Stomatal conductance models for ozone risk assessment at canopy level in two Mediterranean evergreen forests

•Two canopy-level Gs models were tested in two Mediterranean forests (pine and oak).•Canopy-level stomatal ozone flux was estimated in both forests using the Gs models.•Maximum stomatal conductance was 72% higher in Umbrella pine than in Holm oak.•Inclusion of a soil water function improved the performance of the Jarvis-type model.•Soil water status affected m in the BWB model in Umbrella pine but not in Holm oak.

Forests in Mediterranean Europe are located in a hot spot for tropospheric ozone formation. We applied two canopy-level stomatal conductance models (Jarvis-type and Ball-Woodrow-Berry models) to two Mediterranean evergreen forests (Umbrella pine at San Rossore, and Holm oak at Castelporziano, in central Italy), which is essential for assessing ozone impact on forests via the stomatal flux-based approach. Parameterizations of the models was carried out by the Eddy Covariance technique. Both Jarvis-type and Ball-Woodrow-Berry models well explained the observed stomatal conductance and stomatal ozone flux in both forests. Maximum stomatal conductance was 72% higher in Umbrella pine than in Holm oak, leading to higher stomatal ozone flux. Inclusion of a soil water function improved the performance of the Jarvis-type model for the estimation of stomatal ozone flux. We found contrasting results concerning the coefficient m (the slope of the photosynthesis-stomatal conductance relationship) in the Ball-Woodrow-Berry model between the two forests: while m was constant for varying soil water status in the Holm oak forest, it declined with soil drying in the Umbrella pine forest. This may result from higher stomatal sensitivity to drought in Umbrella pine as a response to avoid drought stress. Overall, these results advance our understanding of ozone risk assessment in Mediterranean forests.