Fine spatial resolution mapping of soil organic matter quality in a Histosol profile

Abstract

SummarySoil science lacks a fine spatial resolution imaging technique that is able to measure the quantity and quality of organic matter (OM) for complete soil profiles. We tested whether laboratory Vis‐NIR imaging spectroscopy, together with an unsupervised k‐means classification, can be used to distinguish between different OM fractions in a Histosol profile. A rectangular soil column (22‐cm long) of a folic Histosol (Tangelhumus) was collected from an alpine Norway spruce forest in south‐eastern Germany with a stainless steel box (100 × 100 × 300 mm). A hyperspectral camera (400–1000 nm with 160 bands) with a pixel sampling of 63 × 63 µm was used to acquire the data. We took images of three vertical cuts through the soil profile, each separated laterally by 25 mm. Reference samples were taken at representative locations and analysed for soil organic matter (SOM) quantity and quality with a CN elemental analyser and solid‐state 13C nuclear magnetic resonance (NMR) spectroscopy. Principal component analysis and unsupervised k‐means classifications were used to discriminate between different qualities of OM. We identified three OM fractions based on their reflectance characteristics: living and dead roots with a small degree of decomposition, decomposed particulate OM and decomposed amorphous OM. These fractions were consistent with the morpho‐functional classes of two soil classification systems and can be used for the improved identification of diagnostic horizons. The spectra of the fractions contained additional information on, for example, lignin content and the degree of decomposition. Vis‐NIR imaging spectroscopy is a powerful technique for mapping SOM quality in visually homogeneous organic surface layers.