Spatial and temporal variability of perfluoroalkyl substances in the Laurentian Great Lakes

Christina K. Remucal *ab
aEnvironmental Chemistry and Technology Program, University of Wisconsin–Madison, Madison, Wisconsin, USA. E-mail: remucal@wisc.edu; Fax: +1-608-262-0454; Tel: +1-608-262-1820
bDepartment of Civil and Environmental Engineering, University of Wisconsin–Madison, 660 N. Park St., Madison, WI 53706, Wisconsin, USA

Received 31st May 2019 , Accepted 13th July 2019

First published on 22nd July 2019


Per- and polyfluoroalkyl substances (PFAS) are a diverse group of fluorinated organic chemicals that have been used in industrial and consumer applications since the 1950s. PFAS are resistant to chemical and biological degradation and are ubiquitous in the environment, including in water, sediment, and biota in the Laurentian Great Lakes. This critical review evaluates the spatial and temporal variability of commonly studied perfluoroalkyl sulfonates (PFSAs) and perfluoroalkyl carboxylates (PFCAs) in the Great Lakes by synthesizing data collected in water, surface sediment, sediment cores, lake trout (Salvelinus namaycush), and herring gull (Larus argentatus) eggs. The lowest PFAS concentrations in all matrices are detected in Lake Superior, which is located in the most pristine region of the Great Lakes Basin. In contrast, higher concentrations are observed in Lakes Erie and Ontario, which are more impacted by industrial activity and wastewater discharge. The distribution of individual PFAS compounds also varies across the lakes in response to changes in PFAS sources, with higher proportions of PFSAs in the eastern lakes. Sediment and biota are enriched in long chain PFSAs and PFCAs relative to concentrations in the water column, as expected based on predicted partitioning behavior. Sediment cores and bioarchives consistently demonstrate that PFAS concentrations increased in the Great Lakes from the initial time points until the early 2000s. The available data indicate that PFOS and PFOA concentrations decline after this period in the upper Great Lakes, but are stable in Lake Ontario. However, these trends depend on the lake, the individual compound, and the organism considered.


image file: c9em00265k-p1.tif

Christina K. Remucal

Associate Professor Christy Remucal leads the Aquatic Chemistry group at the University of Wisconsin, Madison and is the Director of the Water Science and Engineering Laboratory. She is a faculty member in the Department of Civil & Environmental Engineering, the Environmental Chemistry & Technology Program, and the Freshwater & Marine Science Program. She holds a BS (2003) in Environmental Engineering and Science from the Massachusetts Institute of Technology and an MS (2004) and PhD (2009) in Civil and Environmental Engineering from the University of California, Berkeley. She completed her postdoctoral research in the Institute of Biogeochemistry and Pollutant Dynamics at the Swiss Federal Institute of Technology in 2012.



Environmental significance

Per- and polyfluoroalkyl substances (PFAS) are a diverse group of fluorinated organic chemicals that have been used in numerous industrial and consumer applications since the 1950s. PFAS are resistant to chemical and biological degradation and are ubiquitous in the environment, including in the Laurentian Great Lakes, which are one of the most important water resources in North America. This review evaluates spatial and temporal trends in perfluoroalkyl sulfonates and perfluoroalkyl carboxylates in water, surface sediment, sediment cores, lake trout (Salvelinus namaycush), and herring gull (Larus argentatus) eggs across the Great Lakes. The concentrations of PFAS increase from Lake Superior (i.e., the most pristine lake) to Lake Ontario (i.e., the most anthropogenically impacted lake) and the distribution of PFAS reflects changes in sources. Temporal trends broadly reflect manufacture and usage patterns of PFAS (i.e., increasing concentrations through the 2000s), although there is limited data on shorter chain compounds.

Introduction

Per- and polyfluoroalkyl substances (PFAS) are a diverse group of fluorinated organic chemicals that are both hydrophobic (i.e., repel water) and lipophobic (i.e., repel lipids/grease).1,2 PFAS have been used in numerous industrial and consumer applications that take advantage of these unique chemical properties, such as water-repellent cookware and firefighting foams, since the 1950s.3–5 PFAS are resistant to chemical and biological degradation due to their strong carbon–fluorine bonds and are poorly removed by many conventional drinking water and wastewater treatment processes.1,3,6–8 Potential health concerns include developmental toxicity, cancer, and bioaccumulation, especially for longer chain PFAS such as perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA; Table 1).4,5,9 As a result of these concerns, the U.S. Environmental Protection Agency included six PFAS compounds in its Third Unregulated Contaminant Monitoring Rule and issued a lifetime health advisory of 70 ng L−1 for combined concentrations of PFOS and PFOA.10 Lower guidelines have been issued for drinking water at the state level (e.g., a drinking water screening level of 8 ng L−1 for PFOS in Michigan)11 and for surface waters in other countries (e.g., 1 ng L−1 for PFOS in the European Union).12 While the production of some PFAS has declined due to voluntary phaseouts and the inclusion of PFOS on the persistent organic pollutants list of the Stockholm Convention,4,5 many other compounds are still produced and novel compounds are routinely identified in many environmental matrices.3,13
Table 1 Commonly studied PFAS in the Great Lakes
Compound Acronym Carbons Structure
Perfluoroalkyl sulfonates (PFSAs)
Perfluorobutane sulfonate PFBS 4 image file: c9em00265k-u1.tif
Perfluorohexane sulfonate PFHxS 6
Perfluorooctane sulfonate PFOS 8
Perfluorodecane sulfonate PFDS 10
[thin space (1/6-em)]
Perfluoroalkyl carboxylates (PFCAs)
Perfluorobutanoate PFBA 4 image file: c9em00265k-u2.tif
Perfluoropentanoate PFPeA 5
Perfluorohexanoate PFHxA 6
Perfluoroheptanoate PFHpA 7
Perfluorooctanoate PFOA 8
Perfluorononanoate PFNA 9
Perfluorodecanoate PFDA 10
Perfluoroundecanoate PFUnA 11
Perfluorododecanoate PFDoA 12
Perfluorotridecanoate PFTrA 13
Perfluorotetradecanoate PFTeA 14


PFAS are ubiquitous in the environment due to their persistence and long history of use.3,6–8,14 PFAS may be introduced to the aquatic environment via point sources, such as industrial or municipal wastewater effluent,15,16 and nonpoint sources, such as runoff from military fire training areas, landfill leachate, and atmospheric deposition.4,16–19 Furthermore, neutral PFAS, including fluorotelomer alcohols, can serve as precursors for ionic perfluoroalkyl carboxylates (PFCAs) and perfluoroalkyl sulfonates (PFSAs).20–22 Compared to many traditional persistent organic pollutants, ionic PFAS have high aqueous solubilities, have low volatility, and typically have acid-dissociation constants (pKa) < 1, resulting in a negative charge at circumneutral pH.3,7,8 Therefore, PFAS are frequently detected in the aqueous phase in the environment. Long chain PFAS (i.e., C ≥ 8 PFCAs and C ≥ 6 PFASs) can partition to biota and sediment,6,9 which tend to be enriched in long chain PFAS relative to the PFAS distribution measured in water.23–25 In contrast, less is known about the environmental fate of shorter chain PFAS, including perfluorobutane sulfonate (PFBS) and perfluorobutanoate (PFBA), which can be formed from compounds that are used as replacements for PFOS and PFOA.1,4,26

The Laurentian Great Lakes comprise the largest freshwater system in the world and are one of the most important water resources of North America.27–29 The Great Lakes include five major lakes that border the United States and Canada: Superior, Michigan, Huron, Erie, and Ontario. Lake Superior, which is located on the northwestern side of the basin and is the furthest upstream, is the largest of the lakes and is surrounded by a relatively rural and forested watershed. In contrast, Lakes Erie and Ontario on the eastern side of the basin are the furthest downstream and are more heavily impacted by industrial activity and urbanization. Legacy pollutants and chemicals of emerging concern are of interest in the Great Lakes because many companies that produce or use industrial chemicals are located in the Basin.30 Furthermore, the wide range of hydraulic residence times (i.e., 2.7 years in Lake Erie and 173 years in Lake Superior)27 suggests that the lakes may respond on different timescales to changes in production and use of persistent pollutants, such as PFAS. The Great Lakes are home to fish that are harvested by recreational anglers and for commercial purposes,29,30 which is a route of human exposure for bioaccumulative compounds. Additionally, elevated PFAS concentrations have been reported in the public water supplies for many communities in the Great Lakes Basin and may be associated with contamination in the Great Lakes in some cases.16,31 The importance of the Great Lakes as a water resource and the depth of study of a wide range of pollutants in water, sediment, and biota in the region make the Basin an ideal location to study the fate of these contaminants.

There are several review articles that are relevant to PFAS and biomonitoring programs in the Great Lakes, in addition to multiple reviews on PFAS in the environment more broadly.2–5,8,9,20,32 Giesy et al.6 provides an overview on PFAS partitioning, analytical techniques, and concentrations of PFAS in water and biota in the Great Lakes, primarily focusing on PFOS and PFOA. Klečka et al.33 and Clement et al.30 review the detection of chemicals of emerging concern in various matrices in the Great Lakes, including PFAS. These reviews are limited in scope in terms of the compounds considered and synthesis of existing data,6,30,33 in part because analytical techniques and data sets have improved greatly since their publication. Gewurtz et al.34 evaluates the concentrations of PFOS, PFOA, and long chain PFCAs in air, water, sediment, fish, and birds across Canada, including several sites in the Great Lakes (i.e., primarily Lake Ontario). Reviews on targeted and non-targeted screening programs for contaminants in archived biological samples, such as the Great Lakes Fish Monitoring and Surveillance Program, provide information on biomonitoring programs and their adaptation to pollutants such as PFAS.29,35

The aim of this review is to assess the spatial and temporal variability of PFAS in the Laurentian Great Lakes in water, sediment, and biota. Data in each matrix is synthesized in order to assess trends in both concentrations and distributions of PFAS in each lake spanning the industrialization gradient from west to east. Additionally, data from long term archives, such as sediment cores and biomonitoring programs, is assessed to investigate how the lakes are responding to changes in PFAS production and use. This review focuses on perfluoroalkyl substances, in which all of the H atoms on the carbon chain have been replaced by fluorine (Table 1),2 because these compounds are most frequently studied in the Great Lakes. However, studies that identify novel PFAS are also noted. Most studies of the Great Lakes report total concentrations and do not distinguish between branched or linear isomers, although data is summarized on the variable partitioning of isomers when available.

Trends in aqueous concentrations of PFAS

PFAS are frequently detected in the aqueous phase due to their negative charge and high aqueous solubility,3,7 making water an important reservoir of PFAS in the environment. This is particularly true for the Great Lakes, which have long hydraulic residence times ranging from 2.7 years (Erie) to 7.5 years (Ontario) to 21 years (Huron) to 62 years (Michigan) to 173 years (Superior).27 Thus, PFAS are expected to be present in the water column of the Great Lakes for extended periods of time, even if inputs of the persistent compounds decrease, which has implications for mobility and fate of PFAS in the aqueous environment.

Despite the importance of the aqueous phase in the fate and transport of PFAS, the available data in the Great Lakes is limited. Nine studies report data on a per sample or location basis, with typically only a few samples per lake and with most studies focusing on a single lake. Most studies quantify PFAS in near surface open water samples, although a few depth profiles exist.36,37 Unless otherwise stated, the concentrations reported below are for near surface (<10 m) open water samples. Finally, it is challenging to compare across studies due to instrumental differences since analytical techniques to quantify PFAS have improved since PFAS were first detected in Lakes Erie and Ontario in 2004.38 Despite these limitations, it is possible to draw conclusions about general concentration ranges, cross-lake variability, and spatial variability of PFAS in the Great Lakes.

