The carbon fixation, sequestration, and emission processes in permafrost regions present a great uncertainty in global C cycle and climate models (BERAC 2010, Roberts et al. 2010. The Arctic has been a C sink for millennia, but it could become an important source of CO2 and CH4 to the atmosphere if a warmer climate leads to the release of vast quantities of stored C in excess of the annual net carbon uptake. The nature, magnitude, and rates of these changes in the C cycle will depend on climate-driven changes in Arctic biogeochemical, vegetation, and hydrological processes, creating a critical feedback loop (Grosse et al. 2011).

D. Graham will lead this team.

Primary Participants:

L. Liang (ORNL) will lead tasks measuring rates and mechanisms of subsurface transport processes and carbon interactions with sediment minerals.

B. Gu (ORNL) will lead tasks to measure rates and mechanisms of soil carbon transformation and carbon-mineral interactions.

D. Elias and T. J. Phelps (ORNL) will lead tasks establishing microbial microcosms and mesocosms to measure rates and modes of microbial transformations.

R. Hettich (ORNL) will lead tasks using metaproteomics to compare microbial activities.

E. Brodie and J. Jannson (LBNL) will lead tasks using metagenomics and microbial community analysis.

S. Hubbard, Y. Wu, T. Kneafsey, and S. Nakagawa (LBNL) will lead tasks studying the impact of freeze-thaw processes on SOM dynamics.

M. Torn (LBNL will lead field tasks for integrative measurements of GHG production and C14 analysis.

W. Riley and C. Koven (LBNL) will coordinate closely with the team on biogeochemical process


Postdoctoral research associates and technical support staff at ORNL and LBNL will assist with each of these tasks.

Biogeochemistry Goal: Develop a quantitative model of organic matter decomposition rates in high-latitude soils with underlying permafrost, as needed to improve predictions of CO2, CH4, and N2O GHG feedbacks on changing Arctic ecosystems.

Close interactions between modeling, field observation, and laboratory measurement teams will build a framework for the initialization, parameterization, testing, and improvement of biogeochemical process representations in multiple modeling scales. Observations and experimental results will reduce uncertainty in models’ GHG response functions attributable to impacts on C turnover times associated with (1) temperature, moisture, and aqueous chemistry (e.g., pH and redox potential); (2) SOM structure, C cascades, and intrinsic turnover times; (3) sorption and physical protection by soil minerals; and (4) freeze-thaw cycles and microbial adaptation. The investigations of fundamental biogeochemical processes will be represented in greatest detail at the fine scale of models, which most closely addresses the spatial scale of core sampling and subsequent experimentation. However, the integrated knowledge from many of these studies will be directly applicable to process representations and parameterizations in the intermediate- and climate-scale models.

Soil and Groundwater Sampling. During the first year, observations and measurements will focus intensively on a single geomorphic unit, a low-centered polygon in BEO Site 1. This wet meadow tundra contains Typic Aquiturbel or Histoturbel soil, with an active layer ~20–55 cm deep (Hubbard et al. 2012). Replicate frozen core samples (approximately 7.7 cm diameter by 1 m depth) will be removed from the center, ridge, and trough regions of this low-centered polygon using a SIPRE auger (redesigned with DOE support) and hydraulic drill (Hughes and Terasmae 1963, Bockheim and Hinkel 2007). Core samples will be shipped frozen to laboratories for the analyses and experiments described below. In addition to intensive sampling at this site, representative core samples will be removed from nearby high-centered and transitional polygonal tundra on the BEO.

A basic suite of analyses will be performed on these samples. To initialize models, organic carbon and nitrogen composition, bulk density, pH, texture, microbial community profiling, and soil water content will be measured in soil and permafrost horizons to a depth of approximately 1 m (Burtt 2011, Johnson et al. 2011). These results will be used to initialize models at all scales.

In collaboration with the hydrology and geomorphology group (Task HG4), groundwater wells or piezometers will be installed to sample water from the surface to the permafrost table. These water samples will be compared with analyses of soil water from core samples. Geochemical measurements performed will include pH; ionic composition and concentration; oxidation-reduction potential; dissolved O2, CO2, CH4, and H2 concentrations; dissolved organic matter (DOM) and particulate organic matter (POM) concentrations; and chemical characteristics (Rinnan and Rinnan 2007); and fluid electrical conductivity measurements. Concentrations of redox-sensitive species such as nitrate-nitrite-ammonium, sulfate-sulfide, Fe(II/III) and Mn(II/IV) will be analyzed in the field or in the laboratory. Interpretation of the biogeochemical wellbore-based measurements will be performed in the context of the microtopography and spatially variable subsurface environmental variables (e.g., soil moisture, temperature) that will be characterized as described in Section IV.3.2, “Hydrology and Geomorphology.”

