Article pubs.acs.org/est
Accumulation of Perfluoroalkylated Substances in Oceanic Plankton Paulo Casal,† Belén González-Gaya,†,‡ Yifeng Zhang,§ Anthony J. F. Reardon,§ Jonathan W. Martin,§ Begoña Jiménez,‡ and Jordi Dachs*,† †
Institute of Environmental Assessment and Water Research, Spanish National Research Council (IDAEA-CSIC), Jordi Girona 18-26, 08034 Barcelona, Catalonia, Spain ‡ Department of Instrumental Analysis and Environmental Chemistry, Institute of Organic Chemistry (IQOG-CSIC), Juan de la Cierva 3, 28006 Madrid, Spain § Division of Analytical and Environmental Toxicology, Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Alberta Canada S Supporting Information *
ABSTRACT: The bioaccumulation of perfluoroalkylated substances (PFASs) in plankton has previously been evaluated only in freshwater and regional seas, but not for the large oligotrophic global oceans. Plankton samples from the tropical and subtropical Pacific, Atlantic and Indian Oceans were collected during the Malaspina 2010 circumnavigation expedition, and analyzed for 14 ionizable PFASs, including perfluorooctanoate (PFOA), perfluorooctanesulfonate (PFOS) and their respective linear and branched isomers. PFOA and PFOS concentrations in plankton ranged from 0.1 to 43 ng gdw−1 and from 0.5 to 6.7 ng gdw−1, respectively. The relative abundance of branched PFOA in the northern hemisphere was correlated with distance to North America, consistent with the historical production and coherent with previously reported patterns in seawater. The plankton samples showing the highest PFOS concentrations also presented the largest relative abundances of branched PFOS, suggesting a selective cycling/fractionation of branched PFOS in the surface ocean mediated by plankton. Bioaccumulation factors (BAFs) for plankton were calculated for six PFASs, including short chain PFASs. PFASs Log BAFs (wet weight) ranged from 2.6 ± 0.8 for perfluorohexanesulfonic acid (PFHxS), to 4.4 ± 0.6 for perfluoroheptanoic acid (PFHpA). The vertical transport of PFASs due to the settling of organic matter bound PFAS (biological pump) was estimated from an organic matter settling fluxes climatology and the PFAS concentrations in plankton. The global average sinking fluxes were 0.8 ± 1.3 ng m−2d−1 for PFOA, and 1.1 ± 2.1 ng m−2d−1 for PFOS. The residence times of PFAS in the surface ocean, assuming the biological pump as the unique sink, showed a wide range of variability, from few years to millennia, depending on the sampling site and individual compound. Further process-based studies are needed to constrain the oceanic sink of PFAS.
■
alternative PFASs12−15 requires continuous monitoring and fate assessment of PFASs in the global environment. The biogeochemistry and long-range transport of POPs depends largely on their physico-chemical properties. The water solubility of ionizable PFASs is higher than that of chlorinated POPs, such as polychlorinated biphenyls (PCBs) or polychlorinated dibenzo-p-dioxins (PCDDs). For that reason, although PFAS neutral precursors undergo atmospheric transport16 reaching remote regions,17,18 oceanic transport of PFASs has been suggested as an important transport vector from source regions to remote marine environments.19−21 Due to the importance of the marine system in the global distribution of PFASs, previous surface seawater monitoring
INTRODUCTION
Perfluoroalkylated substances (PFASs), including perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkyl sulfonic acids (PFSAs), are ubiquitous in the global environment. PFASs have been detected in wildlife and human tissues and have been associated with a broad array of adverse effects, including neurodevelopmental alterations,1 reproductive,2 immunologic3 and metabolic effects.4 The occurrence of PFASs has also been described in remote oceanic and polar regions.5−8 Concerns regarding the persistence, toxicity, and environmental effects9,10 led to the inclusion of perfluorooctanesulfonic acid (PFOS), its salts, and perfluorooctane sulfonyl fluoride (PFOSF), in the list of Persistent Organic Pollutants (POPs) regulated under the Stockholm Convention.11 Perfluorooctanoate (PFOA) is currently under consideration for its inclusion in this list. However, ongoing production of PFOS and its precursors, of PFCAs and their respective precursors, as well as new © XXXX American Chemical Society
