but a small amount is always left behind.

The amount of C entering the soil (CPLAN) or the top 15 cm of soil (CPLAN₁₅) was

(7)

(7a)

Before 1980, the only data recorded about the cattle manure was the application rate. Since 1980, manure was from well-decomposed piles, and samples were analyzed yearly for moisture, total N by colorimetry on a H₂SO₄–H₂O₂ digest (Technicon Industrial System, 1977), and total C by combustion on a LECO Carbon Determinator CR12 (LECO Corp., St. Joseph, MI).¹ On average, manure (FYM; kg ha^-1 yr^-1) contained water at 0.7 ± 0.07 kg kg^-1 (wet basis); with N = 0.024 ± 0.002 kg kg^-1, and C = 0.311 ± 0.06 kg kg^-1 (dry basis). Using these concentrations C inputs via manure (CFYM; kg ha^-1 yr^-1) were

(8)

Total C input to the soil (CTOT; kg ha^-1 yr^-1) or to the top 15 cm (CTOT₁₅; kg ha^-1 yr^-1) was

(9)

(9a)

Soil Sampling and Analyses
No soil sampling was conducted in 1930 when the two rotations, at that time a 4-yr rotation and continuous wheat, were initiated. The first archived Ap horizon soil samples are from 1936 (on Series C), the second from 1938 (Series E). Further samplings occurred in 1957 (Bentley et al., 1971) and in 1968 (Khan, 1969), but no soil samples were archived. Three complete sets of Ap soil samples were taken and archived in 1972, 1979, and 1990.

We retrieved all archived samples (Ap horizons) for the treatments studied (Nil, F, and M) and analyzed them for total C and N using a Carlo-Erba NA-1500 analyzer (Carlo Erba Inc., Milan, Italy). Total C and N concentrations were corrected for moisture content (0.03 kg kg^-1). All soil samples lacked carbonates (pH < 6); consequently, total C values, including any charcoal that may have been present represented the organic C fraction.

In fall 1979, samples were taken at 0 to 0.15 m from all series and plots of the Breton Classical Plots (Cannon et al., 1984), together with horizon depth (cm) and bulk density (D_b, Mg m^-3) in the plots and in an undisturbed area under poplar (Populus spp.) immediately north of the plots. Information from that sampling used in this study includes horizon description and depth, D_b, and concentrations of organic C and total N (Table 2).

(10)

where C_t = total SOC (kg ha^-1) at time t (yr); C₁ + C₂ + C₃ = C_t; F₁, F₂, F₃ = fraction of original SOC in each of the kinetic compartments (C₁, C₂, and C₃), respectively, at time 0; F₁ + F₂ + F₃ = 1; k₁, k₂, k₃ = specific decay rate (yr^-1) for each of the kinetic compartments (C₁, C₂, and C₃), respectively; A_t = annual addition of C (kg ha^-1 yr^-1); R₁, R₂, R₃ = fraction of the added C that is partitioned to each of compartments (C₁, C₂, and C₃), respectively; R₁ + R₂ + R₃ = 1.

The values for F and R each sum to one and those for C sum to C_t, so we defined Compartment 3 in terms of Compartments 1 and 2. Equation [10] then becomes

(11)

Equation [11] has seven parameters to be fitted, two less than Eq. [10].

The above model, herein referred to as the Cum model, describes the cumulative mass of SOC over time. Models should be fitted to data in which the values at time t + 1 are independent of the values at time t. To assure that such a condition is met, incremental data are often used (Jans-Hammermeister and McGill, 1997). Although the data used here meet the condition of independence, we fitted the model to incremental (Change) data as an additional check on the parameter values derived from the cumulative data. Consequently the equation for change in SOC since t = 0 yr (Eq. [12]) is referred to as the Change model and was used to estimate parameter values as above.

(12)

Model parameter values were estimated using C_t at 1938 (t = 0), 1972 (t = 33), 1979 (t = 40), and 1990 (t = 51) together with average annual addition of C for the periods 0 to 33, 0 to 40, and 0 to 51 yr. We realize that this procedure unavoidably hides some variation in addition rates among the three intervals.

