Water quality, geomorphology, and aquatic life assessments for the Olentangy River TMDL evaluation

by 1975- Witter, Jonathan D.

In the second chapter a customized version of the Soil and Water Assessment Tool (SWAT) model was used to predict hydrology and water chemistry constituent fate and transport for the Olentangy River watershed for the time period 1985-2002. The customized version of SWAT simulates a restrictive layer of material in the soil profile and its impact of subsurface drainage, watershed hydrology, and nutrient and sediment transport. Model parameterization and calibration are presented. Predictions of stream discharge were evaluated at three locations with stream discharge records, however only one location (USGS gage at Claridon, Ohio) was not impacted by Delaware Dam which controls discharge. Regression results for annual and monthly observed and predicted discharge at the Claridon gage provided coefficient of determination values of 0.84 and 0.81, respectively. A similar analysis of annual and monthly discharges with the Nash- Sutcliffe statistic provided coefficient of efficiency values of 0.81 and 0.77, respectively. SWAT predictions for water chemistry constituents for the Olentangy River at Claridon and Worthington, Ohio ranged from 0.32-0.37 mg/l, 3.6-3.8 mg/l, and 114-172 mg/l for total phosphorus, nitrate-nitrogen, and total suspended sediment, respectively. Measured water quality from two studies conducted on the Olentangy and Scioto Rivers provided ranges of 0.23-0.49 mg/l, 3.4-4.4 mg/l, and 47-170 mg/l for total phosphorus, nitratenitrogen, and total suspended sediment, respectively. Following calibration of the model ii all land use in the watershed was changed to forest, prairie, or wetlands to predict the potential water quality of pre-European conditions in the watershed. The results show that delivery of sediment, total phosphorus, and nitrate-nitrogen to the stream system has increased by factors of approximately 50, 20, and 3, respectively, for most subwatersheds in the Olentangy River watershed. In chapter 3 the calibrated SWAT model was used to predict the potential water quality benefits of alternative management scenarios and conservation practices in the Olentangy River watershed. Alternative management scenarios and conservation practices that were evaluated include: 1) alternative crop rotations, 2) alternative fertilizer application rates, 3) alternative types of tillage, 4) timing of tillage operations, and 5) buffer strips of various widths. A 33-meter buffer strip was predicted to provide the largest water quality benefits when compared to other alternative management scenarios. Inclusion of winter wheat into corn-soybean crop rotations also provided significant water quality benefits. Another study was conducted to develop a geomorphology assessment index to evaluate dynamic equilibrium at study sites in the Olentangy River watershed. Regional curves were developed using geomorphology data from study sites impacted by a variety of disturbances. The geomorphology assessment index was developed to provide a logical and minimally biased framework to assess dynamic equilibrium at study sites in the Olentangy River watershed. A diagnostic approach that integrated multiple lines of evidence was used to develop the index. Thirty-six sites in the watershed were assessed with the geomorphology index. The geomorphology index indicated that many streams iii in the watershed are incised and further land use change could pose significant threat to the geomorphology of the stream system. In chapter 5 a statistical analysis was conducted to examine the relationships between stream biology and environmental variables such as water quality, habitat, and geomorphology. The spatial location of a study site within the drainage network and its impact on the structure of the biological community was also evaluated. Correlation analyses were performed to develop a parsimonious set of explanatory variables. Regression analysis was used to confirm the significance of the reduced dataset. Canonical correspondence analysis (CCA) was used to elucidate relationships between biology and environmental variability. Partial CCA techniques were used to partition variation among the groups of environmental variables. Results of the correlation analysis identified nine significant environmental variables. The significance of those variables was confirmed with regression analysis. The CCA identified stream gradient, pool quality, stream order, and width to depth ratio as the most important environmental variables that explained the structure of biological communities. Environmental variables were grouped into three categories including geomorphology variables, habitat variables, and spatial location variables. The partial CCA technique was used to partition variation between each of the variable groups. Respectively, 37%, 19%, and 24% of the variation was attributed to geomorphology variables, spatial location variables, and habitat variables. The remaining variation was not explained by the measured variables. iv