Setting ozone critical levels for protecting horticultural Mediterranean crops: case study of tomato.

Seven experiments carried out in Italy and Spain have been used to parameterising a stomatal conductance model and establishing exposure- and dose-response relationships for yield and quality of tomato with the main goal of setting O3 critical levels (CLe). CLe with confidence intervals, between brackets, were set at an accumulated hourly O3 exposure over 40 nl l(-1), AOT40 = 8.4 (1.2, 15.6) ppm h and a phytotoxic ozone dose above a threshold of 6 nmol m(-2) s(-1), POD6 = 2.7 (0.8, 4.6) mmol m(-2) for yield and AOT40 = 18.7 (8.5, 28.8) ppm h and POD6 = 4.1 (2.0, 6.2) mmol m(-2) for quality, both indices performing equally well. CLe confidence intervals provide information on the quality of the dataset and should be included in future calculations of O3 CLe for improving current methodologies. These CLe, derived for sensitive tomato cultivars, should not be applied for quantifying O3-induced losses at the risk of making important overestimations of the economical losses associated with O3 pollution.