	RESULTS AND DISCUSSION

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

 
Site Characteristics
Wyatt (1945) provided a general characterization of gray wooded soils (Gray Luvisols, Cryoboralfs) based on soil survey reports. The top three horizons included a "moss to leaf mould" organic A₀, 5 cm thick; followed by a "brownish to black platy" mineral A₁, 5 cm thick; and a "light gray...platy..." A₂, 23 cm thick. Cultivation of these soils changed the color of surface soil layers to light gray due to the mixing action of plowing (Wyatt, 1945). Howitt and Pawluk (1985) excavated and described, in 1977, a soil profile north of and adjacent to the Breton Plots. The site vegetation was a mixture of trembling aspen (Populus tremuloides Michx.), balsam poplar (Populus balsamifera L.), white birch (Betula papyrifera Marsh.), and white spruce [Picea glauca (Moench) Voss]. The soil profile had an organic layer (LFH) 8 cm thick, and the three A horizons (A_eh, A_e1, and A_e2) had a combined thickness of 13 cm. Results from the 1979 quantitative soil description at an undisturbed site immediately north of the Breton Plots (Table 2) revealed an LFH horizon 10 cm thick and an A horizon 18 cm thick. Given the ranges recorded in 1979 and the descriptions by Wyatt (1945) and Howitt and Pawluk (1985) we infer (i) that the site of the Breton plots is representative of Gray Luvisols and (ii) that the thickness of the A horizon at the beginning of cultivation around 1920 varied between 13 and 18 cm.

From the data for horizon depth and D_b sampled at the end of the 1979 growing season (Table 3) it appears the Ap horizons under the WF rotation were 3 cm thinner than those under the WOBHH rotation. Such a difference is not expected. The thickness of an Ap should be determined mostly by tillage. We can't confirm whether equipment went deeper when soil conditions were better or alternatively if the soil compacted more so the Ap depth is thinner on plots under the WF rotation. Although the depth range of these determinations (12–18 cm) falls within that determined on the native sites just north of the plots (13–18 cm), it is likely that there was some soil movement by either tillage or erosion in the WF rotation. The fallow frequency dictated that these plots were tilled frequently, especially before 1979, to control weeds and to prepare the seedbed. This could have induced some soil movement by tillage implements within the plot area or outside the plot area. The ¹³⁷Cs data reported for these plots by Monreal et al. (1995) are consistent with such movement. Another possibility is that some soil was lost by water erosion. On the basis of the contour lines and slope classes of Fig. 1B determined on the upper half of the experimental area-that includes the plots for which horizon depth and D_b are reported—we conclude that soil transport by water erosion was small if any at all. Wind erosion was not a factor at this site.

Plant Carbon Productivity
There were substantial differences in plant C productivity between the WF and the WOBHH rotations (Fig. 2) . The 91 Mg ha^-1 of aboveground plant C accumulated in WOBHH treatments during the period 1939 to 1994 was 86% greater than the 49 Mg ha^-1 accumulated in WF treatments. There were clear differences in C productivity induced by fertility treatments. In the WF rotation, the Nil treatment had a low (1000 kg C ha^-1 per rotation cycle) but stable aboveground C productivity during the entire study period. Additions of either fertilizer or farmyard manure increased C productivity over the Nil treatment but without clear differences between them. Carbon productivity of both addition treatments varied during the first 35 yr of the study, and a productivity increase occurred during the 1980s followed by a decline early in the 1990s. It is not the intent here to ascribe this variability to climatic events because the values presented are for rotation cycles and not individual years. However, the influence of weather variability or potential S fertilization from local petroleum production on plant productivity cannot be discounted.

View larger version (28K): [in this window] [in a new window]   Fig. 2. Aboveground plant C productivity per rotation cycle at the Breton Classical Plots

Soil Organic Carbon and Total Soil Nitrogen
The average SOC concentration in 1938 (calculated from one sample taken in 1936 and two in 1938) was 13.2 g kg^-1, while that of total soil N was 1.21 g kg^-1, yielding a C/N ratio in Ap horizons of 11 (Table 4).

We emphasize that the changes in SOC just discussed arose mostly as a result of root and manure C additions. The contribution of litter C was small because most of the aboveground plant biomass was removed from the plots. The role of manure in soil C sequestration is discussed below. In agreement with Balesdent and Balabane (1996), Campbell et al. (1991), and Solberg et al. (1998), these results continue to confirm the important contribution of roots to conserving or increasing SOC.

Total soil N followed similar trends as SOC, with the greatest decline in concentration (-0.3 g kg^-1) in the WF–Nil treatment, while the greatest increase (0.8 g kg^-1) was in the WOBHH–M. The management imposed created two groups of soil samples based on their C/N ratios. The WF rotation narrowed the C/N ratio from 11 to 10, while the WOBHH rotation maintained the C/N ratio closer to the original (11), suggesting cropping systems with ability to retain more C.

Soil Organic C Dynamics
Model Fitting
We calculated the annualized rates of plant and manure C additions to soil during three periods from 1939 to 1990 (Table 5). While on average the largest difference in plant C additions among the six treatments varied fourfold, the difference increased up to 10-fold when manure additions were considered.

