Summary
The life histories of fishes from three different ecosystems are compared and discussed with emphasis on species that move between habitats: between estuary and sea in tropical northern Australia; between bottom waters and pelagic waters of the temperate continental slope off Tasmania; and between river and pans in a tropical African floodplain. The advantages and disadvantages of such movements are discussed in relation to proximate factors such as cannibalism, and in an evolutionary context. The reality of divisions between habitats has to be critically examined; whereas some are pronounced, others may be generated mainly by the dimensions and limits of research projects.
The lack of sufficient basic knowledge of the biology of most species, especially in the tropics and deep sea, and the continuing decline of support for basic research are an important constraint to the development of reliable models. It is stressed that further advances in understanding relationships and differences between ecosystems require new commitments to fundamental work, and depend upon interaction between all involved in marine and freshwater research.