Comparison of UV, visible and near-infrared, and mid-infrared spectrometers to estimate maize and sorghum leaf nutrients using dry-intact and ground leaves

Abstract

Spectroscopy has been explored as a potential tool to estimate various nutrients in plant tissues rapidly and nondestructively. Compared to visible and near-infrared regions, ultraviolet and mid-infrared (MIR) regions have been sporadically investigated for predicting various nutrients in plant leaf tissues. Sample grinding is one of the tedious steps recommended for MIR spectroscopy. Still, dry-intact samples may potentially estimate leaf nutrients eliminating the need for grinding, which could reduce the cost and time of analysis. In light of this, five different spectrometers (four portable and one laboratory-only) were compared to estimate eleven macro and micronutrients (N, P, K, Ca, Mg, Zn, S, Cu, B, Fe, Mn) using dry-intact and ground leaf spectra. A total of 154 maize (Zea mays L.) and 308 sorghum (Sorghum bicolor L.) leaves from different genotypes subjected to different abiotic stress treatments, were collected and used to acquire spectral data. Partial least squares regression was used to calibrate models using 70% of the data and validate on the remaining data (30%). Five macronutrients (N, P, K, S and Mg) and two micronutrients (Mn and Fe) were predicted accurately (R2 > 0.6), and the ground scan accuracies were superior to the dry-intact scans for all the spectrometers used. Some of the critical elements (N, P, Mg, Mn, Fe and Zn) were successfully predicted (R2 = 0.6–0.89) with dry intact scans as well. Overall, the results suggested that portable spectrometers can accurately estimate leaf nutrients using ground or dry-intact leaves.