AQUATIC PLANT GROWTH MODEL REFINEMENT FOR THE UPPER MISSISSIPPI RIVER-ILLINOIS WATERWAY SYSTEM NAVIGATION STUDY. Elly P.H.Best1, Gregory A. Kiker1, Beth A. Rycyzyn1, Kevin Kenow2, Jim Fischer3 Shyam K. Nair4 and Dan B. Wilcox5 1 U.S. Army Engineer Research and Development Center, Environmental Laboratory, Vicksburg, MS 39180 2 U.S. Geological Survey, Upper Midwest Environmental Sciences Center, La Crosse, WI 54602, 3 Wisconsin Department of Natural Resources, Upper Midwest Environmental Sciences Center Onalaska, WI 54650, 4The Cadmus Group, Inc., Oak Ridge, TN 37830 5U.S. Army Corps of Engineers, St. Paul District, St. Paul, MN 55101. Simulation models have been developed for two submersed plant species, reproducing vegetatively through tubers. These species, Sago pondweed -a canopy former- and American wildcelery -a non-canopy former, represent the characteristic life forms of submersed aquatic vegetation in the Upper Mississippi River System. The models are based on carbon flow through a vegetation occupying one m2 water column. They take effects of changes in (1) water depth, (2) shading by seston, (3) self-shading, and (4) temperature on plant biomass formation into account. The models have been validated using field data sets from The Netherlands, California, and India. For application to the UMR System the models were expanded with equations accounting for effects of current velocity and shading by epiphyton. Calibration data for these effects were derived from new pertinent field data collected in 2002. The effects of current velocity on plant biomass of American wildcelery were quantified in a study on the Red Cedar River, WI. The potential shading effects of epiphyton on Sago pondweed and Wildcelery plants were quantified using plant samples from UMR-Pool 8. A new field data set composed by data on shoot biomass of both species and environmental factors, collected at selected sites in UMR Pool 8 and Pool 13 in 2001 and 2002, was used for model validation. Output generated by the refined plant growth models agreed largely with the measured data, in that the model output generated for several sites in Pool 8 was within the range of the measured data. Simulated plant development was slightly delayed compared to local plant development. Both increased current velocity and shading by epiphyton decrease plant biomass. The combined effect is expected to be highest at sites with a high current velocity. Current velocities > 0.9 m s-1 prevent growth of both species. Plant density in established vegetations of both plant species, to which the default model input values pertain, is relatively constant. In one-year runs, simulated plant biomass is sensitive to initial values on tuber density and size, which may deviate from the default values. Simulated biomass from runs started from default tuber density/tuber size consistently overpredicted measured plant biomass of Sago pondweed, leading us to believe that in UMR-Pool 8 this vegetation starts from a far lower tuber density (10 m-2) than the default one (240 m-2). Simulated biomass from runs started from the default values agreed with the measured plant biomass of Wildcelery. Interpretation by comparison of simulated and measured data was greatly impeded by the : (1) large variability in measured plant biomass data, (2) lack of measured plant growth curves, tuber density and – size, (3) relative scarcity of measured environmental data requiring large-scale interpolation and derivation of values pertaining to other sites within the same water body. The refined models can be used to explore possibilities to modify existing river management practice, and to implement operational scenario’s aimed at conserving/optimizing submersed aquatic vegetation. Keywords: Mississippi River, aquatic plant growth model, refinement, validation