Per- and Polyfluoroalkyl Substances (PFAS) in Surface Water Near US Air Force Bases: Prioritizing Individual Chemicals and Mixtures for Toxicity Testing and Risk Assessment.

Per- and polyfluoroalkyl substances (PFAS) are a large class of persistent chemicals used for decades in industrial and commercial applications. A key challenge with regard to estimating potential risk to ecological (and human) receptors associated with PFAS exposure lies in the fact that there are many different PFAS compounds and several to many can co-occur in any given environmental sample. We applied a data science approach to characterize and prioritize PFAS and PFAS mixtures from a large dataset of PFAS measurements in surface waters associated with US Air Force Installations with a history of the use of aqueous film-forming foams (AFFFs). Several iterations of stakeholder feedback culminated in a few main points that advanced our understanding of a complex dataset and the larger ecotoxicological problem. First, perfluorooctane sulfonate (PFOS) was often a dominant PFAS in a given surface water sample, frequently followed by perfluorohexane sulfonate (PFHxS). Second, a 4-chemical mixture generally accounted for >80% of the sum of all routinely reported PFAS in a sample, and the most representative 4-chemical mixture was composed of PFOS, PFHxS, perfluorohexanoic acid (PFHxA), and perfluorooctanoic acid (PFOA). We suggest that these results demonstrate the utility of formalized data science analysis and assessment frameworks to address complex ecotoxicological problems. Specifically, our example dataset results can be used to provide perspective on toxicity testing, ecological risk assessments, and field studies of PFAS in and around AFFF-impacted sites. Environ Toxicol Chem 2021;40:871-882.