Spectroscopic assessment of three ecologically distinct organic carbon fractions of mineral soils

Soil carbon (C) is heterogeneous and exists in various forms along a decomposition continuum from labile fast-cycling compounds to more persistent forms of C, which can reside in the soil for centuries to millennia. Methods for soil organic C fractionation aim to account for this complexity by separating soil organic C into distinct groups that exhibit similar turnover. Our aims were to (a) fractionate three mineral soils with small C concentrations (<2.5 % C), different textures and mineralogy using a granulometric method to derive the particulate organic C in macroaggregates (POC _mac), the particulate organic C in microaggregates (POC _mic), and the mineral-associated organic carbon (MAOC), (b) test if mid-infrared (MIR) spectra can discriminate the distinct organic C fractions and characterise the critical organic and mineral functional groups, and c) explore the interactions between the dominant mineral and organic functional groups to elucidate C stabilisation. With a multivariate analyses we found that the MIR spectra use information from mineral and organic frequencies to discriminate the organic C fractions. Closer investigation on specific regions of the MIR spectrum showed that absorptions relating to silicates were more pronounced in the POC _mac and POC _mic fractions and clay mineral absorptions were stronger in the MAOC fraction. There was little organic C in the POC _mic and POC _mac fractions, respectively, and their spectra showed mostly mineralogical features. Most of the organic C in the soils was present as MAOC. The stretching vibration of the bonds in the alkyl CH ₂ molecule was most prominent. However, absorptions from C = C and C = O stretching vibrations, due to alkenes and amides were also present. These molecules are known to form MAOC. We found that the wavenumbers associated with the absorption of alkyl CH ₂ were positively correlated with the absorption of clay minerals, which may be used to infer the mineral association of organic C. Our results show that MIR spectroscopy can characterise the compositional differences between the C fractions and that the spectra could potentially infer C stabilisation in mineral soils.