Soil carbon saturation: Implications for measurable carbon pool dynamics in long-term incubations

Soil Biology & Biochemistry 41 (2009) 357–366

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Soil carbon saturation: Implications for measurable carbon pool dynamics in long-term incubations Catherine E. Stewart a, b, *, Keith Paustian b, c, Richard T. Conant b, Alain F. Plante b, d, Johan Six b, e a

Department of Geological Sciences, University of Colorado, Boulder, CO 80309-0399, USA Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80523-1499, USA c Department of Soil and Crop Sciences, Colorado State University, Fort Collins, CO 80523, USA d Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA 19104-6316, USA e Department of Plant Sciences, University of California, Davis, CA 95616, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 May 2008 Received in revised form 15 October 2008 Accepted 20 November 2008 Available online 6 December 2008

The efficiency of agricultural management practices to store SOC depends on C input level and how far a soil is from its saturation level (i.e. saturation deficit). The C saturation hypothesis suggests an ultimate soil C stabilization capacity defined by four SOM pools capable of C saturation: (1) non-protected, (2) physically protected, (3) chemically protected and (4) biochemically protected. We tested if C saturation deficit and the amount of added C influenced SOC storage in measurable soil fractions corresponding to the conceptual chemical, physical, biochemical, and non-protected C pools. We added two levels of 13Clabeled residue to soil samples from seven agricultural sites that were either closer to (i.e., A-horizon) or further from (i.e., C-horizon) their C saturation level and incubated them for 2.5 years. Residue-derived C stabilization was, in most sites, directly related to C saturation deficit but mechanisms of C stabilization differed between the chemically and biochemically protected pools. The physically protected C pool showed a varied effect of C saturation deficit on 13C stabilization, due to opposite behavior of the POM and mineral fractions. We found distinct behavior between unaggregated and aggregated mineralassociated fractions emphasizing the mechanistic difference between the chemically and physically protected C-pools. To accurately predict SOC dynamics and stabilization, C saturation of soil C pools, particularly the chemically and biochemically protected pools, should be considered. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Agroecosystem Carbon sequestration 13 C labeling Particulate organic matter Soil aggregation Soil incubation Soil carbon saturation Soil organic matter stabilization

1. Introduction

Abbreviations: SOC, soil organic carbon; POM, particulate organic matter; cPOM, coarse non-protected particulate organic matter (>250 mm); LF, fine non-protected POM (lighter than 1.85 g cm3, 53–250 mm); iPOM, microaggregate-protected POM (heavier than 1.85 g cm3, >53 mm in size); magg, microaggregate fraction, (53– 250 mm); mSilt, microaggregate-derived silt-sized fraction (heavier than 1.85 g cm3 2-53 mm); mClay, microaggregate-derived clay-sized fraction (heavier than 1.85 g cm3, <2 mm); NH-dSilt, non-hydrolyzable easily dispersed silt-sized fraction, (acid-resistant 53-2 mm); NH-dClay, non-hydrolyzable easily dispersed claysized fraction, (acid-resistant <2 mm); H-dSilt, hydrolyzable easily dispersed silt-sized fraction, (acid-soluble 53-2 mm); H-dClay, hydrolyzable easily dispersed clay-sized fraction, (acid-soluble <2 mm); NH-mSilt, non-hydrolyzable microaggregate-derived silt-sized fraction (acid-resistant 53-2 mm); NH-mClay, nonhydrolyzable microaggregate-derived clay-sized fraction (acid-resistant <2 mm); H-mSilt, hydrolyzable microaggregate-derived silt-sized fraction (acid-soluble 532 mm); H-mClay, hydrolyzable microaggregate-derived clay-sized fraction (acidsoluble <2 mm). * Corresponding author. Department of Geological Sciences, University of Colorado, Campus Box 399, 2200 Colorado Ave., Boulder, CO 80309-0399, USA. Tel.: þ1 303 735 4953; fax: þ1 303 492 2606. E-mail address: [email protected] (C.E. Stewart). 0038-0717/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.soilbio.2008.11.011

Conservation management practices that decrease soil disturbance and increase the amount of C added to soil generally increase SOC as well as soil fertility. As interest increases in both promoting organic C storage and alternative uses of crop residues, it is crucial to understand the relative stabilization efficiency of added residue C as well as its stability in the soil. Some long-term agroecosystems having treatments with varying C addition levels do not show increased equilibrium SOC stocks at higher C addition levels (Paustian et al., 1997; Huggins et al., 1998a,b; Reicosky et al., 2002), suggesting that soil C content becomes saturated with respect to C inputs at equilibrium (Six et al., 2002; Stewart et al., 2007). Other long-term agroecosystem experiments showed decreased SOC stabilization efficiency in high C compared to low C soils under the same treatments (Campbell et al., 1991; Nyborg et al., 1995). These results suggest that there is a limit to the stabilization of added C in soil (Six et al., 2002; Stewart et al., 2007) and that the further a soil is from saturation (i.e., the greater the saturation deficit), the greater the SOC storage potential (i.e., C saturation level minus current SOC content) (Hassink, 1996; Stewart et al., 2008b). The

358

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

hypothesis of saturation also implies that a greater saturation deficit will result in greater sequestration efficiency of added C (DSOC/DC input) (Stewart et al., 2008b). In contrast, a soil with a smaller saturation deficit (i.e. approaching saturation) will accumulate a lower amount of SOC at a lower efficiency (Hassink and Whitmore, 1997; Stewart et al., 2007, 2008b). As a soil approaches C saturation with increasing C inputs, we hypothesize that decreasing C stabilization efficiency in the whole soil is a function of the behavior of four C pools: the silt þ clayprotected pool (Hassink, 1996; Hassink et al., 1997), the soil microaggregate-protected pool (Six et al., 2002), the biochemically protected pool (Baldock and Skjemstad, 2000), and also a nonprotected C pool (Six et al., 2002). The dynamics and saturation level of the silt þ clay and physically protected C pools are mediated by texture and mineralogy. Biochemical SOM protection occurs through the biochemical recalcitrance of its structure, and is influenced by the type of C input. The non-protected C pool is independent of the other, mineral-related protection mechanisms and the balance of C inputs and decomposition rate. Decreased whole-soil SOC stabilization efficiency has been observed in long-term agricultural experiments with high C input levels and/or high soil C levels (Six et al., 2002; Stewart et al., 2007; Kool et al., 2007; Gulde et al., 2008). Soil fraction data from longterm agroecosystem experiments suggested decreased soil fraction stabilization efficiency at high C contents (as a proxy for C input levels), even in soils that do not show decreased stabilization in the whole soil (Stewart et al., 2007). However, in these studies, C saturation has been evaluated using experiments investigating effects of management (i.e., tillage, fertilization, and crop rotation) on yield or soil C across sites differing in climate, soil texture, mineralogy and decomposition kinetics (Stewart et al., 2007, 2008a). Our objective was to experimentally test whether C saturation deficit and varying C input levels influence soil C stabilization of residue-derived 13C in measurable soil fractions corresponding to the conceptual chemical, physical, biochemical, and non-protected C pools proposed by Six et al. (2002). More specifically, we examined the level of 13C-labeled residue stabilization, for two levels of C additions in A- versus C-horizon soils from seven different sites with a broad range of SOC contents and physicochemical characteristics. We hypothesized that the proportion of stabilized residuederived C would be greater in soils with a larger compared to smaller C saturation deficit (i.e., the C- versus A-horizon) and the relative stabilization efficiency of added C would be higher for the low C input level (1 residue addition) compared to the high C input level (5 residue addition).