View this table: [in this window] [in a new window]   Table 5. Annual rates of plant- or manure-derived C added to a Breton loam during three periods from 1939 to 1990.

View this table: [in this window] [in a new window]   Table 6. Values of At, Ct (Ap horizons at end of period), and t used to derive parameter values for the three compartment model.

We then used the parameterized model and the annualized C input rates for each treatment–period to simulate the change in SOC. The Cum model described 70% of the variation in gain or loss of SOC. Observed changes regressed on predicted changes had a slope (1.02) that was not significantly different from one and the y-intercept (-237 kg ha^-1) was not significantly different from 0 (Fig. 3A) .

View larger version (22K): [in this window] [in a new window]   Fig. 3. (A) Relationship between simulated and observed changes in soil organic C (SOC) (kg C ha-1 yr-1). Change in SOC is the SOC at each sampling date minus SOC at 1939. Simulated SOC at each date was calculated with Eq. [10] using parameter values reported in Table 7 for the Cum model. (B) Mass of SOC was simulated using adjusted values of k to distinguish wheat–fallow (WF) from wheat–oat–barley–hay–hay (WOBHH) rotations and compared with observed values for the Breton plots over 51 years. Adjustment entailed multiplying each k by 1.3 for the WF rotations and dividing them by 1.3 for the WOBHH rotations

We compared the Cum and Change models on the basis of their simulations of total SOC using parameter values shown in Table 7. We accounted for about the same amount of variation in SOC in both cases (71%), but the observed values regressed on predicted values had a y-intercept of -1298 kg ha^-1 for the Change model compared with -864 kg ha^-1 for the Cum model. The above is not a test of either model against independent data. It describes how the parameters obtained from the data when used in a variety of ways estimate the original data set. We used the values Cum model (Eq. [11]) for the remaining analyses.

A Test of Two Hypotheses
One purpose of using a simple model to describe all the variation in SOC was to test the hypotheses: (i) that the amount of C stored or lost is regulated by the initial mass and distribution of SOC among compartments, and by the mass of C inputs as distinct from the nature of the rotation, and (ii) that at this site the decay rates of SOC compartments and the proportions of C entering them did not vary among rotations. The model should fail to the extent that the nature of the rotation, which is tied to timing, quality, or location of C input, and to frequency or intensity of tillage, regulates SOC storage. The model should fail to the degree that decay rates or apportioning of C into SOC compartments vary among rotations and cropping systems.

Overall, the three-compartment model explained slightly more than 70% of the total variation induced by the treatments (C loss of 8300 vs. gain of 16700 kg ha^-1 = 25 Mg ha^-1) (Fig. 3A). Consequently, 70% of the variation in SOC gain or loss under conditions of the same climate and soil across all treatments appears to be related to variation in mass of C input. The remaining 30% appears to be related to variations in decay rates and apportioning C inputs to compartments or to variations in nature, timing, or location of C inputs among the rotations used here. We hypothesized higher decay rates in the WF rotation than in WOBHH because of more frequent soil disturbance and perhaps better moisture regime. To test this hypothesis we multiplied the values of k₁, k₂, and k₃ by factors ranging from 1.1 to 2 to calculate SOC in WF rotations and divided them by the same number to calculate SOC in WOBHH treatments. Adjusting the decay rates up by a factor of 1.3 for WF treatments and down by the same factor for WOBHH rotations caused the predicted values to converge more closely on the observed (Fig. 3B). The regression equation was Observed = 0.99 Predicted - 31.8 kg ha^-1 (r² = 0.85; P = 10^-8). Although the correlation could be further improved by increasing the difference in decay rates between WF and WOBHH, the slopes of the relationships diverged from one, and the y-intercepts increased and differed from zero. This simple test of sensitivity to decay rates suggests that an additional 15% of the variation in SOC contents at this site may be associated with differing decay rates between WF and WOBHH rotations. Apparently, the nature of the rotation influences both amount of C entering soil and its behavior once there.

Maintenance Carbon Inputs
We plotted the 51-yr mean annual change in SOC in the top 15 cm against mean annual addition of C (A_t) (Fig. 4) . For a simplified first-order system with constant rates of addition and decay, such a plot is linear. However, the line appears to be logarithmic, indicating departure from the simplified system.

