Biogeosciences [B]

B22C  MW:2007   Tuesday
Soils: Mechanisms of Carbon Stabilization and Response to Climate Change IV
Presiding: E A Holland, National Center for Atmospheric Research; K A Whittinghill, University of Minnesota

B22C-01 INVITED 

Soil carbon as a metric for longterm ecosystem function

* Harden, J W (jharden@usgs.gov), U.S. Geological Survey, 345 Middlefield Rd, Menlo Park, CA 94025, United States Neff, J C), Geological Sciences Dept. & Environmental Studies, Univ. Colorado, Boulder, CO 80309, United States

Soils are responsible for the processing of carbon, water, nutrients, and toxins and as such are a key component of ecosystem resilience. It follows, then, that ecosystem recovery from disturbances by geomorphic, fire, cultivation or urbanization processes should be controlled in part by soil carbon storage and turnover. Soils however, are also highly heterogeneous and are composed of a range of SOM pools with varying turnover times. We suggest that the relative distribution of carbon through these pools is a key characteristic of ecosystems that are more or less resilient following disturbance. Based on a number of soil types around the world, we propose that the most resilient systems are those with large stocks of carbon that turn over in decadal to century time scales. These systems are typically found on early Holocene to late Pleistocene geomorphic surfaces in a wide range of climatic settings. Using this paradigm for soils and ecosystems, we can generalize as to the vulnerability of soil systems across a number of ecosystems. In boreal forest ecosystems, changes in climate, and fire disturbance make ecosystems highly vulnerable to loss of decade-to-century soil carbon, which resides near the soil surface as free light fraction. In these settings, near to complete loss of the intermediate turnover pools is possible because of limited physical protection of the SOM and has repercussions that range from ecological successional dynamics to global C cycling. Dryland soils are similarly vulnerable to change because, like boreal systems, the bulk of SOM is located in pools of light density carbon that are prone to loss following disturbances such as fire and grazing. This paradigm can also be applied to understanding specific mechanisms that destabilize or stabilize this decade-to-century soil carbon. Combustion, erosion, and physical disaggregation (such as by tillage) destabilize while ponding, erosion mitigation, and no-till aggregation help to stabilize this important pool. These examples help to establish a generalized understanding of ecosystem resilience under various disturbance events.

B22C-02 

Influence of overstory on total soil carbon and soil carbon aggregation in temperate forest plantations in Wisconsin

* Stoffel, J L (jlmartin3@wisc.edu), Department of Forest and Wildlife Ecology, 1630 Linden Drive, 120 Russell Laboratories, Madison, WI 53706, United States Gower, S T (stgower@wisc.edu), Department of Forest and Wildlife Ecology, 1630 Linden Drive, 120 Russell Laboratories, Madison, WI 53706, United States

Soil stores the most carbon of all terrestrial systems and we are only beginning to uncover the role of tree species for storing soil carbon and nitrogen. This study characterized the total soil carbon and nitrogen storage and the storage of soil carbon and nitrogen in various aggregate fractions in forest soils of five different species within the top 15 cm of soil. Forest plantations of 45 year old red pine, white pine, European larch, Norway spruce, and red oak were located at the Coulee Experimental Forest in La Crosse County, WI. The carbon and nitrogen content within free microaggregates differs significantly among species. Additionally IPOM (intra- aggregate particulate organic matter from this fraction differ significantly among species with Norway spruce generally having more soil carbon and nitrogen than other species. IPOM nitrogen in for Norway spruce was significantly greater than European larch and red pine, and red oak had significantly more nitrogen than European larch for the same fraction. Other aggregate fractions were not significantly different among species perhaps suggesting that more time would be needed for the organic matter to be incorporated into large and small macroaggregates. The results of this study suggest that different tree species sequester carbon and nitrogen differently within aggregate classes which may have implications for long term carbon sequestration and potentially selection of species for soil carbon storage.