2. Materials and methods 2.1. Rationale for experimental approach to test C saturation concept Soils with differing C saturation deficits were required to examine the effect of soil C saturation deficit on the stabilization of added C residue. Field-level experiments are not appropriate to directly test C saturation deficit effect due to confounding variation between SOC content and climatic factors that could alter C input as well as decomposition. Therefore, we used laboratory incubations to directly test the effect of saturation deficit and C input level on C stabilization where residue addition, soil characteristics and decomposition factors were controlled. We chose seven long-term agricultural research sites with several soil types and a range of characteristics (e.g. texture and mineralogy) that were all cultivated under continuous corn (Table 1). The sites were located in Sioux City, IA; W.K. Kellogg Biological Station, MIS; Saginaw, MI; Lamberton, MN; Mead, NE; Wauseon, OH; and Melfort, SK. To obtain soils with differing saturation deficits (low and high C soils) we sampled the A- and C-genetic horizons within one profile of each site. The A- and Chorizons of our soils were similar in major physical and chemical properties (e.g. clay content, pH, CEC) except for SOC and N content (Table 1). The sites we chose varied up to an order of magnitude in SOC content between the A- and C-horizon. To test our two hypotheses, we added different amounts (i.e., 1 and 5 average annual C addition under field conditions) of 13 C-labeled wheat straw to both the A- and the C-horizons. This enabled us to trace added C into respiration versus stabilization within soil fractions of each soil. The effect of C saturation deficit on C accumulation is tested directly by comparing the C- and Ahorizon within each addition level. If C saturation deficit influenced the residue-derived C stabilization, the proportion of new C stabilized would be greater in soils with a larger compared to smaller C saturation deficit (i.e., the C- versus A-horizon). The effect of C addition level on the relative stabilization efficiency is evaluated statistically by the difference of the effect size (the difference between A- and C-horizon) between the 1 versus the 5 addition levels. If C addition level influenced C stabilization, the relative stabilization efficiency of added residue C would be greater for the low C input level (1 addition) than the high (5 addition). To assess the stabilization of added residue in a soil suggested to be already at or near saturation (Campbell et al., 1991), we added

Table 1 Basic soil properties of the A- and C-horizons of the seven agricultural sites. Site

Horizon

Texture (g 100 g soil1) Sand

Silt

Clay

pH

Total organic C (g C kg1 soil)

d13C (&)

Total N (g N kg1 soil)

CEC (cmol(þ) kg1)

Base sat (%)

IA

A C

68.5 62.4

21.9 27.5

9.6 10.1

7.27 8.08

11.4 6.2

19.25 24.01

1.0 0.4

15.4 17.2

80.8 ND

KBS

A C

12.0 3.7

30.0 31.3

58.0 65.0

6.63 6.35

9.4 7.8

23.32 22.22

1.0 0.7

23.5 23.1

23.4 ND

MIS

A C

12.2 19.2

19.0 13.5

68.9 67.2

8.21 8.38

14.8 1.9

21.64 24.22

1.8 0.5

36.6 37.4

71.0 ND

MN

A C

39.9 36.1

27.6 31.8

32.4 32.1

6.34 8.57

18.6 6.0

15.98 23.57

1.9 0.3

28.6 37.3

63.4 ND

NE

A C

7.6 8.5

54.6 60.1

37.8 31.3

6.27 7.30

17.8 1.7

14.57 20.76

2.0 0.4

25.6 28.7

91.3 ND

OH

A C

85.2 90.2

8.1 3.2

6.7 6.6

5.66 6.37

10.9 1.2

19.20 24.77

1.0 0.1

9.5 10.9

22.8 ND

SK

A C

9.5 2.1

50.2 20.0

40.3 77.8

6.16 8.56

53.1 9.2

25.19 20.42

5.2 0.9

29.7 47.5

89.9 ND

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

5 the addition of 13C-labeled wheat straw to A- and C-horizon soil samples from the long-term experiment in Melfort, Canada. 2.2. Soil sampling We sampled A- (0–20 cm) and C-genetic horizons (variable depths) in spring 2001 from seven long-term agricultural field experiments (Table 1). Samples were taken from soil pits dug to corresponding horizon depth. Soils were packaged to remain cool and uncompacted during transport to the laboratory. In the laboratory, large rocks, recognizable surface litter, and root material were removed, as samples were gently broken by hand and passed through an 8-mm sieve. Soils were then air-dried, passed through a 2-mm sieve, and stored at room temperature. 2.3. Soil analyses All soils were analyzed for pH, texture, carbonates, field capacity, total C and N content, and base saturation (base saturation was done only on the A-horizon samples). Soil pH was determined in 2:1 water: soil ratio using a digital pH meter (Radiometer, Copenhagen). Soil texture was determined using a modified version of the standard hydrometer method without removal of carbonates or organic matter (Gee and Bauder, 1986) on a 30-g subsample dispersed with 100 ml of 5% sodium-hexametaphosphate solution for 18 h. Total sand content was determined by sieving (53 mm) and clay content was measured by the 2-h hydrometer method. Silt was determined by difference. Soil carbonates were determined by a modified pressure transducer method described by Sherrod et al. (2002). Field capacity was determined on three replicates (50 g) of 2mm sieved soil, wetted slowly with 8 ml of deionized water in glass test tubes, covered with perforated parafilm, and allowed to equilibrate overnight. A subsample from the middle of the column was then weighed, dried overnight in a 105  C oven and weighed again. Field capacity was calculated using the equation:

Field capacityðFCÞ ¼ ðwet weight  dry weightÞ=ðdry weightÞ  100 Soil C and N were determined on ground subsamples using a Carlo Erba NA 1500 CN analyzer (Carlo Erba, Milan, Italy). Cation exchange capacity (for both horizons) and base saturation (for the A-horizons) were determined by the Plant, Soil and Water Testing Laboratory, Colorado State University, Fort Collins, Colorado using the ammonium acetate method at a pH of 7 (Sumner and Miller, 1996). 2.4.

13

C wheat labeling

Spring wheat (Triticum aestivum, AC Teal, var awnless) was continuously labeled with 13C in a 1.22 m  1.37 m  3.90 m airtight Plexiglas chamber. Air was mixed with two fans (2.83 m3 displacement) and humidity was maintained between 70% and 90% with a Frigidaire dehumidifier operated by a humidity controller (Ohmic Instruments Co., model EHC-100). Temperature was maintained between 20 and 30  C by two radiators. Both temperature and humidity measurements were made with a hygrothermometer (Extech instruments Model 45320). Fifty wheat seeds were planted into 36 pots (17.6 l), in a soil mixture of 25% perlite, 25% sand and 50% autoclaved soil. The soil was obtained from the Agricultural Research, Development and Education Center at Colorado State University. Soil was brought to field capacity using 1 l water and 1 l modified Hoagland’s nutrient solution containing Ca, N, K, Mg, P, Na and micronutrients (B, Mn, Zn, Cu, and Mo). Plants were watered 2–3 times a week and the N input varied between 100 and 200 g KNO3 per 18 l solution.

