Ausweisung von symphänologischen Gruppen montaner Grünlandgesellschaften des Osterzgebirges mittels Clusteranalyse

Abstract

Richter, F.; Zöphel, B.: Symphenological species groups of mountain meadows in the Eastern Ore Mountains (Germany) based on cluster analysis. – Hercynia N.F. 39 (2006): 51–68.In this study we tested a method for classifying symphenological groups according to the generative part of the life cycle of species typical for mountain grasslands. We realized a need for harmonising grassland management as mowing, grazing or hay transfer with phenological characteristics during work in a restoration project for mountain meadows. The study is based on weekly recorded phenological data for 54 grassland species at 17 mountain meadow sites in the Eastern Ore Mountains at an altitude of 475–875 m above sea level in the course of the vegetation period 2004. We used a semiquantitative key of the phenological stages of species. Using different criteria for defining groups following hierarchical cluster analysis we classified symphenological groups of flowering, ripening and dispersal behaviour. The analysis included data for specific phenological stages related by the beginning of full flowering of a very common and representative species as well as durations of selected stages. The resulting groups, especially flowering groups, show high accordance with the established symphenological groups from Dierschke (1995). Some species however were found to be classified differing from the suggestions of Dierschke (1995). In several cases the species composition of the ripening and dispersal groups follows another classification than that of the flowering groups. This underlines the expedience of separate groups for fruit stages according to their relevance for nature conservation. Generalized types of fruit phenology were determined by considering classes of durations of ripening and dispersal stages. Their response to grassland management is discussed. It is obvious that for use in practical grassland management time ranges of groups are still too large. One reason is that concerning to cluster analysis method time strictness of groups is equal independent of their real duration. On the other hand wide ranges are the result of different response of phenological characteristics to site factors specific to the species. Also the differing behaviour of individuals and populations in some species complicates a more constricted classification. Due to this the use of phenological indicator species is limited. Thus further studies and improved analysis methods are required to develop well founded and practicable phenologically based concepts in grassland management.