Spatial distribution

The first reports of PFAS in the Great Lakes focused on Lakes Erie and Ontario and, as a result, these lakes have the longest record of aqueous concentration data. PFOS and PFOA were first quantified in the two lakes in samples collected in 2003 by single quadrupole liquid chromatography-mass spectrometry.38 This initial study reports mean PFOS and PFOA concentrations of 43 ng L−1 and 39 ng L−1, respectively, and observes that concentrations in Lake Ontario are 1.7 times higher than Lake Erie on average.38 Concentrations in more recent studies are an order of magnitude lower, with mean surface water PFOS concentrations ranging from 2.8 to 4.5 ng L−1 in Lake Erie39,40 and from 5.0 to 5.8 ng L−1 in Lake Ontario.24,39–41 Similarly, mean PFOA concentrations range from 1.9 to 5.5 ng L−1 in Lake Erie39,40 and from 2.5 to 4.3 ng L−1 in Lake Ontario.39–42 Interestingly, one additional study in Lakes Erie and Ontario reports mean PFOS concentrations of 3 and 4.9 ng L−1, respectively, that agree with the more recent work, while the reported mean PFOA concentrations of 15 and 21 ng L−1 are an order of magnitude higher.43 The concentrations in Boulanger et al.38 and Sinclair et al.43 are likely outliers due to analytical issues, the lack of proper field blanks, and/or potential contamination of sample collection equipment.44,45 However, the existing data support the observation that concentrations of PFOS and PFOA are higher in Lake Ontario than Lake Erie (Fig. 1).38–40,43
image file: c9em00265k-f1.tif
Fig. 1 Mean surface (<10 m depth) concentrations of (a) PFOS and (b) PFOA in Lake Superior (SU), Lake Michigan (MI), Lake Huron (HU), Lake Erie (ER), and Lake Ontario (ON). Points represent the average value and error bars represent the standard deviation of concentrations within each study. Data is compiled from Simcik et al. 2005,36 Scott et al. 2006,42 Sinclair et al. 2006,43 Furdui et al. 2008,40 Houde et al. 2008,24 Scott et al. 2010,37 De Silva et al. 2011,39 and Myers et al. 2012.41 Additional PFOS38 and PFOA38,43 data for Lakes Erie and Ontario are not included as discussed in the text.

PFOS and PFOA concentrations are generally lower in the upstream lakes (i.e., Superior, Michigan, and Huron) compared to the downstream lakes (i.e., Erie and Ontario). Mean surface water PFOS concentrations are 0.2 to 0.3 ng L−1 in Lake Superior,37,39,40 2.0 ng L−1 in Lake Michigan,36,39 and 2.1 to 2.3 ng L−1 in Lake Huron.39,40 Likewise, mean PFOA concentrations are 0.2 to 0.7 ng L−1 in Lake Superior,37,39,40,42 1.8 to 4.1 ng L−1 in Lake Michigan,36,39 and 0.6 to 3.2 ng L−1 in Lake Huron.39,40 Thus, the aqueous concentrations of PFOS and PFOA increase as the watersheds become more urbanized from west to east across the Great Lakes. This trend is true for studies that analyzed samples collected from multiple lakes across the Great Lakes Basin39,40 and is also apparent when available aqueous PFOS and PFOA concentrations quantified in multiple studies using different analytical methods are compared (Fig. 1). These concentrations are lower than the U.S. EPA lifetime advisory of 70 ng L−1 (combined for PFOS and PFOA),10 but higher than the European surface water guidelines (i.e., 1 ng L−1 for PFOS)12 in all lakes except Lake Superior.

While PFOS and PFOA are the most frequently quantified PFAS in aqueous Great Lakes samples and are present in the highest concentrations,23,33,36–38,40 other PFAS have been quantified in multiple lakes. Considering PFSAs, there is data of PFHxS (mean range = 0.01–0.94 ng L−1) in all five lakes.37,39,40 Similarly, C6–C12 PFCAs have also been quantified in the Great Lakes, with available concentrations on PFHxA (mean range = 0.30–2.99 ng L−1),37,39,42 PFHpA (0.07–2.2 ng L−1),36,37,39,40 PFNA (0.03–0.8 ng L−1),37,39,40,42 PFDA (0.04–0.39 ng L−1),37,39,40 PFUnA (<0.01–0.45 ng L−1),37,39,40 and PFDoA (<0.01–1.08 ng L−1).37,39,40 Note that the upper mean ranges of PFUnA and PFDoA are quite high; multiple lakes have mean concentrations on the order of 0.1 ng L−1, suggesting that the longer chain PFCAs are typically present in very low concentrations in the water column. These PFAS generally follow the same spatial trends as observed for PFOS and PFOA, with the lowest concentrations reported in Lake Superior and highest concentrations reported in Lake Ontario (Fig. 2, S1 and S2).

Aqueous concentration and spatial variability data are also available on a limited number of less frequently studied PFAS in the Great Lakes. With respect to shorter chain PFAS that can be formed from the compounds used to replace PFOS and PFOA in industrial applications,1,4,26 mean PFBS concentrations range from 0.14 ng L−1 in Lake Superior to 0.87 ng L−1 in Lake Huron for samples collected between 2005 and 2010 (Fig. S1a).39 The mean PFBS concentrations are 1.9 to 7.3 times lower than the mean PFOS concentrations in each lake.39 A separate study in Lake Superior reports much lower PFBS concentrations of 0.01 ng L−1 for samples collected in 2002.37 PFBA is reported in Lake Ontario in one study with a mean concentration of 5 ng L−1;42 this compound can have additional sources to the environment, such as atmospheric degradation of hydrofluorocarbons.46 Additionally, there is one report of trifluoroacetic acid (TFA) in Lake Ontario, which had a mean concentration of 100 ng L−1 (i.e., ∼2 orders of magnitude higher than the PFOA concentration in the same study).42 There are known non-industrial or wastewater sources of TFA to the environment that may contribute to the high concentrations of TFA observed in Lake Ontario. For example, TFA can have natural sources47 and is also the major photodegradation product of the aquatic lampricide 3-trifluoromethyl-4-nitrophenol which is applied to tributaries around the Great Lakes.48,49 Finally, high concentrations of perfluoroethylcyclohexane sulfonate (PFECHS), which is a corrosion inhibitor used in aircraft, are reported in aquatic samples in one study.39 The mean aqueous concentrations of 0.16–5.7 ng L−1 are higher than PFOS concentrations, suggesting that the less frequently studied compound warrants further attention.

Sources of PFAS

Attempts to identify the sources of PFAS to the Great Lakes also show that sources of PFAS vary from west to east. A mass balance constructed for Lake Ontario indicates that aqueous sources (i.e., surface waters and wastewater effluent) are major inputs of PFOS and PFOA to the lake, while atmospheric deposition is less important.50 This mass balance has been called into question due to the high PFOS and PFOA concentrations reported in the aqueous phase (i.e., an order of magnitude higher than concentrations shown in Fig. 1) and wastewater.44,45,50 However, the atmospheric PFOS concentrations (<LOD – 8.1 pg m−3)50 are similar to other measurements near Toronto, Canada (approximate mean range = 1–8 pg m−3)34,51,52 and a comparison of PFAS distribution in precipitation versus Lake Ontario water also highlights the relative importance of wastewater effluent, street dust, and industrial activities as PFAS sources to Lake Ontario.53 Similarly, the major sources of PFAS to Lake Michigan are also considered to be non-atmospheric based on the use of PFHpA[thin space (1/6-em)]:[thin space (1/6-em)]PFOA as a tracer of atmospheric deposition.36

In contrast, atmospheric deposition becomes more important as a PFAS source in the more pristine watershed of Lake Superior. For example, one study estimates that tributaries are the major source (e.g., contribute 59% of PFOA and PFOS) to Lake Superior, followed by atmospheric precipitation (32–51% for PFHpA, PFOA, PFNA, and PFOS) and direct wastewater discharge (1–8%).37 Likewise, the distribution of PFAS in precipitation and Lake Superior water are similar, confirming that deposition is a major source to the lake.53 Overall, these mass balance studies agree with the general observation that PFAS concentrations are higher in tributaries compared to the open water of the Great Lakes and that concentrations can be particularly elevated in wastewater effluent.37,40,41,50 Across the Great Lakes, the major output of PFAS is aqueous outflow, rather than volatilization or sorption.37,50

The observed shift in a major contribution of atmospheric PFAS sources in the western Great Lakes Basin to increased importance of surface water inputs in the east is reflected by trends in the ratios of selected PFAS compounds. For example, PFOA concentrations are consistently higher than PFOS concentrations in Lake Superior,37,39,40 while the opposite is true in Lake Ontario.38–41 Moving beyond the commonly studied C8 PFAS and combining all available data, a comparison of ratios of C4, C6, and C8 PFSAs versus C6–C12 PFCAs also reveals a general shift towards more sulfonates from west to east.34,39,40 PFSAs are 20% of the considered PFAS in Lake Superior, 29% in Lake Michigan, 35% in Lake Huron, 41% in Lake Erie, and 53% in Lake Ontario (Fig. 2b).


image file: c9em00265k-f2.tif
Fig. 2 (a) Concentration and (b) percent distribution of commonly studied PFAS in aqueous samples in the Great Lakes based on mean concentrations calculated across all studies shown in Fig. 1, S1, and S2.24,36,37,39–43

In-lake variability

There is limited evidence of within-lake spatial variability in the aqueous concentrations of PFAS. For example, PFAS are generally equally distributed in available depth profiles collected in Lakes Superior,37 Michigan,36 and Ontario.42 Similarly, there are small (i.e., typically <20%) differences in surface concentrations in samples collected in the eastern and western basins of Lakes Superior,37 Erie,38,39 and Ontario,38,39 as well as the northern and southern basins of Lake Michigan,36,39 although this data must be interpreted cautiously since there are typically only a limited number of sites in each comparison. Overall, there appears to be little spatial variability within each lake, which is expected given the long hydraulic residence times in the Great Lakes.27

PFAS in sediments of the Great Lakes

An understanding of the distribution of PFAS in sediments of the Great Lakes is important for several reasons. From a mass balance perspective, losses of PFAS by sedimentation are much smaller than outflows.37 However, sorption to sediment represents one of the few losses of PFAS from the water column.20,54 Furthermore, there is evidence of increased concentrations of PFAS in benthic organisms, as discussed below.54–56 Finally, sediment cores are excellent records of historical contamination for other chemicals, such as polychlorinated biphenyls57,58 and mercury,59,60 and it is valuable to assess their utility for serving as records of PFAS trends.

Sediment data in the Great Lakes is available in seven studies. Despite the low number of available studies, nearly all of the studies include both a large number of surface grab samples and at least one dated core, which enables the assessment of spatial distribution and temporal trends. In addition, several studies include data from multiple lakes, with data on Lakes Superior,25 Michigan,25,61 Huron,25 Erie,62 and Ontario,24,41,56,62,63 facilitating a comparison of concentrations across the Great Lakes Basin. Finally, three studies analyzed the same cores in Lake Ontario,41,63,64 providing an opportunity to assess reproducibility and variability in analytical methods.

Partitioning

The unique chemical properties of PFAS, such as the simultaneous hydrophobicity of the perfluorinated chain and hydrophilicity of the ionic head group, make predicting partitioning to solid phases challenging.3,6,8,65 Laboratory studies of PFAS sorption reveal that isotherms are typically linear at lower PFAS concentrations, whereas some non-linearity is observed at higher concentrations.54,66,67 Therefore, a linear distribution coefficient (KD; in units of L kg−1), which is the measured sediment concentration divided by the measured aqueous concentration, is frequently used to assess sorption for the aqueous PFAS concentrations that are typical of the Great Lakes. However, KD values can vary widely for an individual compound because sediment characteristics and solution conditions can influence sorption. For example, the sediment organic carbon content is considered to be a major factor influencing sorption of PFAS.25,64–69 Therefore, KD values may be converted to organic carbon-normalized partition coefficients (KOC) by dividing by the fraction of organic carbon (fOC). This approach is used to compare PFAS sorption across different sediments,8,66–68 although there are some deviations from correlations with organic carbon70 and other mechanisms, such as electrostatic interactions, can influence sorption under some conditions (e.g., sediments with low organic carbon content).6,65,66,69

Molecular structure and chain length also influence the sorption of PFAS to sediment. The relationship between sorption coefficients and chain length is well-established for longer chain PFAS. For example, log[thin space (1/6-em)]KOC increases by 0.5 to 0.8 units per –CF2 group.54,66,68 Similarly, sulfonates sorb more strongly than carboxylates that have the same chain length (e.g., by >0.2 log units).25,54,67 Based on these trends, it is generally assumed that short chain PFAS (i.e., C ≤ 4) have minimal sorption to sediments and the presence of PFBA in sediment samples in the Great Lakes is attributed to porewater concentrations.62 However, sorption of short chain PFAS is not well understood and may not follow the trends expected based on chain length. For example, sorption of PFBA (i.e., C4 PFCA) to sediment is found to be nearly equivalent to that of PFOA (i.e., C8 PFCA) under laboratory conditions, possibly due to ion exchange or steric interactions.68 The uncertainties concerned with sorption of short chain PFAS, coupled with the observed mobility of PFAS in sediment cores,64 introduces challenges in interpreting sediment core data for more mobile PFAS in the Great Lakes.

