Salinity-specific stomatal conductance model parameters are reduced by stomatal saturation conductance and area via leaf nitrogen.

Modeling stomatal behavior is necessary for accurate stomatal simulation and predicting the terrestrial water‑carbon cycle. Although the Ball-Berry and Medlyn stomatal conductance (g) models have been widely used, variations and the drivers of their key slope parameters (m and g) remain poorly understood under salinity stress. We measured leaf gas exchange, physiological and biochemical traits, soil water content and electrical conductivity of saturation extract (EC), and fitted slope parameters of two genotypes of maize growing in two water and two salinity levels. We found m was different between the genotypes, but no difference in g. Salinity stress reduced m and g, saturated stomatal conductance (g), the fraction of leaf epidermis area allocation to stomata (f), and leaf nitrogen (N) content, and increased EC, but no marked decrease in slope parameters under drought. Both m and g were positively correlated with g, f, and leaf N content, and negatively correlated with EC in the same fashion among the two genotypes. Salinity stress altered m and g by modulating g and f via leaf N content. The prediction accuracy of g was improved using salinity-specific slope parameters, with root mean square error (RMSE) being decreased from 0.056 to 0.046 and 0.066 to 0.025 mol m s for the Ball-Berry and Medlyn models, respectively. This study provides a modeling approach to improving the simulation of stomatal conductance under salinity.