Prediction of soil organic matter and clay contents by near-infrared spectroscopy - NIRS

Abstract

Among the soil constituents, special attention is given to soil organic matter (SOM) and clay contents, since, among other aspects, they are key factors to nutrient retention and soil aggregates formation, which directly affect the crop production potential. The methods commonly used for the quantification of these constituents have some disadvantages, such as the use of chemical reactants and waste generation. An alternative to these methods is the near-infrared spectroscopy (NIRS) technique. The aim of this research is to evaluate models for SOM and clay quantification in soil samples using spectral data by NIRS. A set (n = 400) of soil samples previously analyzed by traditional methods were used to generate a NIRS calibration curve. The clay content was determined by the hydrometer method while SOM content was determined by sulfochromic solution. For calibration, we used the original spectra (absorbance) and spectral pretreatment (Savitzky-Golay smoothing derivative) in the following models: multiple linear regression (MLR), partial last squares regression (PLSR), support vector machine (SVM) and Gaussian process regression (GPR). The curve validation was performed with the SVM model (best performance in the calibration based on R² and RMSE) in two ways: with 40 random samples from the calibration set and another set with 200 new unknown samples. The soil clay content affects the predictive ability of the calibration curve to estimate SOM content by NIRS. Validation curves showed poorer performance (lower R² and higher RMSE) when generated from unknown samples, where the model tends to overestimate the lower levels and to underestimate the higher levels of clay and SOM. Despite the potential of NIRS technique to predict these attributes, further calibration studies are still needed to use this technique in soil analysis laboratories.

Soil texture and organic matter prediction using Mehlich‐3 extractable nutrients

“…Hence, the increased R 2 in the prediction model, when adding Mehlich‐3 extractable Na and Cu, and pH to EstCEC, might be related to cations occupying variable charges associated with clay particles. The clay prediction models' accuracy using EstCEC and Mehlich‐3 extractable nutrients (Figure 1; Table 2) produced similar results as previous research that used different prediction methods (Chang et al., 2001; Lazzaretti et al., 2020; Pinheiro et al., 2017; Pirie et al., 2005; Terra et al., 2015), indicating the potential use of routine soil analysis data to predict soil clay content.…”

“…The moderate goodness-of-fit statistics (Figure 4) highlight the complexity of SOM and suggest that soils with similar physical-chemical characteristics and management could improve the prediction accuracy. Lazzaretti et al (2020) reported that soil clay content affects the predictive ability of a model to estimate SOM content by NIRS, suggesting that separating soils into groups with similar characteristics can enhance SOM prediction accuracy. Unlike NIRS, where soil mineralogy and particle size can affect the spectra and the predictions (Dalmolin et al, 2005;Lazzaretti et al, 2020;Sousa et al, 2008), SOM prediction based on routine soil-test information might be more influenced by crop management practices and fertilization.…”

“…SOM contents, including near-infrared spectroscopy (NIRS) (Jaconi et al, 2019;Lazzaretti et al, 2020;Zhao et al, 2021), microscope (Sudarsan et al, 2016), smartphone (Swetha et al, 2020), image acquisition systems, and machine learning algorithms (John et al, 2020). These methods are accurate in their predictions (i.e., R 2 = 0.68-0.98), but special instruments are required for such analyses or calibrated results are available only for a small range of soils and geographic locations.…”

“…However, the length of time needed to complete the measurements of SOM and clay content with standard methods (e.g., LOI and hydrometer) increases the turnaround time for sample analyses in soil testing laboratories as well as the cost of soil analysis and, therefore, might be impractical for some soil test laboratories. Research has sought alternative, indirect measurement methods for determining the clay and SOM contents, including near‐infrared spectroscopy (NIRS) (Jaconi et al., 2019; Lazzaretti et al., 2020; Zhao et al., 2021), microscope (Sudarsan et al., 2016), smartphone (Swetha et al., 2020), image acquisition systems, and machine learning algorithms (John et al., 2020). These methods are accurate in their predictions (i.e., R 2 = 0.68–0.98), but special instruments are required for such analyses or calibrated results are available only for a small range of soils and geographic locations.…”

“…SOM is generally determined by weight loss on ignition (LOI) or estimated from the C content determined by wet or dry combustion (Nelson & Sommers, 1996). While effective, these analytical methods may use hazardous chemical reagents, be laborious and time‐consuming to execute, or both (Demattê et al., 2019; Lazzaretti et al., 2020).…”