SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 May 2008
Published in Soil Sci Soc Am J 72:605-612 (2008)
DOI: 10.2136/sssaj2007.0251
© 2008 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Agricola
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Related Collections
Right arrow Ecosystem Management
Right arrow Soil Organic Matter
Right arrow Biogeochemical Processes

SOIL BIOLOGY & BIOCHEMISTRY

Soil Carbon Saturation Controls Labile and Stable Carbon Pool Dynamics

S. Guldea,b, H. Chunga, W. Amelungc, C. Changd and J. Sixa,*

a Dep. of Plant Sciences, Univ. of California, Davis, CA 95616
b Institute of Geography, Univ. of Bonn, Germany
c Institute of Crop Science and Resource Conservation, Soil Science and Soil Ecology, Univ. of Bonn, Germany
d Lethbridge Research Center, Agriculture and Agri-Food Canada, Lethbridge, AB, Canada T1J 4B1

* Corresponding author (jwsix{at}ucdavis.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Recently, it has been suggested that soil organic C (SOC) does not always respond linearly to increasing C input, thereby limiting the rate and efficiency of C stabilization in soils. Therefore, we postulated that when a soil is exposed to a broad range of C inputs through a range of manure treatments, it will show C saturation behavior and different SOC pools will saturate at different rates. To test this, different SOC pools were isolated by physical fractionation techniques from a long-term agricultural experiment in Lethbridge, Canada. In this experiment, manure has been applied since 1973 at rates of 0, 60, 120, and 180 Mg ha–1 yr–1 (wet weight). In the total mineral soil as well as the small macroaggregates (250–2000 µm), microaggregates (53–250 µm), and the silt plus clay fraction (<53 µm), an increase was observed in SOC contents with an increase in manure application rate to 120 Mg ha–1 yr–1. However, no additional C was sequestered when the manure application rate was augmented to 180 Mg ha–1 yr–1, indicating C saturation in these SOC pools. Large macroaggregates (>2000 µm) were the only water-stable aggregate fraction that increased in C content across all manure input levels. Further physical separation of macroaggregates into subpools by microaggregate isolation showed that coarse (>250 µm) particulate organic matter (POM) was the fraction that accounted most for the increase in C content of the large macroaggregates. Furthermore, the turnover of large macroaggregates increased with increasing manure applications, as indicated by decreased formation and stabilization of intramicroaggregate POM within the large macroaggregates. We conclude that as C input increases, the mineral fraction of a soil saturates and consequently additional C input will only accumulate in labile soil C pools that have a relatively faster turnover.

Abbreviations: cPOM, coarse particulate organic matter • fine iPOM, fine intramicroaggregate particulate organic matter • LF, light fraction • POM, particulate organic matter • SOC, soil organic carbon • SOM, soil organic matter


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils are an important reservoir for C, storing three times the amount of C residing in plants (Schlesinger 1990). Increasing soil organic matter (SOM) can benefit multiple agroecosystem services, including mitigation of CO2 emissions to the atmosphere, reduction of erosion, and functioning as a nutrient reservoir (Paustian et al., 1997b; Robertson and Swinton, 2005). In light of atmospheric CO2 enrichment and degradation of soil quality as a result of intensive management practices, research efforts have been focused on identifying agroecosystem management practices that can increase soil C sequestration and restore soil quality (Paustian et al., 1997b). The potential of soils to sequester C is heavily debated, however, and might be limited (Six et al., 2002; West and Six, 2007). If soils have a finite capacity to store C, modeling efforts aimed at determining the rate and the extent to which soils could act as C sinks have to take this limit into account, especially for high-C soils (Six et al., 2002).

