Chemometrics for mapping the spatial nitrate distribution on the leaf lamina of fenugreek grown under varying nitrogenous fertilizer doses

Abstract

Abstract Excess nitrogen fertilizer use leads to vegetables with high amounts of nitrate content. Consumption of vegetables with high amounts of nitrate is carcinogenic to human beings. In this study, fenugreek plants were grown under varying nitrogen fertilizer doses (0, 50, 100, 150, 200, 250, 300, 350 and 400 kg N/ha). A Vis-NIR hyperspectral imaging (HIS) camera captured images of fenugreek leaves within the 398–1,003 nm spectral range. The spectral data were pre-processed using different pre-processing techniques before the model development. Partial least-squares regression (PLSR) models were constructed with complete spectral data and selected wavelengths. The performance of the PLSR model decreased with pre-processed spectral data, and there was no significant difference compared to the model constructed with raw spectral data (R 2 CV = 0.915, SECV = 591.933, slope = 0.518 and RPDCV = 1.421). The wavelengths 411, 435, 466, 558, 669, and 720 nm were selected as feature wavelengths for representing nitrate content in fenugreek leaves. The performance of the PLSR model constructed with feature wavelengths (SECV = 648.672; RPDCV = 1.482; R 2 CV = 0.869) was non-significant compared with the model developed with raw complete spectral data (SECV = 591.933; R 2 CV = 0.915 and RPDCV = 1.421). Using the complete raw spectral data, the spatial distribution images of nitrate content in fenugreek leaves indicated that the nitrate content was concentrated near and along the midrib up to the apex. The overall results obtained in the present study suggest that VIS-NIR HSI, along with suitable chemometric techniques, can be used for rapid assessment of nitrate content in fenugreek leaves.