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


     

Published online 6 January 2006
Published in Soil Sci Soc Am J 70:287-296 (2006)
DOI: 10.2136/sssaj2004.0363
© 2006 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 Similar articles in ISI Web of Science
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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Agricola
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Related Collections
Right arrow Structure and Properties
Right arrow Soil Organic Matter
Right arrow Other Soil Management

Nutrient Management & Soil & Plant Analysis

Impact of Soil Texture on the Distribution of Soil Organic Matter in Physical and Chemical Fractions

Alain F. Plantea,*, Richard T. Conanta, Catherine E. Stewarta, Keith Paustiana,b and Johan Sixa,c

a Natural Resource Ecology Lab., Colorado State Univ., Fort Collins, CO 80523
b Dep. of Soil and Crop Sciences, Colorado State Univ., Fort Collins, CO 80523
c Dep. of Plant Sciences, Univ. of California, Davis, CA 95616

* Corresponding author (alainfplante{at}hotmail.com)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Previous research on the protection of soil organic C from decomposition suggests that soil texture affects soil C stocks. However, different pools of soil organic matter (SOM) might be differently related to soil texture. Our objective was to examine how soil texture differentially alters the distribution of organic C within physically and chemically defined pools of unprotected and protected SOM. We collected samples from two soil texture gradients where other variables influencing soil organic C content were held constant. One texture gradient (16–60% clay) was located near Stewart Valley, Saskatchewan, Canada and the other (25–50% clay) near Cygnet, OH. Soils were physically fractionated into coarse- and fine-particulate organic matter (POM), silt- and clay-sized particles within microaggregates, and easily dispersed silt- and clay-sized particles outside of microaggregates. Whole-soil organic C concentration was positively related to silt plus clay content at both sites. We found no relationship between soil texture and unprotected C (coarse- and fine-POM C). Biochemically protected C (nonhydrolyzable C) increased with increasing clay content in whole-soil samples, but the proportion of nonhydrolyzable C within silt- and clay-sized fractions was unchanged. As the amount of silt or clay increased, the amount of C stabilized within easily dispersed and microaggregate-associated silt or clay fractions decreased. Our results suggest that for a given level of C inputs, the relationship between mineral surface area and soil organic matter varies with soil texture for physically and biochemically protected C fractions. Because soil texture acts directly and indirectly on various protection mechanisms, it may not be a universal predictor of whole-soil C content.

Abbreviations: CPOM, coarse particulate organic matter > 250 µm in size • d-clay, easily dispersed clay-sized fraction • d-silt, easily dispersed silt-sized fraction • fPOM, fine particulate organic matter 53–250 µm in size • POM, particulate organic matter • µagg-clay, microaggregate-derived clay-sized fraction • µagg-silt, microaggregate-derived silt-sized fraction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SOIL ORGANIC MATTER can be protected from decomposition and stabilized in soils by different mechanisms, including chemical protection by association with mineral surfaces, physical protection by occlusion within aggregates, and biochemical protection by recalcitrance (Jastrow and Miller, 1997; Six et al., 2002; Krull et al., 2003). Chemical stabilization of organic molecules through mineral-organic matter binding is well established (Ladd et al., 1985; Gonzalez and Laird, 2003). Even labile organic material that would otherwise decompose quickly can be protected from decomposition by close association with silt and clay particles (Sørensen, 1972). Analyses synthesizing multiple studies suggest that stabilization capacity is dictated by soil silt and clay content and the surface area and reactivity of mineral soil particles (Hassink, 1997; Kiem et al., 2002; Kiem and Kögel-Knabner, 2002; Six et al., 2002). Several studies have shown that soil texture influences aggregation (Kemper and Koch, 1966; Chaney and Swift, 1984; Schlecht-Pietsch et al., 1994) such that increased clay contents were associated with increased aggregation or aggregate stability. In increasing soil aggregation, soil clay content indirectly affects soil C storage by occluding organic materials, making them inaccessible to degrading organisms and their enzymes. Therefore, soil texture (particularly soil clay content) plays direct and indirect roles in chemical and physical protection mechanisms.