359

Continuous plant 13C labeling was achieved by maintaining an enriched CO2 atmosphere throughout plant growth. Enriched CO2 was produced by adding 1% 13C sodium bicarbonate solution by an automated micropipetter (Hamilton Company, Reno, NE) to a 10 M H2SO4 solution and the 13CO2 pumped into the chamber. Average chamber CO2 concentration was maintained at 350 ppm and a 2% isotopic enrichment. Chamber CO2 was monitored by an infrared gas analyzer (LICOR model LI-800, Lincoln, NE). 2.5. Incubation Each treatment was comprised of four 200-g replicates of A- and C-horizon soils mixed with either 0.26 g (1) or 1.28 g (5) 13C wheat straw (8-mm sized). Samples were slowly wetted to field capacity, and allowed to equilibrate overnight in a refrigerator (4  C). The 2.5-year samples were placed into airtight 3.79-l glass jars and capped with lids containing septa for gas sampling. We measured total respiration every other day for the first month of incubation and monthly thereafter using an IRGA (LICOR model LI6252, Lincoln, NE). Samples were flushed with tank air and blanks were maintained to assure complete flushing between sampling dates. Samples were maintained at field capacity throughout the incubation. Samples were destructively sampled at 2.5 years, gently broken by hand to pass an 8-mm sieve, and air-dried. Then, we 2-mmsieved the soils and removed all of the added 13C wheat straw greater than 2-mm. 2.6. Soil fractionation Separation of the various C pools was accomplished by a combination of physical and chemical fractionation techniques in a simple, three-step process (Fig. 1) modified from Six et al. (2002) and detailed in Plante et al. (2006a) and Stewart et al. (2008a). The first step was the partial dispersion and physical fractionation of the soil to obtain three size fractions: >250 mm (coarse non-protected particulate organic matter, cPOM), 53–250 mm (microaggregate fraction, magg), and <53 mm (easily dispersed silt and clay, dSilt and dClay). Physical fractionation was accomplished by fractionating air-dried 2-mm sieved soil in the microaggregate isolator described by Six et al. (2000). The microaggregate isolator dispersed the >2-mm soil with 50 glass beads in running water over a 250-mm sieve in order to flush released microaggregates and finer particles through the 250-mm mesh screen. Material greater than 250 mm (cPOM) remained on the sieve. Microaggregates were collected on a 53-mm sieve that was subsequently wet sieved by hand for 50 strokes in 2 min (Elliott, 1986) to separate the easily dispersed silt- and clay-sized fractions from the water-stable microaggregates. The suspension was centrifuged at 127  g for 7 min to separate the silt-sized fraction. The supernatant was subsequently separated, flocculated with 0.25 M CaCl2–MgCl2 and centrifuged at 1730  g for 15 min to separate the clay-sized fraction. All fractions were dried in a 60  C oven and weighed. The second step involved further fractionation of the microaggregate fraction isolated in the first step (Plante et al., 2006a). A density flotation with 1.85 g cm3 sodium polytungstate (SPT) was used to isolate fine non-protected POM (LF) (Six et al., 1998). After removing the fine non-protected POM, the heavy fraction was dispersed overnight by shaking with 12 glass beads and passed through a 53 mm sieve, separating the microaggregate-protected POM (>53 mm in size, iPOM) from the microaggregate-derived siltand clay-sized fractions (mSilt and mClay) (Denef et al., 2004). The resulting suspension was centrifuged to separate the microaggregate-derived silt- versus clay-sized fraction as described above. The third step involved the acid hydrolysis of each of the isolated silt- and clay-sized fractions. The silt- and clay-sized fractions from

360

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

2mm sieved soil Microaggregate isolator

Coarse nonprotected POM C >250 µm (cPOM)

Acid Hydrolysis

Clay-sized C < 2 µm (dClay)

µaggregate 53-250 µm C ( µagg)

Non-hydrolyzable (NH-dClay) Acid Hydrolysis

Silt-sized C 2-53 µm (dSilt)

Dispersion

µaggregate protected POM C (iPOM) Silt-sized C 2-53 µm (µSilt)

Hydrolyzable (H-dSilt) Non-hydrolyzable (NH-dSilt)

Density flotation Fine non-protected POM C (LF)

Hydrolyzable (H-dClay)

Acid Hydrolysis

Clay-sized C < 2 µm (µClay)

Hydrolyzable (H- µClay) Non-hydrolyzable (NH- µClay)

Acid Hydrolysis

Hydrolyzable (H- µSilt) Non-hydrolyzable (NH-µSilt)

Fig. 1. Soil fractionation scheme that isolates the four hypothesized C pools; non-protected, physically protected (microaggregate), the chemically protected (silt þ clay), and biochemically protected pools. Modified from Six et al. (2002) to separate silt- and clay-associated C pools.

both the density flotation (mSilt and mClay) and the initial dispersion and physical fractionation (dSilt and dClay) were subjected to acid hydrolysis as described in Plante et al. (2006b). Acid hydrolysis consisted of refluxing at 95  C for 16 h in 25 ml of 6 M HCl. After refluxing, the suspension was filtered and washed with deionized water over a glass-fiber filter. Residues were oven-dried at 60  C and weighed. These fractions represent the non-hydrolyzable C fractions (NH-dSilt, NH-dClay, NH-mSilt and NH-mClay). The hydrolyzable C fractions (H-dSilt, H-dClay, H-mSilt and H-mClay) were determined by difference between the total organic C content of the fractions and the C contents of the non-hydrolyzable fractions. This simple three-step process isolates a total of 16 fractions, some of which are composites of others (e.g., magg is composed of LF, iPOM, mSilt and mClay, and the latter two are each composed of hydrolyzable and non-hydrolyzable portions). This fractionation scheme is based on the assumed link between the isolated fractions and the protection mechanisms involved in the stabilization of organic C within that pool and as described in detail by Six et al. (2002). The non-protected C pool consists of the cPOM fraction, isolated during the first dispersion step, and the LF fraction isolated during the second fractionation step. The physically protected C pool consists of the magg fraction as a whole and the iPOM. The chemically protected pool corresponds to the hydrolyzable portion of the silt- and clay-sized fractions isolated during the initial dispersion (H-dSilt and H-dClay). The biochemically protected pool corresponds to the non-hydrolyzable C remaining in the silt and clay fractions after acid hydrolysis (NH-dSilt and NH-dClay). 2.7. Carbon and

13

 13

Rsample  13 R

standard

13 Rstandard

Qt dt ¼ Qr dr þ Qs ds þ Qb db where Qt, Qr, Qs, and Qb are respired C (mg C kg1 soil) and dt, dr, ds, and db are its isotopic composition (&) from total, residue, soil, and blanks, respectively. Soil and wheat-straw C and 13C were determined on ground subsamples using a Carlo Erba NA 1500 CN analyzer (Carlo Erba, Milan, Italy) coupled with a Micromass VG isochrome-EA mass spectrometer (Micromass UK Ltd., Manchester, UK) (continuous flow measurement). Carbonates were removed prior to analysis by acid fumigation (Harris et al., 2001) modified to a one-half hour fumigation for 3 mg samples. The proportion of residue-derived C stabilized in the soil (f) was calculated using the equation:

f ¼

dt  ds dr  ds

where dt ¼ d13C of the whole soil at time t, ds ¼ d13C of the original whole soil; dr ¼ d13C of the added residue (738.63& both for whole and hot-water extracted residue). The quantity of residue-derived C stabilized in the soil smaller than 2 mm was calculated as

