High-throughput phenotyping for climate-resilient forests: integrating multi-sensor fusion and root-shoot dynamics

Area of Science:

• Forestry science
• Plant physiology
• Remote sensing

Background:

• Compound drought and heat events, exacerbated by climate change, pose significant threats to global forest stability.
• Genomic advancements allow rapid identification of tree genotypes, but phenotyping—characterizing adaptive traits—has not kept pace, creating a bottleneck in understanding stress responses.
• Current high-throughput phenotyping (HTP) primarily focuses on above-ground canopy traits, neglecting the critical role of root systems and the soil-plant-atmosphere continuum (SPAC) in drought resilience.

Purpose of the Study:

• To review the current state of HTP for forest climate resilience.
• To highlight the limitations of canopy-focused phenotyping and the need to investigate below-ground traits.
• To propose a framework integrating aerial sensing with eco-hydrological and modeling approaches to infer root functional strategies.

Main Methods:

• Multi-sensor data fusion combining thermal imaging, Solar-Induced Fluorescence (SIF), hyperspectral remote sensing, and LiDAR on UAVs and robotic systems.
• Application of deep learning models for tasks like tree-crown segmentation and stress classification.
• Development of a constraint-based framework coupling aerial sensor data with eco-hydrological approaches and process-based modeling.

Main Results:

• Integrated sensing approaches effectively detect pre-visual physiological stress, such as changes in stomatal conductance.
• Deep learning models show promise but face challenges with overfitting, transferability, and domain shift in complex forest canopies.
• The proposed framework aims to constrain plausible root functional strategies by linking above-ground signals to below-ground processes.

Conclusions:

• Developing climate-resilient forests necessitates a shift towards below-canopy investigation.
• Integrating above-ground sensing with eco-hydrological models offers a viable approach to understanding root system contributions to drought resilience.
• Future research priorities include standardized protocols, open datasets, and Explainable AI (XAI) to bridge the gap between observed signals and underlying traits.