Environ. Sci. Technol. 2005, 39, 80-84
Temporal Trends of PFOS and PFOA in Guillemot Eggs from the Baltic Sea, 1968-2003 KATRIN E. HOLMSTRO ¨ M , * ,† U L F J A¨ R N B E R G , † A N D A N D E R S B I G N E R T ‡ Institute of Applied Environmental Research (ITM), Stockholm University, SE-106 91 Stockholm, Sweden, and Contaminant Research Group, Swedish Museum of Natural History, P.O. Box 50007, SE-104 05 Stockholm, Sweden
Perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) have recently been identified as ubiquitous environmental contaminants. Although they have been produced for 50 years, little is known about when they first appeared in the environment and how their concentrations have changed over time, particularly in response to the phase-out of PFOS, which began in 2000. In this study temporal trends in the concentrations of PFOS and PFOA in the Baltic Sea marine environment were measured using archived guillemot eggs. Samples collected from Stora Karlso¨ (Sweden) between 1968 and 2003 were received from an environmental specimen bank and concentrations of PFOS and PFOA were analyzed using HPLC coupled to ESI-MS/ MS. PFOA was not detected in any of the samples (LOD 3 ng/g), but there was an almost 30-fold increase in PFOS concentrations in the guillemot eggs during the time period, from 25 ng/g in 1968 to 614 ng/g in 2003 (wet weight). Regression analysis indicated a significant trend, increasing on average between 7 and 11% per year. A sharp peak in PFOS concentrations was observed in 1997 followed by decreasing levels up to 2002, but this cannot be linked to the PFOS phase-out, which occurred at the end of this period.
Introduction Fluorinated organic compounds have been manufactured for over 50 years. They form a diverse group of chemicals used in a variety of specialized consumer and industrial products such as surfactants, polymers, and fire-fighting foams (1). The environmental fate of fluorinated organic compounds received little attention for many years, in contrast to chlorinated and brominated organic substances. Lately, however, in connection with the development of a new method for analysis based on LC/MS (2), several perfluorinated compounds including perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) have been identified in the environment. PFOS and PFOA have been detected in human serum and blood (3, 4), surface water (5-7), freshwater and marine biota (8-13), and even in remote areas such as the Arctic. In particular the discovery of PFOS in the blood of nonoccupationally exposed humans has raised concerns about these chemicals, and in May 2000 * Corresponding author telephone: +46 8 6747136; fax +46 8 6747637; e-mail:
[email protected]. † Stockholm University. ‡ Swedish Museum of Natural History. 80
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 1, 2005
the main producer of PFOS announced that they were phasing out their production of PFOS and PFOA (12). PFOA, however, is still in production by other manufacturers. It is not known if, or to what extent, the discontinuation of PFOS production will result in a decrease in environmental levels. For evaluation of the environmental impact of the withdrawal of PFOS from the market and of any possible future restrictions put on PFOS and PFOA, there is a need for time trend information on environmental levels. Such information is also crucial for understanding the relationship between emissions and environmental levels. We are aware of just one temporal trend study for these chemicals. It employed white tailed sea eagles sampled in eastern Germany (11). The results suggest a possible increase in PFOS concentrations with time based on a comparison of the average concentrations in samples from 1979 and the 1980s (25 ng/g wet wt) with the average from the 1990s (40 ng/g wet wt). However, there were only a few samples for each time period and these had highly variable concentrations, so the statistical significance of the postulated time trend was questionable. The authors recommended the use of other matrixes for assessments of time trends. The aim of this study was to describe the development of concentrations of PFOS and PFOA in the Baltic Sea marine environment over time. To this end we analyzed archived guillemot eggs, which are an excellent monitoring tool that has frequently been used in retrospective studies of organic contaminant levels within the environment (14-18).
Materials and Methods Guillemots as a Monitoring Tool. The highest concentrations of PFOS are found in marine animals from the higher trophic levels feeding on fish (19). Such a species in the Swedish fauna is the guillemot (Uria aalge). Guillemots nest in colonies in the Baltic Sea, which has a heavy anthropogenic influence. They feed exclusively on migratory pelagic fish species, thus integrating concentrations of pollutants from a large part of the Baltic Sea. The advantages of using guillemot eggs as a monitoring matrix have been summarized by Bignert et. al. (20) and include the fact that the guillemot is a nonmigratory species with homogeneous feeding habits which promotes a low random within-year and between-year variability of contaminant levels. Guillemot eggs have been sampled and stored under controlled conditions since the 1960s by the Swedish Museum of Natural History. The guillemot eggs used in this study were collected on Stora Karlso¨ (Figure 1), an island in the Baltic proper with few inhabitants, which holds a large colony of guillemots. The island is located 7 km west of the island of Gotland and 80 km east of the Swedish mainland, far from any local sources of pollution. Collection and Processing of Samples. Guillemot eggs were collected annually in early to mid May and transported to the Swedish Environmental Specimen Bank (www.nrm.se/ mg), located at the Museum of Natural History in Stockholm. Eggs were transported in an intact state thus avoiding contamination. At the specimen bank the yolk and white were removed, mixed and homogenized, and deep-frozen. Until 1991 the eggs were kept at -30 °C. Since 1991 the eggs have been kept at -85 °C. Egg sample aliquots were prepared by the Swedish Environmental Specimen Bank and delivered during 2003. A preliminary study indicated that individual variability in concentrations among the guillemot eggs was low, and it was concluded that analysis of 8 individuals per time period was sufficient. In the first part of the study, eight individual 10.1021/es049257d CCC: $30.25
2005 American Chemical Society Published on Web 12/03/2004
FIGURE 1. The Baltic Sea and its drainage basin in dark gray. An arrow marks the sampling location Stora Karlso1 (geographical coordinates: 56° 53′ N, 18° 38′ E).
