Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Oceanography and Coastal Sciences

First Advisor

Robert P. Gambrell


Removal of reduced manganese in aqueous systems has been described as a two-step process that incorporates a chemical or abiotic step followed by or continued with microbially mediated processes. Field and laboratory studies were conducted to describe the removal of manganese from reservoir tailwaters and partition the abiotic and microbial removal processes. Manganese removal was measured in controlled raceways and in microcosms with four sizes of substrate (cobble, 2.00 to 0.50 mm, 0.50 mm to 0.063 mm, and less than 0.063 mm). Sodium azide was used to inhibit biological uptake. An exponential decay model and a hyperbolic decay model were used to describe removal rates. Field studies verified spatial variability in manganese adsorption to substrate exist downstream from reservoirs and adsorption occurs on a wide range of substrate sizes. Removal rates varied with discharge with a rate near 1 day-1 observed for low flow conditions (0.5 m 3 sec-1) while the removal rate at a higher discharge (0.2 day-1) was considerably lower (0.2 day-1 ). Higher removal rates were obtained in the raceway experiments (1.4 and 1.7 day-1). Removal rates were comparable for cobble and larger size fractions (1.9 to 0.8 day-1) and were similar to raceway rates. Smaller substrate sizes displayed significantly lower removal rates (near 0.3 day -1). Abiotic removal rates (sodium azide inhibition) were significantly lower than rates for uninhibited microcosms. An exponential decay model accounted for more than 90% of the variability for most of the studies (uninhibited) but did not model abiotic alone removal very well (r2 between 0.5 and 0.7). A hyperbolic decay model was fit to the data and resulted in an (r2 from the nonlinear regression greater than 0.95 for nearly all of the microcosms and the field data as well. Initial concentrations and removal rates developed from field and laboratory studies were used as inputs to develop a predictive model that incorporates biological processes which impact manganese removal. This model can be used as a Michaelis-Menton equation which may be more appropriate in describing the two-step manganese removal process than the exponential decay model.