Published online 29 May 2008
Published in Soil Sci Soc Am J 72:970-974 (2008)
DOI: 10.2136/sssaj2007.0327
© 2008 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SOIL & WATER MANAGEMENT & CONSERVATION
Soil Organic Carbon after Twelve Years of Various Crop Rotations in an Aridic Boroll
E. Bremera,*,
H. H. Janzenb,
B. H. Ellertb and
R. H. McKenziec
a Symbio Ag Consulting, Lethbridge, AB, Canada, T1K 2B5
b Agriculture and Agri-Food Canada, Lethbridge, AB, Canada, T1J 4B1
c Alberta Agriculture and Food, Lethbridge, AB, Canada, T1J 4C7
* Corresponding author (ericbremer{at}shaw.ca).
 |
ABSTRACT
|
|---|
Eliminating summer fallow or establishing perennial grass elicited measurable gains in soil organic C (SOC) after just 6 yr in a crop rotation study on an Aridic Boroll in southern Alberta. This study was resampled after 12 yr to determine if SOC increases were continuing with time and to evaluate the impact of alternative crop rotation and fertilizer treatments on SOC. The crop rotation treatments included in this study were fallow-wheat (FW), fallow-wheat-wheat (FWW), fallow-flax-wheat (FXW), legume-wheat (LW), continuous wheat (W) and continuous grass (G). The gain in SOC due to the elimination of fallow was 1.5 Mg C ha–1 after 12 yr, no greater than that observed after 6 yr. Soil organic C was the same for all rotations that included fallow (FW, FWW, FXW). Fertilizer treatments that had the greatest benefit on grain yields of annual crops also tended to increase SOC, although differences were barely detectable. The gain in SOC of unfertilized grass compared to the FW rotation was no higher after 12 yr than after 6 yr (3 Mg C ha–1). Under fertilized grass, in contrast, SOC continued to increase at a rate of approximately 0.5 Mg C ha–1 yr–1. Accumulation of light fraction C accounted for most of the gains in SOC that occurred with elimination of fallow or establishment of grass. These findings suggest that much of the SOC gain due to adoption of C-conserving practices in soils like those of this study may occur early, within the first decade, and consist primarily of decomposable soil fractions. If confirmed, this means that C sequestration in these soils may be comparatively short-lived and vulnerable to future loss.
Abbreviations: FW, fallow-wheat • FWW, fallow-wheat-wheat • FXW, fallow-flax-wheat • G, continuous grass • LW, legume-wheat • SOC, soil organic carbon • W, continuous wheat
|
INTRODUCTION
|
|---|
Efforts to reduce the rate of increase in atmospheric CO2 have stimulated considerable effort to monitor changes in SOC. Evidence for change in SOC of agricultural lands has largely been obtained from long-term agro-ecosystem experiments. However, accurate and precise estimates of change in SOC are difficult to achieve due to the heterogeneous composition, large background, and high spatial variability in SOC (Ellert et al., 2006; VandenBygaart, 2006). Improved estimates may be achieved by increasing the number of long-term studies sampled and the intensity at which they are sampled.
A dryland crop rotation study was initiated in 1992 on a Chin clay loam (Aridic Boroll) in southeastern Alberta to investigate long-term impacts of crop rotation and fertility amendment on crop productivity and soil quality (Bremer et al., 2002). Total SOC after 6 yr was 1.5 Mg C ha–1 higher (P = 0.03) due to elimination of summer fallow and 3.0 Mg C ha–1 higher (P = 0.0001) due to establishment of grass, but was not affected by fertilizer treatment or the interaction of fertilizer and crop rotation treatments. Small differences in SOC were detectable due to the low spatial variability of SOC at this location. The experiment was resampled after 12 yr to determine if SOC gains during the first 6 yr were continuing with time and if impacts of fertilizer treatment or alternative crop rotations were now detectable.
|
MATERIALS AND METHODS
|
|---|
A crop rotation experiment was established in 1992 at the Bow Island Substation of the Alberta Crop Development Center South. The site is located in the shortgrass prairie ecozone and has an average moisture deficit of approximately 400 mm. The land had been broken from native prairie in about 1920 and, based on typical practice in the region, cropped to spring wheat every other year for about 50 yr, with cultivated fallow in the 20-mo period between crops. Irrigation was established at this site in the early 1970s, and irrigated cereal crops were grown at this location for about 20 yr before experiment initiation. In 1991, unfertilized spring wheat was grown without irrigation over the whole site. The soil is a Chin loam (40.3% sand, 30.3% silt, 29.4% clay), an Aridic Boroll (i.e., Brown Chernozem). The topography is flat. Total SOC to 15 cm at experiment initiation was 20.6 Mg C ha–1.
