Incorporating cultivar-specific stomatal traits into stomatal conductance models improves the estimation of evapotranspiration enhancing greenhouse climate management

Abstract

The effect of considering cultivar differences in stomatal conductance (gₛ) on relative air humidity (RH)-related energy demand was addressed. We conducted six experiments in order to study the variation in evapotranspiration (ETc) of six pot rose cultivars, investigate the underlying processes and parameterise a gₛ-based ETc model. Several levels of crop ETc were realised by adjusting the growth environment. The commonly applied Ball–Woodrow–Berry gₛ-sub-model (BWB-model) in ETc models was validated under greenhouse conditions, and showed a close agreement between simulated and measured ETc. The validated model was incorporated into a greenhouse simulator. A scenario simulation study showed that selecting low-gₛ cultivars reduces energy demand (≤5.75%), depending on the RH set point. However, the BWB-model showed poor prediction quality at RH lower than 60% and a good fit at higher RH. Therefore, an attempt was made to improve model prediction: the in situ-obtained data were employed to adapt and extend either the BWB-model, or the Liu-extension with substrate water potential (Ψ; BWB-Liu-model). Both models were extended with stomatal density (Dₛ) or pore area. Although the modified BWB-Liu-model (considering Dₛ) allowed higher accuracy (R² = 0.59), as compared to the basic version (R² = 0.31), the typical lack of Ψ prediction in greenhouse models may be problematic for implementation into real-time climate control. The current study lays the basis for the development of cultivar specific cultivation strategies as well as improving the gₛ sub-model for dynamic climate conditions under low RH using model-based control systems.

Plant phenotyping relevance

栽培実験の単なる生理測定ではなく、気孔形質を組み込んだ蒸発散・気孔コンダクタンスモデルを検証・改良し、温室制御への適用可能性を評価しているため、方法開発・検証が中心である。

Code and data availability

The supplied blocks describe six pot rose experiments, stomatal trait measurements, and MATLAB/R model parameterisation, but contain no data availability statement, public repository deposit, or author code URL. No paper-specific public asset is identified.

No evidence-backed public reproduction asset is currently recorded.