Capturing temporal variation in aquatic ecotoxicological risks: Chemical- versus effect-based assessment

The integration of effect-based and chemical profiling has been advocated to assess the potential ecotoxicological risks posed by chemical mixtures present in aquatic ecosystems. However, the concentrations of contaminants in surface waters can vary greatly over time and space, making it challenging to ensure risk assessment. Although the first results are promising, it has not yet been proven that these combined approaches are also capable of capturing temporal variation in aquatic ecotoxicological risks. The present study aimed to test this by combining passive time-integrative sampling with effect-based and chemical-analytical techniques in agricultural waterways. Silicone rubber sheets and polar organic chemical integrative samplers (POCIS) were deployed in four agricultural water bodies over four consecutive six-week periods. Passive sampler extracts were analysed using a battery of 22 in vitro and in vivo bioassays in tandem with extensive chemical target analysis of 225 compounds. The extracts induced fluctuating bioassay responses over time for all locations during all sampling periods, highlighting the presence of temporal and spatial variation in toxic pressure. A range of compounds, primarily fungicides and herbicides, were detected in the passive sampler extracts during all sampling periods and at all locations at variable concentrations, highlighting the persistent but variable chemical pressure in surface waters in agricultural regions. However, the toxicity observed in the in vitro bioassays could solely be attributed to detected chemicals in 6 % of the cases with those chemicals explaining only 1-16.9 % of the observed effects, indicating that these were predominantly caused by undetected chemicals. Risk assessments based on bioassay responses revealed frequent exceedances of effect-based trigger values at all locations and during all sampling periods. It is concluded that effect-based assessments better capture temporal variations in potential ecotoxicological risks than traditional chemical analyses, but that advanced chemical analysis is needed to explain the bioanalytical response profiles.