A review of the bioaccumulation and adverse effects of PFAS in free-living organisms from contaminated sites nearby fluorochemical production plants

Accumulation and effects in mammals

In 2004, the first pioneer study by Hoff and coauthors used a BACI approach to investigate the accumulation and potential adverse effects induced by exposure to six PFAS, namely PFOA, perfluorononanoate (PFNA), PFOS, perfluorodecanoate (PFDA), perfluoroundecanoate (PFUnDA), and perfluorododecanoate (PFDOA), towards adult individuals of the wood mouse (Apodemus sylvaticus) sampled in the nature reserve of Blokkersdijk, which is located nearby the FCP in Antwerp (Belgium), and at Galgenweel, a reference site selected 3 km from the FCP^[33]. While PFNA, PFDA, PFUnDA, and PFDoDA were measured sporadically in the liver of mice caught in both sampling areas, the hepatic concentrations of PFOS in the mice from Blokkersdijk were higher (0.47-178.55 μg/g ww) compared to conspecifics from Galgenweel (0.14-1.11 μg/g ww). Although no sex dependency of PFOS levels was observed, the bioaccumulation of this compound increased with the individual age. Different biochemical hepatic endpoints, including peroxisomal β-oxidation activity, microsomal lipid peroxidation, and mitochondrial fraction protein content, as well as liver weight and relative liver weight, were significantly higher in Blokkersdijk mice compared to conspecifics from Galgenweel, showing a positive relationship with PFOS concentrations. However, the effects on serum triglyceride, cholesterol, or potassium levels were not PFOS-mediated effects, while the hepatic PFOS concentrations were negatively related to the serum alanine aminotransferase activity. These results first suggest that levels of a specific compound, such as PFOS, can represent a threat to the health status of individuals living nearby a highly contaminated site.

A similar BACI design was replicated in the same study areas to explore PFOS presence and transfer in a restricted terrestrial food chain, with the omnivorous wood mouse as the apical consumer^[34]. Samples of water, soil, fruits of common blackberry and European elder, and different invertebrates (see Section “Accumulation and effects in invertebrates”) were collected to investigate the PFOS trophic transfer. Focusing on the concentrations accumulated in the wood mouse, as expected, the highest PFOS levels were measured in the liver from individuals collected at Blokkersdijk (range 787-22,355 ng/g ww). Although these levels were generally lower than those detected in the 2002 survey performed in the same study area^[33], they were considered very high, as well as higher than those measured in other mammal species^[35]. These data suggest that mice sampled nearby the FCP are directly exposed via the consumption of contaminated water, inhalation of the air, and ingestion of fine particles and/or suffer an indirect exposure via the consumption of PFOS-contaminated plants and invertebrates, resulting in high bioaccumulation^[34].

Accumulation and effects in birds

The BACI design was also applied in biomonitoring studies of PFAS contamination nearby FCP using birds as sentinel organisms. The first investigation explored the accumulation of PFOS and related adverse effects in great tit (Parus major) and blue tit (Cyanistes caeruleus) nestlings from Blokkersdijk reserve (close to the FCP in Antwerp) and Fort IV (control area)^[36]. The concentrations of PFOS measured in the liver isolated from nestling sampled in Blokkersdijk (86-2788 and 317-3322 ng/g ww for great and blue tits, respectively) were significantly higher compared to those measured in conspecifics from the control area (17-206 and 69-514 ng/g ww for great and blue tits, respectively). While no species- and sex-dependencies of hepatic PFOS concentrations were noted, they were significantly and positively correlated with the serum alanine aminotransferase activity and negatively with the serum cholesterol and triglyceride levels in both species. Moreover, in great tit individuals, hepatic PFOS concentrations were positively correlated with the relative liver weight, while in blue tits, the same relationship was noted with hematocrit levels. These results confirm the high bioaccumulation potential of PFOS in two bird species with different trophic strategies and the ability of this compound to alter biochemical parameters and organ development.

To avoid the previous invasive approach of measuring PFOS concentrations in bird nestlings, in 2007, Dauwe and coauthors demonstrated the reliability of using blood as a PFOS accumulation matrix in birds^[37]. PFOS was measured in both the blood and liver of great tit individuals sampled in the same study area, getting values from 553 to 11,359 ng/g ww in the liver and from 24 to 1625 ng/mL in the blood. These PFOS levels are among the highest ever reported in free-living animals, and they exceeded, in most individuals, the hepatic benchmark concentrations for the protection of avian species^[37]. A significant correlation in the PFOS concentrations measured in the liver and blood was found, suggesting the usefulness of blood as a non-destructive matrix for biomonitoring purposes^[37].