View larger version (21K): [in this window] [in a new window]   Fig. 4. Relationship between annual addition rate of C and measured annual change in soil organic C (SOC) over 51 yr

Only a portion of the plant C enters the top 15 cm of soil (CPLAN₁₅). When CFYM = 0 we express CPLAN₁₅ as a fraction (FI) of the total aboveground C (TAC), such that CPLAN₁₅ = TAC x FI. From this and Eq. [3] to [7]

(13)

(14)

Using our root/shoot ratios the value of FI was 0.29 for the WF–Nil, 0.2 for the WF–F, 0.37 for WOBHH–Nil, and 0.31 for WOBHH–F. The higher values for FI in WOBHH rotations compared with the WF rotations reflect differences in RSR. Hence, higher root/shoot ratios of crops translates into lower total aboveground C productivity needed to maintain the same SOC content. The low-yielding rotations (WF) are hit doubly hard. They have the highest requirements for C input because of high decay rates, and the highest required aboveground C productivity per unit of C input because of low root/shoot ratios of cereals. For example, equating the C input needed to maintain SOC with CPLAN₁₅ in Eq. [13], the required aboveground net C productivity (kg ha^-1 yr^-1) is 1730 for WF–Nil and 2578 for WF–F, but only 656 for WOBHH–Nil, and 779 for WOBHH–F. Consequently, based on data for 51 yr, in the absence of manure additions, the aboveground productivity needed to maintain the original SOC content at this site might vary through crop rotation and management by fourfold (2578/656).

Carbon Allocation Arising from Parameter Values for the Model
The parameter values estimated for the model from the data for this site suggest that in 51 yr the system has not reached steady state, and C may continue to accumulate in the passive component. The following discussion shows why.

Mass (M, kg ha^-1), addition rate (A, kg ha^-1 yr^-1), decay rate (k, yr^-1) and turnover time (T, yr) are all related at steady state: M = A/k; T = 1/k; T = M/A, and A = M/T.

At steady state, Eq. [10] reduces to

(15)

Using Eq. [15] and T = M/A

(16)

From the original model parameters (no adjustment between treatments) and Eq. [16], the steady-state turnover time for soil + litter components (C₁ + C₂ + C₃) is 226 yr (0.803/0.2692 + 0.099/0.0037 + 0.098/0.0005). Based on the global budget of Rodhe (1992), global soil C turnover time is 385 yr. Global turnover times for combined soil C + litter C + peat C are estimated to be 28 yr (Holmén, 1992; Rodhe, 1992).

Using 226 yr as the turnover time, the annual input of C needed to maintain the original SOC is 117 kg ha^-1 yr^-1 (26401/226). In contrast, the maintenance calculations above, which used the equation in Fig. 4 from the 51-yr data, estimated a C input of 334 kg ha^-1 yr^-1 to maintain the original SOC. From M = AT and the 51-yr values for maintenance C, the steady-state SOC content will be 334 x 226 = 75484 kg ha^-1, or 2.9-fold greater than in 1939. If, as we estimated, the native SOC was 60 Mg ha^-1, this steady state is 26% higher than the native SOC.

Why does a constant addition of C appear to maintain the original C for 51 yr, but appear to sequester C after that time? It does so mainly because of interactions between parameters F and R. The parameters F describe initial distribution of C among compartments. Parameters R describe subsequent allocation of input C to compartments. For a fixed input, dynamics during the early decades are dominated by F, the long-term trend is dominated by R; and steady-state values are independent of F, depending only on R/k. Given the parameter values in Table 7, component C₂ will lose C with time and component C₃ will gain C.

Given the estimates of F, 75% of the original SOC is in compartment C₂ (active) with only 22% in C₃ (passive). However, at steady state the situation reverses, with (0.098/0.0005)/M = 87% in compartment C₃ and only (0.099/0.0037)/M = 12% in compartment C₂ (See F_SS in Table 7). The trend observed is the balance of losses from C₂ and gains in C₃ (Fig. 5A) .

View larger version (18K): [in this window] [in a new window]   Fig. 5. (A) Carbon content of each of compartments C1, C2, and C3 in the wheat–fallow (WF) rotation using adjusted k values and model-derived additions that maintain original SOC content at 51 yr. See Table 7 and Fig. 3 and 4 for model information. (B) Trend of total soil organic C (SOC) in each rotation as calculated using the three-compartment model with model-derived additions that maintain the original SOC at 51 yr. Additions needed to maintain the original SOC are greater for the WF than the wheat–oat–barley–hay–hay (WOBHH) rotations

Can Manure Additions Sequester Carbon?
Addition of farmyard manure was a key management component leading to SOC increases (greatest two C additions in Table 5). Using data from Buyanovski and Wagner (1999), Schlesinger (1999) argued that manuring is not a valid method for soil C sequestration because of the extra land required to produce the manure. This means crops produced on a unit area of land yield insufficient manure to maintain or increase SOC.

