Ozone risk assessment in three oak species as affected by soil water availability.

To derive ozone (O) dose-response relationships for three European oak species (Quercus ilex, Quercus pubescens, and Quercus robur) under a range of soil water availability, an experiment was carried out with 2-year-old potted seedlings exposed to three levels of water availability in the soil and three levels of O pollution for one growing season in an ozone free-air controlled exposure (FACE) facility. Total biomass losses were estimated relative to a hypothetical clean air at the pre-industrial age, i.e., at 10 ppb as daily average (M24). A stomatal conductance model was parameterized with inputs from the three species for calculating the stomatal O flux. Exposure-based (M24, W126, and AOT40) and flux-based (phytotoxic O dose (POD)) dose-response relationships were estimated and critical levels (CL) were calculated for a 5% decline of total biomass. Results show that water availability can significantly affect O risk assessment. In fact, dose-response relationships calculated per individual species at each water availability level resulted in very different CLs and best metrics. In a simplified approach where species were aggregated on the basis of their O sensitivity, the best metric was POD, with a CL of 6.8 mmol m for the less O-sensitive species Q. ilex and Q. pubescens and of 3.5 mmol m for the more O-sensitive species Q. robur. The performance of POD, however, was very similar to that of POD, and thus a CL of 6.9 mmol m POD and 3.6 mmol m POD for the less and more O-sensitive oak species may be also recommended. These CLs can be applied to oak ecosystems at variable water availability in the soil. We conclude that POD is able to reconcile the effects of O and soil water availability on species-specific oak productivity.