Articles
ESTIMATION OF SWEET CHERRY TREE WATER STATUS BY SPECTRAL REFLECTANCE
Article number
795_114
Pages
711 – 716
Language
English
Abstract
The goal of this research was to correlate leaf spectral reflectance with water status of sweet cherry trees (Prunus avium L.). In 2003 and 2004, stem water potential (ψstem) was measured using a pressure chamber to characterize tree water status.
Individual leaf reflectance was measured between 350 nm and 2500 nm with a FieldSpec FR spectrophotometer assembled to a leaf clip probe.
The raw and smoothed reflectance data and first and second derivative dataset were developed and compared with ψstem. Partial least squares (PLS) with cross validation (CV) was used to determine the region of reflectance wavebands (10 nm resolution) that best explains variation in ψstem. Each model showed correlation coefficients = r (r2=determination coefficient) between 0.71 and 0.75 with standard error of prediction (SEP) between 0.14 and 0.15. Manipulation of the raw reflectance data such as smoothing and first and second derivatives did not contribute significantly to a better performance of the model.
The majority of wavelengths correlated with water status were in the range of visible light: 490 nm to 560 nm (green), 560 nm to 600 nm (yellow), 600 nm to 630 nm (orange) and 630 nm to 700 nm (red). Leaf reflectance in the range between 540 nm to 710 nm showed good potential for rapid screening of plant water status, especially in the range commonly observed in commercial sweet cherry orchards.
Individual leaf reflectance was measured between 350 nm and 2500 nm with a FieldSpec FR spectrophotometer assembled to a leaf clip probe.
The raw and smoothed reflectance data and first and second derivative dataset were developed and compared with ψstem. Partial least squares (PLS) with cross validation (CV) was used to determine the region of reflectance wavebands (10 nm resolution) that best explains variation in ψstem. Each model showed correlation coefficients = r (r2=determination coefficient) between 0.71 and 0.75 with standard error of prediction (SEP) between 0.14 and 0.15. Manipulation of the raw reflectance data such as smoothing and first and second derivatives did not contribute significantly to a better performance of the model.
The majority of wavelengths correlated with water status were in the range of visible light: 490 nm to 560 nm (green), 560 nm to 600 nm (yellow), 600 nm to 630 nm (orange) and 630 nm to 700 nm (red). Leaf reflectance in the range between 540 nm to 710 nm showed good potential for rapid screening of plant water status, especially in the range commonly observed in commercial sweet cherry orchards.
Publication
Authors
A. Antunez, M.D. Whiting, F.J. Pierce, C. Stöckle
Keywords
deficit water stress, Prunus avium, spectrophotometric detection, stem water potential
Online Articles (157)