Soil imVisIR – Quantification of soil organic matter pools in undisturbed soil profiles

Abstract

Organic matter (OM) is an important soil component controlling many chemical and physical properties ranging from nutrient and pollutant sorption to aggregate stability and water holding capacity. Advanced techniques and sophisticated methods have been developed to characterise and separate qualitatively different fractions of soil organic matter SOM(SOM). Despite its importance and the technical progress, two main deficiencies still constrain the comprehensive analysis of OM quantity and quality in soils: 1) Most of the techniques (e.g. solid state 13C NMR spectroscopy or fourier-transformed infrared spectroscopy) are destructive so that the sample is lost after the measurement; and 2) SOM quantity and quality generally show a heterogeneous spatial distribution across many scales which is not considered in an adequate way in most studies since the analytical techniques require substantial amounts of soil material impeding the sampling of small structures. Therefore, a fast imaging technique that non-destructively measures SOM quantity and quality in a non-destructive way with a high spatial resolution on the pedon scale is requiredneeded. We sampled two different sites in a semiarid steppe in Inner Mongolia, China - one continuously grazed site and the other ungrazed since 1979, with both were classified as calcic Chernozems. We expect lLongterm grazing is expected to decrease carbon contents and the most pronounced particulate organic matter fractions (POM) fractions). Stainless steel boxes (100×100×300 mm³) were used to sample undisturbed soil profiles. Two hyperspectral cameras recorded the visible, near and short wave infrared reflectance (400 to 2500 nm in 416 bands) of the profiles with a ground sampling distance of 63×63 µm² per pixel for VisNIR and 256×256 µm² per pixel for SWIR. This procedure was repeated over three vertical slicescuts at a lateral distance of 25 mm through the soil boxes. After each image recording the profile was divided into ten equal squares (each 50 × 50 mm2) and mixed samples were extracted from each square to a depth of 5 mm. Density fractionation is a standard lab approach to separate SOM fractions with different degrees of decomposition and chemical compositions. We used three image processing approaches to spectroscopically identify and quantify different SOM fractions and validate these results with density fractionation, which is a standard lab approach to separate SOM fractions with different degrees of decomposition and chemical compositions.the standard fractionation approach. We show the potential of laboratory imaging Vis-NIR-SWIR spectroscopy for the qualitative and quantitative mapping of SOM fractions on the pedon scale.