• river;
  • floodplain;
  • flow pulse;
  • temperature;
  • ecosystem process;
  • expansion;
  • contraction;
  • biodiversity;
  • conservation;
  • landscape


The flood pulse concept of Junk, Bayley and Sparks is a major contribution to our understanding of river–floodplain interactions and has become an important paradigm in lotic ecology. The concept is based mainly on large tropical lowland rivers. Floodplains may, however, develop in all geographical areas and at different locations along a river corridor. We extend this concept to temperate areas by including information derived from near-natural proglacial, headwater and lowland floodplains. Specific attention is directed to the role of temperature as a major determinant of floodplain ecology. Further attention is directed to the importance of expansion–contraction cycles occurring well below bankfull (‘flow pulse’ versus ‘flood pulse’). Selected examples are presented that highlight the complexity of expansion–contraction events and their consequences on habitat heterogeneity and functional processes. Habitat heterogeneity is mainly a product of shifting water sources, different flow paths and the relative importance of autogenic processes. In different floodplain systems, expansion may enhance habitat heterogeneity (e.g. glacial floodplain) or create homogeneity (e.g. Danubian floodplain). Further, the ecological consequences of episodic flow and flood pulses are discussed. Finally, a landscape approach is suggested in order to document expansion and contraction processes and to elucidate how these processes influence landscape heterogeneity and biodiversity patterns. Such a landscape-based ecosystem model can be applied to rigorously assess the ecological integrity of river–floodplain systems. Copyright © 2000 John Wiley & Sons, Ltd.