A novel optimization approach incorporating non-stomatal limitations predicts stomatal behaviour in species from six plant functional types

The primary function of stomata is to minimize plant water loss while maintaining CO 2 assimilation. Stomatal water loss incurs an indirect cost to photosynthesis in the form of non-stomatal limitations (NSL) via reduced carboxylation capacity (CAP) and/or mesophyll conductance (MES). Two optimal formulations for stomatal conductance (g s) arise from the assumption of each type of NSL. In reality, both NSL could coexist, but one may prevail for a given leaf ontogenetic stage or plant functional type, depending on leaf morphology. We tested the suitability of two g s formulations (CAP versus MES) on species from six plant functional types (C 4 crop, C 3 grass, fern, conifer, evergreen, and deciduous angiosperm trees). MES and CAP parameters (the latter proportional to the marginal water cost to carbon gain) decreased with water availability only in deciduous angiosperm trees, while there were no clear differences between leaf ontogenetic stages. Both CAP and MES formulations fit our data in most cases, particularly under low water availability. For ferns, stomata appeared to operate optimally only when subjected to water stress. Overall, the CAP formulation provided a better fit across all species, suggesting that sub-daily stomatal responses minimize NSL by reducing carboxylation capacity predominantly, regardless of leaf morphology and ontogenetic stage. © 2019 The Author(s).This work was funded by the IdEx programme of the Université de Bordeaux (project USIFlux) and a Marie Skłodowska-Curie individual fellowship (grant no. 653223). Additional support was provided by the French Agence National de la Recherche (grant agreements ANR-13-BS06-0005-01 and ANR-13-BS06-0005-01) and the European Research Council, under the EU FP7 framework programme (grant agreement 338264)