Received: Revised: Accepted: Published: A
November 18, 2016 January 27, 2017 February 14, 2017 February 14, 2017 DOI: 10.1021/acs.est.6b05821 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology has already been reported.7,9,22,23 The occurrence of PFASs in deeper seawater has received little attention in terms of vertical profiles22 and vertical transport.24 The vertical distribution of PFASs will depend on physical transport by eddy diffusion and subducting water masses, 24,25 but also on the their biogeochemistry in the oceanic water column, including how these chemicals partition into planktonic organisms, and how part of the plankton-derived organic matter sequesters PFASs by settling to deep water masses and sediments, a process known as the biological pump. The biological pump has been described as an important vertical transport mechanism for hydrophobic POPs,26,27 but its relevance for PFASs fate and transport cannot be evaluated unless PFAS concentrations are available for phytoplankton and zooplankton (hereafter collectively called plankton). Plankton accounts also for the first step of PFASs transfer to aquatic food webs, and plays a key role on the cycling of hydrophobic POPs in the water column.26−30 However, there are few reports on the bioconcentration and bioaccumulation of PFAS in plankton, and these are mostly limited to limnic ecosystems,31−34 the Baltic Sea,35 and the coastal Arctic.36,37 PFOA and PFOS are generally present in the environment as a mixture of isomers. Isomer-specific analyses of PFOS and PFOA have found applications in environmental forensics, for example to understand questions of degradation,38 oxidation,39 and transport.40,41 Additionally, PFOA isomer profiles have been used to elucidate its manufacturing source in oceanic seawater samples.42 This is possible because PFASs have been manufactured by two primary methods: electrochemical fluorination (ECF) and telomerization. The ECF method was used by 3 M Co., the largest historic global producer, until its voluntarily phase out of PFOS production in 2000. The production of PFOA shifted to the telomerization method in 2002. Hence, ECF and telomerization are referred as the historical and contemporary manufacturing source of PFOA, respectively.41,42 ECF products are composed by a mixture of linear and branched isomers, whereas telomerization generates a pure isomeric product, typically the linear isomer.43 This difference provides the possibility to discern between the historical or contemporary manufacturing source of PFOA in environmental samples such as ocean water. Unlike PFOA, PFOS has primarily been produced by the ECF method, and therefore PFOS isomer profiles in ocean water do not provide manufacturing source information but may reflect selective weathering or cycling of the different isomers. The isomer profiles of PFOS have been previously evaluated in phyto- and zooplankton from freshwater environments,32 but not yet for oceanic plankton. As a necessary step toward a better understanding and modeling of PFASs biogeochemical cycling and transport, the objectives of this study were (i) to provide a first assessment of the occurrence of PFCAs and PFSAs in plankton from the Atlantic, Indian, and Pacific tropical and subtropical oceans; (ii) to elucidate PFOS and PFOA isomer composition in the first levels of the oceanic food web; (iii) to estimate PFASs bioaccumulation factors (BAFs) for oceanic plankton; and (iv) to quantify the biological pump fluxes of PFASs in the oligotrophic subtropical and tropical oceans.
