Antimicrobial resistance in the river Trent under flood and non-flood conditions

A comprehensive sampling campaign with high spatial resolution has revealed the nature of antimicrobial resistance (AMR) in river water under varied hydrological conditions. A 2.6 km stretch of the river Trent in England was sampled at 18 locations. The river was sampled during a lowland flood (in late winter) as waters were rising and receding, and during non-flood conditions (in the following autumn). The sampling encompassed reaches upstream and downstream of a large sewage treatment works (STW) and a sewer overflow (SO) to assess the impact on AMR of treated and untreated sewage discharges. We assessed AMR through enumeration of bacterial and fungal phenotypes, antimicrobial resistance gene (ARG) markers by high-throughput quantitative polymerase chain reaction (qPCR) as well as measuring a suite of physicochemical water quality parameters. The phenotypic AMR methods broadly underperformed with floodwater samples, as high particulate matter appeared to reduce antimicrobial efficacy resulting in false-positive colony growth. Nonetheless, most AMR and water quality parameters showed the influence of the treated sewage discharge which was measurable approximately 2 km downstream. When floodwaters were rising after heavy rainfall, the treated sewage discharge actually diluted AMR levels and improved the microbial water quality of the river (which was impacted by upstream SOs and runoff). The samples taken at sites beside actively discharging SOs recorded (i) the highest levels of total and putative cefotaxime-resistant E. coli, (ii) highest prevalence of colonies resistant to fluconazole antifungal and (iii) significantly increased absolute abundance of ARGs conferring resistance to diaminopyrimidine, disinfectants, sulfonamide and tetracycline drug classes and crAss64 bacteriophage.