Six cropping system treatments (rotations) were included in this experiment, each with two or more fertilizer treatments (Table 1
). All phases of all crop rotations were present each year. Wheat consisted of a recommended cultivar of hard red spring wheat that was planted in late April or early May and harvested in August or September. Recommended legume and flax cultivars were planted and harvested at the same time. All cropped plots were tilled with a heavy-duty cultivator, harrowed and packed before being seeded on the same day. Fallow was maintained free of weeds using herbicides (one to three applications per growing season) or tillage (once or twice per growing season using a heavy-duty cultivator and harrow). Nitrogen application to annual crops consisted of ammonium nitrate banded in late fall at a depth of 7 to 9 cm. Phosphorus application to annual crops consisted of triple superphosphate placed with the seed at the time of seeding. For the fertilized grass treatment, P fertilizer was only applied at the time of seeding at a rate of 65 kg P ha–1, while N fertilizer was broadcast in early spring each year.
The experiment was laid out in a split-plot design with four blocks. Main plots consisted of crop rotation. The two phases of the FW rotation were included as separate main plots and G and LW treatments were blocked together within a single main plot. Subplots (4 by 16 m) consisted of fertility and rotation phase treatment (six subplots per main plot, total of 36 plots per block). Subplot treatments were not randomized between each block.
Grain yields were determined by harvesting 54 m2 with a plot combine. Grass plots were harvested for hay in late June or early July each year. Subsamples were taken of all harvested material and analyzed for moisture and protein concentration. Precipitation was measured at a weather station located 300 m from the experimental site.
Soil samples were obtained in October 2003 with the same methods used in 1997 (Bremer et al., 2002). Four cores (6.7-cm diam.) were taken with a hydraulic soil corer between crop rows in each plot and composited into 0- to 7.5-cm, 7.5- to 15-cm, and 15- to 30-cm depth increments. The total mass of soil within the cored volume was determined for calculation of SOC on an equivalent mass basis. Air-dry samples were stored for 2 yr before processing and analysis. Whole soil samples were ground to <2 mm in a rotating sieve. All crop residues and root material in the soil sample passed through the sieve by the end of the grinding period. Rocks >2 mm made up an average of 0.4% of total soil weight and were removed from soil samples after weighing.
Portions of these 2-mm samples were further ground to <0.125 mm by tumbling in a stainless steel canister with two tumbling bars for 16 h. Total C and N were determined with an automated combustion analyzer (Carlo ErbaTM, Milan, Italy). Soil organic C was determined with the automated combustion analyzer after removing carbonates by adding excess hydrochloric acid to the combustion capsule. Light fraction material (specific gravity <1.7) was isolated from the 0- to 7.5-cm depth by the method described by Janzen et al. (1992). Light fraction C was determined with the automated combustion analyzer and heavy fraction C was determined from the difference between total and light fraction C. Total SOC was calculated on an equivalent mass basis (900 Mg ha–1 to 7.5 cm, 2010 Mg ha–1 to 15 cm) (Ellert and Bettany, 1995).
Statistical analysis of fertilizer treatment effects was conducted separately for each crop rotation using the Proc Mixed procedure of SAS (SAS Institute, 2001), with fertilizer treatment and phase as fixed effects and block as random effect. Contrasts were used to evaluate N and P fertilizer effects. Statistical analysis of rotation effects was conducted using the Proc Mixed procedure of SAS (SAS Institute, 2001), with crop rotation as fixed effect and block as random effect.
|
RESULTS
|
|---|
Precipitation and Crop Yields
The 6-yr period from 1998 to 2003 consisted of 3 yr of average moisture conditions, 2 yr of drought (2000 and 2001), and 1 yr with above-average moisture conditions (2002) (Table 2
). Overwinter precipitation (September–April) was below the long-term mean each year, and averaged only 40% of the long-term mean. Growing-season precipitation ranged from 64 to 263 mm, compared to the long-term mean of 181 mm. Crop yields were substantially reduced in drought years. Fertilizer responses were also reduced by drought, particularly in 2001.
