Beschreibung:
AbstractAimReports on recent changes in high‐mountain vegetation are mostly based on re‐surveys of mountaintop floras. Summits are very specific habitats, however, and detected trends may not necessarily represent alpine vegetation in general. Here, we analyse re‐samples of three prevalent plant communities in non‐summit alpine habitats (snowbeds, two grassland types).LocationNortheastern European Alps.MethodsIn 2013 we re‐sampled a total of 91 relevés of three plant communities from the 1990s. We compared original and re‐samples in terms of species richness (α‐diversity), pair‐wise compositional dissimilarity (β‐diversity) and species pool size (γ‐diversity). Moreover, we calculated mean ecological indicator values (temperature, moisture, nutrient status and soil acidity) as well as potential plant height for each of the 182 relevés. Differences in species pool sizes between the two sampling campaigns were evaluated by randomization tests based on rarefaction curves. In all other cases, differences were compared using paired t‐tests or Kruskal–Wallis tests.ResultsOf all the community attributes analysed, changes were most consistent with respect to increasing mean species' temperature indicator values. Mean potential plant height also increased slightly in all three communities, but differences were statistically significant only in the case of Carex sempervirens grasslands. Changes in species richness varied among communities, with only the Carex firma grasslands accumulating significantly more species at both the local and regional scales. For the other two plant communities, α‐ and γ‐diversity remained constant, but β‐diversity increased. Indicator values other than temperature either did not change or changed idiosyncratically in the different communities.ConclusionsOur results indicate that some long‐term changes occurred in high mountain grassland and snowbed vegetation of the study area. Consistent changes in mean species' temperature indicator values suggest that climate warming was an important driver. However, trends in biodiversity metrics differed substantially between the studied communities. We conclude that recent increase in species numbers commonly reported from temperate and boreal mountaintops cannot readily be generalized to all temperate and boreal high‐mountain vegetation.