Date Available


Year of Publication


Degree Name

Master of Science in Civil Engineering (MSCE)

Document Type

Master's Thesis




Civil Engineering

First Advisor

Dr. Gail Brion

Second Advisor

Dr. Carmen Agouridis


The City of Georgetown, Kentucky relies on a vast karst spring network as a drinking water source. This karst feature has several inputs from sinkholes and streams in the Cane Run Watershed: a watershed associated with a variety of land uses in the recharge area. The recharge area encompasses the area from North Lexington to Georgetown and is composed of urban, suburban, agricultural and industrial usage. A serious water quality issue exists with respect to the impact of fecal contamination within the spring recharge area. Identification of fecal contamination is quantified by microbial indicators adapted from surface water applications: fecal load (E. coli), fecal source (two human-host specific Bacteroides DNA markers) and fecal age (AC/TC ratio). These three criteria are used in a categorical Microbial Source Tracking (MST) model to assign a Sanitary Category Value (SCV) between 0 and 3 for each sample location. Low SCVs (1.5) are associated with high values of fecal load, low fecal age and detectable concentration of human-specific markers. SCV measured during dry weather conditions are indicative of potentially leaking human sewers.

Due to retention and conservation of fecal load (E. coli) and age (AC/TC) microbial indicators in the karstic environment, ambiguous SCV model results cannot pinpoint, with statistical confidence, fecal sources in a karstic environment. Human-host specific genetic markers (HF183 and HuBac) were also detected at all sample sites above limits of detection, indicating steady inflow of fecal material during all sample events. By adding a flow multiplier and expressing HF183 and HuBac values as a load, it was strongly indicated that a human fecal source was entering the groundwater conduit and impacting Royal Spring independent from other upstream fecal sources. Interpretation of these trends, while strongly indicated, cannot be supported with statistical evidence.