An algorithm was first developed to predict expected concentrations (PECs) of antibiotics in wastewater effluent taking into account prescription rates of antibiotics across EU member states, the amount of compound excreted unmetabalised by the human body, and removal rates from wastewater treatment plants. From this, PECs of over 70 antibiotic compounds were estimated. When combined with predicted no-effect concentrations (i.e. proposed limits of concentration for the selection of antimicrobial resistance or AMR) a risk analysis was performed to prioritise the compounds most likely contributing to risk of AMR proliferation in the environment. This data was exploited in two ways. First. using a database of flow rates in receiving waters across the EU to calculate dilution rates, hotspots for AMR were developed at a continental scale. This was further refined with more detailed data from the UK’s National Health Service prescription database, where the percentage of water bodies receiving wastewater effluent at risk for AMR was identified. Finally, at the local scale, daily risk for AMR in the River Foss over the course of a year was established based on prescription practices in York area. Secondly, the prioritised compounds identified by the algorithm were used for the development of an environmentally relevant mixture of 11 antibiotic PECs in UK wastewater effluent that may be used for irrigation of crops.
The PEC mixture was then used in a 4 month mesocosm study using Spring Barley (Hordeum vulgare) as model plant species. Over the course of the experiment, mesocosms were watered twice weekly, to supplement rainfall, with synthetic wastewater effluent containing antibiotics at PEC, 0.1xPEC 10xPEC, 100xPEC, or 0. Soil samples were collected at the start of the study, prior to irrigation with antibiotics, 16 hours after the first antibiotic exposure, 8 weeks of routine antibiotic exposure, and at plant maturity at 16 weeks. The fate of antibiotics was measured in soil-pore water and mesocosm leachate. Plant development endpoints were measured in both below and above ground biomass and included biomass, stem height, tiller and leaf counts, and grain production. Soil samples were collected to monitor microbial community structure and development of AMR genes using high-throughput qPCR. Throughout the experiment, the Skyline 2-D fly-by-wire device, developed at the University of York, was employed to continuously monitor greenhouse gas fluxes including CO2, N20, and methane. Stable isotope analysis was performed on plant leaves pre and post fertilization to measure shifts in the natural abundance of N15 and C13, as indicators for plant stress. A metabolomics approach was conducted to measure differences in the plant extractable metabolome of grain produced under the different exposure regimes.