Whole SOC is commonly separated into labile (active) and stable (passive) pools (Parton et al., 1987). Labile SOM pools are characterized by a rapid turnover, mainly consist of young SOM, and are sensitive to land management and environmental conditions (Parton et al., 1987). Due to these characteristics, labile SOM pools play an important role in short-term C and N cycling in terrestrial ecosystems (Schlesinger, 1990). The most commonly isolated labile C pools are the light fraction (LF) and particulate organic matter (POM). The LF consists mostly of mineral-free, partly decomposed plant debris (Spycher et al., 1983) and appears in soils as free POM (Golchin et al., 1994; Puget et al., 1995; Six et al., 1998). When free POM is further decomposed, it can be incorporated into aggregates, where its decomposition is restricted (Golchin et al., 1994; Puget et al., 1995). In this case, POM becomes part of a less labile pool, but is easily decomposable when it is set free again; hence the degree of protection of labile C is dependent on aggregate turnover. Stabilization of occluded organic matter is greatest when aggregate stability is high and aggregate turnover is slow (Six et al., 1998).

For soils to act as a C sink, however, organic C needs to be stabilized in stable C pools (Paustian et al., 1997a). Of the stable C pools, SOC protected within aggregates is sensitive to land management practices such as manure application, tillage, or crop rotation (Angers and Carter, 1996). Therefore, the amount of C stabilized in protected SOC pools is critical for the determination of the extent to which soils can operate as a global C sink under specific management conditions. Due to different protection mechanisms, the degree and the duration of stabilization of SOC within macroaggregates and microaggregates differ (Tisdall and Oades, 1982). Macroaggregates contain C that (i) functions as transient binding agents holding microaggregates together and (ii) is occluded within microaggregates, which results in greater absolute C contents of macroaggregates than microaggregates (Tisdall and Oades 1982; Elliott 1986). Microaggregates have a lower C storage potential, but as they sequester C in the long term (Skjemstad et al., 1990; Six et al., 2000), the degree of stabilization is greater.

Increased organic matter input to soil through management practices such as animal or green manure application and straw incorporation increases the SOM content (Mikha and Rice, 2004) and enhances soil aggregation (Sommerfeldt and Chang, 1985). Cattle manure largely consists of coarse particles of organic materials (i.e., POM; Aoyama et al., 1999) and the decomposition of this POM results in an increase in microbial activity and production of transient binding agents (Golchin et al., 1994; Puget et al., 1995; Six et al., 1998). Therefore, animal manure application has a positive impact on aggregation (Aoyama et al., 1999), which may lead to a greater content of physically protected SOC.

Most SOC models assume a linear increase in C content with C input, and thus C sequestration can continue regardless of the amount of organic C already contained in each SOC pool (Paustian et al., 1997b). These process-oriented models are based on first-order kinetics in which the decomposition rate is defined as a constant. They have been successfully applied to many agricultural soils, especially to those with low to moderate SOC contents (Paustian et al., 1997b). In contrast, other experiments with soils rich in C have not shown any further increase in SOC following an enhanced C input (Campbell et al., 1991; Solberg et al., 1997) and have been more difficult to model (Paustian et al., 1997b). These findings suggest that there exists a soil C saturation level (Six et al., 2002). Hassink (1997) assumed that the protective capacity of soil to store organic C is limited and that the rate of additional C sequestration depends on the degree to which the protective capacity of the silt plus clay fraction has been reached. Based on this concept, Six et al. (2002) further developed a conceptual SOM model of C saturation that includes physical, chemical, and biochemical protection mechanisms. They defined functional C pools that are measurable and argued that the C sequestration in each of these pools depends on their saturation deficit (i.e., the difference between saturation level and the actual SOC content).

