B53D-01 INVITED 13:40h
Soil Organic Matter Dynamics Following Land-Cover Change in a Subtropical Savanna: Insights from Soil Physical Fractionation and Stable Isotopes
Soil physical structure is an important determinant of soil organic carbon (SOC) storage and turnover due to differential accessibility of SOC to decomposer organisms. Techniques for physical fractionation of soil organic matter in conjunction with isotopic analyses (d13C, d15N) of those soil fractions have been used to (a) determine where organic C is stored relative to aggregate structure, (b) identify sources of SOC, (c) quantify turnover rates of SOC in specific soil fractions, and (d) evaluate organic matter quality. We used these two complementary approaches to characterize soil C storage and dynamics in the Rio Grande Plains of southern Texas where C3 trees and shrubs (d13C = -27 o/oo) have largely replaced C4 grasslands (d13C = -14 o/oo) over the past 100-200 yr. Using a chronosequence approach, soils were collected from remnant grasslands (Time 0) and from woody plant stands ranging in age from 10-130 yr. We separated soil organic matter into specific size/density fractions and determined their C and N concentrations and natural d13C and d15N values. Rates of whole-soil C and N storage in the upper 15 cm of the soil profile averaged 10-30 g C/m2/yr and 1-3 g N/m2/yr, respectively, over the past 130 yr of woodland development. These rates of accumulation have increased soil C and N pool sizes in older wooded areas by 80-200 o/o relative to remnant grasslands. The relative proportions of the free light fraction (density less than 1.0 g/cc) and macroaggregate fraction (greater than 250 um) increased linearly with time following woody plant invasion of grassland. Conversely, the relative proportions of free microaggregate (53-250 um) and free silt+clay (less than 53 um) fractions decreased linearly with time after woody invasion, likely reflecting incorporation of these fractions into macroaggregates. C and N concentrations in all soil fractions increased with time following woody invasion, but most of the C and N accumulated in light (density less than 1.85 g/cc) particulate organic matter (POM) fractions not protected by stable soil structure. Mean residence times (MRTs) of soil fractions were calculated based on changes in their d13C with time after woody encroachment. The shortest MRTs (mean = 30 yr) were associated with all POM fractions not protected within aggregates. Fine POM (53-250 um) within macro- and microaggregates was relatively more protected from decay, with an average MRT of 60 yr. All silt+clay fractions had the longest MRTs (mean = 360 yr) regardless of whether they were found inside or outside of aggregate structure. d15N values of soil physical fractions were positively correlated with MRTs of the same fractions, suggesting that higher d15N values reflect an increased degree of humification. Increases in whole-soil C and N in wooded areas are probably being sustained by greater inputs and relatively slow turnover of POM, perhaps due to the biochemical recalcitrance of POM materials and/or nutrient-water limitations to microbial activity. These results indicate that soil structure may provide a mechanistic explanation for C and N processes and dynamics following land cover changes in terrestrial ecosystems.
B53D-02 14:10h
Dynamics of Biopolymer Turnover in Soil Physical Fractions Following Land-Cover yChange in a Subtropical Savanna
Changes in the apportionment of organic carbon and nitrogen among soil physical yfractions following land-cover shifts are of critical importance to the debate surrounding ythe capacity of terrestrial ecosystems to store or release greenhouse gases. For example, ythe difference between the mean residence times (MRTs) of light particulate organic ymatter (POM) vs. silts and clays is typically quite large, with silt and clay associated yorganic matter having the longest MRTs and the greatest likelihood to contribute to long yterm carbon storage. A few studies in agricultural and forest systems have demonstrated ythat biopolymer chemistry also varies along physical, as well as density, fractionation ygradients. We quantified changes in biopolymer (lignin, suberin and cutin, and yhydrolysable amino acids) chemistry of size and density fractionated soil from the Rio yGrande Plains of Texas where C4 grasslands (d13C = -14 %) have undergone succession yto subtropical thorn woodland dominated by C3 trees/shrubs (d13C = -27 %) over the ypast 150 years. This natural isotopic distinction was used to determine MRTs of free ylight organic matter (density less than 1.0 g/cc), macroaggregate (greater than 250 um), ymicroaggregate (53-250 um) and silt+clay (less than 53 um) fractions (see Liao et al., ythis session) which were then related to their specific biopolymer chemistries. Our yresults illustrate that lignin and aliphatic biopolymers (as measured by hydroxyl fatty yacids) are apportioned differently among size/density fractions and along the successional ychronosequence. Lignin is incorporated into all soil fractions soon after woody yencroachment, whereas aliphatic components are slow to be incorporated in the silt and yclay fractions. The lignin components that do become associated with silts and clays are, yin general, highly oxidized. Differences in foliar chemistry among the plant sources yindicate selective movement of leaf cutins into POM, macro- and microaggregate yfractions, but not into free or intra-aggregate silts and clays. Selected analyses of silt and yclay fractions for hydrolysable amino acids showed differences along the ychronosequence, with total hydrolysable amino acids comprising 30-45% of total ynitrogen. It is possible that amino and phenolic compounds are tightly bound to the silts yand clays (the fractions with the longest MRT) and repel the more hydrophobic and less ywater soluble cutin and suberin monomers, thereby restricting turnover. These results yprovide new insights regarding the interactions between soil structure, chemistry, yturnover, and preservation of soil organic matter. y
