Ökotoptypspezifische Quantifizierung von Ökosystemfaktoren in ihrer Bedeutung für die Beobachtung von Umweltveränderungen im Unterspreewald

Abstract

The Lower Spree Forest (state of Brandenburg), together with the Upper Spree Forest further upstream, is a unique landscape in Central-Europe because of the dendritic drainage pattern of the river Spree and the typical structure of the cultural landscape. The political and economical changes since 1989/90 have led to radical ecological changes, as for example the lack of water during summer months and increasing eutrophication of ground- and surface water. Within the framework of an applied ecological monitoring system, which was initiated in order to register man made ecosystem changes due to surface mining and amelioration, 34 monitoring sites were established. Between 1996 and 1998 groundwater monitoring, sediment drillings, and soil mapping as well as vegetation studies delivered diverse ecosystem data which were managed by means of a complex database. The results of detailed geomorphological and soil mapping were synthesized and brought together in a map showing ecosystem units based on relief structure. Every unit is characterized by similar abiotic and biotic landscape features. The ecosystem units were used as an evaluation base for statistical analysis of field- and laboratory-borne data. Very good results of discriminant analysis show that the classification of the databases with the help of ecosystem units is a practicable and efficient way to transfer monitoring data, registered at selected points, to a spatial extent. The reclassification of a special dataset, including a combination of groundwater, soil, and vegetation data by using discriminant functions, shows that every dataset could be clearly assigned to the empirically found ecosystem unit. Hereby, the computed discriminant functions are exclusively based on geoecological parameters with a high discriminant value. Moreover, the results of multiple correlation analysis show middle to strong correlations between groundwater, soil, and vegetation data and thus express the complex structure and interactions characteristic for the whole ecosystem. They prove that man-made changes of single factors must inevitably lead to consequent changes of further geoecological variables. In addition, Fisher-classification functions were used to develop a new method of monitoring and early recognition of future ecosystem changes. In future, the repeated recording of 14 selected (key-)variables within the framework of the existing monitoring system permits quantitative checking stem unit identification of a monitoring site and can be used to register dynamic elements within the ecosystem structure. The presented investigations are an example for the both useful and necessary combination of qualitative and quantitative working methods as well as for the relevance of geographic contributions to the solution of current environmental problems.