Can mid-infrared (mid-IR) spectroscopy evaluate soil conditions by predicting soil biological properties?

Abstract

A massive suite of dimensions demonstrating the biological properties of soils is required to significantly quantify the holistic view of soil conditions. The value of measuring the full suite of soil conditions with conventional methods will be too costly. We reviewed the skill of mid-infrared reflectance spectroscopy (MIRS) to predict numerous soil biological properties. Results coupled with partial least square regression (PLSR) revealed that mid-infrared reflectance spectroscopy (MIRS) could predict biological activities like soil microbial biomass, soil enzymes, soil respiration, Q10 and others. Based on calibration errors relative to the mean analyte values, the results for biological properties measures are as good as those found in soil chemical and physical capability. The MIR spectroscopy can be utilized to create adjustments for exact determination of compositional boundaries, for example, C and total N, while for soil biological properties, for example, enzyme activities, more effort is necessary to characterize the constraints and usefulness of MIR spectroscopy. Overall, it tends to be utilized in situations where the highest accuracy is not essential. We concluded, that MIR spectroscopy seems to produce more precise calibration for the confirmation of biological properties in soils and be less influenced by sample diversity for some analytes.

Keywords

• Mid-infrared spectroscopy (MIR)
• Soil respiration (SR)
• Q10
• Coefficient of determination
• Root mean square error
• Relative prediction deviation