Understanding Composition and Decomposability of Arctic Soils by Integrating Laboratory Incubations with New Measurements and Models

The large stocks of soil organic carbon (SOC) in the permafrost region are sensitive to changes in global temperature and permafrost thawing. Furthermore, because the relative importance of SOC stabilization mechanisms operating in the permafrost region differ from those of other ecoregions, the composition and potential decomposability of SOC are key uncertainties in models projecting SOC release from thawing permafrost soils. In the search for indicators of decomposability and soil physicochemical properties that could help upscale and improve model predictions across the region, we demonstrated that mid-infrared (MIR) spectroscopy is very sensitive to the degradation state of SOM and is a good predictor of short-term C mineralization from tundra soils. In this study, we are conducting long-term incubations to explore indicators of decomposability and evaluate factors affecting mineralization kinetics for a range of permafrost soils collected from soil horizons occurring at depths from 0 to 3 meters in ice-wedge polygons in Utqiaġvik, AK. The soils were incubated aerobically for over one year at two temperatures (15°C and 5°C). Soil physicochemical properties such as soil particle density, total organic carbon, total nitrogen, and the chemical composition of the SOC were also measured. Results show that MIR is an excellent indicator of decomposability in the long-term. Predictive models of cumulative CO2 production, expressed on a soil mass basis, were derived from the MIR spectra of bulk soils for intervals of 1 to 12 months. In addition, cumulative CO2 production was fitted to the Gompertz model to estimate the maximum CO2 production and decomposition rates of the soils. We found that the higher incubation temperature significantly increased the maximum CO2 production, and decreased the decomposition rate variability among soils. We are currently exploring the influence of soil physicochemical properties on the MIR predictions and the Gompertz model parameter estimates. This information will help constrain process model parameters and help couple indicators of SOM decomposability with geo-referenced data for soil properties and associated environmental predictors to create geospatial assessments and maps, which can serve as benchmarks for models at landscape, regional, and global scales.