Unprotected soil C is not intimately associated with soil mineral particles and is not occluded within aggregates. Unprotected soil C can be defined operationally as free particulate organic matter (POM), which includes rapidly metabolized plant and associated microbial carbohydrates and more recalcitrant molecules derived from resistant plant materials and microbial decomposition products (Golchin et al., 1994; Six et al., 2001). Biochemically resistant C, defined operationally as organic C resistant to acid hydrolysis (Leavitt et al., 1996), is an average of 1300 to 1500 yr older than whole-soil C (Paul et al., 1997; Paul et al., 2001). Even in the presence of cometabolites, specialized enzymes, and optimum environmental conditions, decomposition of this material is slow, resulting in turnover times on the order of centuries to millennia. This pool of organic C is often associated with silt and clay minerals (Paul and Clark, 1989) but is protected from decomposition primarily due to its complex chemical structure rather than by the mineral association.

These observations suggest soil texture affects chemical and physical protection of soil C stocks, whereas unprotected C and biochemically protected C should vary largely independent of soil texture. The principle of soil texture altering soil C levels and decomposition kinetics has been integrated into several biogeochemical models (e.g., van Veen and Paul, 1981; Parton et al., 1987) but has not been fully evaluated across a controlled soil textural sequence. In addition, the means by which texture alters C dynamics in these models reflect only the conceptual chemical protection and do not encompass physical protection mechanisms. There is an increasing demand for new models that incorporate measurable fractions rather than conceptual pools (e.g., Christensen, 1996; Arah and Gaunt, 2001), which has been met with varying degrees of success (e.g., Sohi et al., 2001; Skjemstad et al., 2004). The goal of this work was to examine how soil texture alters the distribution of organic C in physically and chemically defined pools of soil organic matter for the long-term goal of parameterizing the impact of soil texture in a model of soil organic C dynamics based on measurable fractions. While holding C inputs and other factors influencing soil C turnover (temperature, precipitation, litter quality, tillage, etc.) reasonably constant, we evaluated the effects of texture on unprotected and protected soil C stocks. The fractionation scheme outlined by Six et al. (2002) aims to isolate pools of organic C based on physical, chemical, and biochemical mechanisms of protection of organic C and was applied to samples from two in situ soil texture gradients. Specifically, soil texture effects on unprotected and physically protected C are evaluated by testing the hypothesis that particulate organic matter content varies largely independent of soil texture. Soil texture effects on chemically protected organic C are evaluated by testing the hypothesis that silt- and clay-associated soil C stocks are directly related to silt and clay content. The hypothesis that the proportion of silt- and clay-associated C that is nonhydrolyzable does not vary with soil texture is tested to assess the impact of texture on biochemically protected C.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sites and Sampling
Two sites were selected where in situ soil texture gradients could be generated by sampling spatially while keeping all other variables constant. The soils of the Saskatchewan site developed under native grasslands, whereas the soils at the Ohio site developed under native forest. Both sites have been under long-term agricultural production, and each consists of a small localized area with reasonably consistent parent materials, topography, and climate but with significant variability in soil texture. Land management, including tillage type, tillage frequency, tillage timing, and cropping histories, were similar within each area. Analysis of crop yield and aboveground biomass data from the Saskatchewan gradient (McConkey and Brandt, personal communication) showed only minor differences between textures and that interannual variability was greater than the variability across textures. We therefore assumed that crop yields and soil organic C inputs were reasonably constant within each gradient. The result of the site selection is that the dominant factor influencing soil C turnover is soil texture.

Saskatchewan
Soil samples were collected in April 2002 from six farm fields located near Stewart Valley, Saskatchewan, Canada (50°17' N; 107°48' W). All sites were within 20 km of each other and were conventionally tilled under long-term (90+ years) cereal-pulse crop rotation including fallow periods (usually wheat [Triticum aestivum], fallow, and fieldpea [Pisum sativum]). Soils in the area are classified as Aridic Borolls (Septre, Fox Valley, Birsay Haverville, and Birsay Hatton series). Clay mineralogy in this area is dominated by montmorillonite, with some kaolinite and illite (Brierley et al., 1996). With the assistance of staff from Agriculture & Agri-Food Canada, we used soil survey maps and hand texturing in the field to select six locations stratified across a soil texture gradient that represented the treatment levels (i.e., textures). Within each of the six treatment locations, three transects were sampled and treated as replicates. Six surface soil (0–20 cm) cores were collected for each replicate and separated into 0–5 cm and 5–20 cm subsamples in the field. Due to soil compaction of two clayey textures when using a Giddings soil probe, we hand-dug pits and sampled the soil horizontally at the two depths. Only the surface (0–5 cm) subsamples were further analyzed because these samples were expected to demonstrate the largest differences in organic matter protection due to physical protection because of increased wet-dry and free-thaw cycles at the soil surface.