We tested this proposition with data from the Breton site. The key is to focus on maintenance C requirements (values at 51 yr; Equations in Fig. 4) and not on experimental manure application rates. We related CFYM to total aboveground C (TAC) using the fraction of TAC converted to manure (FR)

(17)

We used Eq. [9a] with substitutions from Eq. [13] for CPLAN₁₅ and from Eq. [17] for CFYM and rearranged to yield

(18)

(19)

We set CTOT₁₅ = to maintenance C after 51 yr for WF (505 kg ha^-1 yr^-1) and WOBHH (241 kg ha^-1 yr^-1) rotations and used the RSR for manured treatments in Eq. [18] to calculate TAC for increasing FR. As manure return (FR) increases the net C productivity (TAC) needed to maintain SOC drops (Fig. 6) . Given actual TAC, what FR would be needed to achieve net soil C sequestration?

View larger version (21K): [in this window] [in a new window]   Fig. 6. Relationship between the fraction of net aboveground C (TAC) that is returned as manure (FR) and the net C productivity required to meet C inputs that maintain the original soil organic C (SOC) in the wheat–fallow (WF) and wheat–oat–barley–hay–hay (WOBHH) rotations (from Eq. [18]). The greater inputs needed to maintain the original SOC in the WF rotation compared to the WOBHH rotation in combination with the lower proportion of TAC returned as roots in WF rotations result in a greater TAC needed to maintenance SOC. The actual values of TAC for three WF rotations and the WOBHH–Nil rotation are plotted as horizontal lines. When the WF lines reach the TAC line for the WF rotation they proceed down to the corresponding value of FR. The TAC for the WOBHH rotations are all above the TAC necessary to maintain SOC in the WOBHH rotation so only the point is shown. Values of FR at which the TAC meets maintenance requirements can be calculated with Eq. [19]

If FR > 1 clearly means extra land is needed to produce manure, then is any value of FR < 1 acceptable? No. First, some of TAC (0.1) is already input to the soil, so only 0.9 remains. Second, it is impossible for all TAC to be converted to manure because of conversion to livestock products or CO₂. So how much conversion of TAC to manure is realistic? If 60% of animal feed is oxidized or converted to animal products (Schlesinger, 1999) then 40% might be left for conversion to manure. So FR up to 0.36 [0.9(1 - 0.6)] might be achievable. With the exception of the WF–Nil, all rotations had FR < 0.36. Consequently, the crops produced in them would provide enough manure to achieve net C sequestration with observed TAC. The actual FR for the WF–M treatment was 0.65 (784/1208). This was higher than would normally be expected with all manure produced on site and was higher than necessary (FR = 0.22).

We show a feedback between the fraction of aboveground C that is returned as manure and net C productivity. Increasing the fraction returned reduces the C productivity necessary for a fixed C input to soil, and return of manure increases C productivity. We conclude that return of 30% of the crop C at this site as manure would sustain net soil C sequestration in all WOBHH rotations with net C productivity >400 kg ha^-1 yr^-1; it would sustain net soil C sequestration in all WF rotations with net aboveground C productivity >1000 kg ha^-1 yr^-1.

	SUMMARY AND CONCLUSIONS

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

 
Net aboveground C productivity of the WF system averaged 576 kg ha^-1 yr^-1 without fertilizer, and 1078 with fertilizer, SOC decreased in both; it was 1208 in WF–M, which gained SOC. The WOBHH rotation, had net aboveground C productivity of 853 kg ha^-1 yr^-1 (Nil, maintained SOC), and 1831 (F), and 1714 (M) both of which increased SOC. The range in masses of SOC within the 0- to 0.15-m depth after 51 yr was 25 Mg ha^-1.

About 70% of the variation in SOC content was described using C inputs alone, and 85% by using C inputs and rotation-specific decay rates. Decay rates may vary by a factor of 1.69 between these rotations. Because C inputs are a dominant control on SOC content, it may be more appropriate to use measured crop yields when available than to try to simulate them for use as inputs to SOC models.

After 51 yr the soil in some rotations at this site appears still to be gaining C. Use of a simple kinetic model to describe the data suggests that C is lost from the active compartment and accumulates in the passive compartment. Potential steady-state accumulations of SOC are 2.9 times the SOC in 1939 and 26% above the native SOC content.

