ARTICLE pubs.acs.org/est
Temporal Trends and Pattern of Polyfluoroalkyl Compounds in Tawny Owl (Strix aluco) Eggs from Norway, 1986-2009 Lutz Ahrens,*,†,|| Dorte Herzke,‡ Sandra Huber,‡ Jan Ove Bustnes,§ Georg Bangjord,^ and Ralf Ebinghaus† †
Institute for Coastal Research, Helmholtz-Zentrum Geesthacht, Max-Planck Strasse 1, 21502 Geesthacht, Germany Norwegian Institute for Air Research, FRAM Centre, Hjalmar Johansens gate 14, NO-9296 Tromsø, Norway § Norwegian Institute for Nature Research, FRAM Centre, Hjalmar Johansens gate 14, NO-9296 Tromsø, Norway ^ Oddatunet, 7057 Jonsvatnet, Norway ‡
bS Supporting Information ABSTRACT: Temporal trends of polyfluoroalkyl compounds (PFCs) were examined in tawny owl (Strix aluco) eggs collected in Central Norway over a period of 24 years (1986-2009). Concentrations of 12 PFCs, including C6-C8, C10 perfluoroalkyl sulfonates (PFSAs), perfluorooctane sulfonamide (PFOSA), and C8-C14 perfluoroalkyl carboxylates (PFCAs), were measured, whereas saturated and unsaturated fluorotelomer carboxylates and shorter chain PFSAs and PFCAs were not detected. Perfluorooctane sulfonate (PFOS) was the predominant compound (geometric mean 10.1 ng/g wet weight (ww)), followed by perfluorotridecanoate (PFTriDA) (0.36 ng/g ww) and perfluoroundecanoate (PFUnDA) (0.19 ng/g ww). Significant decreasing concentrations were found for PFOS with an annual decrease of 1.6% (1986-2009), while, conversely, the C10-C13 PFCA concentrations increase significantly with an annual increase P P of 4.2-12% (1986-2009). Consequently, the contribution of PFOS to the PFCs decreased, whereas the contribution of the PFCAs increased over the time. Toxicological implications for tawny owls are limited, but the maximal PFOS concentration found in this stu0dy is about 20 times lower than the predicted avian no effect concentration (PNEC) which suggest adverse effects caused by PFOS are unlikely. However, tawny owls are exposed to a mixture of various PFCs, and PFCA concentrations still increase.
’ INTRODUCTION Polyfluoroalkyl compounds (PFCs) have been widely used as processing additives during fluoropolymer production and as stain repellents in textile and paper products over the past 50 years.1 The production of perfluorooctyl sulfonyl fluoride (POSF, which is a major precursor for perfluorooctane sulfonate (PFOS)) was voluntarily phased out in 2002, but a variety of related PFCs are still produced by manufacturers.2 PFCs have received increasing public attention due to their persistence, bioaccumulative potential,3 and possible adverse effects on human and wildlife.4,5 As a result, perfluorooctane sulfonate (PFOS) has been added to the persistent organic pollutants (POPs) list of the Stockholm Convention in May 2009 resulting in global restrictions on its uses and production.6 In addition, the European Union (EU) prohibited the general use of PFOS and their derivates after June 2008.7 PFCs have been found globally in wildlife, and besides liver and blood samples, bird eggs have commonly been used to determine PFC concentrations.8-21 Previous temporal trend studies indicated mostly increasing PFC concentrations in biota from the marine ecosystem.22-25 Recently, however, a few temporal trend studies have observed a significant decreasing trend of some perfluoroalkyl sulfonates (PFSAs) and perfluorooctane sulfonamide (PFOSA).26-29 In birds, temporal trend studies are limited,8,15,19,24 but long-term data (1968-2003) from Baltic Sea common guillemot (Uria aalge) eggs have shown an almost 30-fold increase in PFOS concentrations until 2002.8 r 2011 American Chemical Society
