Skip to main content
ARS Home » Research » Publications at this Location » Publication #355029
Title: Predicting the decomposability of arctic tundra soil organic matter with mid infrared spectroscopy

Author
item MATAMALA, ROSER - Argonne National Laboratory
item JASTROW, JULIE - Argonne National Laboratory
item Calderon, Francisco
item LIANG, CHAO - Chinese Academy Of Sciences
item Fan, Zhaosheng
item MICHAELSON, GARY - University Of Alaska
item PING, CHIEN-LU - University Of Alaska

Submitted to: Soil Biology and Biochemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/22/2018
Publication Date: 2/20/2019
Citation: Matamala, R., Jastrow, J.D., Calderon, F.J., Liang, C., Fan, Z., Michaelson, G., Ping, C. 2019. Predicting the decomposability of arctic tundra soil organic matter with mid infrared spectroscopy. Soil Biology and Biochemistry. 129:1-12. https://doi.org/10.1016/j.soilbio.2018.10.014.
DOI: https://doi.org/10.1016/j.soilbio.2018.10.014

Interpretive Summary: Due to the cold and wet environment, large amounts of carbon have been preserved in the arctic soils over millennia. There is a need to better understand the potential susceptibility to decomposition and loss to the atmosphere of this soil organic matter. In this study, we investigated the capability of mid infrared spectroscopy (MIR) to quickly predict soil carbon and nitrogen concentrations and the short-term carbon losses of the soil organic matter from tundra soils. Soils collected from four tundra sites on the North Slope of Alaska were incubated atseveral temperatures for 60 days. Mathematical models were used to see how well the spectroscopic data could be to predict the carbon losses from the soils upon incubation. Our results suggest that MIR is a potentially powerful tool for quickly and accurately estimate the short-term carbon losses of tundra soils.

Technical Abstract: Large amounts of soil organic matter have been preserved in the arctic soils over millennia time scales due to the cold and wet environments limiting decomposer activity. With the increase in high latitude warming due to climate change, there is a need to better understand the potential susceptibility to mineralization of this soil organic matter. In this study, we investigated the capability of mid infrared spectroscopy (MIR) to quickly predict soil carbon and nitrogen concentrations and the short-term carbon mineralization potential of soil organic matter of tundra soils. Soils collected from four tundra sites on the North Slope of Alaska were incubated at 1, 4, 8 and 16 °C for 60 days. Total soil organic carbon (SOC) and total nitrogen (TN) concentrations, soluble extractable organic matter (SEOM), 60-day carbon mineralization potential (CMP), and the MIR spectra of the soils were measured. The MIR spectra were calibrated against short-term CMP and the other measurements. Partial least square (PLS) regression models constructed using all spectral data were excellent predictors of SOC and TN concentration of the soils. The calibration model for CMP was optimized upon splitting the dataset into six categories -1) TOC > 10 %, 2) TOC < 10 %, 3) TN < 0.6 %, 4) TN > 0.6 % concentrations, 5) acidic tundra, and 6) non-acidic tundra- that were determined by principal component analysis. The best CMP PLS models were found for soils with TOC < 10 % and TN < 0.6 %. Analysis of the factor loadings and standardized beta coefficients from these two CMP PLS models indicated a small number of influential spectral bands. These bands were associated with clay contents, phenolic OH, aliphatic, silicates, carboxylic acids, and polysaccharides. Our results suggest that MIR is a potentially powerful tool for quickly and reasonably estimating the short-term CMP of tundra soils.