VIS-NIR-SWIR spectroscopy in the evaluation nutritional status irrigated mango in the Brazilian semi-arid region

Abstract

Leaf analysis is an important instrument that assists in the fertilization management of agricultural crops. However, the delay and high cost of conventional methods make its execution difficult. In this context, vis-NIR-SWIR (visible - near infrared - shortwave infrared) spectroscopy emerges as an analysis tool that is faster, more affordable and without major environmental impacts. Thus, the objective of this study was to quantify leaf macro and micronutrients in mango cv. Tommy Atkins cultivated under irrigation in the Brazilian semi-arid region with the use of vis-NIR-SWIR spectroscopy. 159 leaf samples of mango cv. Tommy Atkins coming from three distinct commercial areas were used. For these samples, the reflectance spectra were obtained for a spectral range from 350 to 2500 nm, and the contents of N, P, K, Ca, Mg, B, Zn, Fe, Mn and Cu were determined by the reference methods. Spectral data were initially subjected to preprocessing techniques. Then, multivariate regression models were developed. For all calibrated models, cross-validation was performed and for the models with better performance, external validation was performed. Regressive models had strong predictive performance for P and Ca (R² > 0.70), moderate for N and Zn (0.50 < R² < 0.70), weak for Fe (0.30 < R² < 0.50) and very weak for K, Mg, B, Cu and Mn (R² < 0.30). Therefore, vis-NIR-SWIR spectroscopy is a potential complementary tool to quantify the contents of these nutrients in mango cv. Tommy Atkins.