An additional challenge in assessing sorption of PFAS to sediments is the discrepancy between laboratory-predicted partition coefficients and observations under field conditions. For example, field partitioning estimates are up to an order of magnitude higher than laboratory studies in Lake Ontario and other freshwater systems, indicating that sediments can be relatively enriched in PFAS (Table 2).41,66 Zareitalabad et al.66 calculate average field and lab log[thin space (1/6-em)]KOC values based on all available data in 2013 for a wide range of samples and conditions. They observe that the average log[thin space (1/6-em)]KOC values for field measurements, in which partitioning is calculated based on measured water and sediment concentrations, are 1.2 and 0.9 log units higher for PFOS and PFOA, respectively, than laboratory studies.66 The availability of field-based partition coefficients specifically for the Great Lakes is limited, although one study in Lake Ontario estimates field-based log[thin space (1/6-em)]KOC values that are 2 and 1.5 orders of magnitude higher than the averaged laboratory values for PFOS and PFOA, respectively,41 which agrees with the global average for field measurements.66 In contrast, the field-based log[thin space (1/6-em)]KOC values in Etobicoke Creek (i.e., a tributary of Lake Ontario that was impacted by a fire-fighting foam spill at the Toronto airport) are similar to laboratory measurements (Table 2),70 although it is possible that the tributary sediments were not at equilibrium given the relatively short residence time in the stream and variable aqueous concentrations following the spill. Nevertheless, this conflicting evidence suggests that more accurate measurements of sediment sorption under conditions encountered in the Great Lakes are needed to assess the magnitude of sorption across the Basin.

Table 2 log[thin space (1/6-em)]KOC values (L kg−1) for PFOS and PFOA based on global assessments and measurements in the Great Lakes Basin
Compound Global laboratory average66 Global field average66 Lake Ontario41 Lake Ontario tributary70
PFOS 3 ± 0.7 4.2 ± 1.2 5 ± 2 2.1 ± 0.7
PFOA 2.8 ± 0.9 3.7 ± 0.9 4.3 ± 1.5 1.5 ± 0.1


Spatial variability of PFAS in sediments

As observed with aqueous concentrations, the concentrations of PFAS in surface sediment samples generally increase from west to east. This trend is apparent when the sum of reported mean PFAS concentrations is compared across different studies (Fig. 3a). This comparison is limited to C8–C12 PFCAs and C6, C8, and C10 PFSAs because these long chain PFAS are most readily sorbed to sediment and because they are most frequently quantified in surface sediments in the Great Lakes. On this basis, the sum of long chain PFAS is lowest in Lake Superior (1.3 ng g−1) and Lake Michigan (2.5 ng g−1), higher in Lake Erie (6.1 ng g−1) and Lake Huron (9.2 ng g−1), and highest in Lake Ontario (20.7 ng g−1).
image file: c9em00265k-f3.tif
Fig. 3 (a) Concentration and (b) percent distribution of commonly studied PFAS in surface sediment samples in the Great Lakes based on mean concentrations calculated across all studies shown in Fig. 4, S3, and S4.24,41,56,61–64

The trend in increasing PFAS concentrations in the more urban- and industrially-impacted lakes is also observed for individual compounds. For example, the mean PFOS concentration in surface sediments increases from 0.1 ng g−1 in Lake Superior25 to 0.45–0.7 ng g−1 in Lake Michigan25,61 to 0.9 ng g−1 in Lake Huron25 to 1.7 ng g−1 in Lake Erie62 (Fig. 4a). Studies in Lake Ontario report mean surface PFOS concentrations ranging from 1.8 to 26 ng g−1.24,41,62–64 While there is a large variability in the reported PFOS concentrations in Lake Ontario, the PFOS concentrations are elevated compared to the other Great Lakes in all cases. Similarly, the mean PFOA concentrations are lower than PFOS in surface sediments (Fig. 4b). The lowest PFOA concentrations are observed in Lake Superior (0.1 ng g−1),25 followed by Lake Michigan (<LOD – 0.2 ng g−1),25,61 Lake Huron (0.5 ng g−1),25 and Lake Erie (0.6 ng g−1).62 Sediment PFOA concentrations in Lake Ontario are similar across studies (range = 0.4–2.7 ng g−1) and are typically higher than the values reported in the other lakes.41,56,62–64


image file: c9em00265k-f4.tif
Fig. 4 Mean surface sediment concentrations of (a) PFOS and (b) PFOA in the Great Lakes. Points represent the average value and error bars represent the reported range of concentrations within each study. Data is compiled from Houde et al. 2008,24 De Silva et al. 2009,56 Myers et al. 2012,41 Yeung et al. 2013,63 Codling et al. 2014,61 Guo et al. 2016,64 Codling et al. 2018a,25 and Codling et al. 2018b.62

Long chain PFAS are more likely to be found in elevated concentrations in surface sediment samples in the Great Lakes (Fig. 3b and S3–S5), as expected based on KOC values.54,66,68 For example, long chain PFAS are dominant63,64 and C ≥ 8 PFCAs comprise >80% of the PFCA compounds detected in the sediment in Lake Ontario.63 The proportion of C > 8 compounds is higher in the sediments than in aqueous samples. Comparing Lakes Huron, Erie, and Ontario, which have the most data available, the ratio of PFDA[thin space (1/6-em)]:[thin space (1/6-em)]PFOA is 1.22 ± 1.00 in the sediment25,56,62–64 and 0.19 ± 0.30 in aqueous samples.39,40 Thus, PFDA is nearly equivalent in concentration to PFOA in the sediment, while it is approximately an order of magnitude lower in the water column. For the C > 8 PFAS, concentrations generally increase from west to east (Fig. S3–S5).

The distribution of PFSAs compared to PFCAs varies across the lakes (Fig. 3b). As observed in aqueous samples (Fig. 2b), long chain PFSAs are less dominant in Lake Superior (40% of all long chain PFAS) compared to Lake Ontario (69%). Trends are more variable in the other lakes, in which long chain PFSAs are 58% of all PFAS in Lake Michigan, 29% in Lake Huron, and 45% in Lake Erie. This variability deviates from the clear shift from PFCAs to PFSAs from west to east in water and in biota (discussed below) and may be due to the limited number of available studies (e.g., sediment concentrations of C > 8 PFCAs in Lakes Superior, Michigan, and Huron are only reported in one study).25

Short chain (i.e., more water soluble) PFAS have also been reported in surface sediment samples in the Great Lakes. Mean PFBA concentrations in sediments are <LOD in Lake Superior,25 0.6–1.6 ng g−1 in Lake Michigan,25,61 3.5 ng g−1 in Lake Huron,25 26.2 ng g−1 in Lake Erie,62 and 0.2–14.2 ng g−1 in Lake Ontario (Fig. S4).62,63 Mean concentrations of PFPeA are lower than PFBA and range from 0.1 to 2.5 ng g−1 across the Great Lakes,25,61–63 while mean concentrations of PFHxA range from <LOD to 0.9 ng g−1.25,62,63 Similarly, mean concentrations of PFBS and PFHxS range from 0.2 to 11 ng g−1 and 0.2 to 1.4 ng g−1, respectively (Fig. S3).25,61–64 The presence of the more water soluble PFAS in sediments (e.g., PFBA at concentrations >40 times higher than PFOA in the same study) is attributed to pore water concentrations because these compounds are less likely to sorb and because the sample processing did not exclude pore water.25,62 Thus, the presence of short chain PFAS in sediment samples should be interpreted cautiously.

Additional compounds beyond PFSAs and PFCAs have also been quantified in sediments of the Great Lakes. For example, trace level concentrations (∼1–100 pg g−1) of several polyfluoroalkyl phosphoric acid diesters (diPAPs) and perfluoroalkyl phosphinic acids (PFPiAs) have been reported in sediments in Lake Ontario.64 Extractable organic fluorine (EOF) has also been quantified in sediments in Lake Michigan61 and Lake Ontario.63 In both cases, the fraction of known PFSAs compared to the reported EOF is low. Fourteen known PFSAs comprise 2–44% of EOF in Lake Ontario surface sediments,63 while 25 known PFSAs comprise 0.09% of EOF on average in Lake Michigan surface samples.61

In-lake variability

The analysis of surface samples across several of the Great Lakes provides an opportunity to assess the spatial distribution in PFAS and the role of urbanization in elevated PFAS concentrations. In Lake Michigan, the sum of PFAS in surface sediments is highest in the north and south compared to the center of the lake,61 which may be partially attributed to the gyre that separates the two basins in the winter.71 Additionally, a cluster analysis of the PFAS distribution shows that samples near Chicago and Milwaukee cluster separately from the other 25 samples, suggesting an urban influence.61 Total PFAS concentrations and distributions are more similar across Lakes Superior and Huron, but elevated concentrations are noted in sites that are near airports or more industrially-impacted areas.25

Similarly, concentrations in Lakes Erie and Ontario are higher in sediments that receive inputs from urban watersheds. For example, the highest concentrations in Lake Erie are observed in coastal regions near urban areas and two Areas of Concern (i.e., the Detroit and Maumee Rivers).62 The lowest sediment concentrations in Lake Ontario are consistently observed in the central Mississauga Basin, which is less impacted by urban activities,41,62,63 and elevated concentrations in surface sediments in Lake Ontario are noted in some urban-impacted sites.62 However, higher concentrations are noted in offshore sites compared to nearshore sites in two studies in Lake Ontario.41,63 This unexpected trend is associated with the presence of fine grain sediments and greater sediment accumulation rates offshore.63 Overall, the variability of sediment concentrations compared to the relatively homogeneity of aqueous PFAS concentrations, as discussed above, suggests that sediments may be good records of PFAS inputs for compounds that partition consistently to sediment.

Variability of PFAS concentrations in fish

It is important to consider PFAS concentrations in fish in the Great Lakes for many reasons. First, some compounds, such as PFOS, are known to biomagnify,6,14,55 which means that harvested fish can be a route of human exposure in addition to concerns about ecosystem health. Second, biomonitoring programs operated by the United States and Canada provide opportunities for investigating both spatial and temporal variability of PFAS in fish in the Great Lakes.29,72 These programs maintain frozen archives of lake trout (Salvelinus namaycush) in all five Laurentian Great Lakes and walleye (Sander vitreus) in Lake Erie,29,35 enabling retrospective analysis of emerging compounds of concern. Several data sets exist for fish collected from four39,73 or five23,72 of the Great Lakes, further enabling cross-lake comparison.

Partitioning of PFAS to fish

Multiple partition coefficients are used to compare PFAS concentrations in fish relative to concentrations in other matrices. First, bioaccumulation factors (KBAF; L kg−1) represent the concentration in the organism relative to the concentration in water. Second, biomagnification factors (BMF) are the concentration in the organism divided by the concentration in their food. Finally, trophic magnification factors are the change in concentration with trophic level and are evaluated over multiple trophic levels. Bioaccumulation and biomagnification factors assume that the concentration of individual PFAS in the fish is due to exposure from the same chemical in the water and diet, respectively.39

Bioaccumulation factors are reported for lake trout in all five Laurentian Great Lakes and for walleye in Lake Erie. The log of KBAF increases with chain length for C8–C11 PFCAs, as well as for C6 and C8 PFSAs (Fig. 5; Table S1),23,39 providing evidence that longer chain compounds are more likely to bioaccumulate in fish. The most variability in log[thin space (1/6-em)]KBAF is observed in shorter chain compounds. For example, log[thin space (1/6-em)]KBAF ranges from 1.0 to 3.6 for PFOA in lake trout and walleye (average of all reported values = 2.6 ± 0.9), whereas the range for PFDA is much narrower (3.5 to 4.8; average = 4.0 ± 0.4). This variability is likely attributable to the low concentrations of PFOA measured in fish, as discussed below. As observed for partitioning to sediments, PFSAs are more strongly associated with fish than PFCAs at the same chain length. For C8 compounds, for example, the average bioaccumulation factor of PFOS is 1.5 orders of magnitude higher than PFOA. With the exception of PFOA, the range in log[thin space (1/6-em)]KBAF is quite narrow across the five lakes and two fish species (Fig. 5).


image file: c9em00265k-f5.tif
Fig. 5 Log of bioaccumulation factors for lake trout (filled symbols) and walleye (hollow symbols) versus chain length for (a) PFCAs and (b) PFSAs.23,39 Error bars correspond to reported standard deviations.