Cr ¼ Ct f where Ct ¼ total C content of the soil. 2.8. Expression of a normalized C-accumulation unit

C analysis

The 13C of the respired CO2 was measured with a Micromass VG Optima mass spectrometer (Micromass UK Ltd., Manchester, UK). Results were expressed as

d13 C ¼

where 13R is equal to 13C/12C and the standard is the international Pee Dee Belemnite. Residue-derived C (Qr) was calculated using the equation:



 1000

The residue-derived respiration data indicate that over the 2.5year incubation, there were significant effects of residue addition amount, soil horizon and sample date on the decomposition rate of added 13C wheat straw (Stewart et al., 2008b). Thus in order to assess the relative stabilization of added 13C in the soil fractions, we normalized the 13C residue-derived soil data by dividing it by respired 13C to account for the variation in microbial processing of added residue between treatments (see Stewart et al., 2008b).

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

361

Consequently, the unit we used to express our data is milligram (or microgram) residue-derived soil C (mg residue-derived C respired)1 (Stewart et al., 2008b) and we consider any residue-derived C in a fraction to be stabilized residue-derived C. 2.9. Estimates of C saturation deficit To determine the general effect of C saturation deficit on the stabilization of added 13C residue across sites, we calculated soil C protective capacity (g C kg1 soil) using relationships between soil texture and mineral (silt þ clay) C content (g kg1 soil) developed from Six et al. (2002).

Protective capacity ¼ 0:21  ðsilt þ clay contentÞ þ 14:76 We then estimated C saturation deficit (%) based on SOC content for all sites and horizons as

C saturation deficit ¼ 1 ðSOC content=protective capacity estimateÞ If C saturation dynamics influenced C stabilization of added residue, then as saturation deficit decreases, the stabilization of residue-derived C should also decrease. However, if there is no relationship between C saturation deficit and stabilization of added residue; saturation did not affect C stabilization. 2.10. Statistical analyses Data were analyzed using the PROC MIXED procedure in SASSTAT (SAS Institute Inc., Cary, NC). Treatments (horizon and addition level) and their interaction were included as potential fixed effects and dropped from the model when non-significant at the a ¼ 0.05 level. Contrast statements were used to assess the effect of soil horizon within addition level. Soil C concentration data were analyzed using the PROC MIXED procedure in SAS-STAT (SAS Institute Inc., Cary, NC). The effect of addition level was tested by comparing estimates of effect size (C- minus A-horizon) between the 1 versus the 5 addition levels (1 minus 5) (P < 0.05) using contrast statements. 3. Results

Fig. 2. Stabilized residue-derived C (mg residue-derived C (mg residue-derived C respired)1) in the four theorized C pools capable of C saturation for the A- and Chorizons with the 1 and 5 C addition levels after 2.5 years of incubation. Nonprotected ¼ cPOM þ LF, physical ¼ magg, biochemical ¼ NH-dSilt þ NH-dClay and chemical ¼ H-dSilt and H-dClay fractions. The sites are Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH), and Melfort, SK (SK). Error bars represent standard errors of all fractions combined. Empty bars are the Ahorizon, and dashed are the C-horizon. Treatments are presented in the order 1A, 1C, 5A, and 5C.

biochemically protected pool stabilized more C in the NH-dSilt compared to NH-dClay fractions in most sites (Fig. 4). The physically protected pool represented the largest residue-C stabilized pool, averaging 45% of total stabilized C. Within the magg fraction, we found that 50–85% of total residue-derived C was associated with aggregate-associated hydrolyzable silt and clay fractions (H-mSilt and H-mClay). The iPOM fraction comprised 4– 40% and the non-hydrolyzable fractions (NH-mSilt and NH-mClay) and accounted for only 0–16% of total residue-derived C (Fig. 5). In the physically protected pool (magg fraction) four sites showed significant differences between horizons (Fig. 5) in at least one addition level. The result was not consistent across sites, with two stabilizing more residue-derived C in the C- versus the Ahorizon (MIS and OH) while another two sites stabilized more C in the A-horizon (KBS and NE). The five fractions comprising the

3.1. Saturation deficit effects on C accumulation in C pools After 2.5 years, residue-derived C stabilization was significantly greater in the C- compared to the A-horizon (P < 0.05) whole soil in at least one addition level of all sites, except KBS. At all sites, the majority of added 13C residue was stabilized in the chemically and physically protected pools, together comprising between 55% and 91% of the total residue-C stabilized (Fig. 2). Chemically protected C (H-dSilt þ H-dClay fractions) comprised an average 27% of total residue-C stabilized in the soil. Residuederived carbon accumulation in both fractions of this pool was significantly greater in the C-horizon compared to the A-horizon at all but one site (KBS) in at least one addition level (1 and 5) (Fig. 3 and Table 2). In most sites, the H-dSilt fraction stabilized more added C than the H-dClay fraction (Fig. 3). A small proportion of residue-derived C was stabilized in the biochemically protected pool (NH-dSilt and NH-dClay fractions), ranging from 0 to 23 mg residue-derived C (mg respired residuederived C)1 (Fig. 4). In six soils (IA, MIS, MN, NE, OH, and SK), more residue-derived C was stabilized in the biochemically protected pool (NH-dSilt and NH-dClay) of the C- compared to the A-horizon in at least one addition level (1 or 5 addition levels) (Table 2). A few treatments had the opposite trend (KBS 1 NH-dSilt, and NH-dClay at MIS and SK). As in the chemically protected pool, the

Fig. 3. Stabilized residue-derived C (mg residue-derived C (mg residue-derived C respired)1) in the fractions (H-dSilt and H-dClay) comprising the chemical pool in the A- and C-horizons with the 1 and 5 C addition levels for Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH), and Melfort, SK (SK). Error bars represent standard errors of both fractions combined. Empty bars are the A-horizon, and dashed are the C-horizon. Treatments are presented in the order 1A, 1C, 5A, and 5C.

