Ecological screening assessment report on perfluorooctane sulfonate, salts and precursors: chapter 3

3. Environmental Concentrations

Air

Martin et al. (2002) measured the air in Toronto and Long Point, Ontario for some precursors of PFOS. They found an average N-MeFOSE alcohol concentration of 101 pg/m³ in Toronto and 35 pg/m³ at Long Point. The average concentrations of N-EtFOSE alcohol were 205 pg/m³ in Toronto and 76 pg/m³ in Long Point. These precursors, N-MeFOSE alcohol and N-EtFOSE alcohol, are relatively volatile, especially for such large chemicals, and they have relatively high octanol/water partition coefficients.

Water

In June 2000, PFOS was detected in surface water as a result of a spill of fire-fighting foam from the Toronto International airport into nearby Etobicoke Creek. Concentrations of PFOS ranging from <0.017 to 2210 µg.L^-1 were detected in creek water samples over a 153-day sampling period. PFOS was not detected at the upstream sample site (Moody et al. 2002). Boulanger et al. (2004, 2005) examined concentrations of PFOS in the Great Lakes. Boulanger et al. (2004) analyzed PFOS in 16 water samples taken at 4 meters depth from 4 sampling sites in each of Lake Erie and Lake Ontario. They found measured arithmetic mean concentrations of 31 (sd = 6.9) ng.L^-1 for Lake Erie and 54 (sd = 18) ng.L^-1 for Lake Ontario. The highest value measured was 121 ng.L^-1. A comparison to worldwide surface water concentrations by Boulanger et al. (2004) showed the data to be in a similar range. In a follow up study, Boulanger et al. (2005) calculated steady state concentrations of PFOS in Lake Ontario using a mass balance approach of 32 ng.L^-1 (sd = 14). It was noted in the mass balance study that inflow from Lake Erie and waste water discharges were the primary sources of PFOS to Lake Ontario with particle and gas phase deposition being a negligible portion of the annual inputs. It should be noted that the relative standard deviation on the annual mass flux from waste water discharge is greater than 100%. Therefore, the exact contribution of waste water discharge is inconclusive. Also the amount of PFOS formed from degradation of PFOS precursors is unclear. While it is expected that PFOS precursors are globally distributed within the atmosphere and will primarily enter ecosystems through wet and dry deposition, the work of Boulanger et al. (2005) suggests that there is the potential for point sources of PFOS to outweigh atmospheric deposition at specific sites. However, global distribution of PFOS precursors and degradation to PFOS in the water column is still considered to be the primary route of entry of PFOS into non-industrially impacted freshwaters in Canada.

US data for PFOS are available from one study of six cities. PFOS was detected in quiet water (i.e., a pond) (2.93 µg.L^-1) and sewage treatment effluent (0.048-0.45 µg.L^-1) and sludge (60.2-130 µg.kg^-1 dry sludge) at cities (Port St. Lucie, Florida, and Cleveland, Tennessee) with no significant fluorochemical activities (US EPA OPPT AR226-1030a111). The Port St. Lucie surface water data for PFOS shows a decreasing trend in concentration (from 51.1 µg.L^-1 in 1999 to 1.54 µg.L^-1 in 2001). Therefore, the data may represent a single contamination event to that water system and the decreasing trend may be a result of natural removal processes. PFOS was also detected in drinking water (0.042-0.062 µg.L^-1), surface water (not detected [n.d.] to 0.08 µg.L^-1), sediments (n.d to 0.78 µg.kg^-1 dry sediment), sewage treatment effluents (0.04-5.29 µg.L^-1) and sludge (57.7-3120 µg.kg^-1) and landfill leachate (n.d. to 53.1 µg.L^-1) of four cities that have manufacturing or industrial use of fluorochemicals. Detection limits were 0.0025 µg.L^-1 for water and 0.08 µg.kg^-1 wet weight (ww) for sediment and sludge. Sediment concentrations appear to be approximately 10-fold higher than water concentrations, indicating that there is a tendency to partition from the water to sediment.

In a recent monitoring study near the vicinity of a fluorochemical manufacturing facility located on the Tennessee River (Alabama), PFOS was detected in all surface water and sediment samples collected. The highest concentrations for surface water (151 µg.L^-1) and sediment (5930 µg.kg^-1 ww; 12 600 µg.kg^-1 dry weight (dw)) were found at a location near the point of discharge of a combined industrial effluent. However, the study found that downstream concentrations were not statistically greater than those upstream and concluded that the combined industrial effluent did not significantly affect fluorochemical (including PFOS) concentrations in the main stem of the river. For the upstream reference site (Guntersville Dam), estimated average PFOS surface water and sediment concentrations were 0.009 µg.L^-1 and 0.18 µg.kg^-1, respectively (US EPA OPPT AR226-1030a161). In another study, low levels of PFOS were found throughout a 130-km stretch of the Tennessee River (Hansen et al. 2002). The average PFOS concentration upstream of the fluorochemical manufacturing facility was 0.032 µg.L^-1. This may indicate an unidentified source of PFOS entering upstream.

PFOS was also detected in oceanic waters from the Pacific and Atlantic Oceans and in several coastal seawaters from Asian countries (Japan, Hong Kong, China, and Korea) (Yamashita et al., 2005). PFOS was detected at concentrations ranging from 1.1 - 57 700 pg.L^-1. PFOS was also observed in the North Sea (estuary of the river Elbe, German Bight, southern and eastern North Sea) (Caliebe et al., 2004). The detection of PFOS in oceanic waters suggests another potential long-range transport mechanism to remote locations such as the Canadian Arctic.