In subsequent years, sampling will focus on additional geomorphologically distinct features and interstitial tundra identified from fine-scale elevation and remote-sensing data. The controlling factors for biogeochemical processes will be compared across DTLBs of various ages to parameterize fine- and intermediate-scale models.

SOM Turnover Times and GHG Fluxes in Thawing Soils. In all soils, some organic matter degrades more rapidly than others; therefore, models such as CLM conceptualize belowground C as residing in several interconnected pools with varying intrinsic decomposition rates (turnover times) (Jenkinson and Coleman 2008, Parton et al. 2010). Predicted C storage and fluxes in these models depend critically on how (1) these turnover times and their dependencies on local conditions are formulated, (2) the C cascade is designed, and (3) the belowground N cycle is formulated (Thornton and Rosenbloom 2005).

The physical and chemical differences among these pools are poorly defined (Trumbore 2009, Kleber et al. 2011) because SOM consists of heterogeneous C sources with varying chemical compositions, structural characteristics, degrees of polymerization, water solubility, and mobility (Ping et al. 1997). Most Alaskan permafrost soils are high in organic C but have a low degree of humification compared with temperate soils (Tarnocai et al. 2009). Therefore, a large proportion of SOM can be mineralized in Alaskan soils, including both extractable and nonextractable fractions (Dai et al. 2000). Recent studies have indicated that the CO2 and CH4 released into the atmosphere from these permafrost soils have Δ14C values that are characteristic of older, buried C (Schuur et al. 2009, Wahlen et al. 1989). Therefore, accurately modeling the turnover times of SOM pools will be key to predicting GHG emissions from thawing permafrost. The following measurements will improve the representation of C cycling in Arctic tundra by identifying SOM structural and adsorption properties that control the bioavailability of C pools.

Soils from each horizon will be fractionated for mineralogical analysis and spectroscopic analysis of SOM structure. We will use advanced spectroscopic techniques, such as 2D excitation- emission (EEM) fluorescence and Fourier-transform infrared (FTIR) spectroscopy, to interrogate chemical and structural changes of soil carbon and its degradation rates in response to soil warming and microbial degradation. A proposal for the Pacific Northwest National Laboratory Environmental Molecular Sciences Laboratory (EMSL) high-resolution mass spectrometry and 1H- and 13C-nuclear magnetic resonance user facilities will be developed to extend structural characterization of SOM. The δ13C and Δ14C values of significant SOM pools will be measured in collaboration with the Center for Accelerator Mass Spectrometry (CAMS) at Lawrence Livermore National Laboratory (LLNL) (Guilderson and McFarlane 2009). The ∆14C values of carbon in litter, roots, and organic matter will be used to estimate long-term ecosystem residence times of organic C with depth (Torn et al. 2002, Trumbore 2009).

Dissolved ionic species, such as Fe2+ and Ca2+, facilitate aggregation of organic matter in water. Depending on pH and mineralogy, DOM also interacts with the surface of sediment grains (Gu et al. 1994, Sollins et al. 2009), which could alter its susceptibility to microbial attack and thus its preservation and translocation in the permafrost soil. To test this hypothesis, the interactions between organic matter and sediment minerals will be analyzed. First, intact soil grains will be examined using micro-Raman and FTIR spectroscopies to probe the bonding mechanisms between organic C and mineral phases. Second, in batch experiments, dissolved organic compounds will be extracted and subsequently added to different mineral fractions of permafrost and active layer soils. The extent of sorption and stabilization will be determined. Geochemical conditions such as Eh (aerobic or anaerobic redox potential), pH, and dissolved ionic species are expected to influence C-surface reaction processes and will be considered in the design of these experiments. The bioavailability of both the desorbed and adsorbed compounds, and of the original SOM, will be determined using long-term incubation experiments. By separating carbon into bioavailable desorbed compounds and protected adsorbed compounds in each POM and mineral-associated organic matter fraction, these experiments will distinguish measureable C pools, enabling predictive modeling of the dynamics of SOM degradation and transformation from one C pool to another.