View this table:
[in this window]
[in a new window]
|
Table 2. Precipitation and average yields of stubble wheat, fallow wheat and grass during the second 6-yr period (1998–2003) of the long-term crop rotation study at Bow Island, AB.
|
|
Compared to the first 6-yr period of this study, the average yield in the FW rotation declined 12% during the second 6-yr period, but with no change in fertilizer response (Fig. 1a
). In comparison, the average yield in the W treatment declined 27%, while the average yield gain due to fertilizer application (as a percentage of unfertilized yield) increased from 39 to 71%. The average yield in the G treatment declined 44%, while the average yield gain due to fertilizer application increased from 33 to 108%.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 1. First 6-yr (1992–1997) vs. second 6-yr (1998–2003) effects of rotation and fertilizer treatment on (a) average rotation yield (grain or hay averaged over both cropped and noncropped phases), (b) SOC, and (c) light fraction C. Error bars are standard errors.
|
|
The fertilizer response in the FW rotation was due to both N and P application, with the highest grain yield when both nutrients were applied (Table 3
). Average grain yield in the 40–20 treatment was identical to that in the 20–20 treatment. In contrast, the fertilizer response in the W treatment was almost entirely due to N application. Fertilizer application (40–20 vs. 0–0) increased average yields by 20% in the FWW and FXW rotations, similar to the benefit in the FW rotation. The low rotation yield in the FXW rotation was due to low flax yields.
View this table:
[in this window]
[in a new window]
|
Table 3. Effect of crop rotation and fertility treatments on average rotation yield from 7 to 12 yr (1998–2003) and soil organic carbon (SOC) after 12 yr (2003) of the long-term crop rotation study at Bow Island, AB.
|
|
The highest grain yield in the 0–0 treatment was obtained in the LW rotation (Table 3). The average yield for wheat in this treatment was 2.0 Mg ha–1 yr–1, while pea had an average yield of 1.6 Mg ha–1 yr–1. Wheat yields in this rotation were similar to that of the W treatment receiving 40 kg fertilizer N ha–1 yr–1. Nitrogen fertility in this rotation was increased by 3 yr of annual legume green manures from 1992 to1997 and rotation with annual legumes harvested for seed from 1998 to 2003 (Table 1). Application of P fertilizer increased wheat yields by 13% and pea yields by 4%.
Soil Organic Carbon
The effects of fertilizer treatments on SOC were small for all but the G treatment (Table 3). Phosphorus application increased total SOC to 15 cm by 0.7 Mg C ha–1 (P = 0.07) in the FW rotation, but did not affect SOC in the W treatment. Nitrogen application increased SOC by 1.2 Mg C ha–1 (P = 0.02) in the W treatment, but did not affect SOC in the FW rotation. Fertilizer application had no effect on total SOC to 15 cm in the FWW, FXW, or LW rotations. In contrast, fertilizer application increased SOC by 2.9 Mg C ha–1 (P = 0.01) in the G treatment. This increase in SOC was entirely due to increased light fraction C (Fig. 2
), which, based on visual observation, consisted largely of root fragments in this treatment. Heavy fraction C was not increased by fertilizer application in the G treatment.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2. Difference in light and heavy fraction C to 7.5 cm after 12 yr between selected treatments (upper named treatment less lower named treatment, for example, Wall- FWall). Percentages are the proportion of the total difference accounted for by the difference in light fraction C.
|
|
Fallow-containing rotations had the least SOC and light fraction C, with little difference among rotation type (Table 3). Compared to the FW rotation, total SOC to 15 cm was 1.4 Mg C ha–1 higher in the W treatment and 1.0 Mg C ha–1 higher in the LW rotation. Accumulation of light fraction C accounted for 68% of the increase in SOC caused by eliminating fallow (average of W and LW rotations) (Fig. 2). Heavy fraction C was unaffected by elimination of fallow (Table 3). For monoculture wheat rotations (W, FW, FWW), total SOC to 15 cm was significantly correlated (r = 0.71, P < 0.01) with average grain yield (1992–2003). On average, total SOC increased 2.1 Mg C ha–1 for each Mg ha–1 yr–1 increase in average grain yield.
The G treatment had the most SOC and light fraction C (Table 3). Compared to the FW rotation, the G treatment increased SOC to 15 cm by 3.5 Mg C ha–1 in the unfertilized treatment and 5.5 Mg C ha–1 in the fertilized treatment. Accumulation of light fraction C accounted for 86% of the increase in SOC due to elimination of fallow (average of unfertilized and fertilized rotations) (Fig. 2).
The change in SOC between 6 and 12 yr was negligible (
0.5 Mg C ha–1) in all treatments except the fertilized G treatment (Fig. 1b). In the fertilized G treatment, total SOC to 15 cm increased by 3.5 Mg C ha–1 and light fraction C to 7.5 cm increased by 3.6 Mg C ha–1 between 6 and 12 yr (Fig. 1bc). Heavy fraction C was unchanged between 6 and 12 yr in all treatments (data not presented).
|
DISCUSSION
|
|---|
Our findings indicate that SOC responds quickly to changes in management in semiarid Canadian prairie soils, but that these changes may be largely complete within the first decade.