The objective of this study was to test whether measurable soil C pools indeed show saturation behavior in a soil rich in C under a broad range of C inputs. We examined C sequestration in different soil aggregate fractions in a long-term experiment with animal manure applications between the rates of 0 to 180 Mg ha–1 yr–1 in southern Alberta, Canada. The experiment was established in 1973, and its long duration as well as its high and wide-ranging C input rates makes this field experiment well suited for the study of the concept of C saturation. We hypothesized that manure application enhances the amount of C stored in physically protected C pools. Furthermore, we hypothesized that at high manure application rates, a limit is reached where the C content of these protected C pools does not increase with additional C input, i.e., saturation occurs in stable C pools, and additional C is increasingly allocated to labile C pools.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Site
The experimental site is situated at the Lethbridge Research Center in southern Alberta, Canada. The soil is a well-drained Dark Brown Chernozemic clay loam and has been classified as a Typic Haplustoll (Hao et al., 2003). Originally, this long-term experiment was established to examine the effects of increasing amounts of animal manure on soil properties, crops, and groundwater (Sommerfeldt and Chang, 1985). The experimental design has been described in detail by Sommerfeldt and Chang (1985). Since 1973, solid beef cattle manure, which was stockpiled for 1 yr, has been applied every fall after harvest. The plots have been cultivated under irrigation (on average 148 mm yr–1) and conventional tillage. Mean annual precipitation is 401 mm and the mean annual temperature is 5.2°C. Barley (Hordeum vulgare L.) was grown from 1973 until 1995, canola [Brassica rapa L. subsp. oleifera (DC.) Metzg.] in 1996, corn (Zea mays L.) from 1997 to 1999 (Hao et al., 2003), and then barley until 2005. Manure was applied at four different rates of 0, 60, 120, and 180 Mg ha–1 yr–1 (wet weight; n = 3). These rates correspond to the recommended manure application rate (60 Mg ha–1 yr–1) for irrigated agricultural fields in this part of Canada, plus two and three times this rate (Sommerfeldt and Chang 1985). The manure originated from the same feedlot every year, but the quality of the manure varied. On average, it contained 175 g C kg–1 (dry weight manure; Hao et al., 2003). Manure was the dominant source of soil C input in this experiment (Table 1 ).


View this table:
[in this window]
[in a new window]

  		Table 1. Soil C input, manuring soil organic C (SOC) sequestration efficiency, and soil physical characteristics. One standard error is shown in parentheses.

 
Soil Sampling
The soil samples were taken in August 2005 from experimental plots under the four different manure application rate treatments (4 treatments x 3 replicates = 12 plots). Two soil cores of 15-cm depth and 5.5-cm diameter were taken from each plot. Bulk densities were determined for the individual soil cores and for further analysis, the soil cores from each plot were composited. All soil samples were passed through an 8-mm sieve. Before air drying, large roots and plant residues were removed manually.

Water-Stable Aggregate Fractionation
The separation of four aggregate size classes was performed according to Elliott (1986) by wet sieving (Fig. 1 ). Before sieving, an 80-g subsample of bulk soil was slaked by submerging it in deionized water on top of a 2000-µm sieve for 5 min at room temperature. Water-stable aggregates were then separated by moving the sieve up and down 50 times in 2 min, and the aggregates were collected in an aluminum pan. The remaining soil and water were passed through the 250- and 53-µm sieves, undergoing the same sieving procedure as described above. After oven drying at 60°C, the four aggregate size fractions, namely the large macroaggregates (>2000 µm), small macroaggregates (250–2000 µm), microaggregates (53–250 µm), and the silt plus clay fraction (<53 µm), were weighed. The mean weight diameter was calculated after Van Bavel (1949) to obtain an index for soil stability.


Figure 1
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 1. Physical fractionation scheme: LM = large macroaggregates (>2000 µm), SM = small macroaggregates (250–2000 µm), m = microaggregates (53–250 µm), S&C = silt plus clay fraction (<53 µm), LF = light fraction, HF = heavy fraction, mM = microaggregates within macroaggregates (53–250 µm), cPOM = coarse particulate organic matter (>250 µm), fPOM = fine intermicroaggregate POM (<250 µm), fine iPOM = fine intramicroaggregate POM (53–250 µm).

 
Microaggregate Isolation
The large and small macroaggregates were further fractionated to isolate the microaggregates within the macroaggregates (Fig. 1), following the method of Six et al. (2000). Subsamples of 10 g were slaked for 30 min and then placed on top of a 250-µm mesh screen, immersed in deionized water, and shaken with 50 metal beads (4 mm in diameter) to disrupt all macroaggregates. A continuous flow of water flushed all material <250 µm onto a 53-µm sieve and only coarse POM (cPOM) and sand remained on the 250-µm sieve. The material on top of the 53-µm sieve was wet sieved, as described above, to separate the water-stable microaggregates within macroaggregates from the silt plus clay fraction. All fractions were oven dried at 60°C.

