RESUMO
Ozone uptake through the stomata in tree leaves is an important process for improving air quality by urban trees. Stomatal conductance (gs) is a key determinant of stomatal ozone uptake. The parameterization of gs models for estimating stomatal ozone uptake of trees has mainly been carried out using gs data measured in seedling leaves although the leaf traits may differ between mature trees and seedlings. In the present study, we compared stomatal ozone uptake estimated by gs models parameterised with data from mature trees and seedlings of Zelkova serrata. We measured gs in leaves of mature trees and seedlings of Z. serrata using a leaf porometer for 3-4 growing seasons. The Jarvis-type gs model was parameterised with data from mature trees and seedlings, separately. The maximum gs, and the functions of the seedling gs estimation model regarding the response to air temperature, vapour pressure deficit and atmospheric ozone concentration were the factors inducing lower stomatal ozone uptake. In contrast, the function of the seedling gs estimation model regarding the response to irradiance resulted in a higher estimated stomatal ozone uptake. The estimated stomatal ozone uptake for one growing season (April-September) by the seedling gs estimation model was 27% lower than that by the mature tree gs estimation model. These results indicate that leaf gas exchange traits of Z. serrata were different between mature trees and seedlings, and that estimating ozone uptake in mature tree leaves using a model based on seedling gs measurements results in an underestimation.