B53D-03 14:25h
Land use Effects on Storage, Stability and Structure of Organic Carbon in Soil Density Fractions Revealed by 13C Natural Abundance and CPMAS 13C NMR
The type of land use and soil cultivation are important factors controlling organic carbon storage (SOC) in soils and they can also influence the relative importance, the structure, and the stability of different SOC pools. The objectives of our study were: i) to quantify the SOC stocks in different density fractions (mineral-associated soil organic matter $>$ 2 g cm-3 (Mineral-SOM), free particulate organic matter $<$ 1.6 g cm-3 (free POM), light occluded particulate organic matter $<$ 1.6 g cm-3 (occluded POM$<$1.6) and dense occluded particulate organic matter 1.6 to 2.0 g cm-3 (occluded POM1.6-2.0)) of silty soils under different land use (spruce forest, grassland, maize, wheat), ii) to determine the structure of these SOC fractions by CPMAS 13C NMR spectroscopy, and iii) to analyse the stability of these SOC fractions in the maize soil on the basis of the stable isotope composition of SOC. The SOC concentration in the A horizon increased in the order wheat (12.7 g kg-1) $<$ maize (13.0 g kg-1) $<$ grassland (24.5 g kg-1) $<$ spruce (40.5 g kg-1). The major part (86-91%) of the SOC was associated with the heavy mineral fraction at the grassland, maize and wheat site. In the A horizon of the spruce soil, the particulate organic matter accounted for 52% of the total SOC content. The chemical structure of the soil organic matter (SOM) was influenced by litter quality, the intensity of litter decomposition and the related production and storage of microbially-derived substances. SOM of the acid forest soil was characterized by large amounts of POM with a high content of spruce litter-derived alkyl C. In the biologically more active grassland and maize soil, litter-derived POM was decomposed more rapidly and SOC stocks were dominated by mineral-associated SOM which contained greater proportions of aryl and carbonyl C. The cultivation of the grassland soil induced enhanced mineralization of POM and in particular of mineral-associated SOM. The faster SOC turnover was associated with a relative accumulation of aromatic and carbonyl C structures in the mineral-bound SOM. In all soils, the free particulate organic matter had a smaller proportion of alkyl C and a larger proportion of O-alkyl C than the particulate organic matter occluded in aggregates. The mean age of the SOM in the density fractions of the maize soil increased with increasing aromaticity in the order free POM (22 yr) $<$ occluded POM1.6-2.0 (49 yr) $<$ mineral-associated SOM (63 yr). The results showed that the type of land use influenced the distribution pattern of litter carbon to functionally different SOM pools which represented different stages of SOM decomposition and humification. Additionally, the type of land use influenced the chemical structure of SOM in soil density fractions. Thus, the effect of land use on SOM storage should not only be assessed in terms of total C stocks but also with respect to changes of SOC structure, stability and function.
B53D-04 14:40h
Soil Carbon Stabilization During Tropical Reforestation
Land use change can affect global atmospheric carbon (C) concentrations by changing the quantity and residence time of C stored in plant biomass and in soils. Currently, large areas of agricultural and pastureland in the Neotropics are being abandoned and replaced by secondary forests. The prevalence of secondary forests has focused attention on reforestation as a potential C sink. Our research examines the importance of physical and chemical mechanisms of soil C storage during forest re-growth. Using a long-term successional chronosequence, we sampled replicate pastures, old growth forests, and forests re-growing on pastures abandoned 10, 20, 30, 60 and 80 years ago. We tested a two-compartment, isotopic mixing-model to determine contributions of C$_{3}$-C (forest) and C$_{4}$-C (pasture) to the total soil C pool at each site. There were no significant differences in the relative proportions of C$_{3}$ and C$_{4}$-C when using the isotopic values of fresh leaf litter, forest-floor leaf material or roots. Using site specific$^{13}$C values or average pastures and old-growth forests as end-members also did not affect the estimates of soil C gain and loss. After 10 years of reforestation, half of the pasture-derived C was lost from the top 10 cm of mineral soil. In the following decades, the loss of C$_{4}$-C was slower, following an exponential decay model, so that all pasture-derived C would be lost after 100 years of forest re-growth. Loss of C$_{4}$-C was compensated by a gain of new forest C, resulting in no net change in soil C content down to 1 m depth during 80 years of succession. Although changes in the bulk soil C stock were undetectable, land use change can affect the distribution of different C fractions and hence, residence time. We have used a density fractionation procedure to separate the bulk soil C into a free light fraction, a free occluded fraction released after the disruption of aggregates, and a mineral-associated heavy fraction. The pasture sites had lower quantities of the light fractions as a percentage of total soil mass than the forest sites, suggesting that these fractions may be depleted faster in the pastures. Litter quality may be the controlling factor in determining turnover rates in pastures versus forests, as we found no significant differences in the structural stability of soil macroaggregates with land use, while litter chemistry did differ. Pasture tissues had significantly lower concentrations of hydrophobic compounds and lignin, which play an important role in soil C stabilization. At all sites, the lighter fraction tended to be depleted in $^{13}$C relative to the mineral-associated C, suggesting a forest-C origin or a selective accumulation of plant compounds during decomposition. A better understanding of soil C dynamics during reforestation will enhance our capacity to rehabilitate degraded soils and improve our assessment of their role as C sinks.