B22C-03 

Hydrologic and Vegetative Effects on the Rate of Soil Carbon Accumulation in Restored Midwestern Grasslands

* O'Brien, S L (sobrie1@uic.edu), Biological Sciences, University of Illinois at Chicago, 845 W. Taylor St M/C 066, Chicago, IL 60607, United States * O'Brien, S L (sobrie1@uic.edu), Biosciences Division, Argonne National Laboratory, 9700 S. Cass Ave, Argonne, IL 60439, United States Jastrow, J D (jdjastrow@anl.gov), Biosciences Division, Argonne National Laboratory, 9700 S. Cass Ave, Argonne, IL 60439, United States Gonzalez-Meler, M A (mmeler@uic.edu), Biological Sciences, University of Illinois at Chicago, 845 W. Taylor St M/C 066, Chicago, IL 60607, United States Grimley, D A (grimley@isgs.uiuc.edu), Illinois State Geological Survey, 615 E. Peabody Dr., Champaign, IL 61801, United States

Revitalization of degraded landscapes may provide sinks for increasing atmospheric [CO2], especially where C inputs resulting from substantial belowground productivity are coupled with soil conditions that promote organic matter stabilization. For three decades, a chronosequence of restored prairies in northeastern Illinois has accumulated soil organic C, but the mechanisms controlling the rate and potential C accrual of the system are unknown. We used a repeated measures approach within the chronosequence to explore controls on C accumulation. The rate of soil C accrual was determined for each of four restored prairies and a field of Bromus inermis (a non-native C3 grass), which together represent a range of ages and drainage conditions. Soils were sampled over a 19-y interval at permanent stations in each plot. Stable isotope signatures of the soil C were used to determine accrual rates for C3- and C4-sourced C. Fifteen to 30 y after planting, prairies on poorly drained soils had accumulated more C in the surface 10 cm (6.8-9.4 Mg C ha-1) than better drained prairies (5.2-6.1 Mg C ha-1). Although the better drained B. inermis field apparently accrued C during the initial decade following establishment, it was at steady state with respect to soil C during the 19-y sampling interval. In the 10-20 cm layer, soil C declined initially but recovered with time, resulting in a total accumulation for the surface 20 cm of 5-7 Mg C ha-1 over the 19-y sampling interval. In the prairies, carbon derived from C4 plants appeared to contribute more to total soil C accumulation than C3-derived C. Variation in the rate of soil C accumulation was related to both soil magnetic susceptibility (a proxy for long-term drainage conditions) and proportion of C4-derived C. We hypothesize that the current absence of soil C accrual in the B. inermis field is due to plant community (lower amount and shorter duration of higher quality inputs) rather than moisture conditions. The faster rate of C accrual in poorly drained soils likely results from some long-term combination of reduced C mineralization during seasonally flooded periods and better plant growth during periods of moisture limitation.

B22C-04 INVITED 

Chemical Processes and Thresholds in Hawaiin Soils

* Chadwick, O (oac@geog.ucsb.edu), University of California, Department of Geography, Santa Barbara, CA 93106-4060,

The Hawaiian Islands are a useful natural laboratory for studying soil development particularly those that can be understood using a matrix of chonosequences and climosequences. The islands are formed over a stationary mantle plume and then are carried to the northwest on the Pacific Plate. Thus the islands get older with distance from the hotspot; Kauai has remnant shield surfaces whose lavas date to about 4,000 ky. It is possible to sample soils that are developing on different age flows ranging from a few hundred years to a few million years. Additionally, individual volcanoes are impacted by differing amounts of rainfall depending on location with respect to the northeasterly trade winds. Whereas rainfall over the open ocean near Hawaii is about 700 mm, rainfall over the Islands ranges from 150 to 11,000 mm. Hawaii is minimally impacted by mineral aerosol additions compared to continental areas and this has a significant impact on soil development. More than 100 soil profiles have been sampled along the Hawaii time-climate matrix with some surprising results. For example, in arid soils might be expected to develop smectite clays, but they are rich in halloysite and allophane. Importantly, these same soils show a trend from high-Mg calcite to dolomite as carbonates accumulate within the profiles – this is one of the first documented occurrences of pedogenic dolomite that is not associated with high levels of salts. It appears that lack of smectite formation lowers the incorporation of Mg into silicate clays and increases its incorporation into carbonates. This is an unusual pedogenic process that seems to be enhanced by the lack of substantial amounts of mica in the basalt derived soils. The only mica is in surface horizons that receive dust derived from distant continents. Without mica there is no template to allow smectite clay formation under the rapid wetting and drying regimes encountered in the arid soils. At the same time that halloysite is forming, iron and aluminum oxides tend to move rapidly from poorly crystalline to crystalline forms, which in turn leads to formation of Oxisols under an arid climate regimes – Torrox formation without substantial climate change. By contrast, soils forming in humid environments along the same time trajectory take much longer to go through the same transformations (allophane to halloysite; poorly crystalline goethite to well crystallized goethite; poorly crystalline gibbsite to well crystallized gibbsite). The longer time required for transformation is related to wet rather than wet- dry cycles and interference by organic carbon in the transformation process. Thus whereas it takes about 400,000 years to form a Torrox, it takes more than three times that long to form a humid-zone Oxisol. In Hawaii we have identified several important thresholds in soil properties that have universal applicability: 1. the shift from udic to perudic soil moisture regime is accompanied by reduction related changes in soil properties particularly accumulation of organic matter and loss of iron-bound phosphorus; 2. shift from ustic to udic moisture leads to rapid loss of nutrients with far reaching implications for soil exchange properties and prehistoric land use, 3. the shift from from ustic to aridic soil conditions leads to greater losses of plant nutrients (bases, P, Si) due to greater wind erosion. Based on archeological evidence, it is clear that Polynesians made land-use decisions that incorporated observations of the soil properties associated with these thresholds.