362

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

Table 2 Treatment differences in stabilized residue-derived 13C (mg residue-derived C (residue-derived mg C respired)1) between the C- and A-horizons within the 1 and 5 addition levels (i.e. 1C-1A and 5C-5A) after 2.5 years incubation. Sites are Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH), and Melfort, SK (SK). Soil C pool

Fraction

Chemical

H-dSilt H-dClay

Biochemical

NH-dSilt NH-dClay

Non-protected

IA

KBS

1

5

1

MIS 5

1

15.4* 5.6*

9.2* 2.9

0.5 2.8

14.3 10.0*

4.5* 2.7*

3.5* 1.3*

0.0 0.1

1.2 1.9*

cPOM LF

5.4 7.5*

16.5 6.9*

5.8 14.6

19.0 5.2

46.1* 6.6

Physical

magg iPOM

16.7 47.1*

7.8 1.4

13.9 4.7

Chemical

H-mSilt H-mClay

30.7* 10.5*

0.2 10.8*

9.0 0.8

Biochemical

NH-mSilt NH-mClay

2.9* 5.4*

0.3 3.1*

0.1 1.0

* a

8.5 16.0

9.1* 8.4* 0.7 0.5

66.2* 9.7* 3.2 2.2

MN 5

1

NE 5

OH

1

5

1

SK 5

5

25.9* 0.5

53.1* 6.6*

38.9* 0.1*

42.0* 18.9*

32.3* 18.2*

20.4* 31.7*

9.4* 23.7*

36.2* 4.0*

8.9* 0.6

17.1* 5.5*

18.0* 1.4*

0.0* 3.5*

5.2* 3.1*

4.2* 8.0*

4.9* 8.8*

4.1* 1.4*

27.7* 0.7

19.8 1.4

42.5* 2.9

17.0 41.3

28.7 2.4

25.9 7.4

18.8 10.7

33.4* 7.1

43.6* 13.4

10.8 14.3*

0.7 10.6*

23.5* 11.0*

25.8* 8.9

15.2 6.7

27.0* 4.1

22.7 5.3

6.5 0.5

32.3* 19.2*

12.3* 4.9

8.1* 1.1

8.1* 14.6

2.1 11.7*

5.3 4.8

5.1a 14.1

7.7a 17.4

5.2 1.4

0.6 4.8*

2.1* 5.8*

0.6* 5.0*

1.6* 5.9*

0.9* 2.0*

0.8a 3.7

0.9a 6.3

0.0 0.4

1.9* 0.3

Statistically significant estimates at the 0.05 level. Estimates based on a single, composite sample and not included in statistical analysis – due to small sample size.

physically protected pool (Fig. 2) were combined into aggregatemineral (mSilt þ mClay) versus POM (iPOM) fractions for generality given their similar results (Table 2). The mineral-associated fractions (mSilt and mClay) at four sites behaved similarly to the chemically and biochemically protected C pools, stabilizing more residue-derived C in the C- compared to the A-horizon at four sites (hydrolyzable IA, KBS, MIS, MN; non-hydrolyzable IA, MIS, MN, NE). The iPOM fraction had the opposite response stabilizing significantly more added residue in the A- compared to the C-horizon in three sites (IA, MIS, MN) (Table 2). The response of the physically protected pool was dependent on the relative contribution of POM versus mineral fractions to the whole, with A-horizons generally stabilizing more residue-derived C in the iPOM fraction (Fig. 5). The fractions of the non-protected C pool (cPOM þ LF) comprised between 9 and 46% of total residue-derived SOC. The cPOM fraction generally contained more residue-derived C in the A- compared to the C-horizon, except for the MIS 1 addition and SK 5 addition. There were no significant differences between horizons in the LF, except both addition levels at IA.

deficit and 13C accumulation in either the iPOM or the mineral fractions (silt þ clay) (Fig. 6d). 4. Discussion After 2.5 years of incubation, we found that C saturation deficit and the amount of added C influenced residue-derived C stabilization in measurable soil fractions that correspond to our hypothesized C pools. Carbon saturation behavior was observed in the mineral fractions of most sites, both in the chemically and biochemically protected pools (Fig. 6a), but not in the non-protected C pool. Furthermore, the fractions comprising the physically protected pool could not be combined, as the individual fractions behaved in opposite manners with respect to C saturation with aggregate-protected POM not saturating and some mineral-associated fractions occluded within the aggregates showing evidence of C saturation. Even though some mineral-associated fractions showed C saturation behavior, the dynamics and mechanisms of C saturation differed between the fractions (i.e. occluded versus not

3.2. Residue addition effects on C accumulation in C pools We compared the relative stabilization efficiency between the 1 and 5 addition rates in soil fractions that had greater C accumulation in the C- compared to the A-horizon (Table 3). After 2.5 years, only the chemically protected pool (H-dSilt and H-dClay) had a clear trend, with four sites (IA, MIS, MN, and OH) stabilizing significantly more added residue C in the 1 compared to the 5 residue addition in at least one fraction (Table 3). The remaining fractions showed no general trend in C accumulation with addition level. 3.3. C saturation deficit effect across sites The effect of C saturation deficit was not only apparent within individual soils, but also along the gradient of saturation deficits produced by our range in soils and horizons. In the 5 addition, which had the broadest range in C saturation deficits (due to the SK soil), we found a significant increasing exponential relationship between the saturation deficit and 13C accumulation in the chemically and biochemically protected C pools (Fig. 6a and b). The nonprotected fractions (cPOM and LF) showed no effect of saturation deficit on the accumulation of added 13C residue (Fig. 6c). Physically protected fractions showed no relationship between saturation

Fig. 4. Stabilized residue-derived C (mg residue-derived C (mg residue-derived C respired)1) in the biochemical pool (NH-dSilt þ NH-dClay fractions) for the A- and Chorizons with the 1 and 5 C addition levels after 2.5 years of incubation for Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH), and Melfort, SK (SK). Error bars represent standard errors of both fractions combined. Empty bars are the A-horizon, and dashed are the C-horizon. Treatments are presented in the order 1A, 1C, 5A, and 5C.

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

363

Fig. 5. Stabilized residue-derived C (mg residue-derived C (mg residue-derived C respired)1) in the microaggregate-protected fraction (aggregate POM ¼ iPOM and aggregate-mineral ¼ mSilt þ mClay) for the A- and C-horizons with the 1 and 5 C addition levels after 2.5 years of incubation. The sites are Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH), and Melfort, SK (SK). Error bars represent standard errors of the combined fractions. Empty bars are the Ahorizon and dashed are the C-horizon. Treatments are presented in the order 1A, 1C, 5A, and 5C.

occluded, clay-associated versus silt associated, and chemically versus biochemically protected) of the mineral-associated C pools. 4.1. Chemically protected pools Greater stabilization of residue-derived C in the C- compared to A-horizon in both the 1 and 5 addition treatments suggests that saturation deficit influenced the stabilization of added C in the chemically protected C pool. The observed decreased residuederived C stabilization in the chemically protected C pool (H-dSilt and H-dClay fractions) of soils with a smaller C saturation deficit supported other work that has observed a limit to C stabilization in silt- and clay-sized fractions (Hassink et al., 1997; Roscoe et al., 2001; Jolivet et al., 2003; Dieckow et al., 2006; Stewart et al., 2008a). In a modeling analysis of eight field sites, Stewart et al. Table 3 Addition level effects (i.e. (1C–1A) minus (5C–5A)) in residue-derived stabilized 13C (mg residue-derived C (residue-derived mg C respired)1) between C- and A-horizons of the 1 and 5 residue addition levels after 2.5 years of incubation. Sites include Sioux City, Iowa (IA), W.K. Kellogg Biological Station, Michigan (KBS), Saginaw, Michigan (MIS), Lamberton, Minnesota (MN), Mead, Nebraska (NE), Wauseon, Ohio (OH). Soil C pool

Fraction

IA

KBS

MIS

MN

NE

OH

Chemical

H-dSilt H-dClay

6.16* 2.68

13.76 7.19

40.32* 9.25*

14.22 6.49*

9.75 0.73

10.98* 8.02*

Biochemical

NH-dSilt NH-dClay

1.01 1.34*

1.18 ND

5.74 ND

0.90 4.06*

5.20 0.34

0.61 0.79

Non-protected

cPOM LF

ND 0.52

ND 9.48

ND 34.69

ND ND

ND ND

ND ND

Physical

magg iPOM

38.8 ND

ND ND

40.6* ND

ND ND

ND ND

9.3 1.23

Chemical

H-mSilt H-mClay

ND 0.33

ND 7.65

20.05* 14.31*

ND ND

ND ND

ND 3.34z

Biochemical

NH-mSilt NH-mClay

2.61* 2.27

ND ND

1.54 1.01

0.92 0.86

1.03 ND

ND 2.58

ND – estimates were not determined (ND) due to greater 13C stabilization in the A- compared to C-horizon of that fraction. Treatment estimates are statistically derived and expressed as mg residue-derived C g1 soil. * Statistically significant estimates at the 0.05 level.