Biomagnification factors, which compare fish concentrations directly to concentrations in their prey, are useful for assessing the biomagnification potential of PFAS. BMF values are reported for lake trout in Lake Ontario in two studies, with one study weighting BMF based on the overall lake trout diet55 and a second study calculating individual BMF values compared to alewife (Alosa pseudoharengus), rainbow smelt (Osmerus mordax), and slimy sculpin (Cottus cognatus; Fig. S6).24 BMF values for C9–C14 PFCAs are greater than one for lake trout (range = 1.6–3.4), demonstrating that longer chain PFCAs can biomagnify in this ecosystem. In contrast, a low BMF of 0.4 is reported for PFOA.55 Interestingly, biomagnification factors of PFCAs increase with chain length between C8 and C11 compounds (i.e., as observed for log[thin space (1/6-em)]KBAF; Fig. 5) and then decrease at longer chain lengths. The reason for this unexpected trend is unclear but may be attributable to the lower concentrations of the long chain PFCAs in water (i.e., typically not detected) and sediment (Fig. 3) or their large molecular size. BMF values for PFOS for lake trout relative to their average diet (2.9) and alewife (1.6) also provide evidence of biomagnification of this ubiquitous compound.24 BMF values for lake trout relative to sculpin and smelt are less than one; these bottom-feeding organisms show elevated PFOS concentrations relative to other aquatic organisms, possibly due to sediment contamination of PFOS precursors.24,74

Biomagnification of some PFAS is also observed across food chains in Lake Ontario. Trophic magnification factors range from 0.58 (PFOA) to 4.71 (PFUnA) when calculated over four species (i.e., Mysis, alewife, smelt, and lake trout; Fig. S7).55 As observed with BMF, trophic magnification factors are lower for C12–C14 PFCAs compared to PFUnA. Trophic magnification factors for PFOS are similar across two studies on the same food web in Lake Ontario (3.8 ± 1.0;24 5.88 (ref. 55)), demonstrating the bioaccumulation of PFOS across multiple species. Similar biomagnification of PFOS is observed in aquatic food chains in rivers in Michigan, while PFOA has less biomagnification potential.75

Spatial variability of PFAS in Great Lakes fish

Investigation of PFAS concentrations in fish introduces several complicating factors compared to aqueous and sediment samples. First, different fish species have different diets, which may result in differing exposure to PFAS. Lake trout is most frequently studied across the Great Lakes because it is a top predator fish, although walleye are also frequently studied in Lake Erie (Fig. 6, S8, and S9). Additional data on alewife, sculpin, and smelt from Lake Ontario also exist.24,55,74 Within the same species, there are differences in the tissue analyzed (i.e., whole fish versus skin on fillets versus skin off fillets). Finally, although studies typically analyze homogenates of multiple fish, fish age and length can vary. Although one study notes that total PFAS concentration is proportional to fish weight,23 other studies observe no correlation between PFAS concentrations and fish weight which makes it challenging to correct for variability across studies.73,76,77 The discussion here focuses on whole lake trout homogenates, which are most commonly studied.
image file: c9em00265k-f6.tif
Fig. 6 Mean concentrations of (a) PFOS and (b) PFOA in fish collected in the Great Lakes. Points represent the average value and error bars represent the reported range of concentrations within each study. Data is presented for whole lake trout,23,24,39,55,72,74,78 skin on lake trout fillets,76 combined concentrations for whole lake trout and walleye,72 skin off lake trout fillets,73 whole alewife,24,55,74 and whole walleye.39

The concentrations of PFAS in lake trout ranges widely across the Great Lakes and follows a consistent spatial gradient. Whole lake trout have been analyzed in all five Great Lakes in two studies,23,72 all lakes except Lake Michigan in two studies,39,73 and in Lake Ontario alone in three studies.24,55,78 A summary of all available data shows that the sum of PFAS concentrations in lake trout (C4 and C7–C14 PFCAs and C6, C8, and C10 PFSAs) generally increases from west to east, but is consistently highest in Lake Erie (Fig. 7a).23,72 The total average PFAS concentrations range from 11 ng g−1 in Lake Superior to 24 ng g−1 in Lake Michigan to 46 ng g−1 in Lake Huron. The total PFAS concentrations in lake trout in Lake Ontario are a factor of 2 higher than Lake Huron (92 ng g−1), while the highest concentrations are observed in Lake Erie (136 ng g−1). However, this trend is distinct from aqueous and sediment data, which showed that concentrations are typically highest in Lake Ontario (Fig. 2 and 3). The reason for this discrepancy is unclear but may be influenced by the limited amount of lake trout data available in Lake Erie. Biomonitoring programs in Lake Erie often favor walleye due to the limited spawning population of lake trout in the lake35,72 and it is worth noting that total PFAS concentrations in walleye are lower than those of lake trout in Lake Erie (Fig. 6, S8, and S9).


image file: c9em00265k-f7.tif
Fig. 7 (a) Concentration and (b) percent distribution of commonly studied PFAS in surface whole lake trout collected in the Great Lakes based on mean concentrations calculated across all studies shown in Fig. 6, S8, and S9.23,24,39,55,56,72,74,78

PFOS is found in the highest concentration and detected most frequently in Great Lakes fish.6,23,33,34,39,55,72,73,76–79 On a mass basis, PFOS ranges from 35% of long chain PFAS in Lake Superior to 64% of PFAS in Lake Huron to 80–82% of PFAS in Lakes Erie and Ontario (Fig. 7). The dominance of PFOS in lake trout in the Great Lakes reflects observations in fish collected from other locations. For example, PFOS is detected in 100% of Great Lakes fish samples collected in the 2010 National Coastal Condition Assessment in US waters (n = 157; all five lakes; 18 fish species), with a maximum concentration of 80 ng g−1 in fillets.77 PFOS follows the same spatial distribution as the sum of PFAS, with the highest mean concentrations observed in Lake Erie followed by Lake Ontario (Fig. 7a). The observed shift in increasing concentrations in Lakes Erie and Ontario agrees with broader analyses of fish concentrations, which observe that PFOS concentrations in organisms collected from urban environments are typically higher than rural areas.14,34,76,77 PFOS concentrations are less than an order of magnitude lower than legacy contaminants on a mass basis (e.g., PCBs and mercury) in fish across the Great Lakes.72

The concentrations and spatial distributions of PFAS beyond PFOS are more variable across the Great Lakes. PFOA and PFHpA are essentially constant across the lakes in all available fish species and tissue types (Fig. 6b and S8), which is distinct from the western to eastern spatial gradients observed in aquatic samples (Fig. 1 and S2). In contrast, longer chain PFCAs (C ≥ 9) and other PFSAs (i.e., PFHxS and PFDS) generally increase from Lake Superior to Lakes Erie and Ontario (Fig. S8 and S9), as observed in other matrices. Differences in PFOA and PFHpA are reflective of the low bioaccumulation potential of these compounds (Fig. 5).55,73 On a mass basis, the most dominant PFAS after PFOS in whole lake trout varies widely across the lakes. The western lakes tend to have higher concentrations of long chain PFCAs (e.g., PFTrA is 13% of total PFAS by mass in Lake Superior, PFOA is 18% in Lake Michigan, and PFUnA is 7% in Lake Huron). In contrast, PFSAs are more prevalent in Lakes Erie and Ontario, where PFDS (4.5%) and PFHxS (4.3%) are present in the second highest concentrations after PFOS, respectively.

Less data is available on the shorter chain PFAS in fish. PFBS and PFHxA are infrequently studied in the Great Lakes and are under the detection limit in Lakes Superior and Michigan in skin on lake trout fillets.76 Similarly, PFBS is not detected in any of the 157 fish samples collected from urban rivers in the Great Lakes Basin in the 2010 National Coastal Condition Assessment in US waters.77 Mean PFBA concentrations ranging from 0.8 to 2.0 ng g−1 are reported in lake trout in Lakes Superior, Michigan, and Huron, while a concentration of 2.1 ng g−1 is reported for combined lake trout and walleye homogenates in Lake Michigan.72 Short chain PFCAs, including PFBA and PFPeA, are detected more frequently and at higher concentrations than PFOA (PFBA = 16% of samples with a maximum concentration of 1.3 ng g−1; PFPeA = 32% and 3.0 ng g−1; PFOA = 12% and 0.97 ng g−1) in fish from urban tributaries of the Great Lakes.77

Sulfonate concentrations in Great Lakes fish are typically much higher than carboxylic acids.33,34,39,55,73,75,77,78 When averaged across all available data for whole lake trout, PFSAs (C6, C8, and C10) are 85–88% of total PFAS compared to C4 and C7–C14 PFCAs in Lakes Erie and Ontario (Fig. 7b). The contribution of PFSAs to total PFAS is somewhat lower in Lakes Michigan (61%) and Huron (66%), but still dominate the PFAS concentrations in fish. In contrast, PFASs comprise 37% of total PFAS in lake trout in Lake Superior. The shift from PFCAs to PFSAs from west to east reflect the spatial patterns observed in water (Fig. 2), which indicate that PFAS sources differ across the urbanization gradient in the Great Lakes Basin. The higher contribution of PFSAs to overall distributions in fish relative to those observed in water further demonstrates the increased potential of sulfonates to bioaccumulate. Additionally, enantiospecific analysis indicates that biotransformation of precursor compounds is a source of PFOS to invertebrates and fish, including lake trout, in Lake Ontario.74

In addition to differences in biological accumulation based on chain length and chemical composition of PFAS, isomers of the same chemical formula also display differences in biological samples. Although this review focuses on total concentrations of PFOS and PFOA because these values are typically reported, several studies investigate linear and branched isomers of these common PFAS in the Great Lakes. Houde et al.24 quantifies multiple PFOS isomers in water, sediment, and organisms in a Lake Ontario food chain. The ratio of linear-PFOS to total PFOS is 77% in technical grade PFOS, 43–56% in water, 81–89% in sediment, and >88% in all biota. Similarly, more branched isomers of PFCAs are detected in water compared to sediment and biota in Lake Ontario in two additional studies.56,74 The preferential partitioning of linear-PFOA and linear-PFOS to lake trout is confirmed in a separate study across Lakes Superior, Huron, Erie, and Ontario.39 Collectively, these results suggest that there are different uptake and/or elimination patterns for branched PFOS, as well as for partitioning to sediment,24,74 which introduces complications for distinguishing among PFAS sources by quantifying isomeric distribution.

Less commonly studied PFAS are also detected in fish in the Great Lakes. Perfluoroalkyl phosphonic acids (PFPAs) and PFPIAs are detected in lake trout in Lakes Ontario, Erie, Huron, and Superior (<LOD – 0.032 ng g−1), but are 1–2 orders of magnitude lower than PFSAs.73 Other compounds, such as PFECHS, are detected at concentrations up to 3.7 ng g−1 in lake trout from Lakes Huron, Erie, and Ontario.39 Finally, 30 of 3570 possible novel polyfluorinated compounds are detected using a non-target method in Lake Michigan lake trout,80 demonstrating that there are many compounds yet to be identified and quantified in biota of the Great Lakes.

The presence of PFAS in herring gulls and other biota

While fish are frequently used to assess variability of persistent pollutants in the Great Lakes, PFAS have been quantified in other biota that are located within the Great Lakes Basin or that prey on fish from the Great Lakes. The most commonly studied species is the herring gull (Larus argentatus), which is a top avian predator that is exposed to aqueous contaminants from its diet that is composed of prey associated with aquatic food webs.81,82 Herring gull eggs are collected and stored by Environment Canada as part of the Great Lakes Herring Gull Monitoring Program and are ideal for assessing spatial and temporal variability of PFAS. To that end, there are multiple studies quantifying PFAS in herring gull eggs across the Great Lakes.81–84 Limited data is also available on invertebrates, amphibians, reptiles, other birds, and mammals.