Sediment

Suspended sediment samples were collected annually at Niagara-on-the-Lake in the Niagara River over a 22 year period (1980-2002). PFOS concentrations ranged from 5 to 1100 pg.g^-1 (Furdui et al., 2005, unpublished data). Preliminary findings suggest that PFOS concentrations increased during the study period from < 400 pg.g^-1 in the early 1980s to > 1000 pg.g^-1 in 2002. It was suggested that the presence of PFOS could be due to the fact that the Great Lakes region is heavily industrialized and hazardous waste disposal sites among other sources could potentially contribute to the contamination of Niagara River suspended sediments.

Biota

Appendix 2 presents the levels of PFOS measured in North American and circumpolar wildlife between 1982 and 2005. Recent Canadian Arctic and circumpolar wildlife surveys have detected PFOS and other perfluorinated acids in mammals, birds and fish, including: polar bear, ringed seals, mink, arctic fox, common loons, northern fulmars, black guillemots and fish from various locations in the Canadian Arctic (Martin et al. 2004a; Smithwick et al. 2005a,b). Data are also available for a variety of other species worldwide, including dolphin, turtles, mink, seals, fish-eating birds and oysters (Geisy and Kannan 2002; Kannan et al. 2002a,b).

In Canada, PFOS has been detected in mid- and higher trophic level biota such as fish, piscivorous birds, and Arctic biota far from known sources or manufacturing facilities. Maximum levels of PFOS in liver of Canadian Arctic biota have been reported for mink (20 µg.kg^-1), brook trout (50 µg.kg^-1), seal (37 µg.kg^-1), fox (1400 µg.kg^-1) and polar bear (>4000 µg.kg^-1) (Martin et al. 2004a).

The highest North American or circumpolar concentration of PFOS in mammal tissue reported in the published literature is 59 500 µg.kg^-1 ww in mink liver from USA (Kannan et al., 2005a). The widespread occurrence of PFOS in wildlife worldwide and, in particular, the high concentrations detected in higher trophic level wildlife and the apex predator species, polar bear, are important findings. Smithwick et al. (2005b) reported concentrations of PFOS in a study of polar bears from 7 circumpolar locations (5 North American and 2 European locations). The highest PFOS concentration in Canadian polar bears was 3770 µg.kg^-1 liver (range 2000-3770 µg.kg^-1 liver; mean 2730 µg.kg^-1 liver) found in polar bear from South Hudson Bay (Smithwick et al., 2005b). This data was a re-analysis of polar bear samples from South Hudson Bay conducted by Martin et al. (2004a) which reported concentration in polar bear liver ranging from 1700->4000 µg.kg^-1 liver (mean = 3100 µg.kg^-1 liver, n = 7). The concentrations of PFOS in polar bear liver from the 3 other Canadian locations were: High Arctic 263-2410 µg.kg^-1 liver, mean = 1170; Northwest Territories 982-2160 µg.kg^-1 liver, mean = 1320 and South Baffin Island 977-2100 µg.kg^-1 liver, mean = 1390, respectively (Smithwick et al., 2005b). Polar bears have a very large home range due to their dependence on sea ice for hunting, long-range movements and breeding (Stirling and Derocher 1993). The home range may be 103 000 to 206 000 km² (Ferguson et al., 1999) and young bears may travel up to 1000 km from their mother to establish their home range. Given the very large size of the home range of polar bears, the concentrations in polar bear may reflect integration of exposure over a large geographic area.

PFOS is found in birds worldwide, including birds in Canada and North America (see Appendix 4). PFOS has been found in eagles in the Great Lakes, mallards in the Niagara River, loons in northern Quebec, gulls in the Arctic and in Canadian migratory species from the United States (e.g., common loon in North Carolina). In Canadian or Canada-US migratory species, concentrations have been measured in liver ranging from not detectable to 1780 ppb mean liver PFOS concentration (loon (northern Quebec) and bald eagle, (Michigan)), in blood plasma ranging from <1- 2220 ppb blood plasma in bald eagles and in eggs and egg yolk ranging from 21-220 ppb in double-crested cormorant in Manitoba. In several monitoring studies, PFOS residues in piscivorous water birds were found to have some of the highest liver and serum PFOS concentrations compared to other species (see Newsted et al., 2005). In a study of birds in the Niagara River Region, piscivorous birds (common merganser, bufflehead) contained significantly greater PFOS concentrations than non-piscivorous birds (Sinclair et al., 2005). Preliminary data on temporal trends show an increase in bird PFOS concentrations, in two Canadian Arctic species (thick-billed murres and northern fulmars) from 1993 to 2004 (Butt et al., 2005, unpublished). It is noted that concentrations of PFOS in plasma have been reported in eagle, gulls and cormorants around the Great Lakes and in the Norwegian Arctic ranging from <1 ppb to 2220 ppb. There are no available studies reporting on blood-serum-plasma relationships in wildlife, therefore, it is unclear how concentrations in bird plasma compare to concentrations of PFOS in bird serum. It is also noted that while effect levels are not available on egg or egg yolk basis, the measured concentrations of PFOS in egg or egg yolk of birds in Canada and North America have been reported to range from 21 to 220 ppb.

Worldwide, concentrations of PFOS in plaice (Pleuronectes platessa) liver (7760 µg.kg^-1) from the Western Scheldt estuary (southwestern Netherlands) and ornate jobfish (Pristipomoides argyrogrammicus) liver (7900 µg.kg^-1) from Kin Bay (Japan) are among the highest PFOS concentrations ever reported in wildlife (fish) (Hoff et al. 2003; Taniyasu et al. 2003). These high concentrations may be due to the proximity of a PFOS manufacturing plant (upstream of estuary) and an army base (Kin Bay, Japan) that may use PFOS in fire-fighting operations.