Density Flotation for Light Fraction and Particulate Organic Matter Isolation
The LF was isolated by density flotation (Fig. 1) based on the method of Six et al. (1998). A 3-g subsample of each aggregate size fraction obtained by the water-stable aggregate fractionation was suspended in 45 mL of sodium polytungstate (1.85 g cm–3) and shaken slowly by hand to mix it without breaking up the aggregates. To evacuate air entrapped within aggregates, the samples were put under vacuum for 10 min. Subsequently, they were centrifuged (1250 x g) for 1 h at 20°C. The floating material was considered LF and collected by aspirating onto a 20-µm nylon filter. The LF was rinsed with deionized water to remove any remaining sodium polytungstate and dried at 60°C. The heavy fraction consisted of material that did not float and was likewise dried after removing the sodium polytungstate through three rinses with water. In addition, the fractions obtained from the microaggregate isolation were further separated by density flotation as described above, with the exception of using 2.3 g cm–3 sodium polytungstate solution to separate cPOM from the associated sand (Fig. 1).

For microaggregates occluded within macroaggregates, after floating off the fine intermicroaggregate POM, the fine POM inside the microaggregates within macroaggregates was isolated from the heavy fraction (Six et al., 1998). The heavy fraction was dispersed by shaking with 0.5% sodium hexametaphosphate for 18 h and passed through a 53-µm sieve to collect the silt plus clay fraction (Fig. 1). The material remaining on the sieve consisted of fine intramicroaggregate POM (fine iPOM) and sand.

Calculation of Light-Fraction-Free Weights and Carbon Contents
If LF materials are substantial or even comparable in weight to the aggregate fractions themselves, it is necessary to correct the isolated SOC pools as well as the total mineral soil for both the weight and the C contents of the LF.

Water-Stable Aggregates
The weight proportions of water-stable aggregates were expressed on a LF-free basis as follows:

Formula
where Pn is the aggregate fraction in the total soil (%), Fn is the aggregate fraction (g), LFn is the LF in the aggregate fraction (%), and n = 1 for large macroaggregates, n = 2 for small macroaggregates, and n = 3 for microaggregates. To calculate the LF-free C concentration in each fraction (g kg–1 aggregate fraction), the C contained in the LF was subtracted. Furthermore, the values were based on a per-kilogram LF-free aggregate basis by dividing the proportion of mineral soil contained in every aggregate fraction:

Formula
where Cn is the C content of the aggregate fraction (g kg–1 aggregate fraction), and CLFn is the C content of the LF associated with the aggregate fraction (g kg–1 LF fraction).

Total Soil
The LF-free soil C (g kg–1 mineral soil) was calculated following the same principle, but the aggregate size distribution was taken into account:

Formula
where Csoil is the C content of the bulk soil (g kg–1 soil).

Coarse Particulate Organic Matter
The weight of cPOM (g kg–1 macroaggregates) was corrected for the proportion of LF contained in macroaggregates:

Formula
where Wcn is the cPOM in the macroaggregate fraction (g).

The microaggregates-within-macroaggregates fraction does not contain any LF but only fine iPOM and fine intermicroaggregate POM. These two POM fractions together constitute the fine POM. The C associated with the mineral soil in the microaggregates-within-macroaggregates fraction was determined according to Denef et al. (2004).

Carbon Input Calculations
Estimates of crop-derived C input were needed to calculate the C sequestration efficiency. The efficiency of C sequestration in manured plots was obtained by dividing the difference in SOC levels of manured and control plots by cumulative C input in the manured plots. Empirical equations relating yield to aboveground and belowground residues were used to estimate crop-residue-derived C inputs (see Kong et al., 2005).

