I evaluated diet and food selection of 5 species of spring-migrating female waterfowl including 3 dabbling ducks (Blue-winged teal, Anas discors, Mallard, Anas platyrhynchos, Gadwall, Anas strepera) and 2 diving ducks (Lesser Scaup, Aythya affinis, and Ring-necked duck, Aythya collaris). Diet was evaluated with regards to the proportion of invertebrates and seeds consumed, and compared to forage availability data collected in habitats available to them at 6 study locations throughout the Upper Mississippi River and Great Lakes Region. I found latitude (i.e., stage of migration), longitude, food availability, and date all influenced the diet of spring migrating waterfowl, with some factors having a stronger influence than others. I observed differing diet trends with regard to foraging guild (e.g., dabbling and diving ducks), as each foraging guild was represented by 1 species that was heavily dependant on invertebrates (dabbling duck - Blue-winged teal; diving duck - Lesser scaup) and 1 species that was heavily dependant on seeds (dabbling duck - Mallard; diving duck - Ring-necked duck). The proportion of invertebrate foods in the diet increased throughout spring for all species of waterfowl, suggesting the importance of invertebrate food sources during spring staging. Data from this study provides valuable information to habitat managers and conservationists wishing to improve spring habitat conditions for migrating waterfowl, which likely influences waterfowl productivity.

This thesis details the work completed in order to develop a hydrodynamic and nitrate transport and reaction model for Round Lake, a backwater on UMR Pool 8. This work begins with investigating the fundamentals of nitrogen removal in aquatic ecosystems and reviewing other combined hydrodynamic and nutrient modeling efforts. Field data were gathered to determine model boundary conditions and provide a basis for calibration and validation. Using this data, the flow regime in Round Lake was simulated. CFD applications to model particle residence times and species transport and reaction were used to analyze the effects local hydraulics have on nitrogen removal in the lake. Results demonstrated an ability for CFD to predict spatial variation of nitrate with this ecosystem.

The Upper Mississippi River System (UMRS) is a diverse and dynamic ecosystem that includes the main stem river channel, side channels, backwater floodplains and lakes, islands, wetlands, grasslands, and floodplain forests. The hydrology of this rich ecosystem is one of the key drivers for physical, chemical and biological processes. However, the hydrology and hydraulics of the UMRS has been drastically altered from its natural state as a result of the construction of the locks and dams in the 1930s. Beginning with the Water Resources Development Act of 1986, biologists, ecologists, and engineers have been working to restore the river to a more natural state within the current constraints imposed by the lock and dam system. In an effort to restore rivers to a more natural state, the determination of a hydraulic reference condition is essential to understanding the "why and how" of historical river system function. Understanding the fundamental processes of historical conditions will help prioritize resources and better quantify possible outcomes for riverine restoration. The main goal of this study was to construct a hydrodynamic reference condition for Pool 8 of the Upper Mississippi River System using hydrodynamic computational fluid dynamic (CFD) modeling. The CFD model will provide a better understanding of pre-impoundment flow conditions as compared to post-impoundment conditions today. The numerical model was constructed and developed primarily from a pre-impoundment 1890s topographic map with bathymetric cross-sections in the channels. The 1890s map and other sources from the U.S. Army Corps of Engineers provided historic elevation and hydraulic reference data for model calibration. The calibrated historic model was then compared with a current model of similar scale representing post-impoundment conditions, allowing for quantitative analysis of the differences between the two conditions. Model results indicated large changes in average depth and average velocity between historic and current conditions in certain parts of the pool, while others remained relatively unchanged. For example, velocities decreased in main channel aquatic areas in the lower part of Pool 8 from an average of 0.6 m/s (2.0 ft/s) under historic conditions to 0.1 m/s (0.3 ft/s) under current conditions. In the same part of the pool, however, velocities in contiguous backwater areas remained relatively constant, with most remaining less than 0.25 m/s (0.82 ft/s). Additionally, in the lower part of the pool, discharge distribution between the floodplain areas and the main channel was historically much more dynamic, with flow concentrated in the main and secondary channels at discharges less than 2265 m3/s and in the floodplains at greater than 2265 m3/s. Under current conditions, discharge distribution is much less dynamic, with approximately 2/3 of the total discharge conveyed on the floodplain for all discharges modeled (283 m3/s to 2832 m3/s or 10,000 ft3/s to 100,000 ft3/s).