Evaluating ecohydrological modelling framework to link atmospheric CO 2 and stomatal conductance

Abstract

The establishment of an accurate stomatal conductance (g s ) model in responding to CO 2 enrichment under diverse environmental conditions remains an important issue as g s is the key to understand the plant–water–atmosphere interactions. A better representation of g s is important to reduce uncertainties in predicting the climate change impacts on various ecosystem functions. In this study, we evaluated three most commonly used g s formulations for the estimation of the stomatal response to environmental factors using in situ measurements under different environmental conditions. The three g s models were Leuning's modified Ball–Berry model and two specific cases of the optimization models (i.e., Rubisco limitation model and RuBP regeneration limitation model). On the basis of an analysis of 234 data points obtained from experiments under instantaneous, semicontrolled, and the free-air CO 2 experiment conditions, we found that Leuning's modified Ball–Berry model and RuBP-limited optimization model showed similar performance, and both performed better than Rubisco limitation model. Functional groups (e.g., C 3 vs. C 4 species) and life form (e.g., annual vs. perennial species) play an important role in determining the g s model performance and thus pose a challenge for g s predictions in mixed vegetation communities.