Present wetland management has been developed largely on a trial and error basis (Figure 1 ). The effects of many important environmental variables on wetland productivity 253 are not known, consequently management results have not been predictable with a high degree of accuracy (Weller 1981 ). Many marsh management techniques have been described; however, consistently successful marsh management requires a more comprehensive understanding of the structure and function of wetland systems. Although there have been numerous observational studies, major advances in our understanding will result from tightly controlled experimentation which permits the integration of simultaneous research efforts by a number of different scientific disciplines (Reichle 1975, Weller 1978). Because wetlands are temporally dynamic, this type of multi-disciplinary ecosystem analysis must also span a number of years to document the annual and long-term variability within the system. By better understanding the structure and function of wetlands, managers will be better able to design management techniques and strategies suited to their particular situation and therefore realize greater success in manipulating the productivity of these systems (Figure I).