Recently, the temporal trends of PFCs was investigated in eggs of peregrine falcon (Falco peregrinus) (1974-2007),19 showing increased PFCA concentrations over the entire study period, while PFOS and perfluorohexane sulfonate (PFHxS) leveled off after the mid 1980s. Most studies of PFCs have, however, concentrated on marine species, while studies on terrestrial species are rare. The aim of this study was to examine PFC concentrations and temporal trends in tawny owl (Strix aluco) eggs from Central Norway. The tawny owl is a nonmigratory and highly territorial species, and among the most common predatory birds in Europe. It is relatively omnivorous, but feeds mainly on small rodents, particularly voles, when they are abundant. They may also feed on passerine birds,30 but in our study area they lay their eggs before migrating birds start to arrive. Thus, the concentrations of different contaminants in their eggs will largely reflect the pollution in the local environment.31 The objectives of this study are to (i) determine the concentrations of three PFC classes (i.e., PFSAs, perfluoroalkyl carboxylates (PFCAs), and PFOSA) in tawny owl eggs; (ii) investigate Special Issue: Perfluoroalkyl Acid Received: October 14, 2010 Accepted: December 20, 2010 Revised: December 14, 2010 Published: January 18, 2011 8090
dx.doi.org/10.1021/es103473v | Environ. Sci. Technol. 2011, 45, 8090–8097
Environmental Science & Technology their composition profile; (iii) examine their temporal trends between 1986 and 2009; (iv) calculate annual changes of the PFCs; (v) compare our results with the contamination of PFCs in bird eggs globally; and finally (vi) provide some toxicological implications for tawny owls exposed by PFCs.
’ EXPERIMENTAL SECTION Sample Collection. The study was carried out in the area surrounding the city of Trondheim (N63.42, E10.23) in SørTrøndelag County, Central Norway. Tawny owl eggs have been collected during the period 1986-2009 (n = 107). Trondheim has a population of 170 000 (2009) with no large industry. More than 100 tawny owl nest boxes have been deployed, and annually each nest box was visited twice. The first visit was in early April, shortly after egg laying, and the nest boxes were revisited in the first half of May. At the second visit, the young had usually hatched and all nonhatching eggs were collected. Eggs were placed into polypropylene bags and frozen shortly after collection to avoid blank contamination during collection and storage. The number of samples varied from 1 (1990) to 10 (2004), except for 1988 when no eggs were found. Chemicals. The target analytes include 23 PFCs (i.e., C4-C14 PFCAs; C4, C6-C8, C10 PFSAs; 6:2, 8:2, 10:2 fluorotelomer carboxylate (FTCA); 6:2, 8:2, 10:2 fluorotelomer unsaturated carboxylate (FTUCA); and PFOSA) plus [13C4]-PFOS and [13C4]perfluorooctanoate (PFOA) as mass-labeled internal standards (IS) and branched perfluorodecanoate (bPFDA) as an injection standard (InjS) (for details see Table S1 in the Supporting Information). All solvents and reagents used in this work were of lichrosolv grade, and were purchased from Merck-Schuchardt (Hohenbrunn, Germany). Extraction and Analysis. Egg samples were homogenized and extracted based on the liquid-liquid extraction method described previously.20,32 Briefly, whole egg samples were homogenized in polypropylene (PP) tubes using an Ultraturax disperser (T 25 basic Ultraturrax, IKA, Germany) with stainless-steel dispersing. Aliquots of 1 g of sample material were spiked with 10 ng of an IS mixture. Samples were extracted with 8 mL of acetonitrile for 30 min in an ultrasonic bath. After centrifugation, an aliquot of 1 mL of extract was used for dispersive clean up with ENVI-Carb (100 mg, 1 mL, 100-400 mesh, Supelco, USA) and glacial acetic acid.32 Finally, a volume of 500 μL was transferred to PP vials and 2 ng of an InjS was added. Prior to analysis, aliquots of 150 μL were diluted with 2 mM aqueous ammonium acetate solution (NH4OAc, 50/50, v/v). The separation and detection of PFCs were performed by liquid chromatography (Agilent 1100; Agilent Technologies, Palo Alto, CA and Waters 1525 μ pump and sample manager 2777, Waters Corporation, Milford, MA) with time-of-flight/quadrupole time-offlight high resolution mass spectrometer interfaced with an electrospray ionization source in a negative-ion mode (HPLC-ESI(Q )ToF-MS) (LCT/Q-TOF micro, Micromass, Manchester, England) as previously described.20,33 Aliquots of 50 μL were injected on an ACE C18 column (150 2.1 mm, 3 μm particle size) (ACT, Aberdeen, U. K.) using a gradient of 200 μL/min methanol and water (both with 2 mM NH4OAc). The initial gradient was set at 50:50 methanol/water, then increased to 85:15 methanol/water (hold for 5 min) and further increase to 99:1 methanol/water (hold for 9 min). Full scan (m/z 165-720) high-resolution mass spectra were monitored throughout the chromatograms to test for matrix interferences. Quantification was done using the internal standard method with [13C4]-PFOS and [13C4]-PFOA. In the sample
ARTICLE
extracts PFOS showed more than one peak in the chromatogram, which is a result of the presence of branched isomers resulting from the production process.34 Only the linear PFOS isomer was quantified, because of the lack of calibration standards for the branched isomers. The area of the branched PFOS isomers was compared with the linear isomer to identify sources of exposure. QA/QC. The analytical quality of the laboratory has been approved in interlaboratory studies.35 As standard procedure, laboratory blanks, method detection limits (MDLs), recoveries, and standard reference materials (SRM) were examined. For each sample, a high-resolution full scan spectra was used to control positive detections (typical mass tolerance 50 ppm). The MDLs were calculated as 3 times signal-to-noise for each compound in the egg samples (MassLynx 4.0, QuanLynx) and ranged between 0.01 and 0.05 ng/g ww for individual PFC. No laboratory contamination for any of the analyzed compound was detected (n = 11). The mean recoveries of the IS were 85 ( 11% for [13C4]-PFOA and 97 ( 12% for [13C4]-PFOA in the egg samples (n = 107). Because of the lack of commercial standard reference materials for PFCs a preanalyzed fish tissue sample with origin from the EU project PERFORCE was used to control the performance of the analytical method with every batch of 10 samples (for details see Table S2 in the Supporting Information).35 In addition to our QA/QC protocol, PFC concentrations were compared between fresh laid and nonhatched eggs, which showed no significant differences (for details see Tables S5 and S6 in the Supporting Information). Statistical Methods. Statistical analyses were performed using SPSS for Windows (version 16) and Microsoft Excel at a significance level of R = 0.05. PFCs, which were detected in over 50% of the analyzed samples, were used for the statistical comparison of means and temporal trends. Data were natural-logarithm transformed prior to statistical analysis to meet assumption of normality and homogeneity of variances. Pearson analyses were used for correlations between individual compound concentrations. Temporal trends were depicted by linear regression analysis of logarithmic transformed mean concentrations using ANOVA tests for each analyte separately. Any measured sample reported lower than the MDL was calculated to be 0.5 of the MDL for statistical analysis. Doubling and half-live times were calculated with t1/2 = ln(2)/m, where m represents the slope of the natural-logarithm transformed egg concentration versus time.
’ RESULTS AND DISCUSSION PFC Concentrations and Composition Profiles. An overview of the PFC concentrations in the tawny owl egg samples from 1986-2009 is shown in Table S3 in the Supporting Information. In this study, 12 of 23 target analytes were found (i.e., C6-C8, C10 PPFSAs, C8-C14 PFCAs, and PFOSA). The geometric mean PFSA concentration was 10.4 ng/g wet weight (ww) (median 11.0 ng/g ww, arithmetic mean 13.0 ng/g ww) for the entire study period. PFOS was the predominant compound with a geometric meanPconcentration of 10.1 ng/g ww (median 10.9 ng/g ww). The PFCA concentrations (geometric mean 1.55, median 1.86, and arithmetic mean 2.15 ng/g ww) were by a factor of ∼7 lower compared to the PFSAs. Perfluorotridecanoate (PFTriDA) was the dominating PFCA, followed by perfluoroundecanoate (PFUnDA) with a geometric mean concentration of 0.36 and 0.19 ng/g ww, respectively (median 0.59 and 0.38 ng/g ww, respectively). In contrast to the PFSAs, only the longerchained PFCAs (C8-C14) were detected due to their lower accumulation potential compared to PFSAs.3 8091
dx.doi.org/10.1021/es103473v |Environ. Sci. Technol. 2011, 45, 8090–8097
Environmental Science & Technology
ARTICLE
Figure 1. Composition profiles of PFCs in tawny owl (Strix aluco) eggs from 1986 to 2009 (n = 107).
