A framework for prioritizing contaminants in retrospective ecological assessments: Application in the Milwaukee Estuary (Milwaukee, WI)

Abstract

Watersheds are subjected to diverse anthropogenic inputs, exposing aquatic biota to a wide range of chemicals. Detection of multiple, different chemicals can challenge natural resource managers who often have to determine where to allocate potentially limited resources. Here, we describe a weight‐of‐evidence framework for retrospectively prioritizing aquatic contaminants. To demonstrate framework utility, we used data from 96‐h caged fish studies to prioritize chemicals detected in the Milwaukee Estuary (WI, USA; 2017–2018). Across study years, 77/178 targeted chemicals were detected. Chemicals were assigned prioritization scores based on spatial and temporal detection frequency, environmental distribution, environmental fate, ecotoxicological potential, and effect prediction. Chemicals were sorted into priority bins based on the intersection of prioritization score and data availability. Data‐limited chemicals represented those that did not have sufficient data to adequately evaluate ecotoxicological potential or environmental fate. Seven compounds (fluoranthene, benzo[a]pyrene, pyrene, atrazine, metolachlor, phenanthrene, and DEET) were identified as high or medium priority and data sufficient and flagged as candidates for further effects‐based monitoring studies. Twenty‐one compounds were identified as high or medium priority and data limited and flagged as candidates for further ecotoxicological research. Fifteen chemicals were flagged as the lowest priority in the watershed. One of these chemicals (2‐methylnaphthalene) displayed no data limitations and was flagged as a definitively low‐priority chemical. The remaining chemicals displayed some data limitations and were considered lower‐priority compounds (contingent on further ecotoxicological and environmental fate assessments). The remaining 34 compounds were flagged as low or medium priority. Altogether, this prioritization provided a screening‐level (non‐definitive) assessment that could be used to focus further resource management and risk assessment activities in the Milwaukee Estuary. Furthermore, by providing detailed methodology and a practical example with real experimental data, we demonstrated that the proposed framework represents a transparent and adaptable approach for prioritizing contaminants in freshwater environments. Integr Environ Assess Manag 2023;00:1–21. © 2022 SETAC

Evaluation of Complex Mixture Toxicity in the Milwaukee Estuary (WI, USA) Using Whole‐Mixture and Component‐Based Evaluation Methods

“…Finally, by employing predictive analysis, we were able to bridge component-based and whole-mixture analyses, providing additional weight-of-evidence for chemical/mixture group prioritizations. Overall, our study represents an important complement to other prioritization analyses and complex mixture assessment in Great Lakes watersheds (e.g., Barber et al, 2022;Corsi et al, 2019;Maloney et al, 2022). Indeed, by identifying diverse target sites, individual chemicals, and simplified mixture groups, we have identified additional targets for ecotoxicological evaluation and increased the weight-of-evidence for high-and low-priority compounds and mixtures across these aquatic systems.…”

“…These chemicals included three industrial compounds (4‐tert‐octylphenol, anthraquinone, and bisphenol A), two WWIs (β‐sitosterol and cholesterol), five PPCPs (acetaminophen, desvenlafaxine, metformin, methotrexate, nicotine, and tramadol), two PAHs (benzo[ a ]pyrene and fluoranthene), one pesticide (metolachlor), and two fire retardants (tributyl phosphate and tris(2‐butoxyethyl)phosphate). Interestingly, many of these compounds were also highlighted as high‐ or medium‐priority chemicals in a companion study focused on prioritization of individual contaminants in the Milwaukee Estuary based on detection characteristics, environmental fate properties, ecotoxicity, and predictive relationships with measured effects (Maloney et al, 2022). For example, fluoranthene, benzo[ a ]pyrene, cholesterol, and β‐sitosterol were all flagged as high‐priority chemicals, whereas anthraquinone, acetaminophen, desvenlafaxine, metformin, metolachlor, nicotine, and tramadol were considered medium‐priority chemicals.…”

“…The reference site was at the Great Lakes Toxicology and Ecology Division aquatic laboratory (Duluth, MN, USA). The abbreviation MED (Mid‐Continent Ecology Division) is used to refer to the reference site throughout the manuscript to be consistent with site labeling during the time the study was conducted, and to bridge this work to the companion study and associated databases (Maloney et al, 2022).…”

“…Detailed descriptions of random forest algorithms and their application in ecological classification and regression are available elsewhere (Vitense et al, 2019). Detailed methodology for application of these predictive techniques applied with this data set can be found in our companion paper Maloney et al, 2022). Random forest regression was carried out with the "cforest" function in the "party" R package (Ver 1.3-9; Hothorn et al, 2006;Strobl et al, 2007Strobl et al, , 2008 with ntree (number of trees) = 5000, mtry (# predictors sampled at each node) = ceiling (# predictors/3), and other parameters set to those suggested for constructing unbiased random forests in Strobl et al (2007).…”

“…For example, many of these PAHs have been identified as chemicals of potential concern in the Milwaukee Estuary; Baldwin et al, 2013) and other Great Lakes watersheds (Corsi et al, 2019) based on detection frequency, environmental fate, and/or ecotoxicological potential. Moreover, the two sterols have also been identified as potential highpriority compounds in the Milwaukee Estuary, albeit with significant data limitations that could have influenced the assessment (Maloney et al, 2022). However, a critical evaluation of these findings in the context of published ecotoxicological data demonstrates that some additional factors should be considered Important drivers of mixture effect were identified using apical, nonapical, and pharmacological effect data and cumulative ratio (CR) analyses.…”