In this thesis, I describe a hypothetical journey of a tourist in New Zealand who has arrived on an airplane and is thirsty to explore the country’s spectacular rural and remote environments. My tourist sometimes will have to find her own drinking water. Surface waters in rural areas may be contaminated by waste from livestock, septic tanks, and wastewater pipes. These contaminants can include antibiotics and antibiotic resistant bacteria. When these surface waters are used for drinking and food preparation, they may expose users to antibiotic resistant pathogens, reducing treatment options and delivering poorer treatment outcomes. I was interested in whether these surface waters that are used for drinking water in various rural South Island locations harboured multidrug resistant (MDR) bacteria, and whether associations between resistance and location could be observed.

The microbiological quality of water on long haul international and domestic flights was compared using E. coli as an indicator organism. Next, the number and diversity of MDR E. coli was measured and compared in rural areas across three water sources; treated water (Treated), untreated stream water (Stream), other untreated surface waters such as rivers and tributaries (Other). Comparisons were made between the Marlborough, Banks Peninsula and Wider Canterbury areas. Through a case study in Okains Bay I observed whether there were seasonal effects on MDR E. coli in the main drinking water sourced from the Opara Stream.

E. coli was below the detection limit in the airplane water analysis, however, water from long haul international flights had significantly poorer microbiological quality than domestic airplanes. Across the three regions, 15% of tested isolates were MDR. Among the MDR isolates, 1% were isolated from Treated, 46% were isolated from Other, and 53% were isolated from Stream. No antibiotic enrichment was needed to identify MDR E. coli from water samples. The E. coli population density was not predictive of resistant E. coli occurrence in the environment. In the Okains Bay case study, MDR E. coli were detected at similar frequencies between the sites along the Opara Stream across the four seasons. The residential home had significantly higher MDR frequencies than all the Opara Stream sites.

36% of isolates that were resistant to ampicillin or ciprofloxacin were phenotypically confirmed to be ESBL or AmpC producers. 64% of tested isolates were able to share at least one resistance through horizontal gene transfer. Genetic analysis of 20 isolates indicated that a combination of point mutations, efflux systems and AmpC lactamases are the putative contributors to resistance phenotypes observed in the sequenced isolates.

Previous studies have determined that E. coli concentrations in rural surface waters are often above what is considered safe for drinking. Nevertheless, few studies have investigated the presence of MDR bacteria. The results from my analysis suggest that MDR E. coli and ESBL- and AmpC-producing E. coli are present in surface waters in Marlborough, Banks Peninsula, and the Wider Canterbury area. These MDR E. coli are able to share resistance genes with susceptible bacteria. An important finding with relevance to community safety is that the exposure to, and risk of, catching an infection from a MDR E. coli cannot be estimated using surveys of total E. coli concentrations.