Prediction of the distribution of soil properties in deep Colluvisols in different pedogeographic regions (Czech Republic) using diffuse reflectance infrared spectroscopy

Abstract

Colluvisols form a significant part of the erosion–deposition soil catena mainly in undulating, agriculturally managed landscapes. Due to their sedimentary origin, they are soils with highly variable properties. For such diverse and often extremely deep soils, it is essential to capture the vertical distribution of properties in as much detail as possible. This study demonstrates the potential of the diffuse reflectance infrared spectroscopy (DRIFT) to derive accurate information on the stratigraphy and selected soil properties (soil organic carbon and calcium carbonate contents and aluminium and iron concentrations) of deep colluvial profiles in three regions in the Czech Republic with different dominant soil types (Chernozem, Luvisol, and Cambisol). The normalization of the spectrum using the quartz spectral band was found to be suitable for assessing the carbonate content in Chernozems and Luvisols (coefficients of determination (R2) were 0.89 and 0.88, respectively). The amount of organic carbon can be predicted from such normalized spectra with less confidence (R2 = 0.64). Organic matter quality as an indicator of their stability and maturity was computed from aliphatic compounds spectral bands (2930 cm–1) and aromatics and CO groups band around 1640 cm–1. The highest values of this indicator were found in Colluvisols in the Chernozem region (> 6, while, e.g., in the Cambisol region it is < 1), especially in the layers corresponding to the oldest sediments or buried in-situ horizons. In the Cambisol region, the buried B horizon was identified based on normalized (i) clays and Fe oxides bands and (ii) wider band of OH groups, which in this case is related to the content of iron oxyhydroxides. The study proved that the applied set of spectral parameters is an effective tool for the description of diverse soil parameters and for identification of the boundaries of individual soil layers. It is equally suitable for carbonate-free soils as well as for soils with higher carbonate contents, which usually cause difficulties in the interpretation of soil organic matter spectral parameters. It provides a comprehensive view of the soil and its mineral and organic components and can provide information that is difficult to measure by other methods.