A new method of indicator species analysis is suggested that can be used to demonstrate changes in the species composition of forests resulting from atmospheric deposition of acidifying substances and nitrogen. The method is not dependent on an exact re-localization of sites. 2162 published releves of Fagus sylvatica and Quercus robur-Carpinus betulus forests from two study areas in Germany (S Niedersachsen and Nordrhein-Westfalen) were compiled and divided into releves from before 1975 and more recent releves. Weighted averages of Ellenberg indicator values fbr light (L), moisture (M), reaction (R), and nitrogen (N) were calculated and for each of the four data sets - mesic and dry sites in two areas - N was regressed on R to give a predicted value (N-pred) for any given R. The differences between observed values N-obs and N-pred were calculated separately for old and new releves (Ndev-old and Ndev-new) The same analysis was applied to model data including 16 different scenarios of acidification and/or eutrophication. The model runs resulted in six different types of regression plots for N-dev and R. All models had in common that the regression lines for Ndev-new lay, completely or largely, above the zero line of no change and above the lines for Ndev-old. Also, in the four data sets, Ndev-new was generally larger than Ndev-old, and particularly so in Nordrhein-Westfalen. The values of Ldev-old and Ldev-new indicate that the differences in N-dev are unlikely to be caused by altered forest management, but must result from atmospheric pollution.
