THE EFFECTS OF URBANIZATION ON THE WATER QUALITY AND HYDROLOGY IN THE KASKASKIA RIVER WATERSHED IN SOUTHERN ILLINOIS, UNITED STATES

Author

Charnsmorn Hwang, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2010

Degree Name

Master of Science

Department

Forestry

First Advisor

Williard, Karl

Abstract

The Kaskaskia River is a large river system situated in central and southwestern Illinois serving as a tributary to the Mississippi River. Within the Lower Kaskaskia River Watershed (LKRW), an active urban-rural gradient currently exists in the Metro East area of St. Louis. Such areas of urbanization are particularly vulnerable to stream degradation. This study focuses on the effects of urbanization on water quality parameters within Silver and Richland Creeks, both of which are tributaries to the LKRW. Forty-three catchments within Silver and Richland Creeks were identified as study catchments. Stream water samples were collected within these catchments every two weeks in the dormant season and monthly in the growing season from January 2008 to August 2009. Stream storm samples were collected and stage was recorded within 4 intensively sampled catchments, which were representative of urban, village and agriculture watersheds. Stream samples were measured and analyzed for total suspended solids (TSS), turbidity, pH, specific conductance, total coliform, bacteria coliforms (as total coliform, fecal coliform (FC), and Escherichia coli (EC)), nutrients (orthophosphate, ammonium-N, nitrate-N, chloride, sulfate), and caffeine. GIS was utilized to identify percent urban land cover (LC) at the whole catchment scale and percent impervious surfaces (IS) at the riparian buffer scale. Whole catchment urban land cover (WCULC) was used to designate land cover categories of urban, village, and agriculture watersheds (>10.5%, >1% to 10.5%, and 0% to 1% WCULC, respectively). IS were identified at the riparian scale through manual digitization and classification of structures (i.e. buildings, houses) and roads (i.e. highways, streets) within 10, 30, and 50 m buffer widths. Correlations of water quality variables to percent whole catchment urban LC were comparable to that of riparian-scale IS (RIS). Whole storm Event Mean Concentrations (EMCs) for nitrate-N, orthophosphate, and sulfate were generally significantly lower in urban watersheds compared to village and agriculture watersheds. Stream water levels of EC, FC, and orthophosphate were relatively high. Levels of FC at both baseflow and stormflow and EC at stormflow far exceeded US EPA and IL EPA Review criteria, respectively. During baseflow, stream orthophosphate and nitrate concentrations within urban watersheds were significantly higher than in village and agricultural watersheds. The significant nutrient and bacteria levels in urban streams may be due to inputs via stormwater runoff, wastewater treatment effluent, and home septic systems. Although hydrometric data showed no significant differences among the intensively sampled catchments, the village watershed had much higher mean and larger maximum stream discharge compared to the urban and agriculture watersheds. Caffeine concentration in streams was not a useful indicator of anthropogenic impacts within the LKRW study area. Results from this study demonstrate that watershed managers can utilize WCULC since it is comparable to RIS. In addition, results further illustrate the need for urban best management practices to reduce water quality impacts, such as storm water management, improved wastewater treatment, and maintaining or developing vegetated riparian buffers.