Guidelines provided by OECD and EPPO allow the use of data obtained in greenhouse experiments in the risk assessment for pesticides to non-target terrestrial plants in the field. The present study was undertaken to investigate the predictability of effects on field-grown plants using greenhouse data. In addition, the influence of plant development stage on plant sensitivity and herbicide efficacy, the influence of the surrounding vegetation on individual plant sensitivity and of sublethal herbicide doses on the biomass, recovery and reproduction of non-crop plants was studied. Results show that in the future, it might well be possible to translate results from greenhouse experiments to field situations, given sufficient experimental data. The results also suggest consequences at the population level. Even when only marginal effects on the biomass of non-target plants are expected, their seed production and thereby survival at the population level may be negatively affected. The response of greenhouse-grown wild plant species to herbicide exposure could be related to the response of the same species when grown in the field.

A critical load data base was developed for Europe and Northern Asia using the latest data bases on soils, vegetation, climate and forest growth. Critical loads for acidity and nutrient nitrogen for terrestrial ecosystems were computed with the Simple Mass Balance model. The resulting critical loads are in accordance with critical loads from previous global empirical studies, but have a much higher spatial resolution. Critical loads of acidity are sensitive to both the chemical criterion and the critical limit chosen. Therefore a sensitivity analysis of critical loads was performed by employing different chemical criteria. A critical limit based on an acid neutralizing capacity (ANC) of zero resulted in critical loads that protect ecosystems against toxic concentrations of aluminium and unfavourable Al/Bc ratios, suggesting that ANC could be an alternative to the commonly used Al/Bc ratio. Critical loads of nutrient nitrogen are sensitive to the specified critical nitrate concentration, especially in areas with a high precipitation surplus. If limits of 3¿6 mg N l¿1 are used for Western Europe instead of the widely used 0.2 mg N l¿1, critical loads double on average. In low precipitation areas, the increase is less than 50%. The strong dependence on precipitation surplus is a consequence of the simple modelling approach. Future models should explore other nitrogen parameters (such as nitrogen availability) instead of leaching as the factor influencing vegetation changes in terrestrial ecosystems.