Comparison of three empirical stomatal conductance models in acacia caven (mol.) Under drought conditións

Abstract

Stomatal conductance ( ) is the main controller of the plant-atmosphere interaction and a key feature to understand how natural vegetation copes with drought. Stomatal conductance empirical models base their performance in linking environmental variables to stomatal behavior. However, when facing water scarcity, different considerations rise, such as integrating the effect of soil moisture or plant water status over , variation of physiological and fitting parameters throughout the season or latent heat ( ) partitioning between canopy, soil and bare soil substratum must be taken into account. We studied in an Acacia caven (Mol.) Shrubland Savannah in Central Chile under Mediterranean-type climate conditions, thus presenting clearly distinguishable wet and dry seasons. We compared three completely independent empirical models: (1) The inversion of the Penman-Monteith equation in combination with the Shuttleworth and Wallace model for evaporation from sparse canopies, (2) the Jarvis multiplicative approach and (3) the Farquhar-Ball and Berry model. Models were validated with leaf gas exchange measurements during both wet and dry season. Acacia showed a high synchronicity with soil moisture thus presenting a typical isohydric behavior, proper of this kind of vegetation. Because soil water content was incorporated in the Jarvis approach, best agreement with ground based measurements was achieved with this model. The Farquhar-Ball and Berry model showed a high dependency on photosynthetically active radiation ( ), thus reaching higher values during summer and a low correlation with ground based measurements, especially during the dry season. The inverted Penman-Monteith equation combined with the Shuttleworth and Wallace model presented a correct fit, however scaling issues such as ecosystem heterogeneity and patchiness must be considered when using a “top-down” approach. Empirical models demonstrated to be a good tool for predicting stomatal behavior in this kind of vegetation; nevertheless plant water status or deep soil water availability must be integrated in computation in order to improve model performance.