Fig. 6. Stabilized residue-derived C (mg residue-derived C (mg residue-derived C respired)1) in the chemically, biochemically protected, physically and non-protected C pools in the 5 addition after 2.5 years incubation as a function of the saturation deficit based on estimates from Six et al. (2002).

364

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

(2008a) observed C saturation behavior in the H-dSilt and H-dClay fraction C content as a function of total SOC content at the majority of the sites, suggesting that the chemically protected pool saturated with respect to whole SOC content. Combined with our data, this suggests that sites that are near their C saturation level sequester very little, if any additional C in this pool. The C saturation level of the chemically protected pool is directly influenced by texture and mineralogy. The greater surface area of clay versus silt suggests that the clay-sized fraction should accumulate more added residue than the silt-sized fraction. However, we found that the silt-sized fraction accumulated more residue-derived C than the clay-sized fraction, which may be due to incomplete dispersion of silt-sized microaggregates (Plante et al., 2006a; Stewart et al., 2008a). The greater capacity of the siltcompared to clay-sized fraction to retain C was also observed by Stewart et al. (2008a), who found larger silt – compared to clay – C saturation level estimates in a modeling exercise. Our saturation deficit estimate (Fig. 6) appeared to be a reasonable first approximation and successfully illustrated the effect of C saturation deficit on the stabilization of added residue in the chemically protected pool. The regression equations used by Six et al. (2002) to estimate protective capacity only considered silt þ clay-associated C whereas our estimates of C saturation deficit included the whole-soil C content, and consequently produced a negative value in the case of SK. It is interesting to observe that saturation deficit appeared to be more important than texture in predicting the accumulation of added residue. Two soils with extremely different textures had similarly low C saturation deficits and also low residue-derived C accumulation. The OH A-horizon had less than 15% silt þ clay content and a saturation deficit of 0.02 as did the SK A-horizon with 91% silt þ clay content and a saturation deficit of 0.6. Although differing significantly in texture, both OH and SK accumulated less C than sites having a larger saturation deficit. This is consistent with other work suggesting that the clay in sandy soils is closer to saturation than the clays in clayey soils, evidenced by a negative relationship between clay content of the soil and clay C content (Jolivet et al., 2003; Plante et al., 2006a). Across texture, other researchers have found that soils with a lower silt þ clay content tended to be saturated, while those with greater clay content still had potential to store more C (Carter et al., 2003; Hassink, 1997). 4.2. Biochemically protected pools The biochemically protected pool (NH-dSilt and NH-dClay fractions) showed a strong effect of C saturation deficit with the Chorizon sequestering more residue-derived C than the A-horizon in both the 1 and 5 residue additions (Fig. 4). Further, when expressed as a function of C saturation deficit, a significant increase in stabilization of added residue in the biochemically protected C pool was observed across sites with increasing C saturation deficit in this incubation (Fig. 6b). These results confirm the influence of C saturation deficit on C accumulation in the biochemically protected pool in sites with a broad C saturation deficit. We found evidence supporting the effect of C saturation deficit on C accumulation in soils with a broad C saturation deficit in the biochemically protected C pool. Consequently, one could argue that the chemically and biochemically protected C pools could be combined into one fraction with respect to C saturation. However, other studies that included soils with a smaller range in C saturation deficits, reported mixed results for the biochemically protected C pool and that were not always consistent with those for the chemically protected pool. For example, Stewart et al. (2008a) found that, of the seven sites they examined, the four sites that showed C saturation dynamics in both fractions (silt and clay) of the chemically protected C pool (Alberta, Ohio, Swift Current, Scott),

only Alberta and Scott also demonstrated C saturation behavior for the biochemically protected C pool associated with both the silt and clay fraction. Furthermore, one site, demonstrated C saturation behavior in both biochemically protected C pools, but only one of the chemically protected fractions (Colorado). Hence, the lack of consistency in C saturation behavior in the fractions of the chemically and biochemically protected C pools across sites suggests that expression of C saturation is fraction specific and therefore warrants differentiation between the C pools, despite similar behavior in this study. As with the chemically protected pool, the silt-sized fraction of the biochemically protected generally stabilized more residuederived C than the clay-sized. This could be due to either biochemical or physical protection mechanisms. Differences in hydrolyzability between the two fractions could be due to the carbohydrate concentration (Plante et al., 2006b) of clays being greater than that of the silt-sized fractions (Guggenberger et al., 1994; Amelung et al., 1999; Kiem and Kogel-Knabner, 2003). However, it is also possible that silt-sized aggregates formed in the C-horizon have the capacity to form stable complexes which resist degradation by acid hydrolysis as was seen by greater amounts of added residue C in the non-hydrolyzable silt-sized fraction of the Chorizon soils than of the A-horizon soils.

4.3. Physically protected pools The lack of a consistent effect of C saturation deficit in the magg fraction as a whole has been observed previously and may be due to the contrasting behavior of its composite fractions, the iPOM and mineral-associated fractions (mSilt þ mClay) (Stewart et al., 2008a). The mineral-associated fractions showed a significant C saturation deficit effect for two of the soils, while the iPOM fraction had the opposite, significant relationship. Stewart et al. (2008a) also found contrasting responses between aggregate-associated mineral and POM fractions at Scott, SK. Despite some C-saturation responses in the aggregate-associated mineral fractions, the physically protected pool did not show C saturation when POM was included. Although our data do not support C saturation of the aggregate fraction as a whole, there is some indication that C saturation of aggregates is possible, but requires large amounts of C input to occur. In soils from long-term manure additions, Gulde et al. (2008) found that aggregate fractions showed evidence of C saturation. Mineral-associated fractions have been found to exhibit Csaturation behavior (Hassink and Whitmore, 1997 and others) as a whole. However, the data from this incubation and from eight long-term agroecosystem soils (Stewart et al., 2008a), showed that the aggregate-associated mineral fractions and easily dispersed mineral fractions behaved differently. In this study, the easily dispersed silt and clay showed clear C-saturation dynamics at most sites while some aggregate-associated fractions did not. Unaggregated silt and clay responded to C additions faster than aggregate-mineral fractions, indicating either a different protection mechanism, or a time-lag in residue-derived C accumulation. These results are similar to Stewart et al. (2008a), who found that of the four sites that showed C saturation dynamics in both fractions (silt and clay) of the chemically protected C pool (Alberta, Ohio, Swift Current, Scott), only one (Scott) also demonstrated C-saturation behavior in the aggregated silt and clay fractions. Although Hassink and Whitmore (1997) showed evidence of C saturation of the entire silt þ clay fraction, the lack of consistent behavior between aggregated and unaggregated silt and clay fractions across sites suggests that expression of C saturation is influenced by physical protection mechanisms and therefore warrants differentiation between the C pools. These aggregate-protected fractions in the physically protected C pool may be slow to