Reduced C inputs during the drought years of 2000 and 2001 may have contributed to the lack of SOC change after Year 6. Based on the change in average grain yield and the slope between average grain yield and SOC, drought may have reduced the gain in SOC due to elimination of fallow during Years 7 to 12 by 0.8 Mg C ha–1. However, this effect is counteracted by the reduction in decomposition rate due to drought (Andrén et al., 1993). Periods of drought are normal for this region.
Prior history may have also contributed to the lack of SOC change after 6 yr. Irrigated cereal production during the 20 yr before initiation of this study likely increased total and labile SOC (Entry et al., 2002), and thus the duration of SOC change was probably less than if the experiment had been initiated on land under long-term fallow-wheat management.
The pattern of SOC change in our study was typical for long-term studies on the Canadian prairies. In studies conducted at Swift Current, SK, elimination of fallow increased total SOC by approximately 4 Mg C ha–1, but with no significant difference in SOC gain for sampling dates ranging from 4 to 36 yr (Campbell et al., 1995, 2000a, 2000b). In an 80-yr study conducted on a Typic Boroll at Lethbridge, AB, fallow-induced differences in SOC concentration were generally highest at the earliest sampling dates (Monreal and Janzen, 1993). In a 41-yr study conducted on the same soil, the gain in SOC due to elimination of fallow increased between 3 and 16 yr, but remained constant thereafter (Bremer et al., 1995).
Our results also suggest that the duration of SOC gain depends on the degree of management change, and hence, the amount of SOC eventually stored. This is seen most clearly in the grass treatment: when yields (and C inputs) are limited by nutrients, SOC gains appear to cease after 6 yr, but when supplied with fertilizer, the SOC gains in the second 6-yr period are comparable to those in the first 6 yr.
The management-induced differences in SOC in this study occurred primarily within the light fraction. Similarly, gains in light fraction C accounted for 51% of the gain in SOC due to elimination of fallow and 79% of the gain in SOC due to establishment of native grass in a Typic Haploboroll at Lethbridge, AB (Bremer et al., 1994). In a Typic Haploboroll at Scott, SK, light fraction accounted for a smaller fraction (45%) of the difference in SOC between grassland and cultivated systems than this study, but coarse roots and identifiable plant residues had been excluded from samples (Malhi et al., 2003).
The dominant role of light fraction material for management-induced change in SOC indicates that the dynamics of incompletely decomposed organic residues are the primary driver for SOC change in semiarid Canadian prairie soils (although not all light fraction material consists of incompletely decomposed organic residues). After conversion of cultivated land to grassland, much of the gain in SOC likely consists of live and dead roots (Garten and Wullschleger, 1999). Light fraction material is also concentrated near the soil surface (Malhi et al., 2003), and thus reduced soil mixing under perennial crops or continuous cropping may favor the accumulation of incompletely decomposed organic residues near the soil surface. If the accumulated SOC from C-conserving management practices is mostly in roots and incompletely decomposed organic residues, that implies that the carefully-sequestered C may not be permanently stored–that it is vulnerable to rapid loss under changed practices or conditions in the future.
Changes in labile soil C pools may eventually be reflected in stable C pools. A fraction of SOC decomposes very slowly, if at all, due to chemical and physical characteristics (Jenkinson, 1990). A small fraction of incoming C inputs must be stabilized within these fractions. However, the limited evidence for long-term SOC change (Bremer et al., 1995; Campbell et al., 2000a; Monreal and Janzen, 1993) implies that the proportion of incoming C inputs stabilized is very small and that changes in stable SOC may only be detectable at time-scales of centuries or greater. Statistical power to detect temporal changes in SOC gain in long-term agro-ecosystem studies is often low due to intrinsic variability and limited replication.
