The occurrence of antibiotics in an urban watershed: from wastewater to drinking water.

The presence of 28 antibiotics in three hospital effluents, five wastewater treatment plants (WWTPs), six rivers and a drinking water storage catchment were investigated within watersheds of South-East Queensland, Australia. All antibiotics were detected at least once, with the exception of the polypeptide bacitracin which was not detected at all. Antibiotics were found in hospital effluent ranging from 0.01-14.5 microg L(-1), dominated by the beta-lactam, quinolone and sulphonamide groups. Antibiotics were found in WWTP influent up to 64 microg L(-1), dominated by the beta-lactam, quinolone and sulphonamide groups. Investigated WWTPs were highly effective in removing antibiotics from the water phase, with an average removal rate of greater than 80% for all targeted antibiotics. However, antibiotics were still detected in WWTP effluents in the low ng L(-1) range up to a maximum of 3.4 microg L(-1), with the macrolide, quinolone and sulphonamide antibiotics most prevalent. Similarly, antibiotics were detected quite frequently in the low ng L(-1) range, up to 2 microg L(-1) in the surface waters of six investigated rivers including freshwater, estuarine and marine samples. The total investigated antibiotic concentration (TIAC) within the Nerang River was significantly lower (p<0.05) than all other rivers sampled. The absence of WWTP discharge to this river is a likely explanation for the significantly lower TIAC and suggests that WWTP discharges are a dominant source of antibiotics to investigated surface waters. A significant difference (p<0.001) was identified between TIACs at surface water sites with WWTP discharge compared to sites with no WWTP discharge, providing further evidence that WWTPs are an important source of antibiotics to streams. Despite the presence of antibiotics in surface waters used for drinking water extraction, no targeted antibiotics were detected in any drinking water samples.