High decay rates in the WF rotation necessitate more C input (505 kg ha^-1 yr^-1) to maintain the original SOC in this rotation compared with the WOBHH rotation (241 kg ha^-1 yr^-1); and low root/shoot ratios in WF increase the net aboveground C productivity needed to attain these maintenance inputs (range: 2578 kg ha^-1 yr^-1 in WF–F to 656 in WOBHH–Nil). Increasing the fraction of aboveground C that is returned as manure reduces the C productivity necessary for a fixed C input to soil, and return of manure increases C productivity. Return of 30% of the crop C at this site as manure would sustain net soil C sequestration in all WOBHH rotations with net C productivity >400 kg ha^-1 yr^-1, and in all WF rotations with net aboveground C productivity >1000 kg ha^-1 yr^-1.

Insights from these long-term experiments are pertinent for long-term farm productivity, soil C storage, and the global environment.

	ACKNOWLEDGMENTS

 
We dedicate this manuscript to the memory of Mr. Ben Flesher and family who homesteaded in Breton around 1919 and allowed for the initiation of research work on a piece of their farm. We are deeply indebted to the Flesher family, in particular Mr. Bill Flesher, for continuing the tradition of collaboration. We are grateful to Dr. C.A. Campbell for a thorough review and constructive suggestions for revisions of an earlier version of this paper. We thank J. Brown and H. Puurveen for field assistance; G. Spiers for soil profile descriptions; M. Molina, J. Konwicki, and C. Figueiredo for laboratory analyses; C. Nguyen, K. Cannon, and M.C. Quiroga-Jakas for statistical analyses and data compilation; T. Goddard for topographic survey; and T. Martin for digitizing topographic data. Financial support for the Breton Plots has been received from numerous individuals, institutions, and agencies throughout the years, notably the Breton Plots Endowment Fund, the University of Alberta, the Alberta Agricultural Research Institute, the Canada–Alberta Soil Conservation Initiative, and the Canada–Alberta Environmentally Sustainable Agriculture Agreement, and the Natural Sciences and Engineering Research Council. R.C. Izaurralde and W.B. McGill thank the Office of Science of the U.S. Department of Energy and the U.S. Environmental Protection Agency for support to complete this manuscript.

	NOTES

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

 
¹ Trade names are mentioned here for the benefit of the reader and imply neither our endorsement by their inclusion, nor criticism of similar ones by their omission.

Received for publication January 3, 2000.

	REFERENCES

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

 