Ohio
A second textural gradient was sampled in October 2002 from within a single 400-ha field near Cygnet, OH, located approximately 15 km SW of the Ohio Agricultural Research and Development Center's Northwest Agricultural Research Station near Hoytville, OH (41°0' N; 84°0' W). The soils of the field consisted of Mollic and Aeric Epiaqualfs (Hoytville and Napanee series), all managed under a corn (Zea mays), soybean (Glycine max), oat (Avena sativa) rotation with conventional tillage for the last 20 yr. Clay minerals in these soils are dominated by illites (Collins et al., 2000). Similar to the Saskatchewan site, soil survey maps and hand texturing were used to identify four locations within a single farmer's field, which represented the treatment levels (i.e., soil textures). Three transects within each of the treatment levels were sampled and treated as replicates. Seven surface soil (0–20 cm) cores were collected from each replicate sample and separated into 0- to 5- and 5- to 20-cm subsamples in the field. Only the surface samples were further analyzed; the subsurface samples were archived.

Soil Textural Analysis
Once returned to the laboratory, individual soil core samples were weighed and subsampled for moisture content, and bulk density was determined using the core volume. The six to seven individual cores for each replicate were composited to form the replicate sample. The field-moist soil samples were passed through an 8-mm sieve by gently breaking apart the soil. The samples were air-dried, sieved to 2 mm, and stored at room temperature.

Soil texture was determined using a modified version of the standard hydrometer method without removal of carbonates or organic matter (Gee and Bauder, 1986). Briefly, 30 g of air-dry, 2-mm sieved soil were shaken for 16 h in a 250-mL Nalgene bottle with 100 mL of 5 mg L–1 sodium hexametaphosphate and 10 glass beads (10 mm in diameter). Soil clay contents were determined using hydrometer readings taken at 1.5 and 24 h and appropriate interpolation calculations. Coarse and fine sand contents were determined by pouring and washing the suspension over 250-µm and 53-µm sieves after the sedimentation was completed. Materials retained on the sieves were oven dried at 60°C and weighed. Soil silt contents were determined by difference.

Soil Physical and Chemical Fractionations
Microaggregate Isolation
Microaggregates were isolated using a method described by Six et al. (2000a). Briefly, 50 g of air-dried whole soil were submerged in deionized water for 30 min to promote slaking of macroaggregates and poured onto a 250-µm mesh screen inside a cylinder and reciprocally shaken (120 rev min–1) with 50 glass beads (10 mm in diameter) until the complete disruption of all macroaggregates was achieved. Disruption of microaggregates was prevented by a continuous flow of water that immediately flushed the <250-µm material out of the shaker and onto a 53-µm sieve (Six et al., 2000a). The fraction retained on the 250-µm mesh consisted of coarse (POM and 250–2000 µm sand) and comprised the coarse POM fraction (CPOM >250 µm). The materials retained on the 53-µm sieve were wet sieved by hand for 2 min at approximately 50 cycles per minute, and finer materials were gently washed off to isolate stable microaggregates (53–250 µm). The suspension passing the 53-µm sieve was centrifuged to isolate the easily dispersed silt-sized fraction (d-silt, 2–53 µm). The supernatant was flocculated using 0.25 M MgCl2 and CaCl2 and centrifuged to isolate the easily dispersed clay-sized fraction (d-clay, <2 µm). Fraction suspensions were oven dried at 60°C and weighed. Mass and organic C balances were used to determine the completeness of recovery after the microaggregate isolation procedure.

