Pre-visual detection of iron and phosphorus deficiency by transformed reflectance spectra

Reflectance spectroscopy and strategies for spectral analysis over the visible range from 380 to 780 ran were used to provide diagnostic information on iron (Fe) and phosphorus (P) status of Brassica chinensis L. var parachinensis (Bailey) grown under hydroponics conditions. Leaf reflectance (R) spectra were collected and normalized inner reflectance (NR1) spectra were calculated. The regression coefficients (B-matrix) and variable importance for projection (VIP) in partial least squares regression were used to determine important wavelengths that correlate with total chlorophyll (Chl) content. No single wavelength that showed good correlation with Chl content was found. Therefore, NR1 was transformed into CIELAB color values, which simplified the whole visible spectrum into three values. Our results showed that upon Fe deprivation, plants entered into a deficiency state very rapidly, highlighting the importance of early diagnosis. The direct effect of Fe on leaf Chl content allowed CIELAB color values to be used for pre-visual detection of Fe deficiency 2 days before the appearance of visually distinguishable morphological changes. On the other hand, P-deprived plants showed a marked decline in cellular P levels but remained above critical threshold concentrations after 7 days. The Chl content was not affected by the leaf P content and CIELAB color values showed no difference with control plants. (c) 2006 Elsevier B.V. All rights reserved.