The response functions for CO2 and CH4 production in current land models require additional parameterization of temperature and oxygen controls on SOM decomposition rates in Arctic soils. Microcosms containing homogenized soils from major active layer and permafrost horizons will be incubated at temperatures measured during the thawing shoulder season and summer, as reported by thermistor arrays and archived data from the CALM project (Hinkel and Nelson 2003). CO2 and CH4 production rates will be calculated using biweekly gas chromatography measurements, and changes in the concentration of chemical redox species [Fe(II)/Fe(III), and Mn(II)/Mn(IV)] will be determined at the end of the 6–8 week incubation period. Changes in microbial community composition will be monitored using high-throughput molecular phylotyping in collaboration with the Joint Genome Institute. The isotopic composition of CO2 and CH4 released into the headspace will be analyzed at CAMS to estimate the ecosystem age of carbon decomposed in the active layer vs that from recently thawed permafrost material.

In addition to microcosm incubations, mesocosms will be established using intact cores. A two-stage cooling apparatus will be constructed to gradually thaw soil cores to an equilibrium approximating the vertical thermal gradient during mid-summer. Access ports will permit sampling, and instrumentation arrays will monitor changes in temperature, water content, and chemistry along the vertical profile. Periodic sampling of the headspace, soil, and pore water along the length of the column will be used to measure changes in concentrations of gases and solutes, quality of SOM, and the microbial community during the controlled thaw. These time-series data will be used to parameterize and test 1D models of the soil column.

Methanotrophic bacteria oxidize CH4 to CO2. Methane biogeochemistry models recognize a high level of uncertainty surrounding this important process (Riley et al. 2011). Methane oxidation will be measured in microcosms prepared using soil samples taken from different depths and proximity to the rhizosphere to assess root-stimulated methanotrophic microbial communities (Wagner et al. 2005). Microcosm incubations will also be used to determine response functions to changes in temperature, CH4 and O2 pressure, pH and water saturation. Molecular markers of methanotrophy will be measured using quantitative, real-time polymerase chain reaction (RT-PCR) with degenerate primers specific for methane monooxygenase genes and stable isotope probing, using 13CH4 or 13CH3OH (Kelly and Wood 2010). Measurements of methane oxidation rates will be coupled to metagenomic and metaproteomic analyses to identify changes in methanotroph populations and activity that will parameterize fine-scale models (Graham et al. 2011).

During the summer field seasons, integrated CO2 and CH4 fluxes will be measured from the surface of the intensively studied polygonal tundra. These GHG emissions will be measured using chambers and laser-based infrared gas analysis (Natali et al. 2011). The isotopic composition of collected CO2 will be measured at CAMS to compare the age of mineralized C with the age of SOM C from soil horizons and fractionated pools.

Eddy covariance systems will be used to measure 30-minute average net fluxes of CO2, CH4, latent heat, and sensible heat with a footprint on the order of 100 × 50 m. These measurements will be performed according to standard methods for AmeriFlux type systems (Sachs et al. 2008, Vourlitis and Oechel 1997, Wille et al. 2008, Fan et al. 1992), and the results will be used to derive landscape-scale functions of the entire ecosystem response to light, water, and temperature. These functions will be used, through an up-scaling approach, to derive scale-dependent parameters in the intermediate and climate-scale models.

Some information relevant to our understanding of biogeochemistry dynamics will also be available at the climate modeling scale. Starting in 2012, the NASA Carbon in Arctic Reservoirs Vulnerability Experiment (CARVE) will fly over the North Slope of Alaska with an airborne remote-sensing payload that includes an L-band radiometer/radar and a nadir-viewing spectrometer to measure surface parameters that control gas emissions (i.e., soil moisture, freeze/thaw state, surface temperature) and total atmospheric columns of CO2, CH4, and CO. NGEE Arctic scientists are in discussion with the CARVE PI (C. Miller) to share data, since the NGEE Arctic surface eddy flux data will be very useful for CARVE and the CARVE transects will provide a statistical sampling for testing NGEE Arctic models at the climate scale.

Soil Freeze-Thaw Effects on Decomposition and SOM Distribution. Water and ice are heterogeneously distributed in the subsurface. This spatiotemporal variability of soil wetness and ice volume impacts soil pore water flow, chemistry (including redox potential), tension, and soil temperature. During freezing and the binding of water in ice crystals, ions are expelled and concentrate in the remaining liquid phase (Price 2007). Even at subzero ambient temperatures, liquid water exists within permafrost as a very thin film surrounding sediment and ice. This unfrozen water can facilitate mass transport and retard the thermal response of the active layer or permafrost (Romanovsky and Osterkamp 2000). These extreme conditions support nutrient transport and even microbial activity (Ponder et al. 2008). Therefore, at temperatures near 0°C the soil water freezing point is believed to control the temperature response to microbial activity (Nicolsky et al. 2007, Koven et al. 2011, Matzner and Borken 2008). Models are beginning to represent the temperature and moisture controls on decomposition rates near the soil freezing point, but these response functions require parameterization (Lawrence et al. 2009).