Dispersion of Microaggregates and POM Isolation
Five to six grams of microaggregate (53–250 µm) samples, isolated in the previous procedure, were dispersed by shaking for 18 h with 25 mL of 0.5 g mL–1 sodium hexametaphosphate and 12 glass beads (4 mm in diameter) in 50-mL centrifuge tubes to isolate fine POM (Cambardella and Elliott, 1992) and microaggregate-derived mineral fractions. After shaking, the suspension was poured over a 53-µm sieve and washed thoroughly to isolate the fine POM fraction (fPOM, 53–250 µm), which includes fine sand and fine POM that was originally outside the microaggregates and the occluded POM that is released on dispersion of the microaggregates. The suspension that passed the 53-µm sieve was centrifuged as described previously to isolate the microaggregate-derived silt- and clay-sized fractions (µagg-silt and µagg-clay). Fractions were subsequently oven dried at 60°C and weighed. Mass and organic C balances of the POM isolation procedure were used to determine the completeness of recovery.

The microaggregate-derived fractions were corrected for sand content to determine C concentrations on a true microaggregate basis. Fine sand contents (53–250 µm) of the samples determined during particle-size analysis were subtracted from the mass of the microaggregate fractions determined during the first isolation procedure. The organic C concentrations of the fine POM and microaggregate-derived silt- and clay-sized fractions could then be expressed on a sand-free microaggregate basis:

The sand correction was not applied on a whole-soil basis because we were concerned only with the composition of the microaggregates and the distribution of organic C within them as a function of changing soil clay content.

Acid Hydrolysis
Easily dispersed and microaggregate-derived silt- and clay-sized fractions were subjected to acid hydrolysis to isolate a resistant pool of organic C using a modification of the method described in Paul et al. (1997) without the pretreatment for removal of carbonates. Briefly, 0.5 g of sample was refluxed at 95°C for 16 h in 25 mL of 6 M HCl. When insufficient material was recovered during previous fractionation steps, less material (down to 0.3 g) was used or individual replicates were combined. After refluxing, the suspension was filtered and washed with deionized water over a glass fiber filter. The residue was then washed from the filter into a specimen cup, oven dried at 60°C, and weighed. The proportion of nonhydrolyzable C was determined using the following equation:

which accounts for mass loss during acid hydrolysis and recovery of residues. Mass loss was found to be minor.

Carbon and Nitrogen Analyses
Total C and N analyses were done on the whole soil and each isolated fraction using a CHN analyzer (model LECO CHN-1000; Leco Corp., St. Joseph, MI). Results of soil carbonate determination by the pressure transducer method (Sherrod et al., 2002) indicated that carbonates were not present (data not shown), and thus total C concentrations can be equated to organic C concentrations.

Statistical Analyses
Linear relationships between the sample mass content of various fractions and the organic C contents within these fractions were tested using ordinary least squares linear regression. Linear regression was used in spite of not having an explicit independent variable because we sought only the presence or absence of a relationship and because the error around the X-axis variable was found to be lower than that on the Y-axis variable. The linear relationships were in the form Y = {alpha} + ßX, where X was whole-soil clay or silt content and Y was the organic C content within various fractions. Data from the Saskatchewan texture gradient in Stewart Valley (SK) and the gradient in Hoytville (OH) were generally analyzed separately. Overall comparisons between the sites or between fractions within a site (all textures combined) were done using Student t tests assuming equal variances and were considered statistically significant at P < 0.05. Differences in the response of organic C contents in various fractions to soil texture were tested using standard ANOVA techniques.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Whole-Soil Texture and Organic Carbon Concentration
The range of soil textures at the Saskatchewan site was wider than at the Ohio site (Table 1). Soil clay contents in the Saskatchewan texture gradient ranged from 159 to 603 g clay kg–1 soil and ranged from 249 to 492 g clay kg–1 soil at the Ohio site. Whole-soil organic C ranged from 7.0 to 18.8 g C kg–1 soil at Saskatchewan and from 16.5 to 26.5 g C kg–1 soil at Ohio (Table 1). The relationship between soil texture, as represented by whole-soil silt plus clay content, and the total amount of organic C stored within the soils (Fig. 1 ) was statistically significant in the Ohio (P = 0.012, r2 = 0.48) and Saskatchewan (P = 0.0028, r2 = 0.46) texture gradients.


View this table:
[in this window]
[in a new window]
  		Table 1. Bulk soil characteristics of 0–5 cm surface soil samples (mean ± SD, n = 3).

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 1. Relationships between whole-soil organic C concentration and soil fine fraction (silt + clay) content for the Saskatchewan and Ohio texture gradients.

