Application of vibrational spectroscopy in the discrimination of soil samples

The primary aim of this study was to ascertain whether IR spectroscopy integrated with chemometric data analysis can effectively distinguish the impact of land use on the intensity of humus transformation. Additional focus was put on humic acids (HAs) and fulvic acids (FAs) as representatives of soil organic matter. Humic and Fulvic acids were extracted from 12 soils of different land-use: woodland (W), arable land (A) and fallow (F). The FTIR-ATR technique was applied to collect spectra of HA, FA and soil. The spectral data were then used to construct discriminant models applying the Partial least squares - discriminant analysis (PLS-DA) algorithm. The averaged ATR spectra exhibited differences between FA and HA from the three different environments. The distribution of samples in the LV1/LV3 coordinate system of PLS-DA models revealed a clustering of samples based on their origin; a clear separation between the forest-derived samples and the field- and fallow-originating samples was observed. Similarly to the PLS-DA modeling of the spectral data for the extracts, the forest soil samples are distinguishable from the other two groups. FTIR-ATR spectroscopy combined with chemometric analysis is a valuable tool for distinguishing the origin of the soils from which humic substances were extracted. Moreover, the results obtained are consistent with analyses performed on the soils themselves, proving that humic and fulvic acids accurately reflect the properties of soil organic matter.