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

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 (8.5 × 10 to 2.1 × 10 CFU 100 mL). 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.