Floodplains belong to the most complex and species rich ecological systems in Europe. Man has manipulated and partly destroyed these habitats by: substantial impact upon the hydrological regime of rivers, structural alterations (particularly due to embankments which reduce flood retention areas) and increasing nutrient-and pollution-input. All too easily, floodplains are forgotten as an "integral" part of river ecosystems. The river does not occupy the floodplain all the time, but man nevertheless needs floodplains in order to be able to deal with inundations without harm to himself and his property. Therefore, both the concerned public and responsible authorities require an ecologically compatible and sustainable management of floodplains. Achieving this aim requires an adequate understanding of and suitable tools for assessment and prediction of the ecological consequences of impacts within floodplains. These tools must allow one to sufficiently reflect the complex interactions and interrelationships between abiotic and biotic environmental factors of floodplains. Therefore, indicator systems and predictive models are of extreme importance for the advancement of our knowledge of floodplains as a basis for sound planning and management decisions. Indicator systems serve to describe and assess environmental conditions with the help of combinations of easily recordable indicators that complement one another in their information on ecological site conditions. Appropriate indicators allow avoidance of expensive and time-consuming direct measurement of site conditions. Plant and animal species, which signal certain environmental conditions (natural or anthropogenic) and characterise the sites/ biotopes they live in, can be suitable indicators, provided there is at least some knowledge available of their ecological requirements. Living organisms integrate ecological conditions of a site over certain periods of time, whereas direct measurement of abiotic parameters usually provides information only for single points in time. In highly dynamic ecosystems such as floodplains, the latter necessitates continuous measurements over very long periods of time giving special importance to indicator systems.
The same relationships that underlie the use of species or communities as environmental indicators provide a basis for the derivation of species environment models. The models in turn enable a prediction of changes in community composition or fates of single species under anthropogenic or natural changes in environmental factors, such as the hydrological regime. Whereas many indicator systems already exist for water bodies, not much effort has been made in developing transferable indicator systems for floodplains that are robust in spite of the necessary simplifications in practical work. The project RIVA addressed this challenge in the course of the joint research program "Elbe Ecology" of the German Ministry for Education and Research. RIVA stands for "Transferability and further development of a robust indication system for ecological changes in floodplains". A research team of 30 scientists assessed the relationships between species and communities in floodplains and the complex spatial, structural, and temporal dynamics of key hydrological and pedological factors. They selected grassland, respectively higher plants, molluscs (i.e. snails and mussels), carabid beetles, and hoverflies as model systems, located in the active floodplain of the Elbe River. Based on the relationships detected, they developed an indicator system and models that allow the assessment and prediction of effects of changes in ecological conditions within floodplains.
In this special issue we present selected results of the project, focusing on biological topics. DZIOCK et al. set the stage with a review of existing bio-indicator systems in floodplain systems. They show how the approach chosen for the RIVA project represents an advance over previously existing approaches. A major advance was the derivation, implementation and detailed documentation of a very rigorous and coordinated design for samp-