B53D-05 14:55h
Carbon isotopes and the measurement of biochemically protected fractions
Several chemical fractionation methods purport to isolate soil carbon that is biochemically resistant to decomposition. These fractions are often 1300-1500 years older than whole soil C. The standard model of soil carbon dynamics suggests that biochemcially resistant C should not be lost quickly upon cultivation or land use change, decompose rapidly during laboratory incubations, or accrue as fast as whole soil C. But, in some cases these fractions decompose or build up rapidly. What explains these unexpected changes? What are the implications for modeling measureable soil C fractions? What are the general implications for our understanding of soil C dynamics? We review the literature and evaluate new data from field studies and laboratory incubations to address these important questions.
B53D-06 15:10h
Carbon Structural Investigations of Concentric Layers Within Macro-aggregates From Forest and Agricultural Soils
Much of the current research on the potential of agricultural and forest soils to act as sinks for greenhouse gases focuses on the capacity of the systems to form long-term stabilized fractions of soil organic matter (SOM). One proposed mechanism is that carbon is sequestered within soil aggregate interiors during the aggregation process. Repeated wetting-drying cycles change internal pore geometries and associated microhabitats and create more stable macro-aggregates. Research by Smucker and coworkers (EGU Abstracts, 2004) suggest that the exterior portions of aggregates contain greater concentrations of C and N than their interiors, establishing gradients of \13C values across these aggregates. We present the results of a study to test if there exists molecular evidence of such gradients. Soil samples from forest, conventional tillage (CT) and no tillage (NT) agriculture ecosystems in Hoytville and Wooster LTER sites were gently sieved into various size fractions. Soil macro-aggregates (6.3-9.5mm) were peeled, by mechanical erosion chambers, into concentric layers and separated into exterior, transitional and interior regions. Alkaline CuO oxidation was used to determine the composition of lignin, suberin, and cutin biopolymers to determine changes in source and degradative states of SOM. Preliminary results indicate that both soils show similar relative yields of lignin and hydroxyl fatty acids with a greater abundance of lignin than cutin and suberin acids. Greater abundances (per 100mg organic carbon) of CuO products were observed in the native forest than in either agricultural system. The lignin in the NT agricultural soil was least oxidized, followed by the forest soils, then the CT agricultural soils. For both soils, slight trends in biopolymer concentrations were observed between the exterior, transitional and interior regions of the aggregates from the forest and CT or NT ecosystems.
B53D-07 15:25h
The Biogeochemistry of Soil Fe-Redox Cycling: Is Mottling the Visible End Member of Series of Ubiquitous Organo-Fe Reactions?
The redox cycling of iron is well studied in aquatic systems and in soils that are consistently or frequently anaerobic. Such studies describe a prominent redox cycling of Fe, complete with root-associated Fe-oxidizing bacteria, development of "iron plac" surrounding roots, and the production of enormous amounts of acidity that can weather minerals in a process known as ferrolysis. The objective of our research was to investigate the biogeochemical details of a Fe-redox cycle in a generally well oxygenated, upland soil which has prominent mottles at 1.2 to 1.5-m depth. Within the B horizon of this Kanhapludult soil, we found gray (5Y 8/2) microsites proliferated with roots and mycorrhizal hyphae, enriched in kaolinite (0.73g clay/g soil) and organic carbon with relatively modern 14C age, but relatively low in Fe oxides (5mg DCB-Fe/g soil). These gray microsites contrasted with those that were orange (2.5 YR), which had virtually no roots or fungal hyphae, relatively low clay (0.27g clay/g soil) and organic carbon (which had relatively old 14C age), but remarkably high contents of crystalline Fe oxides (37mg DCB-Fe/g soil). Quinoid compounds were present in both microsites suggesting the possibility of redox reactions with polyphenols. This chemical system is suggested to be biogenically initiated by Rhizosphere additions of organic reductants apparently initiates the Fe-redox cycle, which mobilizes enormous quantities of Fe, which upon oxidation generates significant fluxes of protons that are potentially available for mineral weathering. Such systems may be much more common and significant than we appreciate.