Using Infrared Spectroscopy to Predict Soil Health Indicators in Sodic Soils.

Soil Health in sodic soils is determined by a suite of indicators that include several chemical, physical, and biological properties. Fourier transformed Mid-Infrared spectroscopy (DRIFTS, 4000-400 cm^-1) and Near-Infrared (NIR, 10000-4000 cm^-1) have been used to develop calibrations for soil C, and offer the possibility to quickly predict sodic soil parameters, thus simplifying the evaluation of soil health. This research aims to use NIR and DRIFTS to predict a suite of previously measured shifts in (1) soil properties resulting from anthropogenic erosion that will be used to reclassify Fargo Ryan soil series and (2) chemical, physical, and biological indicators of soil health in a sodic soil under variable land management. Soils were scanned with and without phosphoric acid pretreatment in order to test if predictions for organic C are improved by the removal of inorganic C. The edaphic properties predicted from measured values included: pH, carbonates, total organic carbon (TOC), total organic N (TON), clay, silt, sand, cation exchange capacity (CEC), exchangeable calcium (Ca), exchangeable potassium (K) and electrical conductivity (EC). In addition, phosphatase and β-glucosidase activities were also included in the calibrations. The best calibration R2 were obtained for electrical conductivity and phosphatase. Calibration results show that NIR gives better R2 than DRIFTS for electrical conductivity, and that using spectra from pretreated soils slightly improves the predictions.