Due to the complex sources and fate of perfluoroalkyl substance (PFAS), their source apportionment in the environment remains a challenge. A data set of 11 straight-chain PFAS in 139 samples of fish in the Great Lakes was analyzed using positive matrix factorization (PMF) to investigate their primary sources, whose spatial variations were examined against the surrounding environmental factors. PMF analysis produced five fingerprints. Factor 1 (72% of Σ₁₁PFAS, dominated by PFOS) probably represented emissions from primary sources (such as consumer products) and secondary sources (precursors), and increased in average abundance from west to east across the Great Lakes. Factor 2 (13% of Σ₁₁PFAS) and factor 3 (7% of Σ₁₁PFAS), highly loaded with long-chain PFAS and PFNA, respectively, were thought to represent PVDF manufacture or processing in metal plating. They showed higher contributions in sparsely populated Lakes Superior and Huron. Factor 4 (5% of Σ₁₁PFAS, highly loaded with PFOS and PFHxS) presented hot spots near current and former air force bases, suggesting it was related to aqueous film-forming foams (AFFFs). Factor 5 (4% of Σ₁₁PFAS) contained primarily PFOS and PFOSA, which may imply metabolism of precursors (PFOSA) to PFOS in vivo. Unexpectedly, the spatial trends of the five sources all showed abnormally low values near the more urbanized Chicago and Milwaukee in Lake Michigan, which may be due to their unique wastewater and stormwater infrastructure or may arise from atmospheric transport of precursors. Our study indicated that PMF was an effective tool to identify sources of PFAS in fish despite absorption, distribution, metabolism, and excretion (ADME) processes which might alter fingerprints in fish relative to their surrounding environment.

Passive samplers are useful tools for monitoring hydrophobic, persistent, and potentially bioaccumulative contaminants in the environment. In this study, low density polyethylene passive samplers were deployed in urban-influenced and background nearshore freshwaters of northwestern Lake Ontario and analyzed for a broad range of both legacy halogenated organic contaminants (HOCs) and halogenated flame retardants (HFRs). Non-targeted analysis was conducted for screening additional halogenated substances. For most compounds, concentrations were greatest in the industrialized Hamilton Harbour and more generally at sites that have stronger influences of wastewater effluent discharges and stormwater run-off through rivers and creeks. Polychlorinated biphenyls (PCBs) remain the dominant class of HOCs in water, with dissolved-phase concentrations ranging from 10 to 4100 pg/L (ΣPCBs), followed by polybrominated diphenylethers (ΣPBDEs; 14–960 pg/L) and the organochlorine pesticides (OCPs; 22–290 pg/L). Several non-PBDE brominated flame retardants (nBFRs) and chlorinated Dechlorane-related compounds were detected, with hexabromocyclododecanes (ΣHBCDD; sum of 3 diastereoisomers) the most abundant (1.0–21 pg/L). Non-targeted screening of samples by high resolution mass spectrometry using Kendrick mass defect plots for data analysis indicated that several other halogenated compounds were present in waters at relatively high abundances compared to the flame retardants, based on semi-quantitative estimates. These included methyl-triclosan, four halogenated anisoles (2,4,6-tribromoanisole, dimethyl-trichloroanisole, pentachloroanisole, and pentachlorothioanisole), and pentachloro-aniline. Dissolved-phase methyl-triclosan was estimated to contribute up to approximately 40% of the summed target HOC concentrations. Polyethylene passive samplers provided an excellent medium for both non-targeted screening of HOCs not currently included in monitoring programs and tracking brominated and chlorinated chemicals slated for reductions in uses and emissions through international (Stockholm Convention) and binational (Great Lakes) agreements.

The occurrence and spatial distribution of selected pesticides were investigated along a 200-km reach of the St. Lawrence River (SLR) and tributaries in Quebec, Canada. Surface water samples (n = 68) were collected in the summer 2017 and analyzed for glyphosate, atrazine (ATZ), 8 systemic insecticides (acetamiprid, clothianidin, dinotefuran, fipronil, imidacloprid, nitenpyram, thiacloprid, and thiamethoxam) and some metabolites. Overall, 99% of the surface water samples were positive to at least one of the targeted pesticides. The most recurrent compounds were glyphosate (detection frequency: 84%), ATZ (82%), thiamethoxam (59%), desethylatrazine (DEA: 47%), and clothianidin (46%). Glyphosate displayed variable levels (4–3,000 ng L−1), with higher concentrations in south tributaries (e.g., Nicolet and Yamaska). In positive samples, the sum of ATZ and DEA varied between 5 and 860 ng L−1, and the sum of 6 priority neonicotinoids between 1.5 and 115 ng L−1. From Repentigny to the Sorel Islands, the spatial distribution of pesticides within the St. Lawrence River was governed by the different upstream sources (i.e., Great Lakes vs. Ottawa River) due to the limited mixing of the different water masses. Cross-sectional patterns revealed higher concentrations of glyphosate and neonicotinoids in the north portions of transects, while the middle and south portions showed higher levels of atrazine. In Lake St. Pierre and further downstream, cross-sections revealed higher levels of the targeted pesticides near the southern portions of the SLR. This may be due to the higher contributions from south shore tributaries impacted by major agricultural areas, compared to north shore tributaries with forest land and less cropland use. Surface water samples were compliant with guidelines for the protection of aquatic life (chronic effects) for glyphosate and atrazine. However, 31% of the samples were found to surpass the guideline value of 8.3 ng L−1 for the sum of six priority neonicotinoids.