campaign covered the global tropical and subtropical oceans (north and south Atlantic and Pacific Oceans, as well as the Indian Ocean) between 35°N and 40°S. Details of sample location and ancillary data are given in the Supporting Information (Table S1, SI). The 28 analyzed plankton samples were gathered with a 50 μm mesh net by vertical trawls from 20 m below the deep chlorophyll maximum (DCM) up to the surface. The sampling depth ranged from 40 to 170 m, depending on the site. The plankton samples studied here are composed of phyto- and zooplankton with a trophic level ranging from 1.2 to 1.9.28 Samples were filtered on precombusted and preweighed GF/D filters (Whatman) and immediately stored at −20 °C until further analysis. One liter of seawater from the surface (5 m depth) and the DCM was sampled with 30 L Niskin bottles attached to an oceanographic rosette coupled with a CTD (conductivity, temperature, depth) device, which allowed the in situ determination of the DCM depth. Nomenclature. Fourteen ionizable PFAS analytes were targeted for all samples, including C6, C8, and C10 PFSAs and C4−C14 PFCAs: Perfluorohexanesulfonic acid (PFHxS), PFOS, Perfluorodecanesulfonic acid (PFDS), Perfluorobutanoic acid (PFBA), Perfluoropentanoic acid (PFPeA), Perfluorohexanoic acid (PFHxA), Perfluoroheptanoic acid (PFHpA), PFOA, Perfluorononanoic acid (PFNA), Perfluorodecanoic acid (PFDA), Perfluoroundecanoic acid (PFUnA), Perfluorododecanoic acid (PFDoA), Perfluorotridecanoic acid (PFTrA), Perfluorotetradecanoic acid (PFTeA) (Table S2, SI). Isomeric-specific analysis of PFOS and PFOA was conducted for plankton samples only. The nomenclature for specific PFOS and PFOA isomers is adopted here from previous studies.44 Briefly, using PFOS as an example, linear and perfluoroisopropyl branches are abbreviated as n-PFOS and iso-PFOS, respectively. For monoperfluoromethyl PFOS isomers, the carbon position of the perfluoromethyl is indicated (e.g., 5mPFOS). ∑PFOS and ∑PFOA are used to designate the sum of all isomers, and ∑PFAS is used for the sum of all quantified PFASs. The terms %br-PFOS and %br-PFOA are used to report the percentage of branched isomers to ∑PFOS and ∑PFOA, respectively. Chemical Analysis. The analytical procedure for seawater samples is described in a previous work,5 both surface and DCM seawater extraction and instrumental analysis were done simultaneously. There was no determination of the isomeric composition for seawater samples. The results for surface seawater have been described in a companion work,5 and are used here when needed. In this work, we provide the new data set of concentrations of PFAS in seawater at the DCM depth, and the concentrations of PFAS in plankton. All plankton samples were extracted following a method reported previously for biota samples36 with some modifications to reduce matrix interferences in an ultraclean laboratory at the University of Alberta. Briefly, filters were lyophilized, weighed, and transferred to 15 mL polypropylene tubes. Filters contained approximately 0.1 g of plankton biomass (dry weight). Each sample was spiked before extraction with mass labeled internal standards (Table S2, SI). After addition of 5 mL of acetonitrile, the vials were vortexed, sonicated for 20 min and centrifuged for 10 min at 3000 rpm. The supernatant was collected and the extraction was repeated by adding another 5 mL of acetonitrile. The two extracts were mixed, the solvent was reduced under a gentle nitrogen stream and transferred to a 1.5 mL microcentrifuge tube containing 25 mg of ENVI-Carb
■
EXPERIMENTAL METHODS Sample Collection. Plankton and seawater samples were collected on board the RV Hespérides from December 2010 to July 2011 during the Malaspina 2010 expedition. The sampling B
DOI: 10.1021/acs.est.6b05821 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology and 50 μL of glacial acetic acid. The solution was mixed thoroughly and then centrifuged for 10 min at 12 000 rpm. The supernatant was collected and diluted in 50 mL of HPLC-grade water, followed by solid phase extraction using Oasis WAX cartridges as described for water samples.5 Identification and quantification of PFASs for the plankton samples was performed by Ultrahigh performance liquid chromatography tandem mass spectrometry (UFLC-MS/MS) with a Shimadzu LC-20 AD coupled to a Sciex API 5000 triple quadrupole mass spectrometer (Applied Biosystems-MDS Sciex, Concord, ON, Canada) operating in negative ion mode with multiple reaction monitoring. A 10 μL aliquot of each sample was injected onto an Ascentis Express F5 Column (2.7 μm, 90 Å, 10 cm × 2.1 mm, Sigma-Aldrich, Canada) maintained at 50 °C. Starting mobile phase conditions were 90% water adjusted to pH 4 with acetic acid (A) and 10% pure methanol (B). Initial conditions were held for 1 min, then ramped to 60% B for 2 min, increased to 100% B by 23 min, 100% B by 5 min, then 10% B and equilibrated for 10 min. Parent and product ions used for quantification and confirmation are shown in Table S2, SI. Two calibration curves were made for quantification purposes, one for nonisomer specific