D’Odorico et al. (2025) Deciphering tree drought responses across species: linking leaf water potentials with remote sensing greenness and photoprotection dynamics

Identification

• Journal: Agricultural and Forest Meteorology
• Year: 2025
• Date: 2025-10-01
• Authors: Petra D’Odorico, Dominic Fawcett, Richard L. Peters, David N. Steger, Tobias Zhorzel, Günter Hoch, David Basler, Christian Ginzler, Michael Eisenring, Gaétan Glauser, Roman Zweifel, Arthur Geßler, Ansgar Kahmen
• DOI: 10.1016/j.agrformet.2025.110856

Research Groups

• Swiss Federal Research Institute for Forest, Snow and Landscape research WSL, Birmensdorf, Switzerland
• Department of Environmental Sciences – Botany, University of Basel, Basel, Switzerland
• Technical University of Munich, TUM School of Life Sciences, Freising, Germany
• Neuchatel Platform of Analytical Chemistry, University of Neuchatel, Neuchatel, Switzerland
• Institute of Terrestrial Ecosystems, ETH Zurich, Zurich, Switzerland

Short Summary

This study investigated the drought responses of seven common European tree species in a temperate forest, integrating drone-based multispectral imagery with leaf water potential and pigment measurements. It found that the Photochemical Reflectance Index (PRI) strongly correlated with leaf water potentials, capturing both drought-induced declines and post-rainfall recovery, while the Normalized Difference Vegetation Index (NDVI) primarily detected greening losses in some species but failed to reflect recovery.

Objective

• To assess whether drone-derived greenness (NDVI) and photoprotection (PRI) indicators capture species-specific variation in tree water status and contribute to a mechanistic interpretation of remote sensing signals over seasonal and diurnal timescales.
• Hypothesis 1: Effectively capturing drought stress responses across contrasting species requires a combination of vegetation state (NDVI), function (PRI), and structural indicators (height, crown size).
• Hypothesis 2: Species with lower drought tolerance and experiencing higher drought stress exhibit a higher engagement of the photoprotective system due to higher light dissipation requirements.

Study Configuration

• Spatial Scale: A mature temperate forest at the Swiss Canopy Crane II (SCCII) experimental research site in Hölstein, Basel, Switzerland (47.4386°N, 7.7756°E; 550 m a.s.l.), covering 1.68 hectares. The study focused on 69 target trees from seven common European tree species (Abies alba, Picea abies, Pinus sylvestris, Acer pseudoplatanus, Fagus sylvatica, Carpinus betulus, Quercus sp.) with a diameter at breast height (DBH) greater than 10 cm and heights up to 32 meters. Drone-based imagery had a ground sampling distance of 5 to 7 centimeters per pixel.
• Temporal Scale: Peak growing season of 2023 (July to September). Four measurement campaigns for drone-based multispectral imagery and leaf water potentials (7 July, 31 July, 22 August, 2 September 2023). Leaf pigment samples were collected on 31 July and 22 August 2023. Diurnal drone flights were conducted on 22 August 2023 (approximately 08:00, 12:00, 15:00, and 18:00 Central European Time). Long-term climate data from 1991-2020 were used as a reference.

Methodology and Data

• Models used:
  • Linear Mixed Effects Models (LMMs) for peak growing-season analysis.
  • Linear model: PRIlin = a1⋅PARalbedo + b1 (for low light absorption).
  • Logarithmic model: PRIlog = a2⋅log (b2⋅PARalbedo + c2) (for diurnal PRI light response curves).
• Data sources:
  • Remote Sensing: Drone-based multispectral imagery (Micasense Dual camera on DJI Matrice 300 RTK UAV) for Normalized Difference Vegetation Index (NDVI) and Photochemical Reflectance Index (PRI). Drone-based LiDAR (miniVUX-3 UAV Scanner on DJI Matrice 600) for Canopy Height Model (CHM). Downwelling Light Sensor (DLS) for irradiance data and a grey 60% reflective panel for reflectance calibration.
  • Physiological Measurements: Scholander-type pressure chamber (Model 1000; PMS Instruments) for leaf water potential (Ψleaf). Osmomat 3000 device (Gonotech) for osmolality to determine turgor loss point (Ψtlp). Ultrahigh-performance liquid chromatography with photodiode array detection (UHPLC-PDA) for leaf pigments (xanthophyll cycle pigments, chlorophylls, carotenoids).
  • Environmental Data: MeteoSwiss climate station for long-term climate data. Weather station (Davis Vantage Pro 2) at the SCCII site for local air temperature (Ta), relative humidity (RH), photosynthetic active radiation (PAR), and precipitation. MPS-2 soil sensors (Decagon Devices) for soil water potential (Ψsoil) at 10 cm and 40 cm depth.