The Laurentian Great Lakes are a valuable natural resource that is affected by contaminants of emerging concern (CECs), including sex steroid hormones, personal care products, pharmaceuticals, industrial chemicals, and new generation pesticides. However, little is known about the fate and biological effects of CECs in tributaries to the Great Lakes. In the current study, 16 sites on three rivers in the Great Lakes basin (Fox, Cuyahoga, and Raquette Rivers) were assessed for CEC presence using polar organic chemical integrative samplers (POCIS) and grab water samplers. Biological activity was assessed through a combination of in vitro bioassays (focused on estrogenic activity) and in vivo assays with larval fathead minnows. In addition, resident sunfish, largemouth bass, and white suckers were assessed for changes in biological endpoints associated with CEC exposure. CECs were present in all water samples and POCIS extracts. A total of 111 and 97 chemicals were detected in at least one water sample and POCIS extract, respectively. Known estrogenic chemicals were detected in water samples at all 16 sites and in POCIS extracts at 13 sites. Most sites elicited estrogenic activity in bioassays. Ranking sites and rivers based on water chemistry, POCIS chemistry, or total in vitro estrogenicity produced comparable patterns with the Cuyahoga River ranking as most and the Raquette River as least affected by CECs. Changes in biological responses grouped according to physiological processes, and differed between species but not sex. The Fox and Cuyahoga Rivers often had significantly different patterns in biological response Our study supports the need for multiple lines of evidence and provides a framework to assess CEC presence and effects in fish in the Laurentian Great Lakes basin.

We examine the concentrations and food web biomagnification of three cyclic volatile methylsiloxanes (cVMS) octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6) using aquatic biota collected from Lake Erie. Concentrations of cVMS in biota were within the range reported for other studies of cVMS in aquatic biota. Trophic magnification factors (TMF) were assessed in various food web configurations to investigate the effects of food web structure. TMF estimates were highly dependent on the inclusion/exclusion of the organisms occupying the highest and lowest trophic levels and were >1 for D4 and D5, indicating biomagnification, in only 1 of the 5 food web configurations investigated and were <1 in the remaining 4 food web configurations. TMF estimates for PCB180 were also dependant on food web configuration, but did not correspond with those obtained for cVMS materials. These differences may be attributed to environmental exposure and/or lipid partitioning differences between PCB180 and cVMS.

This study assessed concentrations and investigated potential source regions for PCBs, PBDEs, and organochlorine pesticides in Acadia National Park, Maine, USA. Back-trajectories and potential source contribution function (PSCF) values were used to map potential source areas for total-PCBs, BDE-47, and 10 organochlorine pesticides. The constructed PSCF maps showed that ANP receives high pollutant concentrations in air masses that travel along four main pathways: (1) from the SW along the eastern Atlantic seaboard, (2) from the WSW over St. Louis, and Columbus regions, (3) from the west over Chicago, and Toronto regions, and (4) from WNW to NNW over the Great Lakes, and Quebec regions. Transport of all studied pollutants were equally distributed between the first three pathways, with only minor contributions from the last pathway. This study concludes that the high-pollutant concentrations arriving at ANP do not exclusively originate from the major urban centers along the eastern Atlantic seaboard.

Decades of large-scale production of per- and polyfluoroalkyl substances (PFASs) have resulted in their ubiquitous presence in the environment worldwide. Similarly to other persistent and bioaccumulative organic contaminants, some PFASs, particularly the long-chain congeners, can be biomagnified via food webs, making top predators vulnerable to elevated PFAS exposure. In this study, we measured seven classes of PFASs in bald eagle (Haliaeetus leucocephalus) eggs for the first time. The eggs (n = 22) were collected from the North American Great Lakes in 2000–2012. The ranges of total concentrations of perfluoroalkyl sulfonic acids (∑PFSAs) and perfluoroalkyl carboxylic acids (∑PFCAs) were 30.5–1650 and 5.4–216 ng/g wet weight (ww), respectively. In addition to these traditional PFAS compounds, 6:2 fluorotelomer sulfonic acid (6:2 FTS; median: 15.7 ng/g ww), perfluoro-4-ethylcyclohexanesulfonic acid (PFECHS; 0.22 ng/g ww), and 8-chloro-perfluorooctanesulfonic acid (Cl-PFOS, detected in wildlife for the first time; 0.53 ng/g ww) were also frequently detected. Bald eagle eggs from breeding areas located less than 8 km from a Great Lake shoreline or tributary had significantly greater total PFAS concentrations (∑PFASs) than those from breeding areas located further than 8 km (p < 0.05). In these samples, ∑PFASs rivalled the total concentration of brominated flame retardants, and were significantly greater than those of several other organic contaminants, such as dechlorane-related compounds, organophosphate esters, and flame retardant metabolites.