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

incorporate added residue, and a large quantity of C may be necessary to observe C saturation in this C pool. The iPOM fractions of a majority of soils showed greater C accumulation in the A- compared to the C-horizon, not supporting the effect of saturation deficit on C accumulation in this fraction. This agrees with the results of Stewart et al. (2008a) and Gulde et al. (2008), who observed that all iPOM fractions had a linear best-fit, which suggested C-saturation dynamics did not influence the behavior of that C pool. However, other work has found a maximum amount of occluded POM in the aggregate fraction (Kolbl and Kogel-Knabner, 2004). 4.4. Non-protected Accumulation of residue-derived C in the non-protected pools (cPOM and LF fractions) was greater in the A-horizon compared to the C-horizon soil and showed no evidence of being influenced by C-saturation deficit (Fig. 2). Our data agree with Stewart et al. (2008a) and Gulde et al. (2008), who also found no apparent Csaturation deficit effect on non-protected fractions. Stewart et al. (2008a) observed a best-fit with linear, non-saturating relationship between the cPOM and LF fractions and total SOC content in all eight agroecosystem sites they modeled. However, other researchers have found that LF C did not increase with increasing C additions at Melfort, SK (Six et al., 2002) or under N fertilizer applications (Solberg et al., 1997), which increased C input to the soil. Six et al. (2002) did not suggest a mechanism behind C saturation of the non-protected pool, but observed a lack of C accumulation in the whole soil and non-protected C pool under increasing C inputs in soils with high C contents. Under a scenario of minimized decomposition and maximized C inputs a litter layer would accumulate and its entry into the soil system would depend solely on the decomposition of the litter pool. C-saturation behavior of the non-protected C pool would be independent of mineral-related protection mechanisms. In managed systems, the non-protected C pool would result from C input level through plant production and the specific decomposition rate of the litter layer. 4.5. Is there an order to soil C sequestration? Soils further from saturation sequestered more total added C in chemically or aggregate-protected pools, and those closer to their saturation limit accumulated less total residue-derived C in the non-protected pool. The observed greater C accumulation in the unprotected pools of the A-horizons may reflect the lack of available C storage in the easily dispersed silt and clay fractions due to their lower saturation deficit. Unaggregated and aggregated mineral fractions stabilized added C differentially (Stewart et al., 2008a, and this data), with the unaggregated silt and clay responding to C additions faster than the mineral fractions occluded within aggregates. Once the free mineral protective capacity of a soil has been exceeded, further C accumulation may occur in the aggregate-associated and POM fractions (Gulde et al., 2008; Kool et al., 2007; Carter et al., 2003; Hassink et al., 1997). Other researchers have found that the carbon content of soils near or at their mineral ‘protective capacity’ was influenced by management, but not texture (Carter et al., 2003; Hassink, 1997), suggesting that the importance of the non-protected C pool increases as soil C saturation proceeds (Gulde et al., 2008). 4.6. Conclusions After 2.5 years of incubation, we found that C-saturation deficit and the amount of added C influenced residue-C storage in measurable soil fractions that correspond to the proposed chemical

365

and biochemical C pools. All the fractions comprising each of those pools showed C saturation, but mechanisms of stabilization and C saturation differed. The fractions of the physically protected C pool could not be combined to generalize about aggregate C saturation because the individual fractions behaved in opposite manners with respect to C saturation. The mineral-associated microaggregate occluded fractions demonstrated C-saturation behavior while the iPOM fraction had non-saturating behavior; stabilizing more added residue in the A- compared to the C-horizon. Although mineral-associated fractions have been found to exhibit C-saturation behavior as a whole, we found distinct behavior between the chemically and physically protected mineral fractions. All mineral-associated fractions outside of aggregates demonstrated behavior consistent with C saturation, but only some aggregated fractions did, indicating that unaggregated fractions reached C saturation faster than mineral-associated microaggregate fractions. The difference in behavior between these two fractions indicated an additional protection mechanism rather than just association with silt and clay in residue-derived C accumulation. Therefore, C-saturation behavior of the silt and clay-associated fractions can be generalized, but the mechanism determining the rate of attaining saturation and implications of this saturation are distinct. Lastly, the non-protected pool showed little evidence for C saturation. In summary, the influence of a hypothesized soil carbon saturation limit was observed in soil fractions from a variety of soil taxonomies, textures and climates, suggesting that, in general the chemically and biochemically protected pools are influenced by Csaturation behavior even though the whole soil may not be saturated with respect to C. If the chemically protected pool is filled, added C appeared to accumulate in the physically and non-protected fractions which are inherently less stable and subject to increased decomposition due to changes in management. Therefore, to accurately predict soil C storage under changing scenarios, C-saturation dynamics of soil fractions must be considered. Acknowledgements We thank two anonymous reviewers for comments that helped clarify the manuscript. We would also like to thank Dan Reuss for labeling (re-)construction as well as invaluable laboratory advice. Craig Atteberry, Yasko Matsuoka, Jean Marie Pouppirt, Laurel Beck, Joyce Dickens, Jodi Stevens, Shane Cochran, Sara Moculeski, and Colin Pinney provided laboratory assistance. We thank all the field crews and long-term site managers for their assistance in coordination and field sampling: Jim Gertsma, IA Natural Resource Conservation Service; John W. Doran and Gary Varvel from USDA; Frank Gibbs and Lynn Stuckey, OH Natural Resource Conservation Service; Jeff Strock, Southwest Research and Outreach Center, College of Agricultural, Food, and Environmental Sciences, University of Minnesota; and Shawel Haile-Mariam, Dept. of Crop and Soil Sciences, Michigan State University, East Lansing, MI for. This project was supported by the Office of Research (BER), U.S. Department of Energy and by the Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture. References Amelung, W., Flach, K.W., Zech, W., 1999. Lignin in particle-size fractions of native grassland soils as influenced by climate. Soil Science Society of America Journal 63, 1222–1228. Baldock, J.A., Skjemstad, J.O., 2000. Role of the soil matrix and minerals in protecting natural organic materials against biological attack. Organic Geochemistry 31, 697–710. Campbell, C.A., Lafond, G.P., Zentner, R.P., Biederbeck, V.O., 1991. Influence of fertilizer and straw baling on soil organic matter in a thin black Chernozem in western Canada. Soil Biology & Biochemistry 23, 443–446.