Texture Analysis
In the beginning of the experiment, the soil of the experimental site contained 38% sand and 31% clay. The soil of the feedlot was less sandy and consisted of 20% sand and 45% clay (Gao and Chang, 1996). Inclusion of significant amounts of feedlot soil in the applied manure may cause confounding effects when the texture of the experimental site is affected. Therefore, it was necessary to determine the shift in texture by particle size analysis so that the amount and C content of isolated fractions could be corrected for sand content. Texture analyses were performed according to routine protocols, involving SOM oxidation with H2O2, chemical dispersion with sodium hexametaphosphate, sieving for assessing the sand content, and the hydrometer method (International Organization for Standardization, 1998, Method 11277) for assessing the clay content. All sand-free fractions (g) were then calculated by subtraction of the respective sand content. The amount of microaggregates within macroaggregates was corrected for sand according to Six et al. (2000), and the C contents (g kg–1 sand-free soil or fraction) were determined by dividing by the proportion of sand-free mineral soil or fraction (Six et al., 1998).

Carbon Analysis
Before C analysis, inorganic C was removed by acid fumigation from all fractions following the method developed by Harris et al. (2001). Ground soil samples of 15 mg were weighed in silver foil capsules. After wetting with 25 µL deionized water, the samples were placed in a vacuum desiccator together with a beaker containing 100 mL of concentrated (12 mol L–1) HCl. After 6 h, the samples were taken out and dried at 60°C. Carbon contents were determined using a Carlo Erba (Milan, Italy) NA 1500 elemental analyzer.

Statistical Analyses
We analyzed the soil physical and chemical properties using ANOVA for a randomized block design. The fixed effects in the model were block and manure application treatment. Tukey's honestly significant difference test was used as a post hoc mean separation test (P < 0.05). All statistical tests were performed using SAS 9.1 (SAS Institute, Cary, NC).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Aggregation
Under no manure applications, large and small macroaggregates together accounted for 56 ± 4% of the total soil weight (Fig. 2a ). Different rates of manure applications only slightly affected the aggregate distribution. The proportion of small macroaggregates tended to increase with manure additions (significant only for the 120 Mg ha–1 yr–1 treatment), and the proportions of microaggregates and the silt plus clay fraction tended to decrease when manure was added at 60 Mg ha–1 yr–1, but there was no further decrease at greater application rates (Fig. 2a). Even with significant inputs of clay minerals from the feedlot plot (Table 1), higher manure input did not promote aggregate formation, and the mean weight diameter of the aggregate fractions did not significantly differ between manure application rates (P > 0.05, data not shown).


Figure 2
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 2. (a) Aggregate size distribution and (b) C content of water-stable aggregates and total mineral soil organic C stocks across four manure application rate treatments: LM = large macroaggregates (>2000 µm), SM = small macroaggregates (250–2000 µm), m = microaggregates (53–250 µm), S&C = silt plus clay fraction (<53 µm). Different letters indicate a significant difference between treatment means of manure application rates (Tukey's test, P < 0.05). Error bars indicate one standard error (n = 3).

 
As manures are rich in LF, the weight proportion of LF increased with manure application rates in both bulk soil and aggregate size fractions. The proportion of LF associated with microaggregates was lower than that associated with macroaggregates, but it also increased with greater manure application rates (P < 0.05, Table 2 ).


View this table:
[in this window]
[in a new window]

  		Table 2. Proportion of light fraction in total soil and aggregate size fractions of large (>2000 µm) and small (250–2000 µm) macroaggregates and microaggregates (53–250 µm). One standard error is shown in parentheses.

 
Soil Organic Carbon Stabilization
Total Soil and Water-Stable Aggregates
Overall, less than one-fifth of the added C has been sequestered in the soil during the experimental period (Table 1). The C content of the mineral soil increased significantly with manure application but no significant difference was observed between the two greatest manure application rates (Fig. 2b). Large macroaggregates was the only fraction with significantly increasing C concentration across all treatments; the other isolated fractions did not show a significant C increase between the two (small macroaggregates, P > 0.05) or not even between the three greatest manure application rates (microaggregates, silt plus clay, P > 0.05; Fig. 2b).