 
Organic Carbon Concentrations of Physically Isolated Fractions
Recovery of mass (as CPOM, microaggregates, d-silt, and d-clay) after the microaggregate isolation procedure was 97.9 ± 1.1% (mean ± standard deviation), and total organic C recovery was 96.9 ± 9.8%. Mass recovery (as fPOM, µagg-silt, and µagg-clay) after the POM isolation from microaggregates procedure was 98.8 ± 1.5%, and organic C recovery was 98.5 ± 14.5%. The high variance of C recovery in the POM isolation procedure was due primarily to the high variability in the amount of fPOM obtained, and its organic C concentration was due to interfering sand contents. The two-step physical fractionation scheme was successful in isolating noncomposite pools of organic C (Smith et al., 2002) because C recovery was high and because there was no redundancy in organic C allocation between the fractions. However, their usefulness as modelable, functional pools can be assessed only when their dynamic behaviors become properly described.

The mass distributions of samples after physical fractionation reflect their textural composition (Fig. 2 ). In general, the proportion of mass associated with the coarse (>250 µm) and fine (53–250 µm) sand plus POM fractions decreases with increasing clay content, whereas the mass of the microaggregate-derived silt increased with increasing clay content in the Saskatchewan soils and the mass of the easily dispersed silt-sized fraction increased with increasing clay content in the Ohio soils.



View larger version (58K):
[in this window]
[in a new window]
  		Fig. 2 Soil mass distribution of physical fractions isolated from soils in the Saskatchewan and Ohio texture gradients. Samples are ranked in order from highest to lowest whole-soil sand content (lowest to highest silt + clay content) within sites.

 
For all textures combined, approximately 76% of the organic C in the fractions was associated with mineral fractions (silt and clay-sized fractions); this was nearly the same at both sites (P = 0.99) (Fig. 3 ). Across all textures within the Saskatchewan texture gradient, organic C was greater in the microaggregate-derived silt- and clay-sized fractions than in the easily dispersed fractions (49% of the organic C stock versus 27%, P < 0.001), whereas the reverse was true in the Ohio gradient (30% versus 47%, P < 0.001). Within the Ohio texture gradient, the proportion of the organic C associated with microaggregate-derived or easily dispersed mineral fractions did not seem to differ with increasing soil clay content (P = 0.39 and r2 = 0.075 for microaggregate-derived and P = 0.068 and r2 = 0.29 for easily dispersed mineral fractions), but significant trends were observed in the Saskatchewan gradient (P < 0.001 and r2 = 0.79 for microaggregate-derived and P < 0.001 and r2 = 0.74 for easily dispersed mineral fractions). As the soil clay content increased, the proportion of organic C associated with the microaggregate-derived fractions increased from approximately 38 to 65%, with a concomitant decrease in the easily dispersed mineral fractions from 34 to 18%.



View larger version (34K):
[in this window]
[in a new window]
  		Fig. 3 Organic C concentrations of (a) coarse sand and POM, (b) fine sand and POM, (c) microaggregate-derived silt-sized, (d) microaggregate-derived clay-sized, (e) easily dispersed silt-sized, and (f) easily dispersed clay-sized fractions isolated from soils in the Saskatchewan and Ohio texture gradients. Samples are ranked in order from highest to lowest whole-soil sand content (lowest to highest silt + clay content) within sites.

 
The organic C concentrations of individual isolated fractions showed varying responses to soil texture in the two soils (Fig. 3). In the Saskatchewan texture gradient, statistically significant relationships were found for the fPOM, µagg-silt, and µagg-clay fractions (three out of six fractions). The µagg-silt fraction was the only one with a positive slope, whereas the organic C stocks in the fPOM and µagg-clay fractions decreased with increasing whole-soil silt + clay content. In the Ohio texture gradient, organic C concentrations decreased in the µagg-clay fraction, increased in the d-silt fraction, and showed no significant trends with increasing soil silt + clay content in the remaining fractions.

After correction for sand particles, the mass of microaggregates increased with increasing whole-soil clay content in both texture gradients (P < 0.001, r2 = 0.93 for Saskatchewan and P < 0.001, r2 = 0.76 for Ohio; data not shown). Total sand-free microaggregate-associated C decreased with increasing soil clay content in the Saskatchewan soils, whereas no trend was observed in the Ohio soils (Fig. 4 ). The trend in the Saskatchewan soils is attributable to decreases in fine POM and in the µagg-clay associated C.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4 Organic C concentrations of (a) fine sand and POM, (b) microaggregate-derived silt-sized, and (c) microaggregate-derived clay-sized fractions on a sand-free microaggregate mass basis. Samples are ranked in order from highest to lowest whole-soil sand content (lowest to highest silt + clay content) within sites.

