| Throughout the world, we face a legacy of thousands of degraded rivers and floodplains with limited knowledge about how these systems are affected by development works and what actions are possible to restore them. Few river-floodplain systems better exemplify this situation than the Zambezi Delta in Mozambique. Over the millennia, the annual spread of Zambezi floodwaters over the delta provided recession agriculture, hunting, fishing, and vital natural resources and supported diverse and abundant wildlife populations. Over the past century, however, large dams and other development projects have harnessed Zambezi floodwaters for hydropower and large-scale agricultural development with little concern for downstream livelihoods or ecosystems.; In this dissertation, I hypothesize that Zambezi development has resulted in significant hydrological changes in the delta over the past century and analyze long-term changes in regional runoff patterns, flooding processes, and water balance components. I assess key indicators of hydrological alteration, relating changes in the magnitude, timing, duration, and frequency of river flows to social and ecological changes in the delta.; Second, I hypothesize that hydrological changes have resulted in significant changes in the distribution and dynamics of vegetation in the delta. I compare the historical distribution of vegetation communities of the delta with the current distribution and assess patterns of directional, or non-cyclical, change. I examine the processes of woody species invasion into floodplain grasslands, displacement of flood-tolerant species by more upland species, terrestrialization of floodplain water bodies, displacement of freshwater vegetation by salt-tolerant species, and degradation of coastal mangrove.; Third, I hypothesize that an effective prescribed flooding program can be implemented for the restoration of the delta. Using long-term time series data, I simulate the capacity for different flood release scenarios within the constraints of regional runoff patterns, reservoir design, and hydropower demands. I model the potential to recreate historical flooding patterns and to generate short-duration, high-volume flood releases. I describe steps for defining flood release objectives, determining the structural and financial feasibility of flood releases, developing stakeholder participation and institutional support for releases, selecting a flood release strategy, and establishing a monitoring and evaluation program for the adaptive management of Zambezi waters. |
