Date of Award

May 2016

Degree Type


Degree Name

Doctor of Philosophy



First Advisor

Shangping Xu

Committee Members

Timothy Grundl, Yin Wang, William Kean, Douglas Cherkauer


Bacteriodes, Colloid, DLVO, Enterococcus, Groundwater, Microbial Source Tracking





Jennifer J. Johanson

The University of Wisconsin-Milwaukee 2016

Under the Supervision of Professor Shangping Xu

Groundwater, a primary source of drinking water for nearly half the people in the United States, can be contaminated by pathogenic bacteria from fecal materials causing outbreaks of waterborne illness. Therefore, early identification of the presence of fecal contamination in groundwater can help prevent such outbreaks, and determining whether bacteria originate from human or animal feces can narrow down the location of potential pollution sources, allowing timely remediation and reduced potential for future outbreaks.

Pathogens are found in relatively low concentration in feces leading to difficulties in their detection in groundwater samples. In addition, a wide variety of pathogenic bacteria and viruses may exist in feces making it costly to analyze groundwater directly for all potential pathogens. As a result, groundwater samples are routinely analyzed for non-pathogenic fecal indicator bacteria (FIB), which are used as a proxy for the potential contamination by fecal pathogens. An ideal FIB would be abundant in the source material, easy and inexpensive to analyze, mobile in the subsurface so that it does not lag behind the pathogens, and host-specific to help identify the contaminant source.

Bacteria which can be identified as originating selectively from human vs nonhuman sources (animals) are especially helpful in determining the source of contamination when multiple potential sources are present. Escherichia coli (E. coli) has long been used as a FIB due to its abundance in fecal matter. However E. coli is found in many different hosts, which limits its use for source identification. Recent research has focused on identifying microbial source tracking (MST) bacteria which have markers that are specific to human or animal hosts, and these host-specific markers can be critical in early source identification efforts. This potential for MST is especially promising if combined with the other characteristics of an ideal FIB, such as abundance and mobility in the subsurface.

This research focuses on evaluating the subsurface mobility of two bacteria, Enterococcus faecium (E. faecium) and Bacteriodes fragilis (B. fragilis), in order to better understand their potential use as source-tracking FIB. These bacteria are both abundant in fecal matter and they have shown promise as having human-specific markers. We performed column experiments to compare their subsurface transport through sandy material. Bacteria with relatively high attachment to sand have lower mobility in groundwater and may therefore be less effective as early tracers of fecal contamination

The first part of our research compares two strains of E. faecium; one with and one without Enterococcal surface protein (Esp), a marker which recent research has linked to human sources, to evaluate whether the presence of Esp affects bacterial attachment to sand. The results indicate that in water with neutral pH (~7.2) the presence of Esp is linked to increased attachment to sand, thereby reducing the mobility of the Esp positive E. faecium. Because indicator bacteria should have relatively high mobility, this increased attachment potentially decreases the usefulness of Esp for MST. The results are consistent with calculations using the extended Derjaguin-Landau-Verwey-Overbeek (XDLVO) theory of colloidal attachment, which predicts that attachment in bacteria with Esp should be greater than in those without Esp due to the presence of a higher energy barrier for the bacteria without Esp.

The second part of this research compares the transport of the common aerobic fecal indicator bacteria E. coli, which has had limited success in source tracking, to the much more abundant anaerobic B. fragilis, which has shown promise as a potential MST bacteria. The results indicate that in water with neutral pH and low total ionic strength conditions, both E. coli and B. fragilis have similar attachment to sand, but at high ionic strength, such as may be found in areas near the source of contamination, the B. fragilis has lower attachment (and thus potentially higher mobility) than E. coli. The XDLVO calculations indicate a secondary energy minimum exists at higher ionic strength for both bacteria. This secondary minimum, which is absent at low ionic strength, occurs at a distance of 1 to 20 nm from the sand surface and appears to be the result of compression of the electrostatic double layer. The depth of this energy minimum is greater for E. coli than for B. fragilis, leading to greater attachment in the E. coli than the B. fragilis.