B22C-05 INVITED 

Soil mineral control on organic matter composition across the long substrate age gradient (LSAG) chronosequence in Hawaii

* Kramer, M (mkramer@es.ucsc.edu), UC Santa Cruz, Dept. of Earth and Planetary Sciences, 1156 High Street, Santa Cruz, CA 93106, United States Marin, A (amarin@gmail.com), UC Santa Cruz, Dept. of Earth and Planetary Sciences, 1156 High Street, Santa Cruz, CA 93106, United States Chadwick, O (oac@geog.ucsb.edu), UC Santa Barbara, Dept. of Geography, Santa Barbara, CA 95064, United States

We examined the composition of organic matter in relation to soil mineralogy across the long substrate age gradient (LSAG) chronosequence in Hawaii. Our objectives were to 1) determine the types of SOM that accumulate across lava flows of increasing age, and 2) examine possible mechanisms for SOM stabilization, and 3) assess the relationship between shifts in soil mineralogy and organic matter composition. Sequentially deeper soil samples were collected across 6 LSAG sites of increasing age (300 y - 4.1My). Samples from major diagnostic soil horizons were analyzed for nitrogen stable isotope ratios and carbon composition using solid- state 13C nuclear magnetic resonance (NMR). Abundance of soil metals and mineral type (primary, non- crystalline, and crystalline) was determined using a sequential chemical extraction procedure. Nitrogen stable isotope ratios, C/N and 13C NMR trends in depth profiles were examined in relation to mineral composition and soil metal abundance.

B22C-06 

Soil Carbon Chronosequences From Post-Agricultural Land in Western New England.

* Clark, J D (jclark3@sas.upenn.edu), The University of Pennsylvania Earth and Environmental Science Department, 240 South 33rd Street Room 451, Philadelphia, PA 19104, United States Johnson, A H (ahj@sas.upenn.edu), The University of Pennsylvania Earth and Environmental Science Department, 240 South 33rd Street Room 451, Philadelphia, PA 19104, United States Richter, S (richter2@sas.upenn.edu), The University of Pennsylvania Earth and Environmental Science Department, 240 South 33rd Street Room 451, Philadelphia, PA 19104, United States Art, H W (jclark3@sas.upenn.edu), Williams College Department of Biology, Thompson Biology Laboratory 59 Lab Campus Drive Room 203, Williamstown, MA 01267, United States

We used quantitatively excavated soil pits to sample chronosequences of post-agricultural northern hardwood forest soils in the Hopkins Memorial Forest, Williamstown, MA, to determine how much carbon was lost during the period of agricultural use, and the rates at which C accumulated after abandonment. We developed chronosequences (based on the time of abandonment) for the three main agricultural uses—cultivated cropland, pasture or hay, and woodlot. Active farms served as our theoretical zero time points and old-growth stands in the region served as the likely maximum. Our data show a significant direct relationship between time since abandonment and carbon amount for the organic horizons (Oe and Oa) of plots that were cultivated, hayed or pastured, but not for stands that were formerly woodlots. There was likewise a significant direct relationship between C content and time for the plowed horizons (0-10 cm) of cultivated ground, but not for the top 10 cm of mineral soils that were formerly in hay/pasture or woodlot. Our best estimates suggest that cultivation reduced the C content of plowed soils by 50% to a depth of 10 cm, and that complete recovery of the soil C pool requires about 120 years.

