Plant indicator values of a high-phytodiversity country (Italy) and their evidence, exemplified for model areas with climatic gradients in the southern inner Alps

We tested the new Ellenberg–Pignatti indicator values in Festucetalia valesiacae communities of three valley regions (southern inner Alps: Valle d’Aosta, Valtellina, Valle Venosta/Vinschgau). In these landscape types gradients of all value types exist: light (L), temperature (T), moisture (F/M), nitrogen (N), continentality (C) and soil reaction (R). L- and T-values were extended from 9 to 12 in the database of Pignatti due to the higher intensity of these factors in the area in question; in our dataset this is especially relevant for the L-values.In a spatial approach, we compared DCA results with indicator value results (presence–absence) and with climatic data for eight different plant communities. We found comprehensible results in all cases, also for the extension of the L-factor.In a temporal approach, we compared relevés and their indicator values from two time windows: 1930–1950 (Braun-Blanquet) and 1990–1995 (Schwabe and Kratochwil) for six of the eight different plant communities. Both relevé types were made in the same or in similar localities, but no permanent plots existed. Mostly we found in these communities, which are said to have high “stability”, no significant changes on average for any of the indicator values. In the case of N- and R-values there was a trend of increase in the different datasets, and in some cases this trend is significant.In general, the new database has shown the “robustness” of the Ellenberg indicator value model including the extended L-values, and established its usefulness for a synthetic approach in the studied valley systems. In our systems L-, T-, F/M- and C-gradients can be characterised well by means of indicator values and are an excellent tool for explaining DCA results. Caution is necessary especially when correlations of different factors are intermingled in temporal approaches.