The NITREX project, which encompasses seven ecosystem-scale experiments in coniferous forests at the plot or catchment level in northwestern Europe, investigates the effect of atmospheric nitrogen (N) deposition in coniferous forests. The common factor in all of the experiments is the experimentally controlled change in N input over a period of 4-5 years. Results indicate that the status and dynamics of the forest floor are key components in determining the response of forests to altered N inputs. An empirical relationship between the carbon-nitrogen (CIN) ratio of the forest floor and retention of incoming N provides a simply measured tool through which the likely timing and consequences of changes in atmospheric N deposition for fresh waters may be predicted. In the terrestrial ecosystem, a 50% increase in tree growth is observed following the experimental reduction of N and sulfur inputs in a highly N-saturated site, illustrating the damaging effects of acidifying pollutants to tree health in so me locations. Few biotic responses to the experimental treatments were observed in other NITREX sites, but the rapid response of water quality to changes in N deposition, and the link to acidification in sensitive areas, highlight the need for N-emission controls, irrespective of the long-term effects on tree health. The observed changes in ecosystem function in response to the experimental treatments have been considered within the framework of the current critical-load approach and thus contribute to the formulation of environmental policy.

Dynamic models provide an extension to critical loads by predicting the timescale of chemical recovery to emission reductions. They can also be used to determine the deposition levels required to achieve a prescribed target chemistry within a given timescale and so have direct utility in the formulation of further emission reductions. The MAGIC model has been tested extensively against long-term diatom reconstructions, experimental manipulation experiments and long-term chemistry observations and has in all cases been found to successfully capture the dynamics of chemical change. This also provides confidence in the predictions from the model. Methods for summarising regional model predictions are summarised and a methodology for the derivation of 'target' loads is proposed for use within the Convention. Requirements for further research, including model uncertainty and the link between surface water chemistry and biological receptors, are outlined.