analysis of PFASs, and a second for PFOA and PFOS isomer profile analysis. Target analytes were quantified by internal standard quantification. Quality Assurance and Quality Control (QA/QC). QA/ QC for seawater samples is described in detail elsewhere.5 Average recoveries for the DCM water samples ranged from 76% for PFNA 13C5 to 142% for PFHxS 18O2 (Table S3, SI). Plankton procedural and/or field blanks (i.e., GF/D filters) were analyzed with each batch of 3−4 samples to monitor potential contamination during sampling and extraction (eight blanks in total). The limits of quantification (LOQ) were defined as the mean concentration of procedural and field blanks plus three times the standard deviation of the blank response (Table S4, SI). The LOQ of PFNA was in the same range as concentrations in field plankton samples, thus PFNA was not reported in plankton. For the analytes which were not found in procedural blanks, LOQ was determined as a signalto-noise ratio of three in field samples (Table S4, SI). Triplicate recovery experiments were performed by spiking plankton samples with all target compounds and isomers in addition to the internal standards. Recoveries were in the range from 83 ± 34% for PFHxS to 126 ± 15% for PFOA (Table S4, SI). Recent work has suggested that adsorption of PFAS to filters can be a potential sampling artifact leading to a bias in the measured concentrations.45 The seawater samples were filtered on GF/F filters before dissolved phase PFAS were solid phase extracted. However, it has been shown that 80%, respectively). High relative abundance of PFOS and of other PFSAs was also found in surface and DCM seawater from the South Atlantic (Figure S2, SI). The average PFOS contribution was lower in plankton (21% PFOS) than in seawater (36% PFOS), which is consistent with previous studies.33,36 PFPeA, PFHpA, and PFOA accounted together for more than 65% of ∑PFAS and more than 80% of ∑PFCAs in oceanic plankton, respectively. The long chain PFCAs (PFDA, PFUnA, PFDoA, and PFTeA) where more frequently detected, and accounted for a larger contribution of ∑PFCAs in the northern hemisphere than in the southern hemisphere (Figure S2, SI), but generally accounted for C8), respectively, contrasts with previous reports in aquatic biota,36,48 including plankton in freshwater studies.33,34 Thus, the current results for short-chain PFCAs in marine plankton are surprising and must be taken in consideration in assessments of bioaccumulation potential of short chain PFASs.49 An inverse correlation between plankton concentrations and mean chlorophyll content in the photic water column (mg
Atlantic and Pacific oceans, and maximum concentrations found in the South Atlantic, off the coast of Brazil (Figure 1, Table S7, SI). The average ∑PFAS plankton concentration for all samples was 17 ± 13 ng gdw−1. In samples from the southern hemisphere, average ∑PFSA and ∑PFCA plankton phase concentrations (8.1 ± 13 ng gdw−1 and 14 ± 6.3 ng gdw−1, respectively) were significantly higher (Kruskal−Wallis, p < 0.01) than in samples from the northern hemisphere (0.7 ± 0.1 ng gdw−1 and 6.4 ± 2.9 ng gdw−1, respectively). However, we did not sample all the oceanic regions in the northern hemisphere, such as close to East Asia, where high water concentrations have been reported.20 The concentrations of ∑PFAS in plankton were not significantly different among the Atlantic (mean 18 ± 17 ng gdw−1), Pacific (12 ± 8.4 ng gdw−1) and Indian oceans (18 ± 5.7 ng gdw−1). ∑PFCA and ∑PFSA concentrations were also not significantly different among oceans. However, there were significant differences (Kruskal− Wallis, p < 0.01) when comparing the five oceanic basins (northern and southern basins separately), since plankton from the South Atlantic (22 ± 20 ng gdw−1), South Pacific (20 ± 9.7 ng gdw−1) and Indian ocean (19 ± 5.7 ng gdw−1) had higher ∑PFAS concentrations than in the North Atlantic (7.6 ± 3.2 ng gdw−1) and North Pacific (7.6 ± 3.1 ng gdw−1). There are no previous measurements of PFAS concentrations in oceanic plankton, however, the highest concentrations detected here (for example off-shore Brazil) are comparable to those reported for phyto- and zooplankton from freshwater ecosystems,32,33 or from Arctic coastal seawater.36,37 ∑PFCAs showed similar concentrations at the sampling sites closest to continents (always at more than 200 nautical miles) compared to open ocean regions, without clear spatial patterns otherwise. On the other hand, the global distribution of ∑PFSAs in oceanic plankton followed a pattern similar to previous observations in seawater samples,5 with remarkably high concentrations in the South Atlantic near the Brazilian coast (Figure 1), and generally higher concentrations close to continents, compared to the open oceans. D
DOI: 10.1021/acs.est.6b05821 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Figure 2. Global PFOA and PFOS isomeric pattern as the percentage of the branched PFOA and PFOS to the total PFOA and PFOS. The percentage of branched isomers increases with the distance to the 70° W meridian for the northern hemisphere.