Subpools of Macroaggregates
According to their difference in C sequestration dynamics, the subpools of the small and large macroaggregates also responded differently to manure application. For both small and large macroaggregates, the proportion of occluded microaggregates increased when manure was applied, but decreased again for the 180 Mg ha–1 yr–1 treatment, while the proportion of cPOM increased across all treatments for the large macroaggregates and up to the 120 Mg ha–1 yr–1 treatment for the small macroaggregates (P < 0.05, data not shown). The C content of the silt plus clay fraction of both macroaggregate fractions did not change significantly with increasing manure addition, suggesting that the added fine particles from the feedlot plots did not comprise higher SOM content. In addition, the C contents of microaggregates occluded in large and small macroaggregates were not significantly different between the two greatest C input rates (Fig. 3a ). In contrast, the cPOM-C content of the large macroaggregates increased significantly from the 120 to the 180 Mg ha–1 yr–1 manure application rate (Fig. 3b).


Figure 3
View larger version (20K):
[in this window]
[in a new window]

  		Fig. 3. Carbon content of the fractions isolated by microaggregate isolation across four manure application rate treatments: (a) small macroaggregates (SM, 250–2000 µm) and (b) large macroaggregates (LM, >2000 µm); cPOM = coarse particulate organic matter (>250 µm), mM = microaggregates within macroaggregates (53–250 µm), S&C = silt plus clay fraction (<53 µm). Different letters indicate a significant difference between treatment means of manure application rates (Tukey's test, P < 0.05). Error bars indicate one standard error (n = 3).

 
Ratio of Fine Intramicroaggregate to Coarse Particulate Organic Matter
The ratio of fine iPOM to cPOM drastically decreased for large macroaggregates with increasing C inputs (Fig. 4a ). This decrease is in part due to the increased cPOM-C (Fig. 3b), but is also due to the decrease in fine iPOM (Fig. 4b) with increasing manure input. In contrast, the ratio of fine iPOM to cPOM did not change significantly for small macroaggregates across all manure application rates (Fig. 4a).


Figure 4
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 4. (a) Weight ratio of fine intramicroaggregate (53–250 µm) to coarse (>250 µm) particulate organic matter (fine iPOM/cPOM) and (b) weight proportion of fine iPOM across four manure application rate treatments; LM = large macroaggregates (>2000 µm), SM = small macroaggregates (250–2000 µm). Different letters indicate a significant difference between treatment means of manure application rates (Tukey's test, P < 0.05). Error bars indicate one standard error (n = 3).

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Saturation Behavior of Soil Organic Carbon Pools
Unlike linear increases in C content with greater C inputs, as observed by Gerzabek et al. (1997) or Huggins et al. (1998) after 37 and 120 yr of manure applications at Ultuna, Sweden, and Columbia, SC, respectively, we found no further increase in SOC level in response to greater C input. Therefore, the mineral soil in this experiment, which had a high initial SOC level, has probably reached C saturation levels at high C input rates. This is supported by the finding that the manuring efficiency (i.e., fraction of manure applied that is stabilized in the soil) is low, especially for the mineral SOC. Moreover, the additional input of clay particles from the feedlot plots did not result in higher C contents, suggesting that these minerals were also C saturated. Several studies have found a limited capacity of C sequestration for the silt plus clay fraction (Hassink, 1997; Diekow et al., 2005).

Except for the large macroaggregates, none of the water-stable aggregate fractions showed extra C stabilization at the greatest manure application rates. Hence, our results indicate that the assumption of C saturation behavior is valid for physically protected C pools. Since large macroaggregates comprise only ~20% of the whole soil and are not the dominant aggregate size fraction, they do not determine the C sequestration behavior of the bulk soil. Furthermore, the microaggregates within macroaggregates and the silt plus clay fraction occluded within the large macroaggregates indicate C saturation and only the C content of cPOM continued to increase for all treatments. Thus, cPOM-C is the only fraction responsible for the ongoing increase in C content of large macroaggregates with elevated manure applications.

