Submitted to: Annual Conference on Ecosystems Restoration and Creation
Publication Type: Abstract Only
Publication Acceptance Date: 3/12/2009
Publication Date: 7/20/2009
Citation: O'Donnell, T.K., Anderson, S.H., Baffaut, C., Oster, T. 2009. A Hydrologic Event-based Evaluation of Water Quality Trends in Goodwater Creek Experimental Watershed, Missouri USA: Implications for Watershed Monitoring Strategies and Objective Setting [abstract]. In: Third National Conference on Ecosystems Restoration, July 20-24, 2009, Los Angeles, California. Available: http://www.conference.ifas.ufl.edu/NCER2009/pdfs/Abstract_BOOK.pdf. p. 278.
Technical Abstract: Continued public support for U.S. tax-payer funded programs aimed at reducing agricultural non-point source pollutants depends on clear demonstrations of water quality improvements. However, little is currently known about past watershed-scale effects due to implementation of structural best management practices (BMPs). Effectiveness of structural BMPs, as well as watershed monitoring networks are important information needs to make future efforts more defensible. Watershed-scale assessments of BMP effects using existing monitoring data can simultaneously inform program managers on the type and spatial coverage of physical processes (i.e., monitored variables) critical to tracking future water quality improvements. The objective of this research is to determine if significant linear trends exist between atrazine at three locations in the Goodwater Creek Experimental Watershed (GCEW) and flow, precipitation, and structural BMP implementation chronology between 1993 and 2006. Atrazine is a herbicide of great concern for surface water contamination in the U.S. Midwest. Structural BMPs implemented in the GCEW included grass waterways, terraces, and establishment of permanent vegetation. The GCEW is a 73 km2 instrumented watershed located in the north-central Missouri and currently included in the USDA Conservation Effects Assessment Project (CEAP) watershed assessment studies. Three stream gauges with automated water samplers and ten rainfall gauges provided sub-daily data at varying seasons and years between 1993 and 2006 for three nested watersheds (12, 31, 73 km2). Hydrologic events at each weir were determined by the Hewlett-Hibbert constant slope method. A total of 282 events were identified at the watershed outlet between 1993 and 2006. Atrazine flow-weighted concentrations and loads were determined for each hydrologic event by straight-line interpolation between consecutive chemical samples occurring during the same event. Closer examination of hydrologic events indicated 116 events were sampled during the months of April through June, a critical period for atrazine loss. Variables useful for predicting atrazine trends included event discharge, time to peak discharge during an event, and rainfall contributing to an identified event as well as prior to an event. Results indicated variation associated with atrazine trends was not reduced by including all rainfall gages in linear trends. Effects of BMPs were not consistently shown after 14% of the total watershed area was protected by structural BMPs. However, resulting trends are useful in quantifying minimum reductions in atrazine needed for detection with monitoring networks.