The determination of the molecular composition of plant leaves is essential to assist in nutritional management, whether for cultivated or non-cultivated species. In this sense, the study aimed to apply FTIR technique in combination with chemometrics and ROC analysis for the evaluation of changes in compositional of plant leaves of Physalis angulata and Physalis peruviana due to nitrogen fertilization treatments. Both species were grown under different doses of nitrogen (0, 200, 400, and 600 Kg ha-1) and leaf samples were evaluated using ATR-FTIR. Our results demonstrate that the spectra of both species were influenced by the nitrogen doses. The computed band area from the lipid/amide, lipid/carbohydrates, degree of esterification and calcium oxalate shows nitrogen fertilization due to 400 Kg ha-1 of N treatment is more effective for a better quality of yield. 2D correlation spectral analysis (2DCOS) reveals cellulose and pectin begins changes followed by amide of proteins due to nitrogen treatment in P. peruviana samples. The P. angulata plants shows hemicellulose changes predominating followed by proteins and polysaccharides. The obtained principle component analysis plot and loading values show the Physalis species samples distinctly separated from control with protein and carbohydrates are predominant in influencing separation among them. Receiver operation characteristic analysis shows a higher value of area under the curve reflecting better reliability of the experiments carried out. Hierarchical cluster analysis shows closed separation for a similar group on dissimilarity scale. Thus the use of 2DCOS coupled with chemometrics helps to identify changes in the composition of leaves of physalis species due to nitrogen doses, constituting a fast and precise measuring for the suitable management of this fertilization.
Keywords: 2D correlation spectral analysis; Fourier Transform Infrared Spectroscopy; Mineral nutrition; Physalis angulata; Physalis peruviana.
Copyright © 2020 Elsevier B.V. All rights reserved.