The different trends in C sequestration of small and large macroaggregates can be fully explained by distinct C sequestration potentials of their subfractions. Several studies confirm preferential C sequestration in microaggregates within macroaggregates on increased soil C input (Kong et al., 2005) or after conversion from conventional tillage to no-till (Six et al., 2000; Denef et al., 2004). This is in line with our observation that inside small macroaggregates, additional C is mostly sequestered in the microaggregates-within-macroaggregates fraction. Furthermore, our study suggests that all the subfractions of small macroaggregates reach C saturation at the greatest manure application rates. In contrast to the small macroaggregates where the microaggregates within macroaggregates was the greatest C pool, cPOM was the greatest C pool in large macroaggregates at maximum C input rates but did not saturate.

Hierarchical Carbon Saturation?
Our data (Fig. 5a ) suggest that an asymptotic model is a better fit than the a linear model relating C input to C stabilized at equilibrium, i.e., the total mineral SOC resembles an asymptotic increase. When the mineral SOC pool is further separated into different aggregate fractions, it is evident that the smaller SOC pools saturate before the larger pools (Fig. 5b). Therefore, it seems that C saturation occurs in a hierarchical fashion.


Figure 5
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 5. Change in soil organic C contents in (a) total soil and (b) water-stable aggregate fractions across four manure application rate treatments: LF = light fraction, LM = large macroaggregates (>2000 µm), SM = small macroaggregates (250–2000 µm), m = microaggregates (53–250 µm), S&C = silt plus clay fraction (<53 µm). For every distribution, the curve with a greater R2 value was chosen (curve fitting was performed with Excel 2003).

 
The hierarchical model of aggregate formation developed by Tisdall and Oades (1982) has been continuously modified (e. g. Oades, 1984; Golchin et al., 1994), but appears to have general application for soils where OM is the main stabilizing agent for aggregates (Oades and Waters, 1991; Six et al., 2000). In such soils, C saturation is also expected to occur in a hierarchy of smaller sized to larger physically protected C pools (Kool et al., 2007). Our results support this concept, considering that the C content of large macroaggregates is still increasing at the greatest C input level while all smaller fractions show C saturation behavior. In addition, the C content of the silt plus clay fraction only increased significantly at the lowest manure application rate, and microaggregates and small macroaggregates up to the second lowest manure application rate (i.e., 120 Mg ha–1 yr–1). Therefore, it seems that C saturation follows the same hierarchical order as aggregate formation and that the large macroaggregates will eventually also saturate (Kool et al., 2007).

Changes in Macroaggregate Turnover
Aggregate turnover is critical for C sequestration dynamics, as it determines the degree of stabilization of occluded POM and mineral-associated C (Six et al., 2000; Denef et al., 2004). If aggregate turnover is slow, cPOM will be further decomposed, which will lead to an increase in the amount of fine iPOM (Six et al., 1998). Hence, the amount of fine iPOM and the ratio of fine iPOM to cPOM should increase with decreasing aggregate turnover and have been proposed as indicators for changes in aggregate turnover (Six et al., 2000).