TABLE 1. Concentrations of PFOS and PFOA in Guillemot Eggs (ng/g Wet Weight) sampling year
n
p/ma
PFOS
PFOA
1968 1971 1973 1976 1978 1981 1983 1986 1988 1991 1993 1996 1997 1998 1999 2000 2001 2003
8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 9 9
p m p m p m p m p m p m p p p p m m
25 44 (34-55) 58 162 (32-351) 169 214 (96-309) 459 233 (195-303) 485 411 (344-487) 501 528 (462-652) 1324 834 1023 871 561 (520-601) 614 (551-669)
n.d.b n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.
a p ) Pool of equal aliquots of eggs from 8 individuals, m ) arithmetic mean value of measurements in individual eggs. The concentration range is given in parentheses. b Not detectable at a detection limit of 3 ng/g.
tert-butyl ether (MTBE) was added. The tube was gently turned for 20 min. The organic and aqueous layers were separated by centrifugation and the MTBE fraction was quantitatively transferred to a second PP tube. Another 5 mL of MTBE was added to the aqueous phase and the procedure was repeated. The two MTBE fractions were combined and gently evaporated until dryness using dry nitrogen. Methanol (500 µL) was added, and the extract was filtered through a 0.46-µm PP-filter into a plastic (polypropylene) vial. Instrumental Analysis. The samples were analyzed using high-performance liquid chromatography combined with electrospray tandem mass spectrometry (LC-MS/MS). Aliquots (10 µL) of extracts were injected onto a C18-precolumn (Chrom Tech, 10 × 2 mm, 5 µm HyPurity) followed by a C18-column (Thermo Hypersil-Keystone, 50 × 2.1 mm, 5 µm HyPurity). The flow rate was 0.2 mL/min delivered by a Waters Alliance pump (Waters Corp., Milford, MA). The mobile phase was 10 mM ammonium acetate/methanol with a gradient starting at 40% methanol, increasing to 95% at 5 min, kept for 15 min, and then returning to 40% at 15 min. Total run time was 35 min, including time for conditioning of the column. PFOS and PFOA were measured using selected reaction monitoring (SRM-MS/MS) with argon as a reaction gas, monitoring the transitions 499-99 for PFOS and 413369 for PFOA. MS/MS conditions (Micromass Quattro II, Altrincham, U.K.) were as described elsewhere (10). Quantification. The samples were analyzed in batches and were quantified using extracted matrix standards. Matrix standards were prepared from a large homogenized guillemot egg sample from 1971 with a low concentration of PFOS that was split into 1 mL subsamples. For each batch, three of these subsamples were spiked with PFOS and PFOA standards at three concentration levels and extracted along with the batch. The concentration levels of the spikes were chosen to match the concentration interval of the samples in the batch. Quantification was done using response factors calculated from a calibration curve constructed from the PFOS and PFOA signals in these extracted matrix standards. This procedure also corrects for recovery. PFOA (fluorooctanoc acid) was purchased from Aldrich Chemical Co. and PFOS was purchased from Fluka. Quality Assurance. To monitor the extraction reproducibility, an aliquot of a large guillemot egg homogenate from 1981 was extracted and analyzed in parallel with each batch (control sample). One of these extracts was also repeatedly injected with every batch run on the LC-MS/MS system to monitor the reproducibility of the system (control extract). Laboratory blanks consisting of Milli-Q water were extracted along with each batch of samples. Solvents were run through the LC-MS/MS system every 10 samples. Injection of samples spiked with standard was always followed by solvent runs to avoid memory effects.
Results and Discussion egg samples were analyzed from every fifth year from 1971 to 2001. The study was then expanded and complemented with analysis of pooled samples of 8 individuals from several intermediate years, starting from 1968. Additionally, a peak in concentration suggested by the pooled 1998 sample was further evaluated by analysis of pooled samples of 8 individual eggs from each of the years 1997, 1998, 1999, and 2000 (Table 1). Chemical Analysis. The extraction method was based on ion pairing as described by Hansen et al. (2). Egg samples were first homogenized in Milli-Q purified water (sample/ water ratio 1:5). A 1-mL subsample of the homogenate was transferred to a polypropylene (PP) tube where 2 mL of a sodium carbonate buffer (1 M) and 1 mL of tetrabutylammonium (TBA) solution (10 mM at pH 10) were added. The resulting mixture was vortex mixed, and 5 mL of methyl-
Quality Assurance. The relative standard deviation of the control samples (n ) 5) was 7% for PFOS. The relative standard deviation of the control extract (n ) 9) was 12% for PFOS. This indicates that the precision of the method is determined by instrumental analysis/calibration and not by the sample extraction and cleanup. PFOS was not present in the laboratory blanks or solvent runs. Traces of PFOA were present in several blanks at concentrations of