While previous studies were specifically focused on PFOS, the investigation performed by Groffen and coauthors measured the accumulation of 12 perfluoroalkyl acids [PFAA, namely four perfluoroalkyl sulfonic acids (PFSA) and eight perfluoroalkyl carboxylic acids (PFCA)], as well as the contamination fingerprint, in the eggs of great tit individuals collected along a gradient away from the Antwerp FCP (from 1 to 70 km from the FCP)^[38]. The median concentrations of PFOS, perfluorohexane sulfonate (PFHxS), perfluorodecane sulfonate (PFDS), and PFOA measured in great tit eggs collected in the surroundings of the FCP were 10,380 (extrapolated), 99.3, 47.7, and 19.8 ng/g ww, respectively. Although the concentrations of all compounds decreased with distance from the FCP, the levels measured in the eggs collected close to the FCP were higher than those reported in the eggs of free-living birds collected far from other contaminated sites^[26,39]. PFOS was the main contributor to the PFSA and PFAA (range 63.4%-97.6%) profile at each site, while PFOA was the dominant PFCA close to the FCP and the nearest locations (range 41.0%-52.8%). PFDoDA (37.7%) prevailed in the farthest location. Interestingly, the results from this study show a slight decrease in PFOS levels compared to previous surveys performed in the same study area^[37], likely due to the phasing out of this compound in the 2010s.

A further study by Lopez-Antia and coauthors investigated the accumulation of PFOS in the eggs of three bird species feeding primarily on invertebrates, namely the great tit, the northern lapwing (Vanellus vanellus), and the Mediterranean gull (Larus melanocephalus), sampled in 2006 nearby the Antwerp FCP^[24]. Besides accumulation, the potential adverse effects induced by PFOS exposure in free-living birds were investigated by measuring the alterations of total protein content, cholesterol, triglyceride, and uric acid concentrations in the plasma of great tit adults only. PFOS levels measured in lapwing eggs (46,182 ng/g ww) were higher than those recorded in the other species, as well as with respect to previous surveys performed in the same area^[36-38]. According to Groffen and coauthors, the levels of PFOS accumulated in eggs of lapwing and great tit decreased with distance from the FCP, up to 1700 m from the FCP^[38]. In contrast, no differences in PFOS concentrations were noted in the eggs of the three species collected at 1700 and 5500 m from the FCP. Despite the high PFOS levels measured in the eggs of the great tit, reflecting high levels accumulated by mothers before transferring to their eggs, no changes in plasma parameters were noted in individuals collected at different distances from the FCP^[24].

Additional information on the accumulation of 11 PFCA [i.e., perfluorobutanoate (PFBA), perfluoropentanoate (PFPeA), perfluorohexanoate (PFHxA), perfluoroheptanoate (PFHpA), PFOA, PFNA, PFDA, PFUnDA, PFDoDA, perfluorotridecanoate (PFTrDA), and perfluorotetradecanoate (PFTeDA)] and 4 PFSA [perfluorobutane sulfonate (PFBS), PFHxS, PFOS, and PFDS] in the plasma of adults and nestlings of the great tit collected in the study area previously described^[24,36-38] was reported by Lopez-Antia and coauthors^[23]. In addition, antioxidant and oxidative stress parameters, including total antioxidant capacity, reduced glutathione, superoxide dismutase, catalase and glutathione peroxidase activity, and protein carbonyls, were measured in red blood cells of great tits to check for PFAS-related alterations and to assess potential causal relationships. Five of the eleven PFAA were detected in the highest ever reported concentrations in bird plasma, confirming that Antwerp FCP is a hotspot of PFAS contamination. In contrast to previous studies on birds, this investigation reported that females had higher mean concentrations (and detection frequencies) for PFOS and PFUnDA than males. Moreover, the findings suggest that maternal transfer and dietary intake represent the main routes of exposure for nestlings to PFOS, but not to other compounds. A positive correlation between PFAA concentrations and protein damage was noted in adult birds, while in nestlings, a positive correlation was observed with higher activity of antioxidant enzymes (i.e., glutathione peroxidase and catalase). These results suggest that exposure to different PFAA might imbalance the oxidative status and induce the onset of oxidative damage in free-living birds. In particular, the imbalance of the oxidative status observed in nestlings might translate into detrimental oxidative damage to cellular macromolecules in juvenile or adult individuals. Thus, because of their “sedentary” habits during early post-natal periods, nestlings can be considered as sentinels of local contamination able to point out early-warning signals of biological effects.

