Abstract
Near-infrared spectroscopy (NIR) was used to measure the concentrations of macronutrients [carbon (C), nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg)] and micronutrients [sodium (Na), iron (Fe), manganese (Mn), copper (Cu), and zinc (Zn)] in yerba mate leaves. We collected 270 samples of yerba mate, growing in association with eucalyptus, in different cultivation conditions, in the harvest seasons of 2011 and 2012. Calibration models comprised 216 samples for calibration and internal validation of the models and 54 sample for external prediction. The results allowed determination of the mineral elements, providing values of the correlation coefficient (R) higher than 0.90, on calibration models. The predictive power of the models was verified and showed that NIR has potential to successful quantify P (R = 0.62), Fe (R = 0.75), and Cu (R = 0.89) in this study. On the other hand, for C, K, Na, Mn, and Zn the technique was not efficient.