The decline observed in the ratio of fine iPOM to cPOM, concomitant with the decrease in the amount of fine iPOM in large macroaggregates with increasing manure application rates, suggests that the turnover of large macroaggregates increases when the C saturation deficit of the soil decreases. Manure is mainly composed of coarse particles (Aoyama et al., 1999), thus leading to a selective input of cPOM to the soil. Consequently, more cPOM is incorporated into the large macroaggregates (Fig. 3b). Without an increase in the turnover of large macroaggregates, however, cPOM would be decomposed to fine iPOM and an increase in fine iPOM would be observed (Six et al., 2000). In contrast, we observed a decrease in fine iPOM content, suggesting an increase in the turnover of large macroaggregates that diminishes the stabilization of cPOM into fine iPOM. This can probably be explained by increased microbial activity with additional manure application. Microorganisms derive a majority of their energy from decomposing readily available plant-derived SOM, such as LF and POM (Six et al., 2002), but also from polysaccharides, which are a major component of the transient binding agents holding macroaggregates together (Tisdall and Oades, 1982). Consequently, the decomposition of polysaccharides by microorganisms will result in the destruction of large macroaggregates. The high microbial activity also induces a production of new transient binding agents, however, which leads to the formation of new large macroaggregates (Six et al., 1998; Puget et al., 1995; Golchin et al., 1994). As a result, large macroaggregates will have a faster turnover and will be the most dynamic physically protected C pool. In conclusion, our study suggests that the macroaggregate pool turns over faster when a soil is near its C saturation level.

Importance of Labile Carbon Pools when Mineral Soil Organic Carbon Pools Are Near Saturation
In addition to the increase in cPOM within the large macroaggregates, the LF was also augmented with increasing manure application rate. In the plots without manure applications, the LF constituted ~1.3% of the dry weight of the whole soil, which is comparable to the 0.25 to 2.39% LF (dry weight of whole soil) found by Janzen et al. (1992) in three agricultural sites in southern Saskatchewan. In contrast, the LF increased to 25% of the dry soil weight in the 180 Mg ha–1 yr–1 treatment. This clearly indicates that additional C input to soils close to C saturation results in an accumulation of labile LF. Figure 5a shows that C associated with the LF continued to increase, while C sequestration of the mineral soil showed saturation dynamics. Hence, with a decreasing saturation deficit of the stable C pools, the labile C pools gain more importance and account for an increasing proportion of total soil C. These greater proportions of the labile C combined with a faster turnover of large macroaggregates lead to a more dynamic SOM stock, which can function as an active nutrient sink and source reservoir. Consequently, only a soil close to its saturation level will hold a great reserve of relatively stable SOM and an active pool of relatively labile SOM. Consequently, the dilemma of conserving SOM vs. profiting from SOM cycling (sensu Janzen, 2006) is resolved in soils close to saturation.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Our study indicates that the capacity of chemical and physical protection of SOC can indeed be limited and C saturation can occur. Soil C saturation does, however, seem to follow a hierarchy from smaller to larger sized C pools. Our results also highlight the importance of separation of labile and stable C pools while testing C saturation in temperate ecosystems. If a soil shows increasing C content with increasing C input, the mineral soil might well be already saturated and only labile C pools continue to accumulate C. In this particular study, we found that the LF and cPOM occluded in large macroaggregates were the only fractions whose C content continued to increase with increased manure application, leading to a low efficiency of C sequestration by manuring.

The limited capacity of physically protected C pools to stabilize OM has important consequences for SOC models, because the well-accepted and most-often-used assumption of linearly increasing C content with greater C input must be rejected. Furthermore, the increasing contribution of labile C pools to total C content when the saturation deficit decreases is crucial for the SOM pool's capacity to continue functioning as a sink and source of nutrients. And maybe most importantly, as only stable C pools are relevant for long-term C stabilization, our results indicate that soils cannot act as unlimited C sinks.


   ACKNOWLEDGMENTS
 
Many thanks to Robert Rousseau, who always took time for helping out organizing the laboratory work. Funding for this project was provided by the Office of Science (BER), U.S. Department of Energy, Grant no. DE-FG02-04ER63912.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
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 July 6, 2007.


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




This article has been cited by other articles:


Home page
 Soil Sci.Home page
H. Chung, J. H. Grove, and J. Six
Indications for Soil Carbon Saturation in a Temperate Agroecosystem
Soil Sci. Soc. Am. J., June 18, 2008; 72(4): 1132 - 1139.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Agricola
Right arrow Articles by Gulde, S.
Right arrow Articles by Six, J.
Related Collections
Right arrow Ecosystem Management
Right arrow Soil Organic Matter
Right arrow Biogeochemical Processes

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome