ABSTRACT
Pharmaceutical substances are ubiquitous in the aquatic environment and their concentration levels typically range from ng/L up to several µg/L. Furthermore, as those compounds are designed to be highly biologically active, assessing their impacts on non-target organisms is important. Here, we conducted a mesocosm experiment testing a mixture of five pharmaceuticals (diclofenac, carbamazepine, irbesartan, acetaminophen and naproxen) on fish, three-spined stickleback (Gasterosteus aculeatus). The mixture concentration levels were chosen on the basis of the contamination of the Meuse river in Belgium which had been measured previously during a monitoring campaign undertaken in 2015 and 2016. Three nominal mixture concentration levels were tested: the lowest concentration level mixture was composed by environmentally-relevant concentrations that approximate average realistic values for each pharmaceuticals (Mx1); the two other levels were 10 and 100 times these concentrations. Although no impact on stickleback prey was observed, the mixture significantly impaired the survival of female fish introduced in the mesocosms at the highest treatment level without causing other major differences on fish population structure. Impacts on condition factors of adults and juveniles were also observed at both individual and population levels. Using a modelling approach with an individual-based model coupled to a bioenergetic model (DEB-IBM), we concluded that chronic exposure to environmentally-relevant concentrations of five pharmaceuticals often detected in the rivers did not appear to strongly affect the three-spined stickleback populations. Mechanisms of population regulation may have counteracted the mixture impacts in the mesocosms.