Our study supports the following conclusions: (i) that much of the SOC gain in response to improved practices on semiarid prairie soils occurs within about a decade; (ii) that the duration of SOC gain may be proportional to the magnitude of the gain (i.e., the magnitude of the practice change); and (iii) that much of the SOC gain occurs in labile fractions that are vulnerable to future losses. These conclusions, if confirmed by further study, may constrain somewhat the role of these soils as C sinks for mitigating increases in atmospheric CO2. Further monitoring of this and other studies is required, however, to determine whether the apparent levelling off of SOC gains is a short-term phenomenon, perhaps reflecting current cropping conditions, or a robust, persistent temporal pattern.
|
ACKNOWLEDGMENTS
|
|---|
Allan Middleton and Pat Pfiffner maintain and monitor the long-term crop rotation study at Bow Island, AB. Nancy Lee conducted the bulk of laboratory analysis, with assistance with total C and N analyses by Clarence Gilbertson. Funding for this study was provided by the Conservation and Development Branch, Alberta Agriculture and Food.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication September 7, 2007.
|
REFERENCES
|
|---|
- Andrén, O., K. Rajkai, and T. Kätterer. 1993. Water and temperature dynamics in a clay soil under winter wheat: Influence on straw decomposition and N immobilization. Biol. Fertil. Soils 15:1–8.[CrossRef]
- Bremer, E., B.H. Ellert, and H.H. Janzen. 1995. Total and light-fraction carbon dynamics during four decades after cropping changes. Soil Sci. Soc. Am. J. 59:1398–1403.[Abstract/Free Full Text]
- Bremer, E., H. Janzen, and A. Johnston. 1994. Sensitivity of total, light fraction and mineralizable organic matter to management practices in a Lethbridge soil. Can. J. Soil Sci. 74:131–138.
- Bremer, E., H.H. Janzen, and R.H. McKenzie. 2002. Short-term impact of fallow frequency and perennial grass on soil organic carbon in a Brown Chernozem in southern Alberta. Can. J. Soil Sci. 82:481–488.
- Campbell, C.A., B.G. McConkey, R.P. Zentner, F.B. Dyck, F. Selles, and D. Curtin. 1995. Carbon sequestration in a Brown Chernozem as affected by tillage and rotation. Can. J. Soil Sci. 75:449–458.
- Campbell, C.A., R.P. Zentner, B.-C. Liang, G. Roloff, E.C. Gregorich, and B. Blomert. 2000a. Organic C accumulation in soil over 30 years in semiarid southwestern Saskatchewan–Effects of crop rotations and fertilizers. Can. J. Soil Sci. 80:179–192.
- Campbell, C.A., R.P. Zentner, F. Selles, V.O. Biederbeck, B.G. McConkey, B. Blomert, and P.G. Jefferson. 2000b. Quantifying short-term effects of crop rotations on soil organic carbon in southwestern Saskatchewan. Can. J. Soil Sci. 80:193–202.
- Ellert, B.H., and J.R. Bettany. 1995. Calculation of organic matter and nutrients stored in soils under contrasting management regimes. Can. J. Soil Sci. 75:529–538.
- Ellert, B.H., H.H. Janzen, A.J. VandenBygaart, and E. Bremer. 2006. Measuring change in soil organic carbon storage. p. 25–38. In M.R. Carter and E.G. Gregorich (ed.) Soil sampling and methods of analysis. 2nd ed. CRC Press, Boca Raton, FL.
- Entry, J.A., R.E. Sojka, and G.E. Shewmaker. 2002. Management of irrigated agriculture to increase organic carbon storage in soils. Soil Sci. Soc. Am. J. 66:1957–1964.[Abstract/Free Full Text]
- Garten, C.T., Jr., and S.D. Wullschleger. 1999. Soil carbon inventories under a bioenergy crop (switchgrass): Measurement limitations. J. Environ. Qual. 28:1359–1365.[Abstract/Free Full Text]
- Janzen, H.H., C.A. Campbell, S.A. Brandt, G.P. Lafond, and L. Townley-Smith. 1992. Light-fraction organic matter in soils from long-term crop rotations. Soil Sci. Soc. Am. J. 56:1799–1806.[Abstract/Free Full Text]
- Jenkinson, D.S. 1990. The turnover of organic carbon and nitrogen in soil. Philos. Trans. R. Soc. Lond. B Biol. Sci. 329:361–368.[CrossRef][ISI]
- Malhi, S.S., S. Brandt, and K.S. Gill. 2003. Cultivation and grassland type effects on light fraction and total organic C and N in a dark Brown Chernozemic soil. Can. J. Soil Sci. 83:145–153.
- Monreal, C.M., and H.H. Janzen. 1993. Soil organic-carbon dynamics after 80 years of cropping a Dark Brown Chernozem. Can. J. Soil Sci. 73:133–136.
- SAS Institute. 2001. SAS/STAT, Release 8.01. SAS Inst., Cary, NC.
- VandenBygaart, A.J. 2006. Monitoring soil organic carbon stock changes in agricultural landscapes: Issues and a proposed approach. Can. J. Soil Sci. 86:451–463.
TOP
ABSTRACT
INTRODUCTION