The study by Groffen and coauthors investigated the relationships between the concentrations of 4 PFSA and 11 PFCA in the eggs of great tits collected along a distance gradient from a pollution source and diverse reproductive parameters (i.e., the start of egg laying, clutch size, hatching success, fledging success and total breeding success, eggshell thickness, and body condition of the nestlings)^[40]. High concentrations of PFOS (range 5111-187,032 ng/g ww), PFDoDA (range 1.1-133 ng/g ww), PFTrDA (range below the limit of quantification to 156 ng/g ww), and PFTeDA (range below the limit of quantification to 22 ng/g ww) measured close to the FCP were associated with a reduced hatching success of nests, with at least one egg hatched, thinner eggshells, and increased survival of the hatched chicks. High concentrations of PFDA were associated with reduced hatching success, especially in nests where no eggs hatched, an earlier start of egg laying, and a reduction of total breeding success, mainly caused by the failure in hatching. These findings suggest that exposure to high levels of PFAS can affect the reproduction of birds, with potential negative consequences for the populations of species living in the surroundings of FCP.

Lastly, the recent study by Morganti and coauthors focused on the contamination due to both legacy and emerging PFAS in birds breeding in the surroundings of a perfluoropolymer factory site in the Upper Po Plain (Northern Italy)^[16]. In this study, the accumulation of 13 PFAS, including PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, PFDoDA, PFTrDA, PFTeDA, PFOS, perfluorooctane sulfonamide (FOSA), 8:2 fluorotelomer sulfonate (8:2 FTS), and C6O4, was investigated in the eggs of great and blue tits and European starlings (Sturnus vulgaris) breeding close to the factory and compared to a reference site located in a rural site in Northern Italy, far from any known point source of PFAS release. Overall, the PFAS concentrations in starling and great tit eggs collected nearby the FCP were at least one order of magnitude higher than those collected in the reference site. An unprecedented contamination level of PFOA and C6O4 was recorded in the eggs of these passerine birds. PFOA concentrations in starling eggs collected around the FCP were two orders of magnitude higher than those measured in the tit samples, suggesting species-specific differences in accumulation and maternal transfer of this compound, likely due to different trophic strategies. It was already known that PFOA, used for many years in polytetrafluoroethylene production, heavily contaminates the soil through atmospheric deposition^[41]; therefore, soil organisms were likely enriched with this compound. As the European starling mainly feeds on terrestrial organisms (e.g., earthworms), which accumulate PFOA, the accumulation of this species was higher compared to that of the great tit. Differently, C6O4 is a recently introduced compound and it has not yet heavily contaminated the soil, so its presence in the air through aerosol emissions can result in deposition on local vegetation and the contamination of insects feeding on that vegetation^[16]. This study returned crucial information on the presence and accumulation of emerging PFAS, confirming the role of FCP as hotspots of contamination of these emerging contaminants. However, further studies are necessary to understand the potential toxicity of these new compounds.

Accumulation in amphibians

Only one study investigated the accumulation of PFAS in amphibians collected nearby a production site^[42]. Specifically, the occurrence, tissue distribution, and bioaccumulation of two novel PFAS, chlorinated polyfluorinated ether sulfonic acid (6:2 Cl-PFESA) and hexafluoropropylene oxide trimer acid (HFPO-TA), were measured in black-spotted frogs (Pelophylax nigromaculatus) collected in cities with or without the presence of fluorochemical industries in China. Frogs sampled in cities with large-scale FCP showed higher hepatic levels of ΣPFAS (i.e., 31.22 ng/g ww in Huantai and 54.28 ng/g ww in Changshu) than those from cities without FCP (i.e., 7.68 ng/g ww in Quzhou and 9.91 ng/g ww in Zhoushan). Sex- and age-specific differences in PFAS accumulation were noted, with females showing lower hepatic levels than males, while older frogs accumulated lower levels than younger ones. Concerning tissue distribution, skin, liver, and muscle contributed nearly 80% to the whole-body burden of 6:2 Cl-PFESA in males, while, in females, the ovaries alone accounted for 58.4%.

A recent study performed in a Chinese fluorochemical industrial area^[43] demonstrated the accumulation of PFAS in the liver of the black-spotted frog (Rana nigromaculata; mean ΣPFAS = 163.40 ng/g ww). Measured physiological indices, namely body weight (BW), liver weight, hepatosomatic index, and triglyceride and cholesterol content, resulted in a positive correlation with the hepatic PFAS concentrations, particularly with PFOA, PFOS, and 6:2 Cl-PFESA. In particular, hepatic PFAS levels were positively correlated with body weight, liver weight, and triglyceride content, while PFOS concentrations were only significantly positively correlated with triglyceride content. These results suggest that PFAS exposure might disrupt the lipid metabolism of the black-spotted frog. Moreover, PFAS levels measured in male frogs were higher compared to females and showed stronger correlations with lipid metabolism-related indices, suggesting that males might be more sensitive to PFAS than females. The PFAS mechanism of action on lipid metabolism suggested by the field study was confirmed through a companion laboratory study, which found that PFAS might act as lipid metabolism-disrupting chemicals in frogs^[43].

