Title: Analysis of Stressors Contributing to Hypoxia in Lake Erie Using Deterministic Models

Abstract: Lake Erie continues to experience hypoxia (dissolved oxygen concentrations < 2 mg∙L-1), despite basin-wide reductions in total phosphorus loads intended to limit or eliminate hypoxia, as outlined in the Great Lakes Water Quality Agreement (GLWQA) in 1978.   Despite initial success, periodic hypoxia in the central basin of Lake Erie persisted, and more recently, hypoxic area has enlarged and reemerged as a potential hazard to ecosystem health.  The consequences of hypoxic conditions (e.g., loss of suitable fish habitat and decrease in fish abundance and growth rates) have therefore led to a renewal in interest in understanding the relative contributions to the stressors contributing to hypoxia in Lake Erie.  This dissertation focuses on the impacts of nutrients and hydro-meteorological forcings; although there is some evidence that other potential stressors may contribute to hypoxic severity, such as the role of dreissenids and winter diatom production.

To analyze these stressors, a group of models was developed and applied in multiple management frameworks.  First, a simple dissolved oxygen model was applied in a 1-dimensional, vertically stratified domain.  Meteorological forcings determined temperature and mixing conditions, while the oxygen depletion rate within the water column was adjusted to match observed spatial and temporal dissolved oxygen concentrations; therefore determining if the water column oxygen demand has varied inter-annually, or remained constant.  The model produced an annually varying water column oxygen demand, suggesting that hypoxia is a function of variations in biological activity within the water column (e.g., organic carbon production), and not strictly a meteorological phenomenon.  Second, a more robust lower-food web model was developed and calibrated, using the same 1-dimensional temperature and mixing conditions, however nutrient loads and internal cycling were included to analyze the effects of inter-annual variation in loading magnitude. The analysis suggests that a 46% reduction from the 2003-2011 average total phosphorus load and a 56% reduction from the current GLWQA target would be required to reduce hypoxic area to 2,000 km2.  Finally, several hypothetical scenarios were applied in the lower-food web model, representing variations in load seasonality and meteorological conditions; including two climate warming scenarios.   This work suggests that while there has been significant variation in nutrient loads and annual oxygen depletion rates, meteorological conditions play an important role, due to the duration and intensity of stratification.  The models developed here can provide an estimate of expected hypoxic conditions as a function of loading; however, the forecast estimates must be applied within uncertainty bounds due to the importance of inter-annual meteorological variability.

Committee:

Professor Donald Scavia, Co-Chair

Adjunct Professor Joseph V. DePinto, Co-Chair

Professor J. David Allan

Associate Research Scientist Dmitry Beletsky

Professor George W. Kling