Partitioning of PFAS to birds and mink

Biomagnification factors, which relate concentrations in organisms to concentrations in their food, clearly show that PFOS and longer chain PFCAs biomagnify in multiple species. BMF values for PFOS in mink liver range from 5–10 relative to concentrations in salmon liver75 and 11–23 relative to whole carp from the Great Lakes.14,85 BMF values for PFOS in birds that prey on Great Lakes fish are similar. For example, a value of 5–10 is reported for bald eagle livers relative to concentrations in salmon livers,75 while a value of 8.9 is reported for common merganser livers relative to whole bass.43 Moderate and short chain PFCAs, such as PFOA, are generally considered to have less bioaccumulation potential.75 BMF values for the sum of PFSA compounds in herring gull eggs collected from Strachan Island in the St. Lawrence River is estimated as 9.5 and 2.6 relative to alewife and smelt as prey fish.82 BMF values for the sum of PFCAs are lower (i.e., 6.0 and 1.8 relative to the same fish), but still indicate that longer chain PFCAs can bioaccumulate.82

Spatial distribution of PFAS in herring gull eggs and other biota

PFAS are quantified in numerous species that live in or near the Great Lakes (Table S2). For example, PFAS are detected in invertebrates,24,55,74 turtles,75 frogs,75 bald eagles,75,86 cormorants,14 herring gulls,14,82–84,87,88 tree swallows,89 and mink.75 Reported concentrations of PFOS range from <10 ng g−1 in invertebrates24,74 and tree swallow plasma89 to >1000 ng g−1 in various tissues from mink85 and bald eagles.75 PFOA concentrations are typically much lower and range from <1 ng g−1 in zooplankton56 to 90 ng g−1 in the invertebrate Diporeia.55 However, it is difficult to compare concentrations and distributions of individual compounds across studies since different tissues are often analyzed and it is likely that accumulation of PFAS is tissue-dependent. For example, PFOS concentrations in herring gull tissues range from 4 ng g−1 in brain tissue to 250 ng g−1 in egg yolks in a single study (Fig. S10).88 The proportion of PFAS as PFOS also varies widely and ranges from 16% in herring gull brains to >99% in adipose tissue.88 The remainder of the discussion focuses on herring gull eggs since they are the most widely studied, but it is important to note that the trends are biased toward accumulation in this specific species and tissue.

Total concentrations of PFAS in herring gull eggs follow the same western to eastern gradient observed for water, sediment, and lake trout. PFAS concentrations in herring gull eggs are consistently highest in Lake Ontario and Lake Erie colonies compared to colonies on Lakes Superior, Huron, and Michigan (Fig. 8).81–84 For example, the average total PFAS concentration ranges from 145 ng g−1 in Lake Superior to ∼200 ng g−1 in Lakes Michigan and Huron to 395 ng g−1 in Lake Erie to 464 ng g−1 in Lake Erie. On the same lake, the total PFAS concentrations are typically highest in colonies closest to highly industrialized/urbanized areas (Fig. S11).82,83,87 The observed shift in increasing concentrations in Lakes Erie and Ontario agrees with observations of higher concentrations in organisms located near urban environments in the upper midwestern United States (e.g., mink in northern Illinois85 and tree swallows89 and bald eagles86 in Minnesota), as well as broader analyses of other species.14


image file: c9em00265k-f8.tif
Fig. 8 (a) Concentration and (b) percent distribution of commonly studied PFAS in herring gull eggs collected from 17 colonies in the Great Lakes based on mean concentrations calculated across all studies shown in Fig. 9, S12, and S13.81–84 Only colonies located on islands in the Great Lakes are included. Distributions according to individual colonies is shown in Fig. S11.

As observed in lake trout, herring gull eggs are dominated by PFOS (Fig. 8). The percentage of long chain PFAS present as PFOS is 56% in Lake Superior, 79% in Lake Michigan, 72% in Lake Huron, 78% in Lake Erie, and 84% in Lake Ontario. PFOS follows the same spatial pattern observed for total PFAS (Fig. 9) and increases from an average concentration of 82 ng g−1 in Lake Superior to 390 ng g−1 in Lake Ontario.81,83,84 While most studies report total PFOS, one study quantifies ten linear and branched isomers of PFOS in herring gull eggs collected from 15 sites around the Great Lakes.83 Linear PFOS is dominant in the bird eggs (i.e., 95–98.3% of the quantified PFOS isomers),83 in agreement with observed preferential partitioning of linear isomers to sediment and fish24,56,74 and predictions that linear isomers are more likely to bioaccumulate.8 PFDS is frequently detected in herring gull eggs with concentrations up to 24.8 ng g−1 reported in Lake Erie.81,84 PFDS also increases from west to east across the Great Lakes (Fig. S12). High concentrations of both PFOS and PFDS are noted in other biota in the Great Lakes.85,86


image file: c9em00265k-f9.tif
Fig. 9 Mean concentrations of (a) PFOS and (b) PFOA in herring gull eggs collected from 17 colonies in the Great Lakes. Points represent the average value and error bars represent the reported range of concentrations81,84 or standard deviation82,83 within each study. Data is compiled from Gebbink et al. 2009,82 Gebbink et al. 2010,83 Gebbink et al. 2011,81 and Letcher et al. 2015.84

Longer chain PFCAs are present in much higher concentrations in herring gull eggs compared to C ≤ 8 carboxylates (Fig. 9 and S13). PFUnA is present in the highest concentrations (range = 3.6–59 ng g−1) and comprises 4–15% of total PFAS when averaged across colonies in each lake (Fig. 8).81,82,84 PFTrA is also present in relatively high concentrations, comprising 3–10% of total PFAS in herring gull eggs (concentration range = 2.6–30.8 ng g−1).81,82,84 In contrast, PFOA is present in low concentrations (range = <LOD – 3.7 ng g−1) and is <0.5% of total PFAS in herring gull eggs. Several long chain PFCAs (e.g., PFDoA and PFTeA) show a clear spatial gradient from west to east, but other compounds (e.g., PFOA) are essentially constant in herring gull eggs collected from colonies in the Great Lakes (Fig. 9).

The ratio of PFSAs to PFCAs in herring gull eggs varies across the Great Lakes. PFCAs are more dominant in the upper Great Lakes compared to more industrially-impacted Lakes Erie and Ontario. For example, C7–C14 PFCAs are 42% of total PFAS in Lake Superior, 20% in Lake Michigan, 26% in Lake Huron, 15% in Lake Erie, and 14% in Lake Ontario (Fig. 8b).81–84 The shift in PFAS distribution suggests that different sources of PFAS contribute to loadings in Lake Superior compared to Lakes Erie and Ontario,84 in agreement with trends in aqueous data and mass balances of individual lakes. Overall, the PFAS distribution in herring gull eggs follows the same trends observed in lake trout collected from the Great Lakes (Fig. 7), although the range of PFSA distribution is narrower (i.e., PFSAs are 57–86% of total PFAS in herring gull eggs versus 37–88% in lake trout) which highlights differences in accumulation among species and among different tissues.

Very little data is reported for shorter chain PFAS in biota in the Great Lakes. PFBS is below detection in all herring gull colonies in three studies81,84,88 except for one detection of 1.5 ng g−1 in a colony in the St. Lawrence River.84 PFHxS is frequently detected but at low concentrations (mean concentration = 1.4 ng g−1).81,84 PFBA is not typically included as analyte in studies of biota in the Great Lakes, while PFHpA is typically below detection limit in herring gull eggs.84 Considering other species collected near the Great Lakes, PFBS and PFHpA are not detected in tree swallows from Minnesota and Wisconsin.89 PFHxS concentrations range widely depending on the organism and tissue type (Table S2), while PFBA concentrations of 0.3–0.4 ng mL−1 are quantified in bald eagle plasma.86

Temporal trends in PFAS

Water

There is insufficient PFAS data to evaluate temporal trends in aquatic samples, even when limiting the assessment to the most well-studied compounds (i.e., PFOS and PFOA) in the most well-studied lakes (i.e., Erie and Ontario). The Lake Erie time series is from 2004 to 2007, which is not long enough to determine temporal trends (Fig. S14). Lake Ontario has a longer record of data (2002–2010), but it is still limited. There appears to be a decrease in both PFOS and PFOA concentrations over this time period (Fig. S15), but the trends are not significant at p < 0.05.

Sediment cores

The distribution of PFAS in dated cores in the Great Lakes corresponds with the known historical use of the compounds. The most reliable trends are observed for long chain compounds (i.e., C ≥ 8 PFCAs and C ≥ 6 PFSAs), which partition more readily to sediment, as discussed above. Global production of C8 PFAS increased from the 1970s to early 2000s, followed by a shift towards usage of shorter chain compounds.4,5,20,90 Cores collected in Superior (n = 9), Michigan (n = 2), and Huron (n = 9) in 2011 and 2012 show a general increase in the sum of PFAS concentrations until approximately 2005.25 The same trend is also observed in 8 cores in Lake Michigan, which were collected in 2010 and show exponential increases in PFOS, PFOA, and PFHxS through the late 1990s and early 2000s.61 Trends in Lake Erie cores (n = 7; 2014) are less consistent, which is unexpected given that this lake has the highest sedimentation rates, but also show a general increase in PFAS.62 Many cores in these lakes show a decrease in concentrations of total and individual PFAS compounds after the early 2000s, but this observation is typically limited to a single sample and it is not possible to determine the significance of this trend without additional time points.25,61

The wealth of data in Lake Ontario provides a unique opportunity for comparison across studies because the same samples have been analyzed multiple times using different methods. Two studies analyze the same three sediment cores,41,63 while one of the cores (i.e., from Station 1034) is analyzed a third time in a separate study (Fig. 10).64 Yeung et al.63 report concentrations that are 1.5–2.2 times higher than Myers et al.41 for the same samples, while Guo et al.64 observe that their concentrations are a factor of 2–3 times higher for PFSAs and within ∼30% for PFCAs compared to the original analysis.41 The variability is attributed to differences in sample extraction methods, which highlights issues in reproducibility and emphasizes the need for standard reference samples. However, despite the differences in absolute measurements, the temporal trends observed in the cores are similar across studies, as well as in a fourth study that provides additional data on eight separate cores.62 Overall, PFOS and PFOA increase from the earliest time points (1950s–1970s) until the early to mid-2000s, which corresponds with historical production.41,62,63


image file: c9em00265k-f10.tif
Fig. 10 Concentrations of (a) PFOS, (b) PFOA, (c) the percent of PFAS as PFOS, and (d) the percent of PFAS as PFOA in cores from Lake Ontario. Data is adapted from Yeung et al.63 and Guo et al.64

Physicochemical characteristics and transport also impact PFAS distribution in sediment cores. PFOA is observed in cores in Lakes Superior, Huron, and Michigan prior to 1940s, which is attributed to pore water transport of weakly sorbed PFAS or bioturbation.25,61,62 Likewise, PFBS and PFBA are observed in sediment cores from all five lakes prior to the production period.25,61,62 The presence of highly soluble PFAS in sediment cores complicates interpretation of temporal trends62,64 and shows that sediments are unlikely to be good records of shorter chain PFAS due to their partitioning behavior and mobility. Despite these limitations, some observed trends agree with the increase in PFAS use with time and the shift toward applications using shorter chain PFAS. For example, one study observes a clear increase in PFBA concentrations with time in two out of three cores in Lake Ontario,63 while a second study notes that PFBS and PFBA are becoming increasingly important in sediment samples in Lake Michigan.61

Sediment core records provide additional insight into the historical patterns of PFAS emissions to the Great Lakes. While the absolute concentration of PFOS has increased with time, the proportion of PFOS to total measured PFAS has declined (Fig. 10). For example, PFOS contributes to 96% of quantified PFAS before 1980 and 78% after 2000 (averaged across four cores in Lake Ontario).63,64 At the same time, the contribution of PFOA increases from 2% before 1980 to 6% after 2000. The proportion of PFOS and PFOA in Lake Ontario agree across studies, even though the absolute concentrations differ.63,64 Similarly, analysis of the Niagara River suspended solids archive (1981–2006) suggests that PFOS peaked in 2001 and then decreased (t1/2 = 9 years), whereas PFOA continues to increase throughout the period of record.41 Furthermore, emerging compounds, such as diPAPs and PFPiAs, are only found in the surface sediments of Lake Ontario, confirming the increased use of alternative PFAS in recent years.64 Tabulated core data in Lakes Superior,25 Michigan,25,61 Huron,25 and Erie62 are not available, making it difficult to calculate relative shifts in PFOS and PFOA concentrations in these data sets. However, the difference in proportional inputs and evidence of newer PFAS show that changes in production and usage patterns may be tracked using sediment cores to some extent.