B22C-07 

Carbon accumulation in bogs and fens after permafrost degradation in central Alaska

* Jorgenson, M T (tjorgenson@abrinc.com), ABR, Inc., PO Box 80410, Fairbanks, AK 99708, United States Cater, T (tcater@abrinc.com), ABR, Inc., PO Box 80410, Fairbanks, AK 99708, United States Roth, J (jroth@abrinc.com), ABR, Inc., PO Box 80410, Fairbanks, AK 99708, United States Pullman, E (epullman@abrinc.com), ABR, Inc., PO Box 80410, Fairbanks, AK 99708, United States Racine, C (cracine@crrel.usace.army.mil), Cold Regions Research and Engineering Laboratory, 72 Lyme Road, Hanover, NH 03755, United States Harden, J (jharden@usgs.gov), U.S. Geological Survey, 345 Middlefield Road ms 962, Menlo Park, CA 94025,

In the discontinuous zone in Alaska about 60% of the land area has permafrost and 5% has thermokarst terrain, which is comprised mostly of thermokarst lakes, bogs and fens. The collapse of lowland forests on seasonally saturated soils on ice-rich terrain into permanently saturated ombrotrophic bogs, dominated by Sphagnum and ericaceous shrubs, and into minerotrophic fens, dominated by herbaceous vegetation, can radically alter the carbon balance of boreal ecosystems. To assess the consequences of permafrost degradation on carbon storage in boreal ecosystems, we investigated accumulation rates of soil organic carbon at various stages of vegetation succession in five bogs and six fens on the Tanana Flats in central Alaska. Bogs ranged from 109 to 637 yrs in age based on calibrated radiocarbon dates. The mean carbon stock of bog peat at the oldest successional stage was 42 kg/m2. Fens ranged in age from 50 to 102 conventional radiocarbon years that could not be calibrated to calendar years. The mean carbon stock of fen peat was 20 kg/m2 at the oldest successional stage. Carbon accumulation rates slowed at later successional stages.

B22C-08 

What Determines the N Isotope Composition of Soil Organic Matter Fractions?

* Dijkstra, P (paul.dijkstra@nau.edu), Department of Biological Sciences, Northern Arizona University PO Box 5640, Flagstaff, AZ 86011, United States Coyle, J S), Department of Biological Sciences, Northern Arizona University PO Box 5640, Flagstaff, AZ 86011, United States LaViolette, C M), Department of Biological Sciences, Northern Arizona University PO Box 5640, Flagstaff, AZ 86011, United States Selmants, P C), School of Forestry, Northern Arizona University PO Box 15018, Flagstaff, AZ 86011, United States Schwartz, E), Department of Biological Sciences, Northern Arizona University PO Box 5640, Flagstaff, AZ 86011, United States Hart, S C), School of Forestry, Northern Arizona University PO Box 15018, Flagstaff, AZ 86011, United States Hungate, B A), Department of Biological Sciences, Northern Arizona University PO Box 5640, Flagstaff, AZ 86011, United States

Soil organic matter, tightly bound to the soil mineral phase, exhibits characteristically high 15N/14N ratios. It has been proposed that this high ratio is the result of repeated microbial processing of organic materials. It is well known that animals are 15N enriched relative to their diet. This enrichment is related to N assimilation, dissimilation, and preferential export of the lighter 14N isotope. We investigated whether similar 15N enrichments exist for the soil microbial biomass using a broad range of ecosystems spanning semiarid, temperate and tropical climates, grassland and forests, and over four million years of soil development. We show that 1- the soil microorganisms are 15N enriched relative to other organic and inorganic soil N pools, 2- this enrichment is variable and related to C:N ratio, 3- 15N enrichment correlates positively with net N mineralization, and 4- 15N enrichment increases during soil incubation. We find, using published data, that similar negative relationships between C:N ratios and δ15N can be found for soil organic matter fractions and animal organisms. We propose that the higher 15N/14N ratios of older soil organic matter fractions reflects repeated microbial processing analogous to C and N transformations occurring in animal food webs.