Main Results

• The 2023 growing season experienced a mean air temperature anomaly of +2.28 °C and 33.32% less precipitation than the 1991-2020 average, leading to severe drought conditions.
• The Photochemical Reflectance Index (PRI) significantly correlated with leaf water potentials (Ψleaf pd and Ψleaf md), capturing drought-induced declines and post-rainfall recovery across most species.
• The Normalized Difference Vegetation Index (NDVI) showed significant greenness decline during drought only in Acer pseudoplatanus (treatment), Fagus sylvatica (control), Carpinus betulus (control), and Pinus sylvestris, but did not reflect recovery.
• A linear mixed-effects model combining PRI, NDVI, tree height, and species explained 65% of the variance in predawn leaf water potential (Ψleaf pd) and 70% in midday leaf water potential (Ψleaf md). PRI was the most significant predictor for Ψleaf pd (η² = 45.2%), while species was the strongest predictor for Ψleaf md (η² = 33.9%).
• Tree height had a significant negative effect on water potentials, indicating lower water potentials in taller trees.
• Remotely sensed midday crown-level PRI negatively correlated with leaf-level de-epoxidation state (DEPS) of the xanthophyll cycle (R² = 0.35, p < 0.001), confirming that lower PRI values indicate increased photoprotection.
• Species experiencing higher hydraulic stress (lower water potentials) generally exhibited higher engagement of their xanthophyll cycle, characterized by greater activation rates (PRIrate) and operating ranges (PRIrange) of photoprotection.
• Conifer species (Pinus sylvestris, Picea abies) generally showed less negative Ψleaf and lower non-photochemical quenching (NPQ) requirements, except for Abies alba which displayed increased NPQ demands.
• Quercus sp. maintained positive hydraulic safety margins but showed increased NPQ requirements throughout the day.
• Species with lower drought tolerance (based on climatic distributions) exhibited higher PRIrate (R²=0.84, p = 0.011) and PRIrange (R²=0.66, p = 0.049), excluding Picea abies.
• Shade-tolerant species showed a strong positive correlation with PRIrate (R²=0.86, p = 0.008), excluding Picea abies.

Contributions

• This study provides an integrated framework demonstrating the effectiveness of drone-based greenness (NDVI) and photoprotection (PRI) indicators for assessing and scaling species-specific tree water relations in mixed temperate forests.
• It offers mechanistic insights into remotely sensed signals by directly linking them to leaf water potentials, turgor loss points, and leaf pigment dynamics in mature trees under natural drought and recovery conditions.
• The research introduces a novel approach using diurnal drone flights to derive PRI light response curves and associated metrics (PRIrate, PRIrange), which effectively isolate rapid xanthophyll cycle responses from constitutive pigment pool effects, enabling more accurate cross-species comparisons of photoprotection dynamics.
• The findings highlight the complex interplay between hydraulic and photoprotective functioning and interspecific differences in drought vulnerability, providing valuable information for refining early-warning systems and species-specific drought monitoring strategies.

Funding

• Swiss State Secretariat for Education, Research and Innovation, SERI (Project codes: 22.00419, REF-1131-52104) for contribution to the EU Project FORWARDS.
• SNSF grant FORDROUGHT (Project code: 315230_208197) for Ansgar Kahmen, David Steger, and Tobias Zhorzel.

Citation

@article{DOdorico2025Deciphering,
  author = {D’Odorico, Petra and Fawcett, Dominic and Peters, Richard L. and Steger, David N. and Zhorzel, Tobias and Hoch, Günter and Basler, David and Ginzler, Christian and Eisenring, Michael and Glauser, Gaétan and Zweifel, Roman and Geßler, Arthur and Kahmen, Ansgar},
  title = {Deciphering tree drought responses across species: linking leaf water potentials with remote sensing greenness and photoprotection dynamics},
  journal = {Agricultural and Forest Meteorology},
  year = {2025},
  doi = {10.1016/j.agrformet.2025.110856},
  url = {https://doi.org/10.1016/j.agrformet.2025.110856}
}