• Balesdent, J., and M. Balabane. 1996. Major contribution of roots to soil carbon storage inferred from maize cultivated soils. Soil Biol. Biochem. 28:1261–1263.
• Bentley, C.F., A.M.F. Hennig, T.W. Peters, and D.R. Walker (ed.) 1971. Gray Wooded soils and their management. Bull. B-71-1. 7th ed. Univ. of Alberta, Edmonton, AB, Canada.
• Broersma, K. 1991. The effects of different cropping systems on a Luvisolic soil in the Peace River Region. Ph.D. thesis. Dep. of Soil Science. Univ. of Alberta, Edmonton, AB, Canada.
• Buyanovski, G.A., and G.H. Wagner. 1998. Changing the role of cultivated land in the global carbon cycle. Biol. Fertil. Soils 27:242–245.
• Campbell, C.A., V.O. Biederbeck, and F.G. Warder. 1977. Effects of fertilizer N and soil moisture on yield, yield components, protein content and N accumulation in the above ground parts of spring wheat. Can. J. Soil Sci. 57:311–327.
• Campbell, C.A., G.P. Lafond, R.P. Zentner, and V.O. Biederbeck. 1991. Influence of fertilizer and straw baling on soil organic matter in a thin Black Chernozem in western Canada. Soil Biol. Biochem. 23:443–446.
• Campbell, C.A., R.P. Zentner, H.H. Janzen, and K.E. Bowren. 1990. Crop rotation studies on the Canadian prairies. Publ. 1841/E. Agriculture Canada, Research Branch, Ottawa, ON, Canada.
• Cannon, K.R., J.A. Robertson, W.B. McGill, F.D. Cook, and D.S. Chanasyk. 1984. Production optimization of Gray Wooded soils. Farming for the Future Project no. 79-0132. Dep. Soil Science, Univ. of Alberta. Edmonton, AB, Canada.
• CanSIS. 1999. Canada Soil Information System. Centre for Land and Biological Resources Research, Research Branch, Agriculture and Agri-Food Canada, Ottawa. Available at http://res.agr.ca/CANSIS/_overview.html (verified 15 Nov. 2000).
• Cole, V., C. Cerri, K. Minami, A. Mosier, N. Rosenberg, and D. Sauerbeck. 1996. Agricultural options for mitigation of greenhouse gas emissions. p. 744–771. In R.T. Watson et al. (ed.) Climate change 1995: Impacts, adaptations, and mitigation of climate change: Scientific–technical analyses. Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge Univ. Press, New York.
• Curran, B.S., K.D. Kephart, and E.K. Twidwell. 1993. Oat companion crop management in alfalfa establishment. Agron. J. 85:998–1003.[Abstract/Free Full Text]
• Grant, R.F., N.G. Juma, J.A Robertson, R.C. Izaurralde, and W.B. McGill. 2001. Long-term changes in soil carbon under different fertilizer, manure, and rotation: Testing the mathematical model ecosys with data from the Breton Plots. Soil Sci. Soc. Am. J. 65:205–214.[Abstract/Free Full Text]
• Gu, J. 1988. Carbon and nitrogen assimilation, dinitrogen fixation in fababean (Vicia faba L.) and microbial biomass in soil–plant systems (fababean, canola, barley, and summer fallow) on a Gray Luvisol. MSc thesis. Dep. of Soil Science. Univ. of Alberta. Edmonton, AB, Canada.
• Holmén, K. 1992. The global carbon cycle. p. 238–262. In S.S. Butcher et al. (ed.) Global biogeochemical cycles. Academic Press, Toronto.
• Howitt, R.W., and S. Pawluk. 1985. The genesis of a Gray Luvisol within the boreal forest region. I. Static pedology. Can. J. Soil Sci. 65:1–8.
• Izaurralde, R.C., K.L. Haugen-Kozyra, D.C. Jans, W.B. McGill, R.F. Grant, and J.C. Hiley. 2000. Soil organic carbon dynamics: measurement, simulation and site to region scale-up p. 553–575. In R. Lal, J.M. Kimble, R.F. Follett and B.A. Stewart (ed.) Adv. Soil Sci: Assessment methods for soil carbon. Lewis Publ., Boca Raton, FL.
• Izaurralde, R.C., N.G. Juma, W.B. McGill, D.S. Chanasyk, S. Pawluk, and M.J. Dudas. 1993. Performance of alternative cropping systems in cryoboreal-subhumid central Alberta. J. Agric. Sci. 120:33–41.
• Izaurralde, R.C., N.G. Juma, J.A. Robertson, W.B. McGill, and J.T. Thurston. The Breton Classical Plots dataset. 1996. p. 351–362. In D.S. Powlson et al. (ed.) Evaluation of soil organic matter models using existing long-term datasets. NATO ASI Series I. Vol. 38. Springer-Verlag, Heidelberg, Germany.
• Izaurralde, R.C., M. Nyborg, D.S. Chanasyk, and E.D. Solberg. 1992. Effects of artificial erosion and amendments on soil and crop productivity. Environmental Sustainability Initiative (PFRA-Agriculture Canada). Univ. of Alberta, Edmonton, AB, Canada.
• Jans-Hammermeister, D.C., and W.B. McGill. 1997. Evaluation of three simulation models used to describe plant residue decomposition in soil. Ecol. Modell. 104:1–13.
• 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 Till. Res. 47:181–195.
• Jones, C.A., W.L. Bland, J.T. Ritchie, and J.R. Williams. 1991. Simulation of root growth. p. 92–124. In J. Hanks and J.T. Ritchie (ed.) Modeling plant and soil systems. Agron. Monogr. 31. ASA, CSSA, and SSSA. Madison, WI.
