In a model field study, 12 adult Norway spruce (Picea abies (L.) Karst.) stands located in different mountain areas (Harz - H, Thuringer Wald - TH, Erzgebirge - E) in eastern Germany and having different SO2 and NOx depositions, were investigated. In addition to 30 biochemical parameters of needles from previous year covering many different sections of spruce needle metabolism, soil and needle element contents were measured, as were the standard parameters of forest damage inquiry i.e. discolouration and needle loss.

The prime goal of this model investigation was to establish a general method suitable for characterization of what was termed the ''vitality state'' of the Norway spruce trees. This integrated variable was obtained by application of Spearman rank correlation analysis and discriminant analysis. In the first step of univariate/multivariate statistical analysis the 30 biochemical parameters were reduced on the basis of calculated correlation coefficients (p < 0.05). In the next step the three test site groups (H, TH, E) were characterized and evaluated by discriminant analysis with 12 selected variables. After reduction of variables, 4 biochemical/physiological variables (ADP, D1BC, POD3, SORG) were separated out, which contributed to the scores of the 2 discriminant functions, yielding 3 different states of vitality. Between vitality states and environmental variables (SO2, NOx) a significant correlation consists. The clear-cut relationship between the complex variable ''vitality state'' and the environmental conditions may well offer a chance for early recognition of forest decline in Norway spruce in the future.