Early detection of changes in leaf reflectance of pea plants (Pisum sativum L.) under herbicide action

Abstract

Based on high resolution leaf spectral reflectance data a new technique was developed and applied to detect damages of agricultural plants under the action of low intensity stress factors (herbicides) which at very low concentrations could not be established by the standard biochemical and biometric techniques. Results are presented from a remote sensing study of the peculiarities of the leaf spectral reflectance of pea plants (Pisum salivum L.) treated with atrazine and 2.4-D (2.4 - phenoxyacetic acid) at three low concentrations (0.01 /spl mu/M, 0.1 /spl mu/M and 1 /spl mu/M, respectively 2.15, 21.5, and 215 /spl mu/g/l for atrazine and 2.59, 25.9, and 259 /spl mu/g/l for 2.4-D) as compared to the field dose of these herbicides commonly used in the agricultural practice. The physiological status of the plants was assessed using biometric and biochemical parameters such as length, fresh weight, dry weight and electrolyte leakage. The high-resolution spectral data were obtained using a multichannel spectrometer in the visible and near infrared ranges of the electromagnetic spectrum in 128 channels at a spectral resolution (halfwidth) of 2.6 nm. Using the technique which employs discriminant analysis and other statistical methods we established the presence of statistically significant differences in the arising variations of the leaf spectral reflectance characteristics between control and treated plants in the green (520+580 nm), red and near infrared (690+800 nm) ranges of the spectrum.