• Juma, N.G., R.C. Izaurralde, J.A. Robertson, and W.B. McGill. 1997. Crop yield and soil organic matter over 60 years in a Typic Cryoboralf at Breton, Alberta. p. 273–282. In E.A. Paul et al. (ed.) Soil organic matter in temperate agroecosystems: Long-term experiments in North America. Lewis Publ., CRC Press, Boca Raton, FL.
• Khan, S.U. 1969. Some carbohydrate fractions of a Gray Wooded soil as influenced by cropping systems and fertilizers. Can. J. Soil Sci. 49:219–224.
• Lal, R., J. Kimble, R. Follett, and B.A. Stewart (ed.) 1998a. Management of carbon sequestration in soil. Advances in Soil Science. CRC Press, Boca Raton, FL.
• Lal, R., J. Kimble, R. Follett and B.A. Stewart (ed.) 1998b. Soil processes and the carbon cycle. Advances in Soil Science. CRC Press, Boca Raton, FL.
• McCoy, D.A., and G.R. Webster. 1977. Acidification of a Luvisolic soil caused by low-rate, long-term applications of fertilizers and its effects on growth of alfalfa. Can. J. Soil Sci. 57:119–127.
• McGill, W.B., K.R. Cannon, J.A. Robertson, and F.D. Cook. 1986. Dynamics of soil microbial biomass and water soluble C in Breton L after 51 years of cropping to two rotations. Can. J. Soil Sci. 66:1–19.
• Monreal, C.M., R.P. Zentner, and J.A. Robertson. 1995. The influence of management on soil loss and yield of wheat in Chernozemic and Luvisolic soils. Can. J. Soil Sci. 75:567–574.
• Nyborg, M., E.D. Solberg, S.S. Malhi, and R.C. Izaurralde. 1995. Fertilizer N, crop residue, and tillage alter soil C and N contents after a decade. p. 93–100. In R. Lal et al. (ed.) Soil management and greenhouse effect. Advances in Soil Science. Lewis Publ., CRC Press, Boca Raton, FL.
• Paul, E.A., K. Paustian, E.T. Elliott, and C.V. Cole (ed.) 1997. Soil organic matter in temperate agroecosystems: Long-term experiments in North America. Lewis Publ., CRC Press, Boca Raton, FL.
• Poulton, P.R. 1996. Geesecroft wilderness, 1883–1995. p. 384–389. In D.S. Powlson et al. (ed.) Evaluation of soil organic matter models using existing long-term datasets. NATO ASI Series I, Vol. 38. Springer-Verlag, Heidelberg, Germany.
• Robertson, J.A., and W.B. McGill. 1983. New directions for the Breton Plots. Univ. Alberta Agric. For. Bull. 6:36–41.
• Rodhe, H. 1992. Modeling biogeochemical cycles. p. 53–72. In S.S. Butcher et al. (ed.) Global biogeochemical cycles. Academic Press, Toronto.
• SAS Institute. 1985. SAS user's guide. Version 5 ed. SAS Inst., Cary, NC.
• Schlesinger, W.H. 1999. Carbon and agriculture: Carbon sequestration in soils. Science 284:2095.[Free Full Text]
• Smith, P., J.U. Smith, D.S. Powlson, W.B. McGill, J.R.M. Arah, O. Chertov, K.W. Coleman, U. Franko, S. Frolking, D.S. Jenkinson, L.S. Jensen, R.H. Kelly, H. Klein-Gunnewiek, A.S. Komarov, C. Li, J.A.E. Molina, T. Mueller, W.J. Parton, J.H.M. Thornley, and A.P. Whitmore. 1997. A comparison of the performance of nine soil organic mater models using datasets from seven long-term experiments. Geoderma 81:153–225.[ISI]
• Solberg, E.D., M. Nyborg, R.C. Izaurralde, S.S. Malhi, H.H. Janzen, and M. Molina-Ayala. 1998. Carbon storage in soils under continuous cereal grain cropping: N fertilizer and straw. p. 235–254. In R. Lal et al. (ed.) Management of carbon sequestration in soil. Adv. Soil Sci. CRC Press, Boca Raton, FL.
• Technicon Industrial Systems. 1977. Individual/simultaneous determinations of nitrogen and/or phosphorus in Bd acid digests. Industrial method 334-74w/Bt. Technicon Industrial Systems, Tarrytown, NY.
• United States Army Corps of Engineers. 1993. GRASS 4.1 user's reference manual. U.S. Army CERL, Champaign, IL.
• Wani, S.P., W.B. McGill, K.L. Haugen-Kozyra, J.A. Robertson, and J.J. Thurston. 1994. Improved soil quality and barley yields with fababeans, manure, forages and crop rotations on a Gray Luvisol. Can. J. Soil Sci. 74:75–84.
• Wyatt, F.A., J.D. Newton, and T.H. Mather. 1930. Soil Survey of St. Ann Sheet. Univ. of Alberta Bull. 20. Univ. of Alberta, Edmonton, AB, Canada.
• Wyatt, F.A. 1945. Fifteen years experiments on the Gray Wooded soils of Alberta. Sci. Agric. (Ottawa) 25:626–635.

This article has been cited by other articles:

				B. E. Fronning, K. D. Thelen, and D.-H. Min Use of Manure, Compost, and Cover Crops to Supplant Crop Residue Carbon in Corn Stover Removed Cropping Systems Agron. J., November 7, 2008; 100(6): 1703 - 1710. [Abstract] [Full Text] [PDF]

				S. A. Khan, R. L. Mulvaney, T. R. Ellsworth, and C. W. Boast The Myth of Nitrogen Fertilization for Soil Carbon Sequestration J. Environ. Qual., October 24, 2007; 36(6): 1821 - 1832. [Abstract] [Full Text] [PDF]

C. A. Campbell, H. H. Janzen, K. Paustian, E. G. Gregorich, L. Sherrod, B. C. Liang, and R. P. Zentner Carbon Storage in Soils of the North American Great Plains: Ef