366

C.E. Stewart et al. / Soil Biology & Biochemistry 41 (2009) 357–366

Carter, M.R., Angers, D.A., Gregorich, E.G., Bolinder, M.A., 2003. Characterizing organic matter retention for surface soils in eastern Canada using density and particle size fractions. Canadian Journal of Soil Science 83, 11–23. Denef, K., Six, J., Merckx, R., Paustian, K., 2004. Carbon sequestration in microaggregates of no-tillage soils with different clay mineralogy. Soil Science Society of America Journal 68, 1935–1944. Dieckow, J., Mielniczuk, J., Gonzalez-Vila, F.J., Knicker, H., Bayer, C., 2006. No-till cropping systems and N fertilisation influences on organic matter composition of physical fractions of a subtropical Acrisol as assessed by analytical pyrolysis (Py-GC/MS). Geoderma 135, 260–268. Elliott, E.T., 1986. Aggregate structure and carbon, nitrogen, and phosphorus in native and cultivated soils. Soil Science Society of America Journal 50, 627–633. Gee, G.W., Bauder, J.W., 1986. Particle size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis, Part I. Physical and Mineralogical Methods, second ed. American Society of Agronomy, Madison, WI, pp. 383–411. Guggenberger, G., Christensen, B.T., Zech, W., 1994. Land-use effects on the composition of organic-matter in particle-size separates of soil .1. Lignin and carbohydrate signature. European Journal of Soil Science 45, 449–458. Gulde, S., Chung, H., Amelung, W., Chang, C., Six, J., 2008. Soil carbon saturation controls labile and stable carbon pool dynamics. Soil Science Society of America Journal 72, 605–612. Harris, D., Horwath, W.R., van Kessel, C., 2001. Acid fumigation of soils to remove carbonates prior to total organic carbon or carbon-13 isotopic analysis. Soil Science Society of America Journal 65, 1853–1856. Hassink, J., 1996. Preservation of plant residues in soils differing in unsaturated protective capacity. Soil Science Society of America Journal 60, 487–491. Hassink, J., 1997. The capacity of soils to preserve organic C and N by their association with clay and silt particles. Plant and Soil 191, 77–87. Hassink, J., Whitmore, A.P., 1997. A model of the physical protection of organic matter in soils. Soil Science Society of America Journal 61, 131–139. Hassink, J., Whitmore, A.P., Kubat, J., 1997. Size and density fractionation of soil organic matter and the physical capacity of soils to protect organic matter. European Journal of Agronomy 7, 189–199. Huggins, D.R., Buyanovsky, G.A., Wagner, G.H., Brown, J.R., Darmody, R.G., Peck, T.R., Lesoing, G.W., Vanotti, M.B., Bundy, L.G., 1998a. Soil organic C in the tallgrass prairie-derived region of the corn belt: effects of long-term crop management. Soil & Tillage Research 47, 219–234. Huggins, D.R., Clapp, C.E., Allmaras, R.R., Lamb, J.A., Layese, M.F., 1998b. Carbon dynamics in corn-soybean sequences as estimated from natural carbon-13 abundance. Soil Science Society of America Journal 62, 195–203. Jolivet, C., Arrouays, D., Leveque, J., Andreux, F., Chenu, C., 2003. Organic carbon dynamics in soil particle-size separates of sandy Spodosols when forest is cleared for maize cropping. European Journal of Soil Science 54, 257–268. Kool, D.M., Chung, H., Tate, K.R., Ross, D.J., Newton, P.C.D., Six, J., 2007. Hierarchical saturation of soil carbon pools near a natural CO2 spring. Global Change Biology 13, 1282–1293. Kiem, R., Kogel-Knabner, I., 2003. Contribution of lignin and polysaccharides to the refractory carbon pool in C-depleted arable soils. Soil Biology & Biochemistry 35, 101–118. Kolbl, A., Kogel-Knabner, I., 2004. Content and composition of free and occluded particulate organic matter in a differently textured arable Cambisol as revealed

by solid-state C-13 NMR spectroscopy. Zeitschrift Fur Pflanzenernahrung Und Bodenkunde. (Journal of Plant Nutrition and Soil Science) 167, 45–53. Nyborg, M., Solberg, E.D., Malhi, S.S., Izaurralde, R.C., 1995. Fertilizer N, Crop Residue, and Tillage Alter Soil C and N Content in a Decade. In: Lal, R., Kimble, J., Levine, E., Stewart, B.A. (Eds.), Advances in Soil Science: Soil Management and Greenhouse Effect. CRC Press Inc., Boca Ration, FL, pp. 93–100. Paustian, K., Andren, O., Janzen, H., Lal, R., Smith, P., Tian, G., Tiessen, H., van Noordwijk, M., Woomer, P., 1997. Agricultural soil as a C sink to mitigate CO2 emissions. Soil Use and Management 13, 230–244. Plante, A.F., Conant, R.T., Stewart, C.E., Paustian, K., Six, J., 2006a. Impact of soil texture on the distribution of soil organic matter in physical and chemical fractions. Soil Science Society of America Journal 70, 287–296. Plante, A.F., Conant, R.T., Paul, E.A., Paustian, K., Six, J., 2006b. Acid hydrolysis of easily dispersed and microaggregate-derived silt and clay-sized fractions to isolate resistant soil organic matter. European Journal of Soil Science 57, 456–467. Reicosky, D.C., Evans, S.D., Cambardella, C.A., Armaras, R.R., Wilts, A.R., Huggins, D.R., 2002. Continuous corn with moldboard tillage: residue and fertility effects on soil carbon. Journal of Soil and Water Conservation 57, 277–284. Roscoe, R., Buurman, P., Velthorst, E.J., Vasconcellos, C.A., 2001. Soil organic matter dynamics in density and particle size fractions as revealed by the C-13/C-12 isotopic ratio in a Cerrado’s oxisol. Geoderma 104, 185–202. Sherrod, L.A., Dunn, G., Peterson, G.A., Kolberg, R.L., 2002. Inorganic carbon analysis by modified pressure-calcimeter method. Soil Science Society of America Journal 66, 299–305. Six, J., Conant, R.T., Paul, E.A., Paustian, K., 2002. Stabilization mechanisms of soil organic matter: implications for C-saturation of soils. Plant and Soil 241, 155–176. Six, J., Elliot, E.T., Paustian, K., 2000. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biology & Biochemistry 32, 2099–2103. Six, J., Elliot, E.T., Paustian, K., Doran, J.W., 1998. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Science Society of America Journal 62, 1367–1377. Solberg, E.D., Nyborg, M., Izaurralde, R.C., Mahli, S.S., Janzen, H.H., Molina-Ayala, M., 1997. Carbon storage in soils under continuous cereal grain cropping: N fertilizer and straw. In: Lal, R., Kimble, J.M., Follett, R.F., Stewart, B.A. (Eds.), Management of Carbon Sequestration in Soil. CRC Press, Boca Raton, FL, pp. 235–251. Stewart, C.E., Paustian, K., Conant, R.T., Plante, A.F., Six, J., 2007. Soil C saturation: concept, evidence, and evaluation. Biogeochemistry 86, 19–31. Stewart, C.E., Plante, A.F., Paustian, K., Conant, R.T., Six, J., 2008a. Soil C saturation: linking concept and measurable carbon pools. Soil Science Society of America Journal 72, 379–392. Stewart, C.E., Paustian, K., Conant, R.T., Plante, A.F., Six, J., 2008b. Soil C saturation: evaluation and corroboration by long-term incubations. Soil Biology & Biochemistry 40, 1741–1750. Sumner, M.E., Miller, W.P., 1996. Cation exchange capacity and exchange coefficients. In: Sparks, D.L. (Ed.), Methods of Soil Analysis. Part 3. Chemical Methods. Soil Science Society of America, Inc., Madison, Wisconsin, pp. 1201–1229.