Accumulation and effects in invertebrates

D’Hollander and coauthors first investigated the accumulation of PFOS in earthworms (family Lumbricidae), slugs (order Stylommatophora), millipedes (class Diplopoda), and woodlice (order Isopoda) collected at Blokkersdijk, close to the FCP in Antwerp (Belgium)^[34]. PFOS was detected in all invertebrate taxonomic groups ranging from 28 to 9000 ng/g ww. In addition, the same study measured the concentrations of PFOS in soil (68 ng/g dry weight, dw) and water (22 ng/L) to derive the field biota-to-soil accumulation factors (BSAF), which ranged from 0.11 to 68 for earthworms. At the same time, these data allowed the estimation of the biomagnification factors for the wood mouse (concentration in the liver/concentration in berries), which were as high as 302, while biomagnification factors for invertebrates were lower (≤ 2).

In recent years, Groffen and coauthors explored if the concentrations of 15 PFAA, measured in isopods (Oniscidae) collected at the Antwerp (Belgium) FCP and in four other areas representing a gradient in the distance from the PFAS source (1-11 km), were related to the concentrations in the soil and the eggs of the great tit collected in the same areas^[44]. PFOS and PFPeA were the main PFAA measured in isopods, with mean concentrations of 253 and 108 ng/g ww, respectively. As observed in other studies, the PFAA concentrations decreased with increasing distance from the FCP. As the PFAA concentrations measured in isopods correlated with the concentrations in the soils, as well as in bird eggs, this study demonstrated that isopods can be considered as excellent bioindicators of PFAA contamination in soil and eggs.

Recently, a field study was conducted to elucidate the distribution and partitioning of perfluoroalkyl carboxylic acids (PFCAs; C7-12) in a terrestrial ecosystem nearby a fluoropolymer manufacturing facility in Ohio^[45]. Surface soil (0-6 cm), plants, and earthworms (Lumbricus terrestris) were collected from a field located within a one-mile radius of the FCP. The spatial distribution of PFCAs at this site suggested that both atmospheric deposition and groundwater recharge contributed to the contamination. Soil and plant tissue contamination consisted mainly of PFOA (range 77%-97% of the total PFCA), whereas longer-chain PFCA, such as PFUnDA (18%) and PFDoDA (32%), accounted for relatively higher proportions in earthworms than in soil. However, PFOA in earthworms was the major compound found at concentrations that ranged from 51 to 860 ng/g dw, with a mean value of 270 ng/g dw. The measured concentrations were higher than those reported for corresponding matrices in other locations worldwide, suggesting that fluorochemical manufacturing contributed to the contamination. BSAF (soil-earthworm) was calculated and compared to values derived from literature data and controlled laboratory exposure studies. The estimated BSAF of PFCA in earthworms increased with perfluorocarbon chain length (the values increased by approximately 0.07-0.58 log units for each additional CF₂ group), but the values were substantially lower than those in the controlled laboratory exposure studies. These results suggest that PFCA are efficiently transferred from soil to soil-dwelling invertebrates and that the bioavailability and bioaccumulation factors calculated from field studies can provide more realistic information on PFAS environmental transport and partitioning.

Lastly, Rusconi and coauthors evaluated the impact of a fluoropolymer factory plant in Northern Italy on a river macrobenthic community by a multidisciplinary approach based on combined chemical, ecological, and genetic analysis^[30]. First, a BACI approach was implemented to investigate the stream macrobenthic community composition up- and downstream of the FCP discharge point, whereby the concentrations upstream of FCP discharge were lower than those downstream. The mean sum of 12 PFAA (i.e., PFBA, PFPeA, PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, PFDoDA, PFBS, PFHxS, and PFOS) measured upstream was 35 ng/L (range 0-120 ng/L), while downstream was 2050 ng/L (range 333-8042 ng/L)^[30]. Second, the toxicological implication at the population scale was explored by investigating changes in population genetics of a native benthic organism (Hydropsyche modesta, Trichoptera), with the aim to highlight sub-lethal effects that can accumulate through generations and lead to fitness reduction and, potentially, population extinction. The overall ecological quality of the sites was calculated through the STAR_ICMi multimetric index, which analyzes the community structure by calculating and integrating different biotic indices, such as ASPT, Log10(Sel_EPTD + 1), 1-GOLD, total number of families, number of EPT families, and Shannon-Wiener Index^[46]. The multimetric index did not show any difference between the two sampling points, which were both classified as moderate. Nevertheless, there was some evidence of an ecological impact since the community composition upstream of the discharge point showed the presence of more sensitive taxa, such as Ephemeroptera (e.g., Choroterpes sp.), while the downstream site was dominated by a larger number of taxa generally resistant to moderate organic pollution loads such as Diptera taxa and Echinogammarus sp., potentially caused by genetic drift between exposed and non-exposed populations. However, genetic drift cannot be causally linked to a specific effect induced by PFAS.