Lake trout bioarchives

Fish archives in the Great Lakes enable the assessment of temporal trends in bioaccumulative PFAS. Time series data is limited to whole lake trout homogenates collected in Lake Ontario. Two early studies show clear increases in PFOS and total PFAS in lake trout from 1979–1980 (mean [PFOS] = ∼18 and 43 ng g−1)55,79 until the mid-1990s–2001 (mean [PFOS] = ∼65 and 180 ng g−1, respectively).55,79 However, the increasing trend is not linear, with elevated concentrations observed in late 1980s/early 1990s followed by a decline in later time points that correspond with the onset of zebra mussels in the lake.55,79 However, later data sets show that PFOS and total PFAS concentrations have remained stable since the early 2000s.34,78 This trend is also evident when all available data is combined (Fig. 11).23,39,72 This observation is interesting because PFOS production has declined since the early 2000s.4,5 Therefore, the continued steady concentrations of PFOS in fish suggests either (1) alternate sources (e.g., aqueous film forming foams or wastewater effluent, as well as continued use of PFOS-containing products) or (2) biotransformation of PFOS precursors.16,39,78 There is some evidence of decline in other PFAS, most notably PFDS,78,79 but this compound is less prevalent in fish (Fig. 7) and its decline does not alter overall trends in total PFAS in lake trout.
image file: c9em00265k-f11.tif
Fig. 11 Concentrations of PFOS (circles) and total PFAS (diamonds) in whole lake trout collected in Lake Ontario. Data from Furdui et al. 2008 (ref. 79) and Martin et al. 2004 (ref. 55) was estimated from figures. All other data was summarized from tabulated data.23,39,72,74,78 The 2010 data point was averaged for samples collected in 2008–2012.72 Error bars represent standard deviations of data, which is not available for all studies.

The distribution of PFOS compared to other PFAS is also constant over the period of record in lake trout (Fig. S16). Considering all available data, PFOS contributes to 81 ± 8% of total PFAS in lake trout prior to 2001 and 87 ± 6% after 2001. This trend is contrary to the observed changes in PFAS distribution in Lake Ontario sediment, in which PFOS becomes less dominant with time (Fig. 10). This discrepancy is likely explained by the dominance of PFOS in lake trout compared to other compounds, particularly in Lake Ontario (Fig. 7). Thus, while bioarchives, such as lake trout, may be good records of some PFAS compounds, this record may be limited to compounds that bioaccumulate and are present in high proportions (i.e., PFOS). Therefore, bioarchives may not be suitable for assessing relative changes in sources or concentrations, particularly for shorter chain compounds.

Herring gull bioarchives

Herring gull eggs also provide a useful bioarchive that can be used to assess temporal changes in bioaccumulative PFAS in the Great Lakes. The most detailed study analyzes C6, C8, and C10 PFSAs and C6–C14 PFCAs in herring gull eggs collected in 1990 and 1997–2010 (i.e., 15 timepoints).81 The eggs were collected from Agawa Rock (Lake Superior), Gull Island (Lake Michigan), Chantry Island (Lake Huron), Channel-Shelter Island (Lake Huron), Fighting Island (Detroit River), Weseloh Rocks (Niagara Falls, above the falls), and Toronto Harbor (Lake Ontario), providing good spatial coverage across the Great Lakes. PFOS and the sum of PFSAs generally decreases in herring gull eggs across the Great Lakes between 1990 and 2010, except for the Toronto Harbor site on Lake Ontario which has a positive trend.81 In contrast, the sum of PFCA concentrations generally increases, with a significant positive trend observed in four sites, while PFOA shows a negative trend in sites with sufficient data. A later study analyzes the same compounds in herring gull eggs collected in 2012 and 2013.84 PFSAs are the same or lower than the concentrations measured in 2010,81 whereas PFCAs are either slightly higher or lower.84

Sediment and fish concentration data show clear evidence of increasing PFAS concentrations from the earliest timepoints until the late 1990s/early 2000s (Fig. 10 and 11). Fish records, which are only available for Lake Ontario, show that concentrations of PFOS stabilize after this time period, whereas sedimentary records, which have data for all five lakes, generally do not have enough time points after the early 2000s due to low sedimentation rates to compare trends. Therefore, herring gull eggs provide more temporal resolution in the period of interest, particularly for the upper Great Lakes. The observed increase in PFSAs and PFOS in the Lake Ontario colony81 agrees with the lack of rebound in lake trout concentrations in that lake (Fig. 11).23,39,55,72,74,78,79 Furthermore, the observed decrease in PFOS and PFOA in herring gull eggs from the other four Great Lakes corresponds with the 3M phaseout of C8 PFAS,81 suggesting that biota in these ecosystems may be responding to changing environmental concentrations. Similar declines are observed in bald eagles from the south shore of Lake Superior over a limited time period (i.e., 2006–2011).86 Likewise, declines in PFOS and PFOA are reported in precipitation collected near Lakes Superior, Huron, and Ontario (2006–2018), whereas short chain PFCAs show either no decline or an increase in the most recent timepoint.53 While the herring gull data is not provided in a way to enable calculation of shifts in PFOS and PFOA contribution as done for the sediment data, an overall decrease in the sum of PFSAs and an increase in the sum of PFCAs suggests that sources of PFAS are changing with time to favor PFCAs (i.e., as observed in global analyses9 and in sediments of Lake Ontario; Fig. 10).

Conclusions and need for future research

This critical review evaluates the spatial and temporal variability of PFAS in the Laurentian Great Lakes in water, sediment, and biota. By synthesizing the available data, it is possible to observe trends over broad spatial and temporal scales and across environmentally relevant matrices that are often not possible in individual studies. There is a clear spatial pattern that emerges when comparing concentrations of PFAS across the Great Lakes. The lowest PFAS concentrations in all matrices are detected in Lake Superior, whereas higher concentrations are observed in Lakes Erie and Ontario. Within individual lakes, higher concentrations are observed in sediments that are impacted by industrial activities and wastewater effluent. Similarly, higher concentrations are observed in herring gull colonies that are near urban environments. Collectively, the trends in PFAS concentrations across the lakes demonstrate that urbanization and industrial activities contribute to PFAS loading and that proximity to PFAS sources is important, even in water bodies that have very long residence times.

There are clear trends in the distribution of PFAS across the Laurentian Great Lakes. The western lakes (e.g., Lake Superior) have higher relative proportions of PFCAs, whereas the eastern lakes (e.g., Lakes Erie and Ontario) have higher proportions of PFSAs. This trend is observed in water, lake trout, and herring gulls. The surface sediment data is less clear for Lakes Michigan, Huron, and Erie, although the differences between Lakes Superior and Ontario are consistent with the other matrices. Overall, the shift toward PFSAs in the east likely reflects changes in PFAS sources across the Great Lakes (i.e., higher contributions of atmospheric deposition in Lake Superior compared to more aqueous sources in Lake Ontario).37,50

The distribution of PFAS varies across matrices. Water has higher relative concentrations of shorter chain PFAS, whereas long chain compounds are more prevalent in sediment in biota. For example, PFOA is a major fraction of PFAS in water, but is typically present in very low concentrations in lake trout and herring gull eggs. The opposite trend is observed for PFOS, which is dominant in biota. These trends reflect predicted partitioning behavior of PFAS, which is driven by changes in chain length and functional group. Furthermore, the partitioning trends are similar for both lake trout and herring gulls, even though herring gull egg concentrations are roughly a factor of three higher than those observed in whole lake trout, which is reflective of differences in partitioning among individual species and tissues.

Temporal trends can be examined in sediment cores and bioarchives of lake trout and herring gull eggs. Both sediment core and lake trout data indicate that PFOS concentrations generally increased from the 1950s until the early 2000s, which is consistent with the production patterns of the compound. Lake trout data indicate that PFOS concentrations in Lake Ontario stabilized after the early 2000s. Herring gull egg data also demonstrate a stabilization in PFOS in Lake Ontario, but eggs collected from colonies in other lakes show declining concentration trends. Furthermore, the proportion of PFAS present as PFOS has declined with time in sediment cores. The temporal trends agree with global monitoring information on PFAS, which suggest that there is a general decrease in PFOS and increase in PFCAs globally.9 While these long term data sets are valuable for assessing the ability of ecosystems to respond to changes in PFAS inputs, it is important to note that they are biased towards compounds that partition to sediment or biota and that are present in high concentrations. In other words, these records are invaluable for assessing trends in PFOS concentrations, but less reliable for PFOA and short chain PFAS.

There are many areas that need future study in order to better understand the partitioning and fate of PFAS in ecosystems such as the Great Lakes. First, there is a clear need for aquatic measurements of PFAS in all Great Lakes over a longer time period to further elucidate temporal trends, particularly for shorter chain PFAS that are less likely to partition to sediment and biota. Second, despite the limitations in the types of compounds that can be tracked by long term archives (i.e., sediment cores and bioarchives), these data sets are still valuable for many compounds and continued monitoring is needed. In particular, more recent sediment core data is needed to determine if PFAS concentrations across the Great Lakes are declining (i.e., as observed in some herring gull colonies) or stabilizing (i.e., as observed in lake trout in Lake Ontario). Fourth, there are many PFAS compounds that are known and detected in other environments,3 but not included in many studies. While there is debate about which compounds should be monitored and which compounds are most important for understanding ecosystem health,91 we should consider expanding our analyte list beyond common PFSAs and PFCAs. Finally, within the Great Lakes, Lakes Erie and Ontario have very rich data sets, while other lakes, such as Lake Michigan, are very understudied and warrant future attention.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

