Bioaccumulation of per- and polyfluoroalkyl substance in fish from an urban river: Occurrence, patterns and investigation of potential ecological drivers.

Per- and polyfluoroalkyl substances (PFAS) are ubiquitous in aquatic environments and a recent shift toward emerging PFAS is calling for new data on their occurrence and fate. In particular, understanding the determinants of their bioaccumulation is fundamental for risk assessment purposes. However, very few studies have addressed the combined influence of potential ecological drivers of PFAS bioaccumulation in fish such as age, sex or trophic ecology. Thus, this work aimed to fill these knowledge gaps by performing a field study in the Seine River basin (France). Composite sediment and fish (European chub, Squalius Cephalus) samples were collected from four sites along a longitudinal transect to investigate the occurrence of 36 PFAS. Sediment molecular patterns were dominated by fluorotelomer sulfonamidoalkyl betaines (i.e. 6:2 and 8:2 FTAB, 46% of ∑PFAS on average), highlighting the non-negligible contribution of PFAS of emerging concern. C-C perfluoroalkyl carboxylic acids, perfluorooctane sulfonic acid (PFOS), perfluorooctane sulfonamide (FOSA) and 10:2 fluorotelomer sulfonate (10:2 FTSA) were detected in all fish samples. Conversely, 8:2 FTAB was detected in a few fish from the furthest downstream station only, suggesting the low bioaccessibility or the biotransformation of FTABs. ∑PFAS in fish was in the range 0.22-3.8 ng g wet weight (ww) and 11-140 ng g ww for muscle and liver, respectively. Fish collected upstream of Paris were significantly less contaminated than those collected downstream, pointing to urban and industrial inputs. The influence of trophic ecology and biometry on the interindividual variability of PFAS burden in fish was examined through analyses of covariance (ANCOVAs), with sampling site considered as a categorical variable. While the latter was highly significant, diet was also influential; carbon sources and trophic level (i.e. estimated using C and N stable isotope ratios, respectively) equally explained the variability of PFAS levels in fish.