The composition profile P was dominated by PFOS with a contribution of ∼83% to the PFCs. The other PFSAs had only a low contribution of ∼2%, while PFOSA, which is a precursor compound of PFOS,36 was only ∼0.2% contributing to the P PFCs. The PFCAs, PFTriDA andPPFUnDA, account for ∼4.9% and ∼3.1%, respectively, to the PFCs. The dominance of PFTriDA and PFUnDA within the PFCA pattern observed in the tawny owl eggs might reflect the influence of the terrestrial food chain by the degradation of their precursor compounds originated from atmospheric transport.26,37 It has been interpreted that the degradation of fluorotelomer alcohols (FTOHs) is the main source of exposure for this pattern,38 while the marine ecosystem seems to lead to a different PFCA pattern dominated by perfluorononanoate (PFNA).20,23,25 However, the odd-even pattern can be changed during the bioaccumulation process.3,39 Interestingly, the contribution of PFOS decreased over time from ∼96% P (1990) to ∼63% (2009), whereas the contribution of the PFCAs increased at the same time from ∼3.2% (1990) to ∼34% (2009) (Figure 1). This decreased contribution could be caused by changes of the emission rates of PFSAs, PFCAs, and their precursors and the influence of ecological factors such as feeding habits or changes in food supply.2,27,40 The isomeric profile of the individual PFCs was compared to identify sources for the exposure of the tawny owls. The electrochemical fluorination (ECF) products contain a mixture of branched (20-30%) and linear isomers and the PFCA pattern showed an almost symmetric distribution of PFCA homologues around PFNA with all chain lengths from C4-C15 (phased out in 2002).2 The current fluorotelomer-based products (TM) contain exclusively linear homologues and usually caused a formation of predominantly even carbon numbered homologues, dominated by PFOA.2 The FTOHs from the TM process can be degraded to PFCAs.38 The absence of branched PFCA isomers in the samples indicates that the TM process could dominate as a source. In contrast, PFOS had a relatively constant contribution of branched isomers over the time (16.0 ( 7.1%, Figure S1 in the Supporting Information) indicating that the exposure originated from the ECF production process. However, branched isomers showed lower blood depuration half-lives which indicate that the isomeric profile can be changed during the bioaccumulation process.39 The role of ecological factors and mechanism of bioaccumulation and metabolism, which can influence the concentration levels, pattern, and isomeric profile of PFCs, are not yet fully understood. But overall, changes in the pattern over time and
different isomeric profile indicate that PFCAs and PFOS might originate from different sources. This is supported by absence of a correlation between the PFOS and PFCAs (p > 0.05, Table S4 in the Supporting Information), while individual PFCAs were significantly correlated with each other (p < 0.01, except for PFTriDA and perfluorodecanoate (PFDA)). The positive correlation among PFHxS, PFDA, and PFUnDA (p < 0.05) indicates that these compounds might have the same origin. In terms of local source contributions, PFCs can be originated from the city of Trondheim with a population of 170 000, whereas there is no fluorochemical industry in the sampling area. Another possibility is that PFCs originated from atmospheric deposition of distance sources (i.e., urban and industrial areas in Europe). However, more studies are required to further elaborate and quantify the contribution from each source type. Global Comparison with Other Bird Eggs PFC Measurements. Previous studies focused mainly on marine bird species,8-17 while only a few studies reported PFC concentrations in terrestrial bird species.18-21 PFOS was generally the dominant PFC in bird eggs, followed by PFUnDA, while other PFCs were found in lower concentration levels or were not analyzed in the past. A global comparison of PFOS and PFUnDA concentrations in marine and terrestrial bird species is shown in Figure 2. PFOS and PFUnDA concentrations in bird species from the Antarctica and North Pacific Ocean are generally 1-2 orders of magnitude lower than in Europe, North America, and Asia. PFC composition profiles and contamination levels are species specific. For example, a comparison among different bird species in Norway and Sweden showed that PFOS and PFUnDA concentrations in common guillemots9 and peregrine falcons19 were about 1 order of magnitude higher than in common eiders (Somateria mollissima)20 and tawny owls (this study), while the concentrations in the latter two species were in the same range. The different concentrations might be caused by different exposure pathways and diet habits. Common guillemots and peregrine falcons feed mainly at high trophic levels, which might be higher contaminated with PFCs than the diets of tawny owls and common eiders which are mainly rodents and invertebrate aquatic animals, respectively.30,41,42 Another reason could be the different half-lives of PFCs in the different species. Interestingly, the pattern in common guillemot eggs changed depending on the location.9 Thus, PFOS is the dominant PFC in Norway and Sweden, while PFUnDA had higher concentrations than PFOS in Iceland and the Faroe Islands. Higher PFUnDA than PFOS concentrations were also found in Gentoo penguin (Pygoscelis papua) and Adelie penguin (Pygoscelis adeliae) from Antarctica,17 8092
dx.doi.org/10.1021/es103473v |Environ. Sci. Technol. 2011, 45, 8090–8097
Environmental Science & Technology
ARTICLE
Figure 2. PFOS and PFUnDA concentrations in bird eggs from species living mainly in the marine/aquatic (blue) and terrestrial (red, italic) environment in ng/g ww: * this study, a ref 8, b ref 9, c ref 10, d ref 18, e ref 11, f ref 12, g ref 21, h ref 13, i ref 19, j ref 14, k ref 15, l ref 20, m ref 16, n ref 17.
