Document Type

Dissertation - Open Access

Award Date


Degree Name

Doctor of Philosophy (PhD)

Department / School

Agricultural and Biosystems Engineering

First Advisor

Rachel McDaniel


attachment, cattle excess, E. coli, sediment, variability, water quality


Fecal indicator bacteria (FIB), including E. coli, are the number one cause of water quality impairments in the United States according to the USEPA. FIB are used as a predictor to identify the possible presence of pathogens in waterbodies. E. coli is a useful indicator of gastrointestinal (GI) related illnesses from contact with fresh water. While surface water is routinely monitored for water quality, streambed sediments are rarely considered as a source of FIB to the overlying water column. This study focuses on understanding the variation of E. coli concentrations in streambed sediments and the potential impact of sediment sources on microbial water quality impairments. A total of five sites were monitored, including four sites located on Skunk Creek and one site located on Six Mile Creek, both of which are tributaries to the Big Sioux River in eastern South Dakota. The Skunk Creek monitoring sites are abbreviated as Sk1, Sk2, Sk3 and Sk4 and cover approximately three miles along a stream reach. Sk1 is the site farthest upstream and has direct cattle access, while the other three sites are under seasonal riparian area management which is a cattle exclusion-based management practice. The monitoring site at Six Miles Creek is abbreviated as SM and a swine facility located near the site is a possible source of microbial contamination in the creek. A total of 25 sediment samples were collected from each of the five sites by creating a 5×5 grid. The detailed spatial grid provided insight into the variation of E. coli in sediment throughout the stream cross section. Samples were processed two times, within 8-hour and 24-hour of sample collection, to understand temporal stability of E. coli in sediment samples and the uncertainty related to sediment storage. Samples were also analyzed for sediment texture. The results showed high spatial variation of E. coli in bed sediment, ranging from 4 to 997 CFU g-1 (8.9×102 to 2.1×105 CFU 100 mL-1). The highest and lowest E. coli concentration and variability was observed at Sk1 and Sk4, respectively. Pockets of high E. coli concentrations were measured at all sites, and were typically located at the edge of the stream with the exception of Sk1 where the high pockets of E. coli were located in the middle of the stream, likely due to direct deposition of fecal matter by cattle. Due to the high variability of E. coli in stream sediment, large sample sizes of 4 to 65 samples were required to appropriately represent E. coli in sediment with a moderate margin of error (E= 60 CFU g-1) at a 95% confidence interval. Sediment holding times up to 24-hour after collection will not result in a statistically significant change in a majority of cases (80%). There was a strong correlation between the quantity of fine particles and E. coli concentration in sediment, though the direction was inconsistent. The four monitoring sites in Skunk Creek were also monitored over the course of two recreation seasons and surrounding a series of storm events. Three to five sediment samples were collected from May to October to analyze seasonal variation for a two-year period (2017-2018). Sk2 and Sk4 were selected for storm event monitoring which included nine samples collected in a 3×3 grid from both sites prior to, after, and five to seven days after the storm events. A significant reduction of E. coli concentrations in sediment were observed in 2018 compared to 2017 at the cattle crossing site. All sites showed higher E. coli in the late season (August to October) compared to the early season (May to July). Among the four storm events monitored, only one resulted in a noticeable hydrological response and a corresponding significant increase in E. coli concentrations in both water and sediment samples. The highest median E. coli concentration was observed at Sk3 (85 CFU g-1) followed by Sk1 (53 CFU g-1), while Sk2 and Sk4 showed significantly lower E. coli concentrations in the sediment. Higher attachment rates of E. coli were observed in the sediment as compared to the water column. The sediment attachment rate to settleable particles ( > 0.004 mm) ranged from 37% to 78% and was highest at Sk2 and Sk3. Phenotypic antibiotic resistance was measured at Sk1 and Sk2, and Sk2 demonstrated a significantly higher proportion of E. coli isolates resistant to ampicillin, tetracycline and sulfisoxazole, while erythromycin resulted in no significant change. In addition, the average organic content ranged from 4.2% to 8.2% and, overall, the organic content had little correlation with E. coli concentrations in sediments. However, there was a significant positive relationship between sediment organic matter and sediment E. coli concentrations taken from the middle of the stream at Sk1 and Sk2, possibly due to direct fecal deposit from cattle. The findings from this study will expand the knowledge regarding sediment sources of water pollution which can be useful in developing water quality monitoring projects and strategies. Additionally, the information can be incorporated into the development of microbial fate and transport models to better predict the contribution of streambed sediment on poor water quality.

Library of Congress Subject Headings

River sediments -- Environmental aspects -- South Dakota.
Escherichia coli.
Water -- Pollution.
Water quality.



Number of Pages



South Dakota State University



Rights Statement

In Copyright