m−3) was only significant for PFHxA (r2s = 0.19, p < 0.05), even though with a very low percentage of variance explained. Negative correlations between plankton concentrations and biomass have been reported for other POPs such as PCBs and dioxins28 and are thought to be due to interactions between atmospheric deposition and biogeochemical controls like biomass dilution, the biological pump and degradation.26,28,30 The fact that these correlations are generally not observed for most PFASs is consistent with a larger contribution of oceanic transport versus atmospheric deposition, a PFAS partitioning to plankton not consistent with the hydrophobicity paradigm, and the stronger PFAS persistence in the water column compared with other POPs. Isomer Profiles of PFOA and PFOS in Plankton. To our knowledge, this is the first time that the isomer profiles of PFOA and PFOS are reported in open ocean plankton. The nPFOA isomer was the predominant PFOA isomer in all samples, with a mean contribution of 81 ± 10% of total PFOA, followed by 5m-PFOA (10 ± 10%), iso-PFOA (6.4 ± 2.8%) and 4m-PFOA (2.2 ± 4.2%). The high occurrence of branched PFOA isomers in plankton samples substantiates the significance of historical ECF as the manufacturing source responsible for the oceanic occurrence and distribution of this compound. In fact, the relative abundance of the isomers was comparable to the rank order in historical ECF manufacturing (n-PFOA > iso-PFOA > 5m-PFOA > 4m-PFOA) in approximately one-third (36%) of the samples. Nevertheless, although the average relative abundance of n-PFOA and % brPFOA was similar to that of historical ECF PFOA (78% nPFOA, 22% br-PFOA46), there was high spatial variability in the percentage of n-PFOA ranging from 55% near Europe’s West coast to 98% in Australia’s South coast (Figure 2). The lowest average %brPFOA contributions were found in the Indian Ocean (13 ± 5.2%) (Figure 2), and overall being significantly lower in the southern hemisphere than in the northern hemisphere (Kruskal−Wallis, p < 0.05). PFOA branched isomers where detected in all the oceans at a much higher frequency (89%, 96%, and 39% of the samples for isoPFOA, 5m-PFOA, and 4m-PFOA, respectively) than in previous studies for biota.32 The %n-PFOA range (55 to 98%) was wider than that reported in fish from the Taihu Lake32 (92−100%) or Lake Ontario50 (99−100%), presumably owing to the wide spatial coverage in the current study. These differences in planktonic PFOA isomer composition are likely related to the historic origin of PFOA, where higher % brPFOA is associated with more contribution from historic ECF PFOA manufacturing. Although it has generally been suggested that the preferential accumulation of n-PFOA versus brPFOA in organisms makes biota inappropriate to perform source apportionment from the isomeric pattern,32 there was a significant (p < 0.01) linear correlation between %brPFOA in
