Assessing the impacts of watershed indexes and precipitation on spatial in-stream E. coli concentrations

作者：Pandey Pramod K^*; Soupir Michelle L; Haddad Monica; Rothwell James J

来源：Ecological Indicators, 2012, 23: 641-652.

DOI：10.1016/j.ecolind.2012.05.023

• In-stream E. coli concentrations
• Watershed indexes
• Rainfall
• Nonpoint pollution
• Watershed scale modeling

Pathogen contamination of waterbodies, which is often identified by the presence of pathogen indicators such as Escherichia coli, is a major water quality concern in the United States. Reducing in-stream pathogen contamination requires an understanding of the combined impacts of land cover, climatic conditions, and anthropogenic activities at the watershed scale. In this study these factors are considered by assessing linear relationships between in-stream E. coli water quality data, watershed indexes, and rainfall for the Squaw Creek Watershed, IA, USA. The watershed indexes consider the undisturbed land cover which encompasses the natural land cover area, wetlands, and vegetated stream corridors, and the disturbed land cover extent which includes areas receiving manure from confined animal feeding operations (CAFOs), tile-drained areas, and areas in cropped and urban land. In addition to disturbed and undisturbed land, we also calculated indexes for barren land and slope. Bivariate analysis was used to assess the linkage between E. coli concentrations, watershed indexes and the cumulative rainfall 15, 30, 45, and 60 days prior to water sample collection. To predict in-stream E. coli concentrations, we developed multivariate regression models, and predictions were compared with observed E. coli concentrations at 46 sampling locations over four sampling periods in two years. Results show that areas receiving manure, wetlands, drained land, and cropped land all influence in-stream E. coli concentrations significantly (p %26lt; 0.001). The coefficient of determination was higher when indexes were corrected using the cumulative rainfall 30 days prior to the sampling event. Model skill varied from 0.29 to 0.55. More than 95% of the predictions across all spatial locations fall within one order of magnitude of the observed values. This Geographic Information System (GIS) based approach for predicting in-stream E. coli concentrations appears to be a useful technique for assessing the impacts of land management on water quality. Published by Elsevier Ltd.

• 出版日期2012-12