Using pyrolysis molecular beam mass spectrometry to characterize soil organic carbon in native prairie soils
Magrini, K.; Follett, R.; Kimble, J.D.vis, M.; Pruessner, E.
Soil science 172(9): 659-672
The goal of this study was to test if analytical pyrolysis coupled with molecular beam mass spectrometry and multivariate statistical analyses could provide a rapid and accurate methodology to identify and quantify soil carbon fractions and understand the fundamental chemistry that distinguishes these fractions. We analyzed soil organic carbon (SOC) contained in well-characterized agricultural soils with pyrolysis molecular beam mass spectrometry (py-MBMS) and then determined correlations between the mass spectra and associated soil characterization data. Both soil carbon chemistry and the organic forms in which SOC is sequestered (soil microbial biomass (SMBC), particulate organic matter carbon (POM C), and mineral-associated carbon (Cmin C)) were assessed by multivariate statistical analyses to discover existing correlations and if they could be developed into estimative models. The sample set consisted of well-characterized soils collected from native prairie sites in the western U.S. Corn Belt and Great Plains: 11 sites located within 8 midwestern states (CO, NE, IA, ND, MT, TX, MO, and MN). Sample characterization parameters included site, depth, %SOC, POM C, Cmin C, SMBC, and SOC calendar age (determined from 14C age). Correlations were found for samples collected across this large geographic region (at or greater than 0.90) for SOC, POM C, Cmin C, and SMBC. Soil organic carbon calendar age derived from radiocarbon-14 dating could be estimated for ustollic soils from MT, NE, and CO. These soils also contained deeper and younger eolian layers, whose ages were correctly estimated with this technique. The Py-MBMS analysis additionally showed that soils developed from water-sorted sediments on a tilled-floor lake plain (lacustrine soils) were significantly different from the other samples.