Hypothesis-driven experimental research is necessary for natural resource management

Effective management of natural resources must be grounded in a solid scien tific understanding of the ecosystem and its responses to natural and human -induced stress. Such an understanding does not arise easily from observati onal data and models that are not substantiated by experimental data. Cause -and-effect relationships are more easily documented when observations and/ or models are supplemented by hypothesis-driven experimental research. In t his paper we present three examples from south Florida where hypothesis-dri ven experimental research has been combined with observational data collect ion to address specific resource management questions. These include resear ch to determine: (I)the cause of cattail expansion in the Everglades; (2) a threshold phosphorus concentration for the Everglades; and (3) optimal sal inity criteria for Florida estuaries. In each case, the results have led to a better understanding of ecosystem function and more sound guidance for r esource managers than was possible without the hypothesis-drive experimenta l research. Resource managers need to recognize the merits of this holistic approach to environmental science and management if we are to have success in reversing detrimental human impacts on natural ecosystems.