|Reddy, K - MISSISSIPPI STATE UNIV|
Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: May 9, 2000
Publication Date: November 1, 2000
Citation: TARPLEY, L., REDDY, K.R., SASSENRATH COLE, G.F. REFLECTANCE INDICES WITH PRECISION AND ACCURACY IN PREDICTING COTTON LEAF NITROGEN CONCENTRATION. CROP SCIENCE. 2000. Interpretive Summary: Nitrogen deficiency and other stresses cause predictable changes in plant and leaf development and composition, and indirectly cause changes in the spectral distribution of radiation reflected from leaves of stressed plants. We present a systematic method to determine the regions of the reflectance spectrum that provide good precision and accuracy for predicting cotton leaf nitrogen concentration when other stresses are present. A set of calibration leaves was grown to contain a range of nitrogen concentration under otherwise optimal conditions. For each leaf, the reflectance values were obtained for a number of narrow bands positioned throughout the reflectance spectrum. All possible ratios among the narrow bands' reflectance values were created, then related to the leaves' total nitrogen concentrations to produce many prediction equations for total leaf nitrogen. A separate set of leaves from field-grown plants was used for testing the prediction equations. Ratios involving reflectance values between 700 and 720 nm (the "red-edge") and values between 755 and 1000 nm (a region of high leaf reflectance) provided superior precision and accuracy in prediction of cotton leaf nitrogen concentration. The method contributes towards developing optimal procedures for interpreting particular stresses of plants through analysis of their leaf reflectance spectra.
Technical Abstract: Diagnostic methods assaying leaf optical properties can aid rapid site-specific screening of crop nitrogen status. A set of calibration curves relating many 1.5-nm band reflectance ratios to leaf N concentration was established from plants grown in sunlit plant growth-chambers (SPAR units) and at a range of nitrogen levels. Predicted concentrations were compared by regression to the actual concentrations in a validation set of field-grown leaf samples. Only those ratios that combined a red-edge measure (700 or 716 nm) with a waveband of high reflectance in the very near infrared region (755-920 and 1000 nm) provided good precision (correlation) and accuracy (one-to-one relationship between predicted to actual values). Indices that included a chlorophyll-based reflectance feature also had good precision, but were less accurate than those obtained from the red-edge/very-near-infrared- reflectance ratios.