 
When the organic C concentrations of individual fractions are expressed on a per mass of fraction basis rather than on a per mass of soil basis, microaggregate-derived and easily dispersed silt- and clay-sized C concentrations significantly decreased with increasing content of the fraction in the soil (Fig. 5 and 6) , although the relationship in the µagg-silt fraction from the Ohio gradient was weaker (P = 0.085, r2 = 0.27). When compared with each other, these trends did not differ between the easily dispersed versus the microaggregate-derived fractions.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5 Relationships between organic C concentrations (g C kg–1 fraction) in the easily dispersed and microaggregate-derived silt-sized fractions and whole-soil silt contents for the (a) Saskatchewan and (b) Ohio texture gradients.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 6 Relationships between organic C concentrations (g C kg–1 fraction) in the easily dispersed and microaggregate-derived clay-sized fractions and whole-soil clay contents for the (a) Saskatchewan and (b) Ohio texture gradients.

 
Acid Hydrolysis of Physically Isolated Fractions
The proportion of organic C remaining after acid hydrolysis treatment of whole-soil samples was slightly higher (P = 0.084) in the Ohio soils than in the Saskatchewan soil (Table 2). Nonhydrolyzable C increased with increasing whole-soil clay content in both soils (P = 0.007 in the Saskatchewan texture gradient and P = 0.006 in the Ohio gradient). No relationships between whole-soil clay content and the nonhydrolyzable C of the silt- and clay-sized fractions from either soil were observed, regardless whether the fractions were microaggregate derived or easily dispersed. No differences in the proportion of nonhydrolyzable C between easily dispersed and microaggregate-derived fractions were observed for the silt-sized (P = 0.12) or clay-sized (P = 0.46) fractions of both soils. Overall, the silt-sized fractions had higher proportions of nonhydrolyzable C than the clay-sized fractions (P < 0.001).


View this table:
[in this window]
[in a new window]
  		Table 2. The proportion of nonhydrolyzable carbon (%) in various mineral fractions (mean particle-size distribution [g kg–1] SD, n = 3).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Soil Texture Effects on Whole-Soil Carbon Concentration
Although the degree of association of SOM with soil mineral surfaces (particularly soil clays) has long been recognized as a mechanism for the stabilization of organic C, whole-soil clay content is not always a good predictor of whole-soil organic C concentration. The direct evidence for the long-term effect of soil texture on organic C storage is derived primarily from correlations in soil databases and is inconsistent. Nichols (1984) found a strong correlation (r = 0.86) between soil clay content and organic C concentration in the Southern Great Plains. Percival et al. (2000), however, found that soil clay content explained little of the variation in organic C accumulation (r2 < 0.05) in New Zealand. The current study showed a significant relationship between soil texture and soil organic C concentration at Ohio and Saskatchewan. The slope of the relation from the Saskatchewan soils was slightly lower than in Ohio, although the difference was not statistically significant (P = 0.27). Lower organic C concentrations and the slightly lower slope in the Saskatchewan gradient are likely due to the lower C input levels in this gradient. Wheat straw plus grain yields in the general vicinity of the Saskatchewan texture gradient average between 1.8 and 5.3 Mg ha–1 (Campbell and Zentner, 1993), whereas corn grain yields in the area near the Ohio texture gradient average between 6 and 10 Mg ha–1 (Dick et al., 1997). These differing crop types and yields result in contrasting organic C inputs to the two gradients. These were estimated to be 90 to 140 g C m–2 yr–1 in Saskatchewan and 310 to 420 g C m–2 yr–1 in Ohio, based on crop yield data and formulas relating yields to above- and belowground biomass similar to those reported by Kong et al. (2005). In the wheat production system in Saskatchewan, C inputs are relatively low due to low mean annual precipitation and the inclusion of fallow periods in the crop rotations. In Ohio, greater mean annual precipitation and continuous corn or corn-soybean rotations result in more C being returned to the soil. Therefore, even if there might be an increased capacity for stabilization due to greater soil clay content, the relationship between whole-soil C concentration and clay content was not fully expressed in the SK gradient because there is little C input available to be stabilized.

