Identifying factors which modify the effects of ambient ozone on white clover (Trifolium repens) in Europe

This paper describes results of experiments carried out by participants of the UN/ECE ICE-Crops to investigate the dose response of plants to long-term ozone exposure. The data were analysed to determine the influences of climatic and geographical factors upon this dose response. Members of the ICP-Crops carried out experiments at 12 sites in Europe, covering a geographical range from Jokioinen, Finland, to Milan, Italy, to generate the database used in this study. The effects of ozone were determined using N-(2-(2-oxo-1-imidazolidinyl)ethyl)-N'phenylurea (ethylene diurea; EDU) as a protectant against ozone damage. Thus, the ratio of biomass in EDU-treated plants to untreated plants indicated the severity of ozone damage. Simple linear regression produced a low r(2) of 0.16 for the dose response. Analysis of the residuals identified three groups in the data depending on their mean residual values. Re-plotting the data into separate dose responses for these three groups produced improvements in the r(2) values to 0.33, 0.36 and 0.36, for Groups 1 to 3, respectively. The data were subsequently used to train artificial neural network (ANN) models to determine the factors that caused the scatter within the dose response. Vapour pressure deficit and temperature were found to be important influences on the dose response whereas longitude, year and altitude were weaker influences. Inclusion of these factors in the ANN model increased the r(2) of the model to 0.79. (C) 1998 Elsevier Science Ltd. All rights reserved.