Large, temperature-controlled column experiments will be designed (aligned with Task HG2) to examine the effects of freeze-thaw processes on SOM degradation mechanisms and rates, N dynamics, and associated GHG production under different soil characteristics (texture, porosity), liquid water and ice contents, and freeze-thaw cycle characteristics. The ~1 m long columns, which will consist of continuous core of active layer sediments underlain by permafrost, will be subject to several freeze-thaw cycles to mimic in situ environmental conditions. Sampling ports will be installed along the length of the column and at the effluent to nondestructively assess the key hydrogeological, biogeochemical, and geophysical properties and their transformations over space and time, including water content, temperature, pressure, pH, DOM, gas flux, NO3– and NH4+, 14C isotopic signatures, and geophysical attributes (dielectric constant and complex resistivity). Parallel columns will be set up to obtain complementary measurements using destructive sampling and computerized tomography (CT) scanning (see Appendix XIV.2), including: ice content, δ13C and H/D ratio of methanogen precursors and products.

The freeze-thaw column experiments are expected to be especially useful for (1) quantifying the mechanisms and rates of SOM decomposition as a function of local environment (e.g., moisture, water phase, solutes, pH, temperature, texture); (2) identifying the vertical location within the column where CO2, CH4, and dissolved organic carbon (DOC) are being produced; (3) characterizing the interacting C and N dynamics and impacts of moisture and temperature on microbial activity at the freeze-thaw boundary; (4) quantifying microbial activity and functional speciation associated with SOM degradation, competition for resources, and local conditions; (5) assessing how vertical variability in rates and mechanisms compares with integrated measures, such as effluent fluxes; and (6) quantifying the geophysical signatures of environmental characteristics that control organic carbon degradation (i.e., water content, water phase, soil textures) and the distribution of organic carbon degradation products (i.e., gas bubble volume, dissolution/ precipitation), which is a necessary step prior to the use of these methods at the field scale for nondestructive biogeochemical-hydrological monitoring. The response functions for these freeze-thaw experiments will be incorporated directly into model decomposition algorithms for models at all spatial scales.

Carbon-Nitrogen Interactions. Nitrogen availability is predicted to be a strong control on plant photosynthesis, growth, and respiration in the Arctic. Thawing permafrost could release nitrogen into the active layer, stimulating plant growth and microbial activity (Keuper et al. 2012). Parameterizing and improving models’ response functions to nitrogen speciation in the active layer and thawed permafrost is a priority for representing couplings of the carbon and nitrogen cycles (Xu et al. 2011, Thornton et al. 2007). Reports of N2O greenhouse emissions suggest that microbial nitrification and denitrification can be significant mineralization processes in some environments (Elberling et al. 2010). In cooperation with the vegetation dynamics group (Task V3), the biogeochemistry group will identify key components of the Arctic nitrogen cycle that respond to thawing permafrost. In addition to the measurements of organic and inorganic nitrogen species in soil horizons and permafrost, microcosms and soil columns (described above) will be used to investigate nitrogen mineralization and mobilization. Microcosms of active layer soil amended with 15N-labeled tracer substrates like NH4+, NO3–, N2, nucleosides or amino acids will be established. Mass spectrometry will be used to measure 15N incorporation into microbial biomass. Metagenomic and metaproteomic analyses will be used to identify microbial populations that rapidly assimilate these labeled nitrogen sources (Mackelprang et al. 2011, Banerjee and Siciliano 2012). If significant amounts of N2O are detected from samples, then nitrification and denitrification processes will be interrogated.

Phase 1 Deliverables

  • Geochemical and microbial characterization of permafrost core samples obtained from study sites for model parameterization.
  • Temperature response function for GHG production in soil columns and microcosms developed and compared with field measurements.
  • Characterization of SOM pools and turnover times to develop a predictive model of SOM decomposition and availability.
  • Response function for soil freeze-thaw effects on SOM transport, decomposition, and GHG production developed using soil columns.

Measurements of key nitrogen species in core samples and microcosms to parameterize models and prioritize N cycle studies.