Monitoring Eastern White Pine Health by Using Field-Measured Foliar Traits and Hyperspectral Data

Area of Science:

• Forestry
• Plant Pathology
• Remote Sensing

Background:

• Canopy foliar traits are key indicators of plant health and ecosystem dynamics.
• Eastern White Pine (EWP) in the Northeastern USA is threatened by white pine needle damage (WPND).
• Timely detection of forest diseases is crucial for effective management.

Purpose of the Study:

• To investigate the use of hyperspectral data and foliar traits for WPND detection in EWP.
• To evaluate the efficacy of remote sensing techniques for early disease identification.
• To compare the performance of models using spectral vegetation indices versus combined data.

Main Methods:

• Utilized field-measured leaf traits and hyperspectral remote sensing data.
• Applied parametric and non-parametric methods for WPND detection.
• Developed and compared Random Forest (RF) models for classification.

Main Results:

• The RF model using only spectral vegetation indices (SVIs) achieved 87% accuracy and a Kappa coefficient (K) of 0.68.
• The combined approach of field traits and remote sensing yielded 77% accuracy with K=0.46.
• Remotely sensed SVIs proved more effective for classifying asymptomatic and symptomatic EWP.

Conclusions:

• Hyperspectral remote sensing, particularly SVIs, shows high potential for accurate WPND detection in EWP.
• This technology offers a valuable tool for early disease identification and improved forest management.
• Remote sensing is increasingly vital for monitoring forest health amidst rising pest and pathogen threats.