Applying MAR Analysis to Identify Human and Non-Human Fecal Sources in Small Kentucky Watersheds

Abstract

The recurrence of reports citing water quality impairments in watersheds is evidence that tools are needed to identify pollution sources and facilitate restoration efforts such as implementing total maximum daily limits (TMDLs) or best management practices (BMPs). Fecal bacteria in surface waters are one of the most commonly cited impairments to water quality. This study evaluated microbial source tracking (MST), specifically multiple antibiotic resistance (MAR) analysis, as a management tool to differentiate nonpoint source pollution into source groups. A library containing Escherichia coli (E. coli, EC) and fecal streptococci (FS) isolates from poultry (EC n = 282, FS n = 650), human (EC n = 152, FS n = 240), wildlife (EC n = 17, FS n = 43), horse (EC n = 79, FS n = 82), dairy cattle (EC n = 38, FS n = 42), and beef cattle (EC n = 49, FS n = 46) sources was created. The MAR analysis was conducted on the isolates using a profile of seven antibiotics. The antibiotic signatures of unknown source isolates from Elkhorn and Hickman Creek watersheds were evaluated against the library to determine the contributions of potential fecal inputs from the respective sources. Correct classification was >60% when analyzed at the human and non-human-level of classification. On a watershed basis, both watersheds produced similar results; inputs from non-human sources were the greatest contributors to nonpoint source pollution. The results from the multiple antibiotic resistance (MAR) analysis revealed that the information produced, coupled with knowledge of the watershed and its associated land uses, would be helpful in allocating resources to remediate impaired water quality in such watersheds.