Appraisal of Three Proximal Sensing Systems to Estimate Macronutrient Contents of Detached Soybean Leaves

Abstract

ABSTRACT Overapplication of fertilizers is common and may lead to plant toxicity and pollution of water resources in agriculture. A practical method is needed to estimate nutrient stress quickly, easily, and economically for a sustainable management. This study dealt with the prediction of leaf moisture content (MC) and macronutrient contents including nitrogen (N), phosphorus (P), and potassium (K) of soybean leaves using three different optical instruments of chromameter, chlorophyll meter and fluorometer. Forty-five leaf samples from a commercial soybean field were obtained. Color parameters (L*, a*, b*), SPAD (soil plant analysis development) values and quantum yield (QY) values were quantified for each leaf sample. Leaf and soil samples were analyzed using standard leaf and soil chemical analysis procedures. Correlation analysis, principal component analysis (PCA) and partial least square regression (PLSR) were used for data analysis. High correlation was found between two color parameters (L* and b*) with MC and macronutrient contents (r ≥ 0.65) and also, among SPAD, QY, MC, and macronutrient contents (r ≥ 0.70). The prediction models were evaluated based on the RMSEP (root mean square error of prediction) and R 2 (coefficient of determination) values. The results showed that soybean leaf N content can be estimated using all three optical instruments with chromameter giving slightly better results (RMSEP = 0.23%, R 2 = 0.80). Concerning the MC, chlorophyll meter gave slightly better prediction performance (RMSEP = 2.69%, R 2 = 0.73) than the other two instruments. Hence, results suggest that these three optical instruments can be used to assess the macronutrient contents of soybean leaves quickly, easily, and economically.