Polychlorinated biphenyls (PCBs) in dated sediment cores from the East China Marginal Seas (ECMSs) and the chronology of the net fluxes to sediments were analyzed. The accumulation of 27 PCBs (ΣPCBs) in the ECMS sediments is about 5–26 ng cm⁻², with the net depositional fluxes of ΣPCBs 10 times lower than those observed in the Great Lakes during the 1960s–1970s. Exponential increases in PCB deposition to the ECMS sediments since the 1990s were observed, which closely follows the fast growth of PCB emissions from industrial thermal processes and e-waste related sources in China. Recent PCB fluxes to the study sites in the ECMSs and the Great Lakes are comparable; the former surged forward with a rising tendency, while the latter showed continued decline after the late 1970s. Due to the different PCB application histories and sources between the two regions, the ECMS sediments may remain as a net sink for land-derived PCBs, while sediments in the Great Lake may have been acting as a secondary source releasing PCBs to water. A higher proportion of trichlorobiphenyls in the ECMS sediments than the Great Lakes was indicated, which may imply the net transport of atmospheric PCBs from China.

In this study, we measured the concentrations of 58 flame retardants (and related compounds) in bald eagle (Haliaeetus leucocephalus) egg and plasma samples from the Michigan. These analytes include polybrominated diphenyl ethers (PBDEs), novel flame retardants (nFRs), Dechlorane-related compounds (Decs), and organophosphate esters (OPEs). A total of 24 paired eaglet plasma and egg samples were collected from inland (IN, N = 13) and the Great Lakes (GL, N = 11) breeding areas from 2000 to 2012. PBDEs were the most abundant chemical group with a geometric mean of 181 ng/g wet weight (ww) in egg and 5.31 ng/g ww in plasma. Decs were barely found in plasma samples, but they were frequently found in eggs (geometric mean 23.5 ng/g ww). OPE levels were comparable to those of PBDEs in the plasma but lower than those of PBDEs in eggs. Dec and PBDE concentrations were significantly higher in GL than in IN (p < 0.05). The ratio of egg to plasma concentrations (lipid normalized) varied with chemicals and correlated with the chemical's octanol-water partition coefficient. The lipid normalized bald eagle egg and plasma concentrations from Lake Superior and Huron were one to three orders of magnitude higher than concentrations measured in composite lake trout (Salvelinus namaycush) from the same lake, implying that they biomagnify in the environment.

To better characterize the transport of neonicotinoid insecticides to the world's largest freshwater ecosystem, monthly samples (October 2015–September 2016) were collected from 10 major tributaries to the Great Lakes, USA. For the monthly tributary samples, neonicotinoids were detected in every month sampled and five of the six target neonicotinoids were detected. At least one neonicotinoid was detected in 74% of the monthly samples with up to three neonicotinoids detected in an individual sample (10% of all samples). The most frequently detected neonicotinoid was imidacloprid (53%), followed by clothianidin (44%), thiamethoxam (22%), acetamiprid (2%), and dinotefuran (1%). Thiacloprid was not detected in any samples. The maximum concentration for an individual neonicotinoid was 230 ng L⁻¹ and the maximum total neonicotinoids in an individual sample was 400 ng L⁻¹. The median detected individual neonicotinoid concentrations ranged from non-detect to 10 ng L⁻¹. The detections of clothianidin and thiamethoxam significantly increased as the percent of cultivated crops in the basins increased (ρ = 0.73, P = .01; ρ = 0.66, P = .04, respectively). In contrast, imidacloprid detections significantly increased as the percent of the urbanization in the basins increased (ρ = 0.66, P = .03). Neonicotinoid concentrations generally increased in spring through summer coinciding with the planting of neonicotinoid-treated seeds and broadcast applications of neonicotinoids. More spatially intensive samples were collected in an agriculturally dominated basin (8 sites along the Maumee River, Ohio) twice during the spring, 2016 planting season to provide further information on neonicotinoid inputs to the Great Lakes. Three neonicotinoids were ubiquitously detected (clothianidin, imidacloprid, thiamethoxam) in all water samples collected within this basin. Maximum individual neonicotinoid concentrations was 330 ng L⁻¹ and maximum total neonicotinoid concentration was 670 ng L⁻¹; median detected individual neonicotinoid concentrations were 7.0 to 39 ng L⁻¹.