C. K. R. acknowledges funding from ETH Zurich and resources from Eawag during her sabbatical stay.

References

  1. S. D. Richardson and T. A. Ternes, Water analysis: Emerging contaminants and current issues, Anal. Chem., 2011, 83(12), 4614–4648 CrossRef CAS PubMed .
  2. R. C. Buck, J. Franklin, U. Berger, J. M. Conder, I. T. Cousins, P. de Voogt, A. A. Jensen, K. Kannan, S. A. Mabury and S. P. van Leeuwen, Perfluoroalkyl and polyfluoroalkyl substances in the environment: Terminology, classification, and origins, Integr. Environ. Assess. Manage., 2011, 7(4), 513–541 CrossRef CAS PubMed .
  3. F. Xiao, Emerging poly- and perfluoroalkyl substances in the aquatic environment: A review of current literature, Water Res., 2017, 124, 482–495 CrossRef CAS PubMed .
  4. L. Ahrens, Polyfluoroalkyl compounds in the aquatic environment: A review of their occurrence and fate, J. Environ. Monit., 2011, 13(1), 20–31 RSC .
  5. A. B. Lindstrom, M. J. Strynar and E. L. Libelo, Polyfluorinated compounds: Past, present, and future, Environ. Sci. Technol., 2011, 45(19), 7954–7961 CrossRef CAS PubMed .
  6. J. P. Giesy, S. A. Mabury, J. W. Martin, K. Kannan, P. D. Jones, J. L. Newsted and K. Coady, Perfluorinated Compounds in the Great Lakes, in The Handbook of Environmental Chemistry, R. A. Hites, Springer-Verlag, Berlin, 2006, vol. 5, Part N, pp. 391–438 Search PubMed .
  7. M. F. Rahman, S. Peldszus and W. B. Anderson, Behaviour and fate of perfluoroalkyl and polyfluoroalkyl substances (PFASs) in drinking water treatment: a review, Water Res., 2014, 50, 318–340 CrossRef CAS PubMed .
  8. S. Rayne and K. Forest, Perfluoroalkyl sulfonic and carboxylic acids: a critical review of physicochemical properties, levels and patterns in waters and wastewaters, and treatment methods, J. Environ. Sci. Health, Part A: Environ. Sci. Eng., 2009, 44(12), 1145–1199 CrossRef CAS PubMed .
  9. M. Houde, A. O. De Silva, D. C. G. Muir and R. J. Letcher, Monitoring of perfluorinated compounds in aquatic biota: An updated review, Environ. Sci. Technol., 2011, 45(19), 7962–7973 CrossRef CAS PubMed .
  10. U.S. Environmental Protection Agency, Lifetime Health Advisories and Health Effects Support Documents for Perfluorooctanoic Acid and Perfluorooctane Sulfonate, U.S. Environmental Protection Agency, Washington, DC, 2016,EPA-HQ-OW-2014-0138 Search PubMed.
  11. Michigan PFAS Action Response Team Human Health Workgroup, Public health drinking water screening levels for PFAS, Michigan Department of Health and Human Services, Division of Environmental Health, 2019 Search PubMed.
  12. EU, Directive 2013/39/EU of the European Parliament and of the Council of 12 August 2013 amending Directives 2000/60/EC and 2008/105/EC as regards priority substances in the field of water policy, in EU Environmental Quality Standards, 2013 Search PubMed .
  13. Z. Wang, J. C. DeWitt, C. P. Higgins and I. T. Cousins, A never-ending story of per- and polyfluoroalkyl substances (PFASs)?, Environ. Sci. Technol., 2017, 51(5), 2508–2518 CrossRef CAS PubMed .
  14. J. P. Giesy and K. Kannan, Global distribution of perfluorooctane sulfonate in wildlife, Environ. Sci. Technol., 2001, 35(7), 1339–1342 CrossRef CAS PubMed .
  15. E. Houtz, M. Wang and J.-S. Park, Identification and fate of aqueous film forming foam derived per- and polyfluoroalkyl substances in a wastewater treatment plant, Environ. Sci. Technol., 2018, 52(22), 13212–13221 CrossRef CAS PubMed .
  16. X. C. Hu, D. Q. Andrews, A. B. Lindstrom, T. A. Bruton, L. A. Schaider, P. Grandjean, R. Lohmann, C. C. Carignan, A. Blum, S. A. Balan, C. P. Higgins and E. M. Sunderland, Detection of poly- and perfluoroalkyl substances (PFASs) in US drinking water linked to industrial sites, military fire training areas, and wastewater treatment plants, Environ. Sci. Technol. Lett., 2016, 3(10), 344–350 CrossRef CAS PubMed .
  17. E. F. Houtz and D. L. Sedlak, Oxidative conversion as a means of detecting precursors to perfluoroalkyl acids in urban runoff, Environ. Sci. Technol., 2012, 46(17), 9342–9349 CrossRef CAS PubMed .
  18. M. Loewen, T. Halldorson, F. Wang and G. Tomy, Fluorotelomer carboxylic acids and PFOS in rainwater from an urban center in Canada, Environ. Sci. Technol., 2005, 39(9), 2944–2951 CrossRef CAS PubMed .
  19. N. L. Stock, V. I. Furdui, D. C. G. Muir and S. A. Mabury, Perfluoroalkyl contaminants in the Canadian Arctic: evidence of atmospheric transport and local contamination, Environ. Sci. Technol., 2007, 41(10), 3529–3536 CrossRef CAS PubMed .
  20. K. Prevedouros, I. T. Cousins, R. C. Buck and S. H. Korzeniowski, Sources, fate and transport of perfluorocarboxylates, Environ. Sci. Technol., 2006, 40(1), 32–44 CrossRef CAS PubMed .
  21. D. A. Ellis, J. W. Martin, A. O. De Silva, S. A. Mabury, M. D. Hurley, M. P. Sulbaek Andersen and T. J. Wallington, Degradation of fluorotelomer alcohols: a likely atmospheric source of perfluorinated carboxylic acids, Environ. Sci. Technol., 2004, 38(12), 3316–3321 CrossRef CAS PubMed .
  22. J. W. Martin, D. A. Ellis, S. A. Mabury, M. D. Hurley and T. J. Wallington, Atmospheric chemistry of perfluoroalkanesulfonamides: kinetic and product studies of the OH radical and Cl atom initiated oxidation of N-ethyl perfluorobutanesulfonamide, Environ. Sci. Technol., 2006, 40(3), 864–872 CrossRef CAS PubMed .
  23. V. I. Furdui, N. L. Stock, D. A. Ellis, C. M. Butt, D. M. Whittle, P. W. Crozier, E. J. Reiner, D. C. G. Muir and S. A. Mabury, Spatial distribution of perfluoroalkyl contaminants in lake trout from the Great Lakes, Environ. Sci. Technol., 2007, 41(5), 1554–1559 CrossRef CAS PubMed .
  24. M. Houde, G. Czub, J. M. Small, S. Backus, X. Wang, M. Alaee and D. C. G. Muir, Fractionation and bioaccumulation of perfluorooctane sulfonate (PFOS) isomers in a Lake Ontario food web, Environ. Sci. Technol., 2008, 42(24), 9397–9403 CrossRef CAS PubMed .
  25. G. Codling, S. Hosseini, M. B. Corcoran, S. Bonina, T. Lin, A. Li, N. C. Sturchio, K. J. Rockne, K. Ji, H. Peng and J. P. Giesy, Current and historical concentrations of poly and perfluorinated compounds in sediments of the northern Great Lakes - Superior, Huron, and Michigan, Environ. Pollut., 2018, 236, 373–381 CrossRef CAS PubMed .
  26. Z. Wang, I. T. Cousins, M. Scheringer and K. Hungerbühler, Fluorinated alternatives to long-chain perfluoroalkyl carboxylic acids (PFCAs), perfluoroalkane sulfonic acids (PFSAs) and their potential precursors, Environ. Int., 2013, 60, 242–248 CrossRef CAS PubMed .
  27. F. H. Quinn, Hydraulic residence times for the Laurentian Great Lakes, J. Great Lakes Res., 1992, 18(1), 22–28 CrossRef .
  28. D. Egan, The Death and Life of the Great Lakes, W. W. Norton & Company, 2017 Search PubMed .
  29. B. S. Crimmins, H. B. McCarty, S. Fernando, M. S. Milligan, J. J. Pagano, T. M. Holsen and P. K. Hopke, Integrating non-targeted and targeted chemical screening in Great Lakes fish monitoring programs, J. Great Lakes Res., 2018, 44(5), 1127–1135 CrossRef CAS .
  30. R. E. Clement, E. J. Reiner and S. P. Bhavsar, Organohalogen contaminants of emerging concern in Great Lakes fish: a review, Anal. Bioanal. Chem., 2012, 404(9), 2639–2658 CrossRef CAS PubMed .
  31. X. C. Hu, A. K. Tokranov, J. Liddie, X. Zhang, P. Grandjean, J. E. Hart, F. Laden, Q. Sun, L. W. Y. Yeung and E. M. Sunderland, Tap water contributions to plasma concentrations of poly- and perfluoroalkyl substances (PFAS) in a nationwide prospective cohort of U.S. women, Environ. Health Perspect., 2019, 127(6), 67006 CrossRef PubMed .
  32. D. M. Whitacre, Reviews of environmental contamination and toxicology, Rev. Environ. Contam. Toxicol., 2010, 207, vii–viii CrossRef .
  33. G. Klečka, C. Persoon and R. Currie, Chemicals of emerging concern in the Great Lakes Basin: an analysis of environmental exposures, in Reviews of Environmental Contamination in Toxicology, Springer New York, New York, NY, 2010, vol. 207, pp. 1–93 Search PubMed .
  34. S. B. Gewurtz, S. M. Backus, A. O. De Silva, L. Ahrens, A. Armellin, M. Evans, S. Fraser, M. Gledhill, P. Guerra, T. Harner, P. A. Helm, H. Hung, N. Khera, M. G. Kim, M. King, S. C. Lee, R. J. Letcher, P. Martin, C. Marvin, D. J. McGoldrick, A. L. Myers, M. Pelletier, J. Pomeroy, E. J. Reiner, M. Rondeau, M.-C. Sauve, M. Sekela, M. Shoeib, D. W. Smith, S. A. Smyth, J. Struger, D. Spry, J. Syrgiannis and J. Waltho, Perfluoroalkyl acids in the Canadian environment: multi-media assessment of current status and trends, Environ. Int., 2013, 59, 183–200 CrossRef CAS PubMed .
  35. B. S. Crimmins, J. J. Pagano, M. S. Milligan and T. M. Holsen, Environmental mass spectrometry in the North American Great Lakes fish monitoring and surveillance program, Aust. J. Chem., 2013, 66(7), 798–806 CrossRef CAS .
  36. M. F. Simcik and K. J. Dorweiler, Ratio of perfluorochemical concentrations as a tracer of atmospheric deposition to surface waters, Environ. Sci. Technol., 2005, 39(22), 8678–8683 CrossRef CAS PubMed .
  37. B. F. Scott, A. O. De Silva, C. Spencer, E. Lopez, S. M. Backus and D. C. G. Muir, Perfluoroalkyl acids in Lake Superior water: trends and sources, J. Great Lakes Res., 2010, 36(2), 277–284 CrossRef CAS .
  38. B. Boulanger, J. Vargo, J. L. Schnoor and K. C. Hornbuckle, Detection of perfluorooctane surfactants in Great Lakes water, Environ. Sci. Technol., 2004, 38(15), 4064–4070 CrossRef CAS PubMed .
  39. A. O. De Silva, C. Spencer, B. F. Scott, S. Backus and D. C. G. Muir, Detection of a cyclic perfluorinated acid, perfluoroethylcyclohexane sulfonate, in the Great Lakes of North America, Environ. Sci. Technol., 2011, 45(19), 8060–8066 CrossRef CAS PubMed .
  40. V. I. Furdui, P. W. Crozier, E. J. Reiner and S. A. Mabury, Trace level determination of perfluorinated compounds in water by direct injection, Chemosphere, 2008, 73(1), S24–S30 CrossRef CAS PubMed .
  41. A. L. Myers, P. W. Crozier, P. A. Helm, C. Brimacombe, V. I. Furdui, E. J. Reiner, D. Burniston and C. H. Marvin, Fate, distribution, and contrasting temporal trends of perfluoroalkyl substances (PFASs) in Lake Ontario, Canada, Environ. Int., 2012, 44, 92–99 CrossRef CAS PubMed .
  42. B. F. Scott, C. A. Moody, C. Spencer, J. M. Small, D. C. G. Muir and S. A. Mabury, Analysis for perfluorocarboxylic acids/anions in surface waters and precipitation using GC-MS and analysis of PFOA from large-volume samples, Environ. Sci. Technol., 2006, 40(20), 6405–6410 CrossRef CAS PubMed .
  43. E. Sinclair, D. T. Mayack, K. Roblee, N. Yamashita and K. Kannan, Occurrence of perfluoroalkyl surfactants in water, fish, and birds from New York State, Arch. Environ. Contam. Toxicol., 2006, 50(3), 398–410 CrossRef CAS PubMed .
  44. J. A. Field, S. Simonich and D. Barofsky, Comment on “Detection of perfluorooctane surfactants in Great Lakes water” and “Mass budget of perluorooctane surfactants in Lake Ontario”, Environ. Sci. Technol., 2005, 39(10), 3883–3884 CrossRef CAS PubMed .
  45. B. Boulanger, K. C. Hornbuckle, J. L. Schnoor, J. Vargo and A. M. Peck, Response to comment on “Detection of perfluorooctane surfactants in Great Lakes water” and “Mass budget of perluorooctane surfactants in Lake Ontario”, Environ. Sci. Technol., 2005, 39(10), 3885–3886 CrossRef CAS .
  46. J. J. MacInnis, K. French, D. C. G. Muir, C. Spencer, A. Criscitiello, A. O. De Silva and C. J. Young, Emerging investigator series: a 14-year depositional ice record of perfluoroalkyl substances in the High Arctic, Environ. Sci.: Processes Impacts, 2017, 19(1), 22–30 RSC .
  47. H. Frank, E. H. Christoph, O. Holm-Hansen and J. L. Bullister, Trifluoroacetate in ocean waters, Environ. Sci. Technol., 2002, 36(1), 12–15 CrossRef CAS PubMed .