Soil Texture Effects on Unprotected and Physically Protected Carbon
Few researchers have reported on the effect of soil texture on the amount of POM. Although they found no relationship between whole-soil organic C and soil texture in a series of cultivated Cambisols, Kölbl and Kögel-Knabner (2004) found that the amount of organic C present as POM occluded in aggregates increased with increasing soil clay content. In contrast, they found that the amount of organic C present as free POM was not related to clay content. We did not distinguish between free and occluded POM in the current study but separated the total POM based on size. In terms of mass, the coarse (>250 µm) and fine (53–250 µm) POM fractions decreased with increasing clay content in the Saskatchewan and Ohio soils. This is more likely attributable to changes in sand content of the fraction rather than the POM itself. Organic C concentrations in the POM fractions showed no relationship with texture, except the fPOM fraction in the Saskatchewan gradient, which showed a slight decrease. These results seem to contradict those previously reported in the literature, which generally showed increases in total POM with increasing soil clay content (e.g., Needelman et al., 1999). Our results are consistent with those reported by Franzluebbers and Arshad (1997), who showed no response in total POM with soil texture. It is likely that a direct causal relationship between soil clay content and POM-associated C does not exist but that the relationship is indirect through the effects of soil clay on aggregation. In the Saskatchewan texture gradient, the combined results of increased microaggregate mass and decreased organic C with increased soil clay content suggest a trend similar to that observed in the silt- and clay-associated organic C. The increased microaggregation diluted the associated C because the silt- and clay-sized materials comprising the microaggregates had decreasing organic C concentrations. In addition, our results suggest that with reasonably similar C inputs, increasing microaggregation due to greater soil clay contents across the gradient diluted the amount of fine POM rather than promoting its increased retention.

Examination of the total clay recovery from the microaggregate and POM isolation procedures reveals lower clay contents than those determined in the particle-size analysis. This suggests that dispersion was incomplete during the physical fractionation and that clays are likely to be found in silt-sized microaggregates. Complete dispersion of the soil was not the objective of the physical fractionation but does limit the inferences that can be made concerning organic matter associated with the silt-sized fraction. The microaggregate-derived, silt-sized fraction increased dramatically in weight and C associated with it across the Saskatchewan gradient, whereas in the Ohio gradient the easily dispersed silt-sized fraction increased (Fig. 2). Thus, the structural unit in which C is accumulated seems to differ between the two texture gradients. We suggest that this is related to the overall difference in the range of textures between the two gradients; the soil textures at the Ohio site were sandier than at the Saskatchewan sites. In the less sandy soils of Saskatchewan, there was a higher capacity for the formation of stable microaggregates within macroaggregates (Oades, 1984; Six et al., 2000a) than in the coarser soils of OH. Therefore, organic C was preferentially protected in silt-sized aggregates occluded within the microaggregates, whereas silt-sized aggregates, with the associated occlusion of C, would form directly outside of microaggregates at the Ohio site.

Soil Texture Effects on Chemically Protected Carbon
The most surprising results of the study were those of the silt- and clay-associated soil organic C. Soil texture is often used as a surrogate for surface area and reactivity, particularly when the mineralogy of the clays are similar. The limited surface reactivity of a soil is provided as evidence for the potential existence of a limited stabilization capacity (Hassink and Whitmore, 1997; Six et al., 2002). The second hypothesis of this study was that silt- and clay-associated soil C contents would be directly related to whole-soil silt and clay content, such that any increases in total soil C could be attributed to increases in silt and clay contents, which themselves would have similar organic C concentrations. What we found instead were decreasing concentrations of organic C within the silt- and clay-associated fractions and therefore a mixed response in terms of soil C concentrations. In other words, given reasonably consistent organic C inputs within each site, the silt- and clay-associated organic matter became diluted with greater amounts of silt and clay across the gradients. A soil that has reached its stabilization capacity would have a constant organic C concentration in the silt- and clay-associated fractions, and thus any changes in texture would be reflected in the whole-soil saturation capacity. The soils in the current study seem to be far removed from such a stabilization capacity.

