Neue Ernährungskennwerte für die forstlichen Hauptbaumarten in Mitteleuropa - Eine statistische Herleitung aus VAN DEN BURG'S Literaturzusammenstellung

Since decades, foliar nutrient concentrations are used as a basis for estimating the nutritional condition of forest stands. Unfortunately, the current reference system in Germany for the evaluation of the nutritional status (WOLFF and RIEK, 1997) is incomplete with respect to the main forest tree species. To improve the information base, the literature compilation made by VAN DEN BURG (1985, 1990), which contains a huge amount of nutritional threshold values, was entered into a database and evaluated with statistical methods. The new nutrition levels derived from this dataset for the main forest tree species in Central Europe should provide an exceptional basis for the interpretation of the results of the second German forest soil survey BZE II. The systematic relation of threshold values to plant reaction (deficiency symptoms, reaction to fertilizers, growth) is a unique trait of VAN DEN BURGS literature compilation (Fig. 2, Fig. 3) and allowed for the definition of up to seven critical concentrations for species specific nutritional thresholds (.Tables 1 to 9). VAN DEN BURG'S compilation of literature is a hand written manuscript of more than 700 pages. Since the tables of VAN DEN BURG both, spans and only single values are given, data had to be adapted accordingly. We attributed the column averages in the presence of only one value but also recorded boundaries between two columns if the appropriate values had been provided. In total, the VAN DEN BURG data contains 4066 records from nine tree species in Europe. Before the definition of critical concentrations could be set, the data base was thoroughly reviewed and corrected for outliers (Fig. 4). Most data from VAN DEN BURG'S literature compilation were distributed in the range of normal nutritional supply. In order to classify this range in a reliable way, we calculated the lower and upper limit of the normal range by adding and subtracting one standard error from the mean of the normal range (Fig. 5). The division into three sub-ranges was attained by respectively halving the span between the upper limit and the mean and the lower limit and the mean in the normal nutritional range (column 6). Deficiency and surplus ranges were similarly deduced (Tables 1-9). The available data for Scots pine (Table 1) and Norway spruce (Table 2) were sufficient to additionally calculate tables of nutrient threshold values for young and mature stands. It was also possible to differentiate between common oak and sessile oak (Table 4, Table 5). The threshold values for oak plausibly reflect the differences in nutrient demand between these two species. Additionally, critical values for silver fir (Table 6), larch (Table 7), sycamore maple (Table 8), and European ash (Table 9) could be provided. In comparison to WOLFF and RIEK (1997), the information base for the nutrition of forest trees species could be substantially improved. For pine, spruce, beech and oak trees the threshold values for deficiency which were derived from VAN DEN BURG are often higher than the deficiency values used for BZE I, thus leading to a more cautious interpretation of nutrient concentrations. As shown for stands of spruce and oak growing on gravel beds of the Lech River (Fig. 1, Fig. 6), the categorization of foliar potassium (K) concentrations to nutrient thresholds is more plausible when using the values derived from VAN DEN BURG. The presentation of the new nutrition levels for the main forest tree species in Central Europe should inspire further steps of testing and developing critical nutrient concentrations.