Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: August 8, 2005
Publication Date: November 6, 2005
Citation: Calderon, F.J., Venterea, R.T. 2005. Diffuse reflectance mid-infrared spectroscopy (drifts) of soils under different tillage and crop rotation regimes. Agronomy Abstract. Presented at the 2005 ASA.CSSA-SSSA International Annual Meeting. Nov.6-10, 2005. Salt Lake City, UT Interpretive Summary: In this study, we tested the usefulness of diffuse reflectance mid-infrared and near infrared spectroscopy to discern between soils from different crop rotations and tillage treatments. Additionally, we determined if we could develop calibrations for soil variables using both spectral ranges. We have found that while multivariate (PCA) analysis of whole spectra may not result in a clear separation of agronomic treatments, use of specific wavelengths may discriminate between high C soils and low C soils. We have shown that the analysis of soils, which are mostly of mineral composition, benefits from the deliberate selection of useful wavenumbers.
Technical Abstract: Diffuse reflectance mid-infrared spectroscopy (DRIFTS) has been used successfully to quantify soil organic carbon and soil biological properties. In this study we aimed to use DRIFTS as well as diffuse reflectance near infrared spectroscopy to test if spectral differences can be used to characterize management practices such as tillage or crop rotation in a Minnesota soil. In addition, we used Partial Least Squares Regression to test if calibrations could be developed for soil properties such as total C, total N, Denitrification Enzyme Activity (DEA), pH, Dissolved Organic C (DOC), extractable NH4, extractable NO3, moisture, and bulk density. The experiment included a total of 42 samples including two depths a(0-10 cm, and 10-20 cm), three tillage treatments (moldboard plowing, chisel plowing, and no till), and two crop rotations (corn after soybean and continuous corn). Each sample was a composite of 3 soil cores from each plot, and each treatment combination was replicated three times. Samples from an adjacent uncultivated field were also included. The samples were air-dried, ball-milled, and then scanned in the mid-IR and NIR. Principal Components Analysis (PCA) of the whole Mid-IR spectra was not useful to clearly separate the samples according to treatment or compositional parameters. However, restricting the PCA analysis to use wavenumbers from 2800 to 3000 achieved a separation according to soil C. Absorbance at the region between 2800 and 3000 is due to organics (primarily C-H bonds), and this may explain why soils with different C contents are resolved according to absorbance at this wavenumbers. It is important to remember that soils are made up of 95% or more mineral material, so spectroscopic techniques aiming to characterize soil organic or biological properties can benefit by concentrating efforts on useful spectral ranges.