Variability of E. coli in streambed sediment and its implication for sediment sampling

Highlights

• The highest E. coli concentration was observed in the upstream site of direct fecal deposit compared to the downstream SRAM sites.

• There was no significant difference in sediment E. coli concentration between the edge and the middle of the stream.

• Sediment samples can be analyzed within 24-h of sample collection in most cases (80%).

• A single grab sample may not be able to provide the data of E. coli concentrations in sediment without substantial error.

Abstract

E. coli is the number one cause for water quality impairments in rivers and streams in South Dakota and the United States. Stream bottom sediments can be a reservoir for bacteria, including pathogenic organisms and fecal indicator bacteria (FIB), due to the favorable conditions provided by sediments for survival. Despite this, little is known about the variability of E. coli in sediments which should be considered when developing a sampling regime. This study examines the spatial variability of E. coli in sediment across the stream cross-section, the temporal stability of E. coli in sediment samples, and the implications for sediment sampling and processing. Five locations were sampled for sediment E. coli along two tributaries to the Big Sioux River in eastern South Dakota, four along Skunk Creek (Sk1, Sk2, Sk3, and Sk4), and one in Sixmile Creek (SM). In Skunk Creek, site Sk1 has direct cattle access where the other three sites (Sk2, Sk3, and Sk4) are under Seasonal Riparian Area Management (SRAM), a strategy that limits the cattle access to the stream. E. coli concentrations in the sediment ranged from 4 to 997 CFU g⁻¹ (8.5 × 10² to 2.1 × 10⁵ CFU 100 mL⁻¹). The highest and lowest E. coli concentrations observed were at sites Sk1 and Sk4, respectively. The E. coli concentration decreased from the upstream cattle crossing site (Sk1) through the downstream SRAM sites. Analyzing the stream cross-section analysis revealed no significant difference in E. coli concentration between the edge and the middle of the stream. Sediment samples can be held up to 24 h after sample collection in refrigerated conditions (37 °F) in the majority of cases (80%) without significant changes in E. coli concentrations. The sample size analysis indicated the spatial variability of E. coli across the stream cross-section is high and a single grab sample may not be able to provide adequate representation of E. coli concentrations in sediment without substantial error. The findings provide insight for designing E. coli monitoring projects and promote the awareness of unconventional sources of microbiological water quality impairment which are often overlooked.

Introduction

Fecal indicator bacteria (FIB), including E. coli, are the leading cause of impairments in rivers and streams within the United States. Elevated E. coli concentrations alone are responsible for impairments in over 100,000 miles of rivers and streams in the U.S. (USEPA, 2016). Conventional sources of water quality impairment include runoff from manure applied agricultural fields, direct fecal deposition, leaching from failed septic systems, and wastewater treatment plant outflows (Garzio-Hadzick et al., 2010). The presence of FIB in streams is used to indicate fresh fecal contamination (Garzio-Hadzick et al., 2010; Wade et al., 2003), and thus the risk of pathogen exposure (Haack, 2017). However, FIBs have the ability to survive and persist in the environment for a long time (Anderson et al., 2005) meaning that their presence does not always indicate fresh contamination. FIB can live in the environment for a few hours (Haack, 2017), days (Haack, 2017; Anderson et al., 2005), weeks (Haller et al., 2009; Jamieson et al., 2005), or even months (Haack, 2017; Garzio-Hadzick et al., 2010; Czajkowska et al., 2005) depending on the suitability of the environmental conditions.

Different environmental factors influence the survival and persistence of FIB in the environment including temperature, nutrient availability, salinity, oxygen levels, predation, and UV radiation (Craig et al., 2004; Hughes, 2003; Thomas et al., 1999; Davies et al., 1995). Sediment is one such environment that provides favorable conditions for FIB survival (Jamieson et al., 2005; McElhany and Pillai, 2010), with increased nutrient and carbon availability (Davies et al., 1995), reduced exposure to sunlight (Koirala et al., 2008), and protection from predatory protozoans (Jamieson et al., 2005; (Decamp and Warren, 2000). Survival times of FIB increases from days or weeks in the water column (Haack, 2017; Garzio-Hadzick et al., 2010; Czajkowska et al., 2005) to weeks or months in sediments (Garzio-Hadzick et al., 2010; Haller et al., 2009; Anderson et al., 2005; Czajkowska et al., 2005).

Higher concentrations of FIB are often observed in sediment when compared to the overlying water column (Kovacic et al., 2011; Edge and Boehm, 2011; Jamieson et al., 2005; An et al., 2002). For example, Van Donsel and Geldreich (1971) found fecal indicator bacteria concentrations were up to 1000 times greater in the sediment than in the water column. Similarly, Pandey et al. (2018) reported the average E. coli concentration in sediment was 47 to 386 times higher than the average E. coli concentration in the water column.

Sediment disturbance results in the resuspension of particles along with bacteria in the sediment. Disturbances can be caused by storm events (Pandey and Soupir, 2014; Fries et al., 2006; Jamieson et al., 2005; Crabill et al., 1999), recreational activities (Abia et al., 2017; An et al., 2002), animals crossing the river or stream (Abia et al., 2017; Sherer et al., 1988) and commercial activity (Pettibone Irvine et al., 1996; Irvine and Pettibone, 1993; Liou and Herbich, 1976). FIB concentrations in the water column can increase 8 to 88-times due to sediment disturbances (Abia et al., 2017; Pandey and Soupir, 2014), thus substantially contributing to water quality impairments (Abia et al., 2017; Nagels et al., 2002; Sherer et al., 1988).

To accurately predict FIB concentrations in the water column, the resuspension of FIB from streambed sediments must be considered (Rehmann and Soupir, 2009) as these reservoirs of FIB can substantially increase FIB concentrations in the water column. Artificial flood studies have been performed to isolate the impact of sediment disturbances on microbiological water quality (Muirhead et al., 2004; Nagels et al., 2002). For example, Muirhead et al. (2004) released water from a supply reservoir to generate an artificial flood and exclude runoff sources of fecal contamination. They measured E. coli concentrations in the water column up to two orders of magnitude higher than background concentrations, increasing from 10² to 10⁴ CFU 100 mL⁻¹, indicating that bed sediments alone can result in concentrations that exceed water quality standards when disturbed. The single sample maximum standard for E. coli in limited contact recreation waters is 1178 CFU 100 mL⁻¹ (USEPA, 2016). This release of FIB from the sediment to the water column may be incorrectly interpreted as recent fecal contamination rather than resuspension which may lead to an inaccurate perception of water quality.

To better understand sediments as a source of FIB, sediments must be quantified and monitored for concentrations of FIB; however, little information exists to help inform sampling design. Therefore, the goal of this study is to assess the variability of E. coli in streambed sediment and the implications for sampling and reporting results. The main objectives of the present work are to (i) assess the spatial variability of E. coli in sediment across the stream cross-section, (ii) determine the number of samples required for a representative E. coli concentration in sediment, and (iii) evaluate some of the uncertainty in sampling and processing samples. This information will be helpful to design sediment sampling regimes and to understand the impact of unconventional sources of water quality impairments, which are often overlooked.