  48. D. Ellis and S. Mabury, The aqueous photolysis of TFM and related trifluoromethylphenols. An alternate source of trifluoroacetic acid in the environment, Environ. Sci. Technol., 2000, 34(4), 632–637 CrossRef CAS .
  49. M. B. McConville, T. D. Hubert and C. K. Remucal, Direct photolysis rates and transformation pathways of the lampricides TFM and niclosamide in simulated sunlight, Environ. Sci. Technol., 2016, 50(18), 9998–10006 CrossRef CAS PubMed .
  50. B. Boulanger, A. M. Peck, J. L. Schnoor and K. C. Hornbuckle, Mass budget of perfluorooctane surfactants in Lake Ontario, Environ. Sci. Technol., 2005, 39(6), 1920 CrossRef CAS .
  51. L. Ahrens, T. Harner, M. Shoeib, M. Koblizkova and E. J. Reiner, Characterization of two passive air samplers for per- and polyfluoroalkyl substances, Environ. Sci. Technol., 2013, 47(24), 14024–14033 CrossRef CAS PubMed .
  52. S. Genualdi, S. C. Lee, M. Shoeib, A. Gawor, L. Ahrens and T. Harner, Global pilot study of legacy and emerging persistent organic pollutants using sorbent-impregnated polyurethane foam disk passive air samplers, Environ. Sci. Technol., 2010, 44(14), 5534–5539 CrossRef CAS PubMed .
  53. S. B. Gewurtz, L. E. Bradley, S. Backus, A. Dove, D. McGoldrick, H. Hung and H. Dryfhout-Clark, Perfluoroalkyl acids in Great Lakes precipitation and surface water (2006–2018) indicate response to phase-outs, regulatory action, and variability in fate and transport processes, Environ. Sci. Technol., 2019 DOI:10.1021/acs.est.9b01337 .
  54. C. P. Higgins and R. G. Luthy, Sorption of perfluorinated surfactants on sediments, Environ. Sci. Technol., 2006, 40(23), 7251–7256 CrossRef CAS PubMed .
  55. J. W. Martin, D. M. Whittle, D. C. G. Muir and S. A. Mabury, Perfluoroalykl contaminants in a food web from Lake Ontario, Environ. Sci. Technol., 2004, 38, 5379–5385 CrossRef CAS PubMed .
  56. A. O. De Silva, D. C. G. Muir and S. A. Mabury, Distribution of perfluorocarboxylate isomers in select samples from the North American environment, Environ. Toxicol. Chem., 2009, 28(9), 1801–1814 CrossRef CAS PubMed .
  57. A. Li, K. J. Rockne, N. Sturchio, W. Song, J. C. Ford and H. Wei, PCBs in sediments of the Great Lakes: distribution and trends, homolog and chlorine patterns, and in situ degradation, Environ. Pollut., 2009, 157(1), 141–147 CrossRef CAS PubMed .
  58. D. Hu, A. Martinez and K. C. Hornbuckle, Sedimentary records of non-Aroclor and Aroclor PCB mixtures in the Great Lakes, J. Great Lakes Res., 2011, 37(2), 359–364 CrossRef CAS PubMed .
  59. C. Marvin, S. Painter and R. Rossmann, Spatial and temporal patterns in mercury contamination in sediments of the Laurentian Great Lakes, Environ. Res., 2004, 95(3), 351–362 CrossRef CAS PubMed .
  60. R. F. Lepak, R. Yin, D. P. Krabbenhoft, J. M. Ogorek, J. F. DeWild, T. M. Holsen and J. P. Hurley, Use of stable isotope signatures to determine mercury sources in the Great Lakes, Environ. Sci. Technol. Lett., 2015, 2, 335–341 CrossRef CAS .
  61. G. Codling, A. Vogt, P. D. Jones, T. Wang, P. Wang, Y. L. Lu, M. Corcoran, S. Bonina, A. Li, N. C. Sturchio, K. J. Rockne, K. Ji, J.-S. Khim, J. E. Naile and J. P. Giesy, Historical trends of inorganic and organic fluorine in sediments of Lake Michigan, Chemosphere, 2014, 114, 203–209 CrossRef CAS PubMed .
  62. G. Codling, N. C. Sturchio, K. J. Rockne, A. Li, H. Peng, T. J. Tse, P. D. Jones and J. P. Giesy, Spatial and temporal trends in poly- and per-fluorinated compounds in the Laurentian Great Lakes Erie, Ontario and St. Clair, Environ. Pollut., 2018, 237, 396–405 CrossRef CAS PubMed .
  63. L. W. Y. Yeung, A. O. De Silva, E. I. H. Loi, C. H. Marvin, S. Taniyasu, N. Yamashita, S. A. Mabury, D. C. G. Muir and P. K. S. Lam, Perfluoroalkyl substances and extractable organic fluorine in surface sediments and cores from Lake Ontario, Environ. Int., 2013, 59, 389–397 CrossRef CAS PubMed .
  64. R. Guo, D. Megson, A. L. Myers, P. A. Helm, C. Marvin, P. Crozier, S. Mabury, S. P. Bhavsar, G. Tomy, M. Simcik, B. McCarry and E. J. Reiner, Application of a comprehensive extraction technique for the determination of poly- and perfluoroalkyl substances (PFASs) in Great Lakes Region sediments, Chemosphere, 2016, 164, 535–546 CrossRef CAS PubMed .
  65. C. P. Higgins and R. G. Luthy, Modeling sorption of anionic surfactants onto sediment materials: an a priori approach for perfluoroalkyl surfactants and linear alkylbenzene sulfonates, Environ. Sci. Technol., 2007, 41(9), 3254–3261 CrossRef CAS PubMed .
  66. P. Zareitalabad, J. Siemens, M. Hamer and W. Amelung, Perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) in surface waters, sediments, soils and wastewater – a review on concentrations and distribution coefficients, Chemosphere, 2013, 91(6), 725–732 CrossRef CAS PubMed .
  67. L. Ahrens, L. W. Y. Yeung, S. Taniyasu, P. K. S. Lam and N. Yamashita, Partitioning of perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS) and perfluorooctane sulfonamide (PFOSA) between water and sediment, Chemosphere, 2011, 85(5), 731–737 CrossRef CAS PubMed .
  68. J. L. Guelfo and C. P. Higgins, Subsurface transport potential of perfluoroalkyl acids at aqueous film-forming foam (AFFF)-impacted sites, Environ. Sci. Technol., 2013, 47(9), 4164–4171 CrossRef CAS PubMed .
  69. R. L. Johnson, A. J. Anschutz, J. M. Smolen, M. F. Simcik and R. L. Penn, The adsorption of perfluorooctane sulfonate onto sand, clay, and iron oxide surfaces, J. Chem. Eng. Data, 2007, 52, 1165–1170 CrossRef CAS .
  70. E. Awad, X. Zhang, S. P. Bhavsar, S. Petro, P. W. Crozier, E. J. Reiner, R. Fletcher, S. A. Tittlemier and E. Braekevelt, Long-term environmental fate of perfluorinated compounds after accidental release at Toronto Airport, Environ. Sci. Technol., 2011, 45(19), 8081–8089 CrossRef CAS PubMed .
  71. D. J. Schwab and D. Beletsky, Relative effects of wind stress curl, topography, and stratification on large-scale circulation in Lake Michigan, J. Geophys. Res., 2003, 108(C2), 3044 CrossRef .
  72. D. J. McGoldrick and E. W. Murphy, Concentration and distribution of contaminants in lake trout and walleye from the Laurentian Great Lakes (2008–2012), Environ. Pollut., 2016, 217, 85–96 CrossRef CAS PubMed .
  73. R. Guo, E. J. Reiner, S. P. Bhavsar, P. A. Helm, S. A. Mabury, E. Braekevelt and S. A. Tittlemier, Determination of polyfluoroalkyl phosphoric acid diesters, perfluoroalkyl phosphonic acids, perfluoroalkyl phosphinic acids, perfluoroalkyl carboxylic acids, and perfluoroalkane sulfonic acids in lake trout from the Great Lakes region, Anal. Bioanal. Chem., 2012, 404(9), 2699–2709 CrossRef CAS PubMed .
  74. B. J. Asher, Y. Wang, A. O. De Silva, S. Backus, D. C. G. Muir, C. S. Wong and J. W. Martin, Enantiospecific perfluorooctane sulfonate (PFOS) analysis reveals evidence for the source contribution of PFOS-precursors to the Lake Ontario foodweb, Environ. Sci. Technol., 2012, 46(14), 7653–7660 CrossRef CAS PubMed .
  75. K. Kannan, L. Tao, E. Sinclair, S. D. Pastva, D. J. Jude and J. P. Giesy, Perfluorinated compounds in aquatic organisms at various trophic levels in a Great Lakes food chain, Arch. Environ. Contam. Toxicol., 2005, 48(4), 559–566 CrossRef CAS PubMed .
  76. M. C. Williams and C. S. Schrank, Perfluorinated compounds (PFCs) in fish from Wisconsin's major rivers and Great Lakes, Fisheries Management Administrative Report No. 83, Wisconsin Department of Natural Resources, 2016, pp. 1–12 Search PubMed .
  77. L. L. Stahl, B. D. Snyder, A. R. Olsen, T. M. Kincaid, J. B. Wathen and H. B. McCarty, Perfluorinated compounds in fish from U.S. urban rivers and the Great Lakes, Sci. Total Environ., 2014, 499, 185–195 CrossRef CAS PubMed .
  78. S. B. Gewurtz, A. O. De Silva, S. M. Backus, D. J. McGoldrick, M. J. Keir, J. Small, L. Melymuk and D. C. G. Muir, Perfluoroalkyl contaminants in Lake Ontario lake trout: detailed examination of current status and long-term trends, Environ. Sci. Technol., 2012, 46(11), 5842–5850 CrossRef CAS PubMed .
  79. V. I. Furdui, P. A. Helm, P. W. Crozier, C. Lucaciu, E. J. Reiner, C. H. Marvin, D. M. Whittle, S. A. Mabury and G. T. Tomy, Temporal trends of perfluoroalkyl compounds with isomer analysis in lake trout from Lake Ontario (1979–2004), Environ. Sci. Technol., 2008, 42(13), 4739–4744 CrossRef CAS PubMed .
  80. S. F. Baygi, B. S. Crimmins, P. K. Hopke and T. M. Holsen, Comprehensive emerging chemical discovery: novel polyfluorinated compounds in Lake Michigan trout, Environ. Sci. Technol., 2016, 50(17), 9460–9468 CrossRef PubMed .
  81. W. A. Gebbink, R. J. Letcher, C. E. Hebert and D. V. Chip Weseloh, Twenty years of temporal change in perfluoroalkyl sulfonate and carboxylate contaminants in herring gull eggs from the Laurentian Great Lakes, J. Environ. Monit., 2011, 13(12), 3365 RSC .
  82. W. A. Gebbink, C. E. Hebert and R. J. Letcher, Perfluorinated carboxylates and sulfonates and precursor compounds in herring gull eggs from colonies spanning the Laurentian Great Lakes of North America, Environ. Sci. Technol., 2009, 43(19), 7443–7449 CrossRef CAS PubMed .
  83. W. A. Gebbink and R. J. Letcher, Linear and branched perfluorooctane sulfonate isomer patterns in herring gull eggs from colonial sites across the Laurentian Great Lakes, Environ. Sci. Technol., 2010, 44(10), 3739–3745 CrossRef CAS PubMed .
  84. R. J. Letcher, G. Su, J. N. Moore, L. L. Williams, P. A. Martin, S. R. de Solla and W. W. Bowerman, Perfluorinated sulfonate and carboxylate compounds and precursors in herring gull eggs from across the Laurentian Great Lakes of North America: temporal and recent spatial comparisons and exposure implications, Sci. Total Environ., 2015, 538, 468–477 CrossRef CAS PubMed .
  85. K. Kannan, J. Newsted, R. S. Halbrook and J. P. Giesy, Perfluorooctanesulfonate and related fluorinated hydrocarbons in mink and river otters from the United States, Environ. Sci. Technol., 2002, 36(12), 2566–2571 CrossRef CAS PubMed .
  86. W. T. Route, R. L. Key, R. E. Russell, A. B. Lindstrom and M. J. Strynar, Spatial and temporal patterns in concentrations of perfluorinated compounds in bald eagle nestlings in the upper midwestern United States, Environ. Sci. Technol., 2014, 48(12), 6653–6660 CrossRef CAS PubMed .
  87. W. A. Gebbink, R. J. Letcher, N. M. Burgess, L. Champoux, J. E. Elliott, C. E. Hebert, P. Martin, M. Wayland, D. V. C. Weseloh and L. Wilson, Perfluoroalkyl carboxylates and sulfonates and precursors in relation to dietary source tracers in the eggs of four species of gulls (Larids) from breeding sites spanning Atlantic to Pacific Canada, Environ. Int., 2011, 37(7), 1175–1182 CrossRef CAS PubMed .
  88. W. A. Gebbink and R. J. Letcher, Comparative tissue and body compartment accumulation and maternal transfer to eggs of perfluoroalkyl sulfonates and carboxylates in Great Lakes herring gulls, Environ. Pollut., 2012, 162, 40–47 CrossRef CAS PubMed .
  89. C. M. Custer, T. W. Custer, P. M. Dummer, M. A. Etterson, W. E. Thogmartin, Q. Wu, K. Kannan, A. Trowbridge and P. C. McKann, Exposure and effects of perfluoroalkyl substances in tree swallows nesting in Minnesota and Wisconsin, USA, Arch. Environ. Contam. Toxicol., 2013, 66(1), 120–138 CrossRef PubMed .
  90. A. G. Paul, K. C. Jones and A. J. Sweetman, A first global production, emission, and environmental inventory for perfluorooctane sulfonate, Environ. Sci. Technol., 2009, 43(2), 386–392 CrossRef CAS PubMed .
  91. C. A. McDonough, J. L. Guelfo and C. P. Higgins, Measuring total PFASs in water: the tradeoff between selectivity and inclusivity, Curr. Opin. Environ. Sci. Health, 2019, 7, 13–18 CrossRef .

Footnote

Electronic supplementary information (ESI) available: Tables S1, S2 and Fig. S1–S16 are included. See DOI: 10.1039/c9em00265k

This journal is © The Royal Society of Chemistry 2019