Potential for biological effects of per- and polyfluoroalkyl substances in Great Lakes tributaries and associations with land cover and wastewater effluent

Surface water concentrations of per- and polyfluoroalkyl substances (PFAS) and potential for resulting biological effects were estimated in a study using polar organic chemical integrative samplers (POCIS) from 60 tributary sites within 20 watersheds in the Great Lakes Basin in 2018. Sites represented a range of urban to agricultural, forested, and wetland land uses and included a gradient of wastewater treatment effluent from zero to 44% of annual streamflow. Several sites also had airport influence. Twenty-one of 32 targeted PFAS compounds were detected in POCIS samplers, of which, 16 had available POCIS sampling rates, enabling time-weighted water concentration estimates and comparison with available effects data. Estimated water concentrations were compared with published water quality guidelines (available for nine PFAS), effect concentrations reported in primary literature within the ECOTOX Knowledgebase for apical endpoints (10 PFAS) and nonapical endpoints (10 PFAS), and in vitro high-throughput screening data from the U.S. Environmental Protection Agency Toxicity Forecaster (ToxCast; 14 PFAS). Based on a conservative evaluation approach that was also weighted for persistence and limitations in available toxicological information, five individual PFAS, including perfluorooctanesulfonic acid, perfluorohexanesulfonic acid, perfluorobutanesulfonic acid, perfluorooctanoic acid, and perfluorononanoic acid were identified as warranting additional investigation. Possible increased potency of PFAS mixtures over individual chemical effects, estimated by summation of exposure-activity ratios (EARs) for chemicals that influence common ToxCast assays and specified gene targets, indicated that EAR values increased up to 5.6-fold over individual chemicals, with up to 14 chemicals contributing to mixture effect predictions. Potential for biological effects from PFAS, as estimated by summed exposure-activity ratios, were correlated with urban land use and the proportion of streamflow contributed by wastewater effluent.