plankton and degrees of longitude from 70° W for the northern hemisphere (Figure 2). Specifically, %brPFOA increased from Central America toward Europe in the North Atlantic Ocean, and toward Hawaii in the North Pacific Ocean. This spatial pattern is consistent with a previous assessment of the % brPFOA in surface seawater from the North Atlantic, which reported a lower contribution of branched isomers along the 70° W meridian, reaching a minimum near Narragansett Bay (Rhode Island northeast coast).42 These spatial trends suggest that the relative contribution of historical ECF PFOA is currently lower near the North American continent than in the open northern hemisphere oceanic basins. This is probably the result of contemporary nPFOA production or use in North America (i.e., the source region by oceanic and atmospheric transport). However, % brPFOA presents a narrower range and lower abundance in North Atlantic seawater (6.0−24%)42 than in plankton (14− 45%). Other processes may also contribute to the large variability of the isomeric pattern in plankton, leading to % brPFOA values well above the 22% contribution from ECF production lots, such as (i) preferential loss of n-PFOA during PFOA transport to the open ocean, for example due to the biological pump or transfer to the food web, (ii) potential differential isomeric transformation of PFOA-precursors in the atmosphere, or (iii) PFOA-precursors degradation in the water column. As found for PFOA isomers, n-PFOS was the predominant PFOS isomer in all samples, with a mean contribution to total PFOS of 87 ± 5.7%, followed by iso-PFOS (7.2 ± 2.0%), 5mPFOS (3.0 ± 1.3%), 4m-PFOS (2.4 ± 1.0%), 3m-PFOS (1.3 ± 1.0%), and 1m-PFOS (1.6 ± 0.7%). All PFOS isomers were found in the majority of samples: n-PFOS (100% detection frequency), iso-PFOS (96%), 5m-PFOS (89%), 4m-PFOS (86%), 3m-PFOS (89%), with the exception of 1m-PFOS (25%). Similar contributions of n-PFOS to total PFOS in phytoplankton (84%) and zooplankton (80%) were reported in Taihu Lake, China.32 PFOS was mainly produced by the ECF process and, therefore, differences in isomer composition in plankton are less likely attributable to different PFOS manufacturing techniques but rather these are likely indicative of preferential isomer uptake and/or excretion by plankton, and of other biogeochemical processes in the water column. The % brPFOS in historical ECF PFOS is close to 30%,41 greater than observed in most plankton samples. The n-PFOS relative abundance was significantly higher (thus lower %brPFOS) in the northern (91 ± 5.5%) than in the southern hemisphere (84 ± 4.6%) (Kruskal−Wallis, p < 0.01). This may suggest older PFOS sources for the northern hemisphere, due to the higher discrepancy in %brPFOS between the plankton from the North Atlantic and North Pacific oceans and the ECF original production lots. E
DOI: 10.1021/acs.est.6b05821 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology
Table 1. Comparison of Bioaccumulation Factors (BAFs, L kg−1) for PFAS in Oceanic Plankton with Previous Measurements and Modelled BAFsa sampling area
group/size
reference
n
PFHxS
PFOS
PFHxA
PFHpA
PFOA
PFNA
PFDA
PFUnA
PFDoA
South Korea rivers South Korea rivers South Korea rivers Mai Po Marshes Taihu lake Taihu lake Taihu lake Baltic Sea
phytoplankton Microzooplankton Meso-zooplankton phytoplankton phytoplankton phytoplankton zooplankton zooplankton
4 4 4 3 3 17 17 4
nr nr nr 1.76 nr nr nr 1.3
2.3 3.5 3.5 2.22 2.3 2.5 2.4 2.3
nr nr nr nr nr nr nr nr
nr nr nr nr nr nr nr nr
nr nr nr 2.46 1.4 1.94 1.47 2.3
3.2 nr 3.2 3.21 2.7 2.9 2.4 nr
nr nr nr 2.88 2.7 1.6 1.5 1.8
nr nr nr 3.65 4.2 3.1 2.5 nr
nr nr nr nr 3.6 nr 3 nr
North Atlantic Ocean South Atlantic Ocean Indian Ocean
plankton
Lam et al. 2014 Lam et al. 2014 Lam et al. 2014 Loi et al. 2011 Fang et al. 2014 Xu et al. 2014 Xu et al. 2014 Gebbink et al. 2016 this study
6
