Parameters Associated with Fecal Indicator Bacteria Concentrations in the Water Column and Bottom Sediments of Stroubles Creek, Virginia, USA

Abstract

Abstract. High levels of fecal indicator bacteria (FIB) are the leading cause of surface water-quality impairments in the United States (USEPA, 2012). Under the 1972 Clean Water Act, states are required to address identified impairments through the development of Total Maximum Daily Load (TMDL) restoration plans. Due to the technical difficulties and considerable expense associated with continuous monitoring of pathogen indicators, water quality models are commonly used in the TMDL process to identify the relative contributions of various watershed sources of FIB and to develop associated restoration plans. The ability to develop robust and effective TMDLs is therefore directly dependent on the quality of models; however, efforts to model bacteria fate and transport are notoriously plagued by uncertainty (Benham et al., 2006; Shirmohammadi et al., 2006). Several recent critical reviews have remarked that most currently available microbial water quality models represent bacteria as “free” or dissolved phase contaminants with near-neutral buoyancy (Jamieson et al., 2004), although a growing number of studies indicate that a significant amount of bacteria are particle-associated (Bai and Lung, 2005; Fries et al., 2006; Jamieson et al., 2005; Krometis et al., 2009; Rehmann and Soupir, 2009; Russo et al., 2011). This is of potentially significant concern, as these models therefore fail to account for sediment accumulation or resuspension of bacteria. Bacteria associated with sediments are also reported to survive longer in the environment and so potentially pose human health risks for an extended period of time (Droppo et al., 2009). Current implementation of more complex environmental processes within existing models of bacteria transport in receiving waters is limited by scant available observational data for calibration/validation purposes. Past attempts to incorporate sediment-bacteria interactions within watershed-scale modeling packages have relied completely upon field or lab-scale observations of bacteria partitioning between the free and particle-associated phase (Bai and Lung, 2005; Cho et al., 2010; Jamieson et al., 2005; Kim et al., 2010; Russo et al., 2011). No identified peer-reviewed studies have included verification of model outputs via comparison with field observations of both water column and bottom sediment concentrations. This present study details the results of a monitoring regimen focused on recording E. coli and enterococci levels in the sediments and waters of an intensely instrumented urban stream; this reflects the first step in a broader effort to assess the impacts of incorporating more complex bacteria transport processes within watershed models on predictions of in-stream concentrations and associated community health risk. The specific objective of the work presented here is to identify potential correlations between FIB levels in the water column and underlying sediments with climatic and other physio-chemical water quality parameters in Stroubles Creek (Blacksburg, VA). Stroubles Creek is currently included on the Commonwealth of Virginia’s 303(d) impairment list due to elevated E. coli concentrations, with an associated required TMDL. A segment of the creek is continuously monitored for a variety of hydrologic and water quality measures by the Virginia Tech Stream Research, Education And Management (StREAM) Lab with high-resolution equipment, making it a unique study site. Concentrations of two FIB (E. coli and enterococci) in the water column and underlying sediments from three sites (Figure 1) along the creek were monitored weekly for one year (02/2012 – 01/2013). Water and sediment FIB concentrations were then statistically compared to various climatic and water quality parameters collected in real-time by the StREAM Lab, including rainfall, discharge, temperature, specific conductivity, pH, turbidity, and dissolved oxygen. Antecedent precipitation values (12-hr, 24-hr, 48-hr, and 72-hr rainfall) were calculated based on hourly rainfall data. Preliminary comparisons of the data collected over the study period indicates that both E. coli and enterococci concentrations are highest during the late spring, summer, and early fall month (May through October); during this period the stream is essentially in continuous violation of both the applicable geometric mean and statistical threshold value standards (Figure 2 and 3). It is worth noting that the bed sediments collected following collection of these water column samples harbored 10-400 times higher concentrations of FIB than the water column on the same volume basis (data not shown). The statistical analysis results indicated increases in FIB concentrations in both the water column and bed sediments were strongly correlated with increases in antecedent precipitation, increases in temperature, and decreases in dissolved oxygen. Increases in rainfall, and specific conductivity had minor effect on FIB concentrations in the water column, but were strongly correlated with decreases in FIB concentrations in the bed sediments (Spearman rank correlation test, alpha=0.01). Continuing analyses should provide insights to watershed managers and those researching water quality model applications.