Related Experiment Video
Updated: Jul 7, 2025

A Telemetric, Gravimetric Platform for Real-Time Physiological Phenotyping of Plant–Environment Interactions
Published on: August 5, 2020
Can Grapevine Leaf Water Potential Be Modelled from Physiological and Meteorological Variables? A Machine Learning
Miguel Damásio1,2, Miguel Barbosa3, João Deus1
1INIAV I.P., Instituto Nacional de Investigação Agrária e Veterinária, Polo de Inovação de Dois Portos, Quinta da Almoinha, 2565-191 Dois Portos, Portugal.
Accurate plant water status monitoring is crucial for grapevines facing climate change. Stomatal conductance (gs) effectively predicts predawn water potential (Ψpd), enabling precise irrigation strategies.
Area of Science:
- Viticulture and Climate Science
- Plant Physiology and Water Relations
Background:
- Global climate change intensifies heatwaves and drought, impacting viticulture worldwide.
- Precision irrigation, guided by reliable water status indicators (WSIs), is essential for Mediterranean viticulture.
- Traditional leaf water potential (Ψleaf) measurement is invasive and time-consuming.
Purpose of the Study:
- To identify key variables correlated with grapevine water status.
- To determine which variables best predict leaf water potential (Ψleaf).
- To develop accurate, non-invasive methods for monitoring plant water status.
Main Methods:
- Field study with five grapevine varieties in Alentejo, Portugal, under full irrigation (FI), deficit irrigation (DI), and no irrigation (NI) treatments.
- Monitoring included stomatal conductance (gs), predawn (Ψpd) and stem water potential (Ψstem), and thermal imaging.
- Machine learning regression models (ExtraTrees, Gradient Boosting) trained on meteorological, thermal, and gs data to predict Ψpd.
Main Results:
- Stomatal conductance (gs) and predawn water potential (Ψpd) showed differential responses to irrigation treatments.
- Mid-morning and mid-day stem water potential (Ψstem) could not distinguish between treatments.
- gs demonstrated the strongest correlations with other WSIs and the best predictive capability for Ψpd.
- Ensemble machine learning models achieved high accuracy in predicting Ψpd (R2 > 0.83).
Conclusions:
- Stomatal conductance (gs) is a highly effective predictor of grapevine predawn water potential (Ψpd).
- Machine learning models utilizing meteorological, thermal, and gs data offer a robust, non-invasive approach to monitor plant water status.
- These findings support the implementation of precision irrigation strategies in viticulture under changing climatic conditions.
More Related Videos
Related Concept Videos
Adaptations that Reduce Water Loss
Light Acquisition
Responses to Drought and Flooding
Precipitation Processes
Precipitation Gravimetry
In determining nickel by gravimetric analysis, a precipitant of ethanolic dimethylglyoxime is added to a hot nickel salt solution. This is quickly followed by the dropwise addition of dilute ammonia solution until precipitation occurs. A...
Xylem and Transpiration-driven Transport of Resources

