Optimizing stomatal conductance for maximum carbon gain under water stress: a meta-analysis across plant functional types and climates

P>1. Quantification of stomatal responses to environmental variables, in particular to soil water status, is needed to model carbon and water exchange rates between plants and the atmosphere. 2. Models based on stomatal optimality theory successfully describe leaf gas exchange under different environmental conditions, but the effects of water availability on the key optimization parameter [the marginal water use efficiency (WUE), lambda = partial derivative A/partial derivative E] has resisted complete theoretical treatment. Building on previous optimal leaf gas exchange models, we developed an analytical equation to estimate lambda from gas exchange observations along gradients of soil water availability. This expression was then used in a meta-analysis of about 50 species to investigate patterns of variation in lambda. 3. Assuming that cuticular water losses are negligible lambda increases under mild water stress but decreases when severe water stress limits photosynthesis. When cuticular conductance is considered, however, lambda increases monotonically with increasing water stress, in agreement with previous theoretical predictions. Moreover, the shape of these response curves to soil water availability changes with plant functional type and climatic conditions. In general, lambda is lower in species grown in dry climates, indicating lower marginal WUE. 4. The proposed parameterization provides a framework to assess the responses of leaf gas exchange to changes in water availability. Moreover, the model can be extended to scale leaf-level fluxes to the canopy and ecosystem level.