Soil Texture Effects on Biochemically Protected Carbon
Acid hydrolysis is often proposed as a chemical means of isolating a fraction of biochemically or microbially resistant organic matter. We originally hypothesized that the proportion of silt- and clay-associated C that is nonhydrolyzable would not vary with soil texture. Our results supported this hypothesis because no significant differences in proportions of nonhydrolyzable C were found across textures or between the sources (microaggregate-derived versus easily dispersed) of the silt- and clay-sized fractions. Acid hydrolysis is incapable of accounting for physical and chemical protection mechanisms because organic matter sorbed to mineral surfaces from easily dispersed fractions was equally susceptible to acid hydrolysis compared with those from microaggregate-derived reactions. However, our results showing that organic matter associated with silt-sized fractions was more resistant to acid hydrolysis than clay-associated organic matter support reports in the literature (Tiessen and Stewart, 1983; Anderson and Paul, 1984; Christensen and Sorensen, 1985; Six et al., 2000b) that concluded that silt-associated organic matter is the most stable fraction. Therefore, the relative contribution of silt and clay to the soil texture might alter the total amount of nonhydrolyzable C, and thus the size of the biochemically protected pool does indeed vary with texture. This supports the practice of some models of soil organic matter dynamics to use soil texture to modify the proportion of organic matter in the stable pool. However, this outcome may be confounded by the relative amounts of C associated with the silt- and clay-sized fractions.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
We found no significant relationship between soil texture and unprotected (coarse- and fine-POM) organic C. The role of texture in the physical protection of soil C seemed to differ between the Saskatchewan and Ohio texture gradients. In Saskatchewan, a higher capacity to form stable microaggregates resulted in increased organic C in the microaggregate-derived silt-sized fraction with increasing soil silt and clay content. In the Ohio gradient, the easily dispersed silt-sized fraction showed the greater response. Mineral-associated, or chemically protected, organic C was strongly affected by the soil silt and clay contents, as would be expected through the role of surface properties. As the amount of silt or clay increased, the concentration of organic C associated with easily dispersed and microaggregate-associated silt or clay fractions decreased. Biochemically protected (nonhydrolyzable) C increased with increasing clay content in whole-soil samples, but the proportion of nonhydrolyzable C within silt- and clay-sized fractions was unchanged. Our results suggest that soil texture, as represented by soil clay or silt + clay content, may not always be a good predictor of whole-soil organic C content. This is likely because soil texture affects organic C storage through direct and indirect mechanisms. For an assumed constant level of C inputs within each of the texture gradients observed, the relationship between mineral surface area, as expressed by soil silt and clay content, and soil organic matter seem to vary according to the mechanisms by which the organic matter is stabilized in the soil, whether by predominantly physical, chemical, or biochemical protection.


   ACKNOWLEDGMENTS
 
The authors thank Brian McConkey and Kelsey Brandt from Agriculture & Agri-Food Canada for site assistance in Saskatchewan; Matthew Davis, Frank Thayer, and Nathan David from the Ohio Agricultural Research and Development Center for site assistance in Ohio; and Shane Cochran, Joyce Dickens, Mike Katz, Sarah Moculeski, and Jodi Stevens for laboratory assistance during the soil fractionations. This project was supported by the Office of Research (BER), U.S. Department of Energy Grant no. DE-FG03-00ER62997, and Grant no. DE-FG02-04ER63912.

Received for publication November 23, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
G. Yoo and M. M. Wander
Tillage Effects on Aggregate Turnover and Sequestration of Particulate and Humified Soil Organic Carbon
Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 670 - 676.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
C. E. Stewart, A. F. Plante, K. Paustian, R. T. Conant, and J. Six
Soil Carbon Saturation: Linking Concept and Measurable Carbon Pools
Soil Sci. Soc. Am. J., January 25, 2008; 72(2): 379 - 392.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
A. N. Kravchenko, G. P. Robertson, X. Hao, and D. G. Bullock
Management Practice Effects on Surface Total Carbon: Differences in Spatial Variability Patterns
Agron. J., October 3, 2006; 98(6): 1559 - 1568.
[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 Similar articles in ISI Web of Science
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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Agricola
Right arrow Articles by Plante, A. F.
Right arrow Articles by Six, J.
Related Collections
Right arrow Structure and Properties
Right arrow Soil Organic Matter
Right arrow Other Soil Management

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