Characterization of leaf surface phenotypes based on light interaction.

Area of Science:

• Plant Science
• Remote Sensing
• Biophysics

Background:

• Leaf surface phenotypes offer insights into plant health, stress adaptation, and breeding potential.
• Non-invasive identification of these phenotypes is crucial for high-throughput phenotyping.
• Surface phenotypes influence leaf optical properties, impacting biochemical parameter retrieval and species identification.

Purpose of the Study:

• To characterize leaf surface phenotypes using spectral measurements and microscopic observation.
• To establish a correlation between spectral data and distinct leaf surface types.
• To assess the potential for remote identification of leaf surface phenotypes.

Main Methods:

• Characterization of 349 leaf samples using polarized light reflectance at Brewster's Angle and microscopy.
• Identification of four primary phenotypes: glossy wax, glaucous wax, high trichome density, and glabrous.
• Application of quadratic discriminant analysis for spectral classification with random training/testing splits.

Main Results:

• Four distinct leaf surface phenotypes were identified and used as ground truth.
• Spectral classification achieved an average correct classification rate of 72.9%.
• Microscopic assessment of cell attributes did not significantly correlate with spectral measurements.

Conclusions:

• Leaf surface phenotypes can be reliably correlated with remotely obtainable spectral measurements.
• Remote identification of leaf surface phenotypes enhances leaf optical modeling and biochemical estimations.
• Phenotyping leaf surfaces supports plant breeding and aids in plant health monitoring.