Figure 3. Temporal trends of PFSAs in tawny owl eggs from Norway, 1986-2009. The plots display the geometric means (circles) and the median (squares) together with the individual analysis (small dots), the 95% confidence intervals of the geometric means, and a seven-point running mean smoother (dashed line) (n = 107).
whereas in industrial areas (i.e., Europe, North America, and Asia) PFOS was generally dominant (Figure 2). In addition to biological factors, the different pattern may be caused by the different transport pathways. No direct PFC production exists in Norway, thus, the PFC pattern found in the tawny owl eggs might be more influenced by long-range transport in the atmosphere, while industrial areas are more influenced by local sources.37,43 Temporal Trends of PFSAs. PFOS declined significantly in tawny owl eggs between 1986 and 2009, while PFHxS, PFHpS, and PFDS reached a peak in the mid of 1990s and subsequently decreased after ∼1997 (Figure 3, Table 1). PFDS increased significantly between 1986 and 1997 (p < 0.005) and subsequently decreased between 1997 and 2009 (p < 0.013). PFHxS and PFHpS were detected in 66% and 83% of the samples, respectively, but no
significant temporal trend could be observed. These decreasing trends resulted in annually decreasing percentage concentrations of 4.2% for PFHxS, 3.9% for PFHpS, and 6.6% for PFDS (19972009). The PFOS concentration decreased by 1.6% per year between 1986 and 2009 (p = 0.034), and 2.4% per year between 1997 and 2009, which indicate an enhanced decreasing trend over time. The different temporal trends of PFSAs (PFOS vs PFHxS, PFHpS, and PFDS) imply changes of the emission patterns over time. One reason might be that PFOS was phased out in 2002, whereas the other PFCAs were not restricted in their production and uses.2,40 It is interesting to note that PFOS showed periodically high geometric mean concentrations for the years 1987, 1991, 1994, 1996, 2000, 2005, and 2008, which could be due changes of climate variables and feeding conditions. 8093
dx.doi.org/10.1021/es103473v |Environ. Sci. Technol. 2011, 45, 8090–8097
Environmental Science & Technology
ARTICLE
Table 1. Geometric Mean and 95% Confidence Interval (CI) of PFCs (ng/g ww), Annual Rate of Change, Doubling Times, and Half-lives (In Italics) (Years) in the Whole (1986-2009), Early (1986-1997), and Late (1997-2009) Sampling Period in Tawny Owl Eggs (n = 107)a analyte
PFHxS
PFHpS
PFOS
PFDS
PFDA
PFUnDA
PFDoDA
PFTriDA
1986-2009 geometric mean
0.05
0.08
10.1
0.06
0.20
0.19
0.12
0.36
95% CI
0.04-0.06
0.07-0.1
8.85-11.5
0.05-0.08
0.16-0.26
0.14-0.27
0.09-0.16
0.26-0.48
change per year (%)
0.3
-1.7
-1.6
1.5
4.2